tag:blogger.com,1999:blog-15570684963030094092024-03-24T15:27:17.060+01:00soundcheck's - audio@visedigital audio revisited - another perspectivesoundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.comBlogger37125tag:blogger.com,1999:blog-1557068496303009409.post-7304970223124686192024-01-30T11:55:00.002+01:002024-01-30T11:55:32.077+01:00Raspberry Pi 5 - NEWS: Allo HATs revived<p>Just to let you know.</p><p>I had opened a Trouble Ticket related to the serious Allo driver</p><p>issues (https://github.com/raspberrypi/linux/issues/5843) on RPi5</p><p>about 3 weeks ago.</p><span><a name='more'></a></span><p><br /></p><p>Basically all HAT devices were affected that've been making use of the Allo drivers (some companies are piggybacking on these existing drivers to avoid writing their own drivers). The main issues were related to I2C communication with attached devices. I2C is needed to e.g. control volume or filters on the HATs supporting it - intelligent HATs. The RPi5 caused major changes in that I2C department.</p><p>For about 3 weeks there's been troubleshooting ongoing to get these Allo HATs back to life on a RPi5.</p><p>It's done now. I've got my little ones back to work. Perhaps it'll take some time that the still hot corrections make it into the field out there. Be patient.</p><p>Since the whole audio stack got reshuffled there might be a chance that other devices that were failing might face a 2nd life too.</p><p><br /></p><p>Enjoy.</p><p><br /></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com3tag:blogger.com,1999:blog-1557068496303009409.post-85675123773798229452024-01-12T19:58:00.001+01:002024-01-12T19:58:18.692+01:00Raspberry Pi 5 - The audio interface situation<p>I am usually doing a little regression testing on new hardware and new software, simply to see if my own operating system works flawless.</p><p><br /></p><div class="separator" style="clear: both; text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjrfStzQW537qVSJk100KvJ3aIarRLXu2lEHg-yxMVzQOt2qoIsLOSqjf3ih9B8v8zeKdfdZtViGje7Dr9f8LCwOtx0E_Dao0IvN0_fq6YTxm3tyc8jtViN11iv06pI6hYjI5oT-EOVLayWtwDjGPRijUbt4cJGGDqXCTzg0666fo5h8EAA95qdJWvb2MBX/s4030/pi5-dacs.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="3022" data-original-width="4030" height="300" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjrfStzQW537qVSJk100KvJ3aIarRLXu2lEHg-yxMVzQOt2qoIsLOSqjf3ih9B8v8zeKdfdZtViGje7Dr9f8LCwOtx0E_Dao0IvN0_fq6YTxm3tyc8jtViN11iv06pI6hYjI5oT-EOVLayWtwDjGPRijUbt4cJGGDqXCTzg0666fo5h8EAA95qdJWvb2MBX/w400-h300/pi5-dacs.jpg" width="400" /></a></div><br /><p><br /></p><p><br /></p><p>Over years this business was stable. With the introduction of RPi5 things changed.</p><span><a name='more'></a></span><p><br /></p><p>The brandnew <a href="https://www.raspberrypi.com/news/rp1-the-silicon-controlling-raspberry-pi-5-i-o-designed-here-at-raspberry-pi/" target="_blank">RPi5 RP1 I/O controller</a> brought in a lot of new options. And lot of changes too. In conjunction with the RPI5 there have been serious changes on the OS side.</p><p>Which means especially 3rd party devices and 3rd party device drivers might face some serious challenges. </p><p>Now. I did some testing over the last couple of days.</p><p>Let me quickly summarize the situation. </p><p>Basic DACs, like USB DACs, basic I2S HAT DACs or interfaces should work.</p><p>However. </p><p>Certain HAT DACs that require I2C communication for configuration and management purposes, seem to have serious problems. My entire Allo collection is simply not working. I am in touch with RPi folks over that at github.</p><p>There are manufacturers out there using the Allo device drivers as base for their DACs. These will also fail. </p><p>One example is the IMO excellent DAC HAT from Innomaker. It's still sold all over the place at around $/€60. It's their DAC Pro HAT with ESS Sabre DAC. The issue: Innomaker is using the Allo Katana driver. Consequence: It doesn't work on the RPi5 - at least for now.</p><p>There are scenarios where the old DACs simply no longer work. E.g. if the HATs onboard firmware would have to be updated. If the manufacturer disappeared - like Allo. Or if a manufacturer simply do not want to grant lifetime support for their devices.</p><p>If things can be fixed on the OS side this could become a community effort in some cases. In other cases the pretty skilled RPi folks help out. In other cases you simply end up with a bricked device.<br />Or you stick to your old RPi platform.</p><p>This situation now shows that one of the huge RPi selling points over all these years - the well working 3rd party HAT business - gets currently challenged. I hope that the RPi folks get their act together and come up with a certain compatibility layer for the ten-thousands of old and still useful devices out there.</p><p>Now. If you wondering why your HAT is not working on the RPi5 - you got a clue.</p><p>Enjoy</p><p>PS:</p><p>I'd love to have some more common DACs at hand - especially those that'd need a dedicated driver with I2C communications - from e.g. iqaudio, hifiberry and audiophonics to be able to do a little more testing. Unfortunately I can't afford and I am not willing to buy all these gadgets for testing purposes only. It was it is.</p><div><br /></div><p><br /></p><p><br /></p><p> </p><p><br /></p><p><br /></p><p><br /></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0tag:blogger.com,1999:blog-1557068496303009409.post-86662394226867819342024-01-12T11:10:00.008+01:002024-01-19T10:45:17.071+01:00Raspberry Pi 5 - M.2 NVME SSD<p>
Why M.2 NVME SSD on a RPi5??? </p><p>We might intend to create an even better audio streaming giant !?!? ;)</p><p>I am always looking for increasing efficiency on the system. Efficiency? An M.2-NVME SSD attached via PCIe - it can't get any faster - except you use 100% RAM drives. </p><p>The usual USB
route for SSDs adds another layer and a Sata/USB or M.2/USB conversion on top. There's
a good chance that we can increase the efficiency of the overall system by
going the M.2 route. Beside that the USB DAC remains the only device attached to RPi USB. And perhaps we can save some $ on USB filter gadgets. Yep. Plenty of good reasons.</p>
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEizgn5DpKyoIvWYMKbtLHpZBQ4djo-CM09flzMdTocCLhV_ga-fwy6S0XcMVJejdiOtpUp3zCz4DeKFDDh0k9VJvDgt0vmyqVA9DP-CSvppuknB93qHzw1hd51EBR3IhXNBrDsUSvYKIpzvuyIVArUBj1TolhxUilfNqeuZHgILi9ObFO_cRAvS_Ennzo5y/s2872/pi5-pineberry-m2.jpg" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="2872" data-original-width="2872" height="320" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEizgn5DpKyoIvWYMKbtLHpZBQ4djo-CM09flzMdTocCLhV_ga-fwy6S0XcMVJejdiOtpUp3zCz4DeKFDDh0k9VJvDgt0vmyqVA9DP-CSvppuknB93qHzw1hd51EBR3IhXNBrDsUSvYKIpzvuyIVArUBj1TolhxUilfNqeuZHgILi9ObFO_cRAvS_Ennzo5y/s320/pi5-pineberry-m2.jpg" width="320" /></a>
</div>
<br />
<span><a name='more'></a></span><p><br /></p>
<p>
The Raspberry Pi 5 offers a M.2 port right next to the Wifi antenna. It allows
to attach a M.2 NVME SSD HAT. You need a carrier board for your SSD to go that
route. I currently use a board from
<a href="https://pineberrypi.com/products/hat-top-2230-2240-for-rpi5" target="_blank">Pineberry</a>
and a 2230 SSD. There are more of these kind of boards out there. E.g. from <a href="https://shop.pimoroni.com/products/nvme-base?variant=41219587178579" target="_blank">Pimoroni</a> (on order). Most of these boards are hard to come by in this early phase and many of them are WiP (work-in-progress). Meaning. You'll see updates popping in. </p>
<p>
The RPi 5 allocates a single PCIe lane to that I/O port. The foundation
guarantees PCIe Gen2 5Gbit/s over that port. A not officially supported and not certified Gen3
mode allows even 10Gbit/s. The foundation says operation might get unstable in
Gen3 mode. I'll of course give it a try anyhow. ;)</p>
<p>
The RPi5 requires solid cooling. Therefore I wouldn't like to attach the
carrier board on top (HAT) of or underneath the RPi. The situation: PI heat is also spreaded towards the bottom - the PI circuit board also acts as a cooling device and not to forget the SSD also generates quite some heat. If you put all that underneath the RPI you might create a jammed heat bubble at the bottom.</p>
<p>These considerations led to a setup as shown below:</p>
<p><br /></p>
<div class="separator" style="clear: both; text-align: center;">
<a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjpXNJJPJ3Fk_kTy1m-kmC-ZUpgg91oAjdwQlQhYqIr8IHTCX2F1HmWwh0Ao2mmwpe8hg6rE4_mJ6yzTrDCRJMFyTiEqsJaWSQDicFjPGIf0vZjkDs4Yyq4GOjQYQKFGhRkkzAAv2LPDjJxdYHightFVRuAZCblFVmyJVvHcyekKZy7btJAP01zh2mxxev8/s4032/pi5-pineberry-m2-2.jpg" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="3024" data-original-width="4032" height="240" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjpXNJJPJ3Fk_kTy1m-kmC-ZUpgg91oAjdwQlQhYqIr8IHTCX2F1HmWwh0Ao2mmwpe8hg6rE4_mJ6yzTrDCRJMFyTiEqsJaWSQDicFjPGIf0vZjkDs4Yyq4GOjQYQKFGhRkkzAAv2LPDjJxdYHightFVRuAZCblFVmyJVvHcyekKZy7btJAP01zh2mxxev8/s320/pi5-pineberry-m2-2.jpg" width="320" /></a>
</div>
<br />
<p><br /></p>
<p>
As you can see I put the carrier board on stands. A slight problem with my
case of choice from Geeekom is that I have to dismount the case for connecting the cable
and the FFC cable gets bend quite sharply. That's not good, not good at all. Sharp bends are causing impedance issues on these high-speed links. The foundation specifies a length of 50mm max for the cable! And an impedance of 90R. Unfortunately I haven't seen any HAT manufacturer offering FFC cables as spare parts. These can break easily. Please ask for it - if you order a HAT. The more people ask, the earlier we can order them. </p><p>CAUTION: You don't want to put these flimsy onboard FFC connectors under stress!</p>
<p>And... I already changed my setup to accomplish exactly a lower stressed link. I couldn't live with it</p><p><br /></p><div class="separator" style="clear: both; text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhd-pWuss3hHMef44QI_5X7jHpq7ns2pqBoqW6ycxWqdP4AoNpkvk6yKL5rwdHjnSTKSxjStNUB3mcyw8fn-ngxZoKIjgNGBflp0S8zC-ZW8wMBSDS1u7pRr2_7bXWdbA6lxr0QvbFHUnkDcY71X2uKZgd0g9QepyU3d3pWMhHOCZHX7LtFR7ST9FdXJ_9H/s4030/pi5-nvme-the-notch.jpg" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="3022" data-original-width="4030" height="240" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhd-pWuss3hHMef44QI_5X7jHpq7ns2pqBoqW6ycxWqdP4AoNpkvk6yKL5rwdHjnSTKSxjStNUB3mcyw8fn-ngxZoKIjgNGBflp0S8zC-ZW8wMBSDS1u7pRr2_7bXWdbA6lxr0QvbFHUnkDcY71X2uKZgd0g9QepyU3d3pWMhHOCZHX7LtFR7ST9FdXJ_9H/s320/pi5-nvme-the-notch.jpg" width="320" /></a></div><br /><p>Introducing: The Notch</p><p>I took out my Dremel and cut out a notch from my GeeeKom case. Now the cable floats freely.<br />And I can mount and dismount the cable easily without dismounting the heatsink. Keep in mind. Dismounting the heatsink might lower the cooling efficiency! These cooling pads work best on<br />the first application. Any more mounts/dismounts might lower the efficiency of these pads! </p><p>I further mounted the whole stuff on a wooden base and put some space between the bottom fins and the base. </p><p>Don't let that flimsy FFC (flexible flat cable) wriggle around too much. You can easily stress it out.</p><p>The good part is that up to 5W of power can be sourced over the ribbon cable.
There's no need for another external power supply. You'd need to know the power consumption of your SSD to see if it works.</p>
<p>
The bad part is that you can not choose between external OR internal powering with the Pineberry board.
If you attach power through the GPIO header you'd face two concurrent power
paths. </p><p>CAUTION: If you just attach power to the HAT in above setting (I did - because Pineberry told me the board would recognize the power source - cable or header) - you'll power the PI through the FFC cable!<br />You better avoid that to happen!</p><p>Now. There are two more aspects to consider. First. Cooling of SSDs is pretty much standard nowadays.<br />Pretty much every external case you buy offers SSD cooling pads and uses the case as cooling device. None of the M.2 HATs I've seen offers a cooling option. Perhaps the manufacturers should put a thought on that. Second. This open cabling and open HAT board is all but a great solution when it comes to EMI/RFI and direct access. A well covered USB-M.2 case and cable would look like a much preferred solution to me here.</p><p>What I also don't like on the Pineberry board is that you can not turn off the LEDs. The guy from Pineberry simply avoided to answer my related question. </p><p>Let's think a bit about the handling. Before you start the mounting exercise you might first ask yourself how to
get the OS on the SSD!!!<br />And how to run the backups afterwards. My
Waveshare-IO board that I used in the past, allowed me to boot the CM4 in OTG gadget mode from the PC. This way it
was easy to install and backup anything from the M.2-SSD without unmounting the
SSD. It's different here.
</p>
<p>I see two options.</p>
<p></p>
<ol style="text-align: left;">
<li>
You dismount the SSD, attach it to a M.2-USB adapter and run the
installation or backup
</li>
<li>
You boot from an USB or SD drive with a second image and transfer data back
and forth from that drive to the M.2 SSD
</li>
</ol>
<div>Yes. It's gonna be annoying. I'm just saying...</div>
<div><br /></div>
<div><br /></div>
<div>
Now. To be able to boot the M.2 SSD some minor configurations need to be done.
The official related documentation can be found over at
<a href="https://www.raspberrypi.com/documentation/computers/raspberry-pi-5.html" target="_blank">the RPI Docs pages</a>. I'd like to quickly summarize it.
</div>
<div><br /></div>
<div>1. Make sure you've got the latest eeprom firmware installed.</div>
<div>2. Make sure you've got the latest system update incl. latest kernel installed.
</div>
<div><br /></div>
<div>3. We then need to edit the /boot/firmware/config.txt file </div>
<div><br /></div>
<div>In /boot/firmware/config.txt we need to add</div>
<p></p>
<blockquote>
## enable the external PCIe Gen 2.0 output<div>dtparam=pciex1</div>
</blockquote>
<blockquote>
<div>
## force Gen 3.0 speeds (optional) - remove the # in the following line
</div>
<div>#dtparam=pciex1_gen=3</div>
</blockquote>
<p></p>
<div><br /></div>
<div>4. The rpi-eeprom needs to be edited. Following two new lines are required</div>
<blockquote><div>BOOT_ORDER=0xf146 </div>
<div>PCIE_PROBE=1 </div></blockquote>
<p></p>
<p>The RPi will now try to boot the M.2 (6) first, then USB (4) and then the SD (1).</p>
<p>
If you want to use option 2. of my "how to get the data on and off the SSD"
topic, meaning, you want o boot from USB to backup the NVMe drive you need to do some configurations.</p><div>1. The rpi-eeprom needs to be edited. </div><div><br /></div><blockquote><div>BOOT_ORDER=0xf164 </div><div> </div></blockquote><p></p><div>Yep. You need to remove (for now) the PCIE_PROBE=1 entry. Otherwise the NVME will not be shown after USB boot. The RPi then first
tries to boot the USB. </div>
<p>To edit the eeprom accordingly you simply run </p>
<blockquote>
<div>sudo -E rpi-eeprom-config --edit</div>
<div>sudo reboot</div>
</blockquote>
<p><br /></p><p>NOTE: You can't use an USB drive installation that's been the exact copy of the NVME drive.<br />Both drives will than have the same UUIDs - Unique Identifiers - configured, which will cause confusion on the system during the boot process!</p><p> </p><p>Once you're finished with the backup or restore you need to edit the config once more.</p><h2 style="text-align: left;">Conclusion</h2><p>The whole thing IMO still is WiP (Work-in-Progress). HATs are updated as I am writing these lines (I just learned that from Pineberry), and the RPi kernel folks running this or that regression and update on the kernel and firmware part. </p><p>Having a naked board hanging around is not such a great setting. </p><p>The whole backup and restore process is a bit flaky and annoying.</p><p>Attaching an external power supply almost smoked my setup. It's IMO still worth to consider external powering for offloading the PI rails. </p><p>Bottom line. I'd closely follow the developments before hitting the "buy" button. If you can wait a little - I'd say from Q2-3/2024 onwards things might become more attractive and stable. (I know - manufacturers are watching.)</p><p>However. </p><p>The - my - main goal - to increase the system efficiency and to offload the USB bus - was achieved. I am quite happy with that result. </p><p>BUT. All the effort and restrictions around the whole project - at this point in time - IMO is not worth the hassle.</p><p>Enjoy.</p><p> </p>
<p><br /></p>
soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0tag:blogger.com,1999:blog-1557068496303009409.post-74482425131366524662024-01-08T09:22:00.003+01:002024-01-08T09:46:02.628+01:00LMS-squeezelite - great corrections on the way<p>Are you running squeezelite with a super large replay buffer??? </p><p>I am recommending such a setup in public for several years - since the beginning of squeezelite-times to be exact. I've been running and suggesting full-track-ram-caches even before squeezelite times all over the places and even got disbarred from the JR River Media Center forum for proposing such a BAD thing - around 15??? years back. Meanwhile full-track caching is a pretty common way to handle tracks in the streaming world. </p><p>Truth to be told. There were some slightly annoying flaws associated to full-track-ram-caching in the LMS-squeezelite environment... ...and there's been a recent change. And I thought it is worth to write about it.</p><span><a name='more'></a></span><p><br /></p><p>Just for you to recall. What full-track-ram-caching does is following. An entire track gets pre-processed and loaded into a RAM cache on squeezelite - the client - at the beginning of replay. That takes just a few (single digit) seconds and causes quite a high CPU load during these few seconds.</p><p>The advantages are at least twofold</p><p></p><ol style="text-align: left;"><li>You don't see continues network load, which you'll see if you run small buffers with squeezelite</li><li>You lower your base CPU load during playback. In my case instead of 2-3% continuous I see something between 0 and 1%. Substantially lower! </li><li>(For the die-hard environmentalists out there: You can even save some power.)</li></ol><div>Overall we're seeing a lower impact on the chain during replay. A higher efficiency - the main goal of most of the optimizations you'll find on this blog. </div><div><br /></div><div>Limitation: If you run a very low key streaming client with limited RAM this won't work. </div><div><br /></div><div>Now. For years there has been a weird behaviour associated to this setup.</div><div><ol style="text-align: left;"><li>Somehow the playlist management of LMS got mixed up. LMS couldn't handle a fully <br />buffered track. It messed up the playlist pointers. E.g. if you would have used "Add Next", the server would not have played that next track. There have been more of related issues.</li><li>Another issue was that long tracks were cut off a couple of seconds before replay ended</li></ol><div>These issues "were" annoying. Still I could live with it.</div></div><div><br /></div><div>There've been several reportings and discussions about the subject over at squeezelite forums and github. For a long time these issues were simply ignored. Recent developments in streaming, where the bulk-loading of tracks into a cache becomes common practice might have changed the thinking. I also communicated above advantages more than once out there.</div><div><br /></div><div>A fellow developer aliased as philippe44, who's contributing quite a lot to LMS, recently took the challenge to look into the subject. He dissected the whole process on LMS and squeezelite. He had to learn how this LMS dinosaur (never touch a running system - especially if it is called LMS) works in this crucial area. </div><div><br /></div><div>I don't know how much time it took him to figure it all out. Some long lonesome hours...</div><div><br /></div><div>He finally found several related issues. On LMS AND squeezelite. And yes - he figured out potential solutions, got these implemented and made them public.</div><div><br /></div><div>Yesterday I installed the latest LMS-github version and the updated squeezelite.</div><div><br /></div><div>And... ...Wow, It works. The entire weird behaviour is gone. What a relieve. A late Christmas gift. What a great achievement. I love it. I would have never expected that someone ever touches this lower communication layer of LMS and squeezelite. It's extremely tricky. You could mess up (ten-)thousands of streaming installations out there. And now it works flawless. I've been hoping for this for years.</div><div><br /></div><div>These specific corrections of course impact just a small group of die-hard tweakers who are following my tracks for now. I am pretty sure though this group of people will love it. Some other might even feel tempted to jump on that train. ( Also highly recommend for the audiophiles out there to give it a try)</div><div><br /></div><div>Most of you have to wait until these corrections make it into public. You'd at least need a squeezelite version 2.x and a nightly LMS version later than January 7th 2024. Those of you who run my optimized squeezelite versions from github can update their installations already now. I merged the changes yesterday.</div><div><br /></div><div>Enjoy.</div><div><br /></div><div><br /></div><div><br /></div><p></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com1tag:blogger.com,1999:blog-1557068496303009409.post-21160460705018449202024-01-05T18:43:00.008+01:002024-01-06T10:54:00.557+01:00Raspberry Pi 5 - a new audio streaming giant<p>
Since mid of November 2023 I own the recently launched Raspberry Pi5 with 8GB
RAM. Not that I was unhappy with the RPi4 based setup. Though after looking at the
specs and knowing what higher performance gear can bring to
the table, there was no way not to try the new toy. </p><p>Note: Before you run and order the device, you might want to read the whole article. In any case also order a cooling case at the same time! The RPi5 needs cooling!!! <br />
</p>
<p></p>
<div class="separator" style="clear: both; text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhNGyltxR9bl9a6_Nt10AGWc3Pw6nx6tOZsH578Hb7sgAgKJuwqy6EVquH1JwrVsmgn9axYuzUhNXz8Z9cqVDNevyg_oM6-5IuH9uDT2_rzC7diuDergsdMsdZiSMm6XDlb48-KB86bhicdXUUqoS7OhyqFULh0Eyc9NdVqhcyANvVCSB7C3g7d7BdCq8q4/s2160/pi5-1.jpg" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="2085" data-original-width="2160" height="309" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhNGyltxR9bl9a6_Nt10AGWc3Pw6nx6tOZsH578Hb7sgAgKJuwqy6EVquH1JwrVsmgn9axYuzUhNXz8Z9cqVDNevyg_oM6-5IuH9uDT2_rzC7diuDergsdMsdZiSMm6XDlb48-KB86bhicdXUUqoS7OhyqFULh0Eyc9NdVqhcyANvVCSB7C3g7d7BdCq8q4/s320/pi5-1.jpg" width="320" /></a></div><br /><span><a name='more'></a></span><p><br /></p><p>If you look at the specs you'll see that the RPi5 betters the RPi4 on
almost all aspects. Here are my key picks:
</p>
<p></p>
<p></p>
<ul style="text-align: left;">
<li>around twice as fast as the RPi4</li><li>fast LPDDR4X RAM</li>
<li>it comes with 4 PCIe lanes now</li><li>it allows to attach a M.2 NVME HAT via single PCIe lane</li>
<li>2 x USB3.0 ports that offer simultaneous 5Gbit/s transfers</li><li>a better motherboard - that's needed to cope with the higher speeds and throughput</li><li>( the vastly improved GPU performance won't affect headless audio operations)</li>
</ul>
<p></p>
<p>What's not really nice:</p>
<p></p>
<ul style="text-align: left;">
<li>IO port change - ethernet went to the left (DejaVu?)</li>
<li>runs at higher temperatures - passive cooling still possible though</li>
<li>higher power consumption under load</li>
<li>same power consumption @idle - modern devices should run lower!</li>
<li>still USB3.0</li>
<li>still no Wifi 6</li><li>Wifi antenna as poor as the RPi4 antenna and still no option to go external</li>
<li>due to cooling needs, working with HATs is becoming tricky</li>
</ul>
<p></p>
<p></p>
<p>The software side:</p>
<p></p><ul style="text-align: left;"><li>RPiOS Bookworm supports the RPi5 out of the box - status: WIP (work-in-progress) </li><li>the
firmware is still WIP</li><li>some features are not available (e.g. LED off) </li><li>things got - as expected - a little different and a bit more complicated with this device and RPi OS Bookworm</li><li>(I am not sure which audio OSes properly support RPi5 already now)</li></ul><p></p>
<p>It took me about a month to get my own software setup (OS, rt-kernel, custom apps, optimizations)
properly shaped up. And there's still stuff to do. </p>
<p>Power:</p>
<p>The early announcements were shocking. It seemed none of the old power supplies could be used anymore. In real life things look a bit different - at least for my use-case. With SD-card and a basic boot and clock setup, I see around 750+ mA short term
peak during boot. Around 550mA @idle. Which is very close to the RPi4. The community, including me, thinks it should
be less for a modern chip. The RPi designers response was something like "give
us some time, Rome was not built in a day". OK. That's RPi world. Let see what's gonna happen.</p>
<p>
When booting from SSD via USB-Sata cable I had to add around 200mA to the
bill.
</p>
<p>
The power chip seems to limit USB current to 600mA on non-compliant power
supplies. You can disable that safety limitation manually in config.txt (I
did) if you know how.
</p>
<p>I btw attached the RPi5 to me Allo Shanti 3A output. That works rock solid.</p>
<p><br /></p><p>Temperature:</p>
<p>
@1500MHz idle and the GeeekPi passive cooling case I ran @ around 35 °C. @2300MHz and force_turbo=1 I
hit around 54°C @21°C room temperature. </p><p>That's why I am quite happy with the passive cooling for my use-case. I don't want electrical fan-noise mess up my power rails.</p><p><br /></p>
<h2 style="text-align: left;">RPi 5 as audio streamer</h2>
<p>My RPi 5 audio chain setup looks currently like this:</p>
<p></p>
<ul style="text-align: left;">
<li>RPi5 8GB RAM with GeeekPi case</li>
<li>Samsung SSD hooked up to USB3 via SATA-USB adapter</li>
<li>Allo Shanti 5V/3A supply</li>
<li>iFi USB filters on DAC and SSD</li>
<li>
IanCananda ShieldPi Pro (GPIO filter and buffer) attached via 90° GPIO
adapter
</li>
<li>Gustard A18 hooked up to USB3</li><li>my own OS, realtime-kernel, custom LMS, custom squeezelite, optimizations</li>
</ul>
<p></p>
<p>If I talk sound experience in this article, it'll relate to above setup.</p><p>What's next on the hardware side:</p>
<p></p>
<ul style="text-align: left;">
<li>M.2 NVME HAT </li>
</ul>
<p></p>
<div class="separator" style="clear: both; text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhVjmNU1Fy8jiD2dr2S3zHjp8aygMBlxY9bz7-QQG6W2Lr0ZIlPSNQyY0X786y4k6zPhclXfMfAvTkm8NDgut85N-Q6VIJbG8sYyUI3_1bQkq0OH4RnGGjeN5Ik___giLpFEA4QIO8D2YhwF719VUWG_LOLzs8qtbFoBtQn0QFyxGdWqUg7bMd3eJIl3OkM/s2872/pi5-pineberry-m2.jpg" imageanchor="1" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="2872" data-original-width="2872" height="320" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhVjmNU1Fy8jiD2dr2S3zHjp8aygMBlxY9bz7-QQG6W2Lr0ZIlPSNQyY0X786y4k6zPhclXfMfAvTkm8NDgut85N-Q6VIJbG8sYyUI3_1bQkq0OH4RnGGjeN5Ik___giLpFEA4QIO8D2YhwF719VUWG_LOLzs8qtbFoBtQn0QFyxGdWqUg7bMd3eJIl3OkM/s320/pi5-pineberry-m2.jpg" width="320" /></a></div><br /><p><br /></p><p>Using a M.2 NVME SSD might surpass the USB SSD setup. On my CM4 with M.2 NVME it's been the best option. I am looking forward to give that a try. But first I need to get a little more used to my USB-Sata setup. One step at a time.</p><p><br /></p><p>HINT:</p><p>If you intend to build a streamer I strongly recommend to look for a GPIO power filter and buffer (see image above). It'll not just filter noise from your power rail, it'll also stabilize the power rail. Usually it means more stable RPi performance under changing load conditions. It IMO even betters RPi installations with high quality power supplies like the Allo Shanti. </p><p>Here are the ESR specs for the IanCananda ShieldPi: ESR: 2mΩ@200KHz, 0.3mΩ/@2Mhz, 1mΩ@20MHz, 4mΩ/@200MHz . That's pretty impressive.</p><p>And for the audiophiles around: You might even hear it. I do.</p><p><br /></p>
<h2 style="text-align: left;">Conclusion </h2><p>All I can say. On the first glance not much has changed. It's still a RPi we're talking about.</p><p>Many people will ask - "Why swap? My RPi4 runs at 2% CPU load while replaying!?!?" And that's correct. For basic streaming purposes the doubled speed is not required. I btw can confirm the doubled speed. Compiling the Logitechmedia Server went down from about 20 to 10 minutes. ;)</p><p>Do not forget though. All parts of a device and chain contribute to the audio performance. Overall quality, efficiency and speed of a device can play a big role in the game. </p><p>Still you need to try it yourself. Systems are different, expectations are different. </p><p>Finally. </p><p>What about perceived sound quality - any changes on that front ?</p><p>From an audiophile perspective the new RPi5 based setup beats my RPi4/CM4 setups hands
down. I consider the new setup a worthwhile step up on the audiophile ladder. What I achieved exceeded my expectations.<br />The sound stage became deeper, wider, cleaner, more 3D/airy. The low-end sounds ended up to be pretty impressive clean and dynamic. Overall the sound once again became more neutral and transparent. </p><p>Yep. It seems there's no end to it. It somehow always gets a little better.</p><p>Keep in mind though. I am running an exclusive platform that's highly optimized for audio. There's no other RPi based SW platform I am aware of out there that would be able to challenge my setup. If you decide to go for an RPi5 you might experience lesser gains. </p><p>In my experience every little change still makes an audible difference, even on my rather modern DAC and also the new RPi. If you change the CPU clock frequency (I did), if you use a different SSD (I did) , asf, asf - you might hear slight differences. Different platform setups will perform differently though. Keep that in mind!</p><p>All I can say. From my audiophile perspective I consider the whole project a major
success already now. There's simply no looking back. My entire RPi4 fleet is sent to retirement.</p>
<p><br /></p>
<p>Enjoy.</p>
soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com2tag:blogger.com,1999:blog-1557068496303009409.post-63301012114373882422023-02-02T16:24:00.004+01:002023-02-03T15:21:21.478+01:00Chattin' with Chat - my first contact<p>ME and ChatGPT - some interesting, some funny dialogs I'd like to share</p><p>A pretty amazing machine that ChatGPT. Enjoy.</p><span><a name='more'></a></span><p><br /></p><p><span></span></p><div><b>ME:</b></div><p><b>Can you tell me a sad good night story about an audiophile entering the audiophile heaven ?</b></p><p><br /></p><p>ChatGPT:</p><p>Once there was an audiophile who lived a life filled with passion for music. He spent countless hours listening to his favorite records, testing different audio systems and tweaking them to perfection.</p><p><br /></p><p>One day, he passed away and found himself at the gates of audiophile heaven. He was overjoyed to see that the place was filled with the most beautiful and perfect sound systems he could ever imagine.</p><p><br /></p><p>As he walked through the halls, he noticed that no one was playing any music. He approached one of the angels and asked why. The angel replied that they were waiting for the perfect audiophile to come and bring the music to life.</p><p><br /></p><p>The audiophile was filled with excitement as he realized that he was the one they were waiting for. He picked up a record and put it on the turntable, eager to hear the first notes of the music.</p><p><br /></p><p>However, as soon as he dropped the needle onto the record, he heard a crackling sound. He quickly realized that his hearing was not what it used to be and he was no longer able to hear the music with the same clarity and perfection as before.</p><p><br /></p><p>The audiophile was heartbroken as he realized that he could never fully enjoy the beautiful sound systems of audiophile heaven. He spent the rest of his days there, listening to the crackling sound, feeling sad that he could never experience the full beauty of the music again.</p><p><br /></p><p>And so, the audiophile entered the audiophile heaven, but never truly found peace, as he was never able to fully immerse himself in the beauty of the music once again.</p><p><br /></p><p>*********************************************************************************</p><div><b>ME:</b></div><p><b>Can you tell me a funny good night story about an audiophile entering the audiophile heaven ?</b></p><p><br /></p><p>ChatGPT:</p><p>Sure! Here's a story:</p><p><br /></p><p>An audiophile died and went to audiophile heaven. When he arrived, he was greeted by St. Peter who asked if he had any special requests. The audiophile replied, "Yes, I want to hear the most perfect sound in the universe."</p><p><br /></p><p>St. Peter smiled and led the audiophile to a door. The audiophile opened the door and was greeted by the most beautiful music he had ever heard. The sound was so perfect that tears streamed down his face.</p><p><br /></p><p>Suddenly, the audiophile realized that he was standing in front of a large speaker. He looked to his side and saw that there was a small speaker next to him. He looked back at St. Peter and asked, "What's this?"</p><p><br /></p><p>St. Peter replied, "This is the auxiliary speaker. It's just in case you need to adjust the bass."</p><p><br /></p><p>The audiophile smiled and said, "I never knew heaven had such good sound." And he settled in for an eternity of perfect sound.</p><p><br /></p><p>The end! I hope you found this story amusing.</p><p><br /></p><p>*********************************************************************************</p><div><b>ME:</b></div><p><b>Can you tell me a joke about a Golden Ear meeting an engineer talking about the latest greatest digital analog converter (DAC) ?</b></p><p><br /></p><p>ChatGPT:</p><p>Sure! Here's a joke for you:</p><p>A Golden Ear walks into a room where an engineer is talking about the latest greatest digital analog converter (DAC). The Golden Ear interrupts and says, "Excuse me, but I have a highly refined sense of hearing and I can tell you that this DAC is the best one I've ever heard." The engineer raises an eyebrow and says, "Really? How can you tell?" The Golden Ear replies, "I can just hear the difference, it's obvious." The engineer then asks, "Well, can you tell me the difference between this DAC and a cheaper one?" The Golden Ear thinks for a moment and then says, "Well, um, the price is definitely different."</p><p><br /></p><p>*********************************************************************************</p><p><b><i>Let's try something different</i></b></p><p>*********************************************************************************</p><div><b>ME:</b></div><div><b><br /></b></div><div><div><b>Can you write a technical review about a Topping D90 DAC versus a SMSL SU-9 DAC ?</b></div><div style="font-weight: bold;"><br /></div><div>ChatGPT:</div></div><div><br /></div><div>The Topping D90 and the SMSL SU-9 are both highly regarded digital-to-analog converters (DACs) that are popular among audio enthusiasts.</div><p><br /></p><p>The Topping D90 is a high-end DAC that boasts a powerful XMOS XU208 chip and supports a wide range of file formats, including PCM, DSD, and MQA. It has a built-in amplifier and supports up to 32-bit/384kHz audio playback. It also has a sleek, compact design and multiple connectivity options, including USB, coaxial, and optical inputs.</p><p><br /></p><p>The SMSL SU-9 is also a high-quality DAC that uses the same Sabre Hi-Fi 9038 Pro chip as the Topping D90. It supports up to 32-bit/384kHz audio playback and offers both coaxial and optical inputs, as well as a USB input that supports both USB Audio Class 2.0 and DoP (DSD over PCM). The SU-9 also has a compact design and a built-in amplifier.</p><p><br /></p><p>Both the Topping D90 and the SMSL SU-9 are highly rated for their sound quality and build quality. However, some users might prefer the Topping D90 for its wider file format support and its more versatile connectivity options, while others might prefer the SMSL SU-9 for its slightly more affordable price and its use of the same Sabre Hi-Fi 9038 Pro chip as the D90. Ultimately, both are excellent choices for anyone looking for a high-quality DAC.</p><p>*********************************************************************************</p><div><p><b>ME:</b></p><p><b>What means SNR, THD+N, DNR in the audio world ? </b></p><p><br /></p><p>ChatGPT:</p><p>SNR stands for Signal-to-Noise Ratio, which is a measure of the ratio of the desired signal to the background noise in an audio system.</p><p><br /></p><p>THD+N stands for Total Harmonic Distortion plus Noise, which is a measure of the total amount of harmonic distortion and noise present in an audio system, expressed as a percentage of the signal amplitude.</p><p><br /></p><p>DNR stands for Dynamic Range, which is a measure of the difference between the highest and lowest levels of an audio signal that can be recorded or played back without introducing unwanted distortion or noise.</p><p><br /></p><p><b>ME:</b></p><p><b>What is SINAD ?</b></p><p><br /></p><p>ChatGPT:</p><p>SINAD stands for Signal-to-Noise and Distortion Ratio, which is a measurement of the quality of an audio signal. It is calculated as the ratio of the desired signal level to the sum of all other electrical noise and distortion present in the signal. SINAD is expressed in decibels (dB) and is used as a figure of merit to assess the performance of an audio system, such as an amplifier or ADC (analog-to-digital converter). The higher the SINAD value, the higher the quality of the audio signal, as there is less noise and distortion present in the signal.</p><p><br /></p><p><b>ME:</b></p><p><b>What is jitter and phase noise and how is it measured ?</b></p><p><br /></p><p>ChatGPT:</p><p>Jitter and phase noise are types of noise that can occur in digital signals, particularly in clocks that drive digital circuits.</p><p><br /></p><p>Jitter refers to the small variations in the arrival time of consecutive clock pulses. This variability can cause problems in digital circuits, as it can result in incorrect data recovery and decreased system performance.</p><p><br /></p><p>Phase noise is a type of jitter that refers to the short-term fluctuations in the phase of a clock signal. Phase noise is particularly important in applications that require high-frequency signals, such as high-speed digital communication systems and frequency synthesizers.</p><p><br /></p><p>Both jitter and phase noise can be measured using various techniques, including:</p><p><br /></p><p>Time interval error (TIE) measurement, which measures the deviation of individual clock pulses from a ideal clock waveform.</p><p>Allan deviation measurement, which measures the stability of a clock signal over time.</p><p>Phase noise measurement, which measures the spectral density of phase fluctuations as a function of frequency offset from the carrier frequency.</p><p>The unit of measurement for jitter and phase noise is usually expressed in picoseconds (ps) or hertz (Hz). Lower values indicate lower jitter and phase noise, which results in better signal quality.</p><p><br /></p><p><b>ME:</b></p><p><b>Which level of jitter in picoseconds is considered to still be audible?</b></p><p><br /></p><p>ChatGPT:</p><p>The level of jitter that is considered to be audible is a subject of ongoing research and debate in the audio industry. However, most experts agree that jitter levels below 1 nanosecond (ns) are generally inaudible to the average listener. Some studies have suggested that jitter levels as high as 10 ns can still result in audible effects, while others have shown that jitter levels as low as a few picoseconds can be audible in certain circumstances.</p><p><br /></p><p>It's important to note that the audibility of jitter is highly dependent on the specific audio system, as well as the listener's own perception and listening conditions. In general, high-end audio systems that use high-quality clock sources and minimize jitter are less likely to produce audible jitter-related effects.</p><p><br /></p><p>**********************************************************************************</p><p>That's it for now. IMO a pretty good start. </p><p><br /></p></div>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0tag:blogger.com,1999:blog-1557068496303009409.post-37129525170100109802023-02-01T19:04:00.011+01:002023-02-08T14:43:17.654+01:00Touch Toolbox on github<p>Once in a while I am still receiving inquiries of fellow audio folks asking for my Squeezebox Touch Toolbox.</p><p>Squeezebox Touch, a device that's been sold until 2011, if I recall correctly. The Toolbox is pretty much of same age. Time is flying by. It feels like a different life back than. </p><div class="separator" style="clear: both; text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg195O3LcNy4wsOXNQ1wxve8ZTZdjBHSkhM4-pr-HOrrlLp3fhmzKjg2lqL9ZfgT3MlX5pqr-Yy18JR8eivtx26nhGW-dS3fxJMKzm7vaHFWM1RM0AB42e6tf7rutL-nSkXzyRgaftf7C0b9amefF45Q5rHS1loPh6ByeQequEbIwDi99oNeVUuYIQIjg/s2016/SBT-20230207-SC.jpg" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="1512" data-original-width="2016" height="240" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEg195O3LcNy4wsOXNQ1wxve8ZTZdjBHSkhM4-pr-HOrrlLp3fhmzKjg2lqL9ZfgT3MlX5pqr-Yy18JR8eivtx26nhGW-dS3fxJMKzm7vaHFWM1RM0AB42e6tf7rutL-nSkXzyRgaftf7C0b9amefF45Q5rHS1loPh6ByeQequEbIwDi99oNeVUuYIQIjg/w320-h240/SBT-20230207-SC.jpg" width="320" /></a></div><br /><p><br /></p><p>Fun fact. I am sure that this ancient Toolbox will still have a slight impact on modern DACs. Yep. Unfortunately we still face similar challenges on our digital gear after all these years.</p><span><a name='more'></a></span><p>The last toolbox version that I had out in the wild, was Touch Toolbox 3.0. That one I now uploaded to <a href="https://github.com/klslz/tt" target="_blank">github</a>. </p><p>You'll find </p><p></p><ul style="text-align: left;"><li>a related document (background and setup) and </li><li>the toolbox files </li></ul>up there. <p></p><p>You should know how to download the stuff from git. Make sure to also apply the server tweaks I've been pointing out within the Audio Streaming Series.</p><p>Of course, there's no support for the Toolbox from my side. Consider the upload and the rough document review and update a favour from my side. I hope it all works out for you.</p><p>Enjoy.</p><p><br /></p><p>Update 2023 - Feb - 7: </p><p>A little story.</p><p>I found my old Squeezebox Touch buried under a wild heap of old electronics. It looked pretty frayed.<br />Shall I give it a try !?!? Hmmh. I got curious. Hooked up to 5V the screen came up. Wow. Great after 10y. Within seconds it turned off again. WHAT !?!? Aah. I recalled. That's been the Toolbox feature "Screen OFF". Obviously I left the Touch Toolbox on when I trashed the device about 10 years ago. Next thought. Let's do a HW reset to default to factory settings. After pushing the button, nothing happened. Hmmh. Quickly, it dawned on me. That HW reset button got broken 10 years back from all the numerous resets I had to run during the Toolbox design phase. Q: Shall I open it up, perhaps I can run the reset from inside. !?!? Yep. Why not. T's been a 30 minute effort - wasted! Now I could look at the broken switch. Great. I then recalled that I had implemented a SW initiated factory-reset in the Toolbox. Ok. Let's try it that way.<br />Next task: ssh login. Puh.<br />Nice. The IP did show up on the net after power-up. Good. However. sshing into it didn't work at all from my Fedora machine. That's not good. I received all kind ssk key, cipher error messages. Why? <br />Obviously the SBT ssh package is pretty dated @ 12+ years lifetime. After googling and trying all kind of ssh options to find out if there's on option to get into the damned device, I just managed to get more and more errors away by entering another after another parameter. I finally gave up on the message: <i>Bad server host key: Invalid key length. </i>So far I was 2h on the project. While running the task I learned that my Fedora is running the latest openssh version and they rather recently increased the minimum key length to 2048. Grrr. </p><p>What about Debian based Raspberry PI OS !?!? Debian usually lags a couple of releases/years behind on the software side compared to Fedora. Fired my RPi up and gave it try. Bingo. From Raspberry Pi OS sshing into the SBT worked. Once in ten years it's of advantage to work on a usually outdated platform.<br />I ran the reset and all settings were back at factory-default. I did the initial setup. And hooked the SBT up to my Adams. </p><p>Let's play some music! A radio stream. Several of the radio stations that were on the SBT are not longer accessible. Finally music started playing when selecting Radio Paradise. Above image is proves it.</p><p>Meanwhile, a day later, the Fedora problem still kind of bugging me, I found the solution to also do the ssh login from Fedora. Below you'll see the result (the ssh command is actually just one line):</p><p>###########################################################################</p><p>SBT Login:</p><p>ssh -o KexAlgorithms=+diffie-hellman-group1-sha1 \<br /> -o HostKeyAlgorithms=+ssh-rsa \<br /> -o PubkeyAcceptedAlgorithms=+ssh-rsa \<br /> -c aes128-cbc \<br /> -o PubkeyAuthentication=no \<br /> -o PreferredAuthentications=keyboard-interactive,password \<br /> -o RSAMinSize=1024 \<br /> root@192.xxx.x.xxx</p><div>user: root<br />pword: 1234</div><div><br />#############################################################################</div><div><br /></div><div><br /></div><div>Enjoy.</div><p><br /></p><p><br /></p><p> </p><p> </p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0tag:blogger.com,1999:blog-1557068496303009409.post-78511625754354008772023-01-30T12:46:00.018+01:002023-08-04T11:29:49.865+02:00My new desktop streaming system (DIY)<!-- Google tag (gtag.js) -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-N3ZZ82CSQH"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'G-N3ZZ82CSQH');
</script>
<p>The winter season, a great time to do some hardware work. Doesn't happen that often around here - hardware work. <br />Why not getting rid of a bare board DAC and a bare RPi and build a nice looking desktop streaming solution!?!? That idea triggered this little project...</p><div class="separator" style="clear: both; text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjX04LAol8q85urwrYP2FqVYQRxiwS5pcT8jJuVoYPEQzpmDPm-7dSeAdcXEWTLPyrD-6doyKmfKixggW1SVlbjEq4JeJc05WU6USiFVco6ecRLoPBw5lMlvr6P44A0CbtMq00KD7I0_p4h7jURo0Yl9JHtUgMj_H76xvcC78yyUFAiCjUXLVCJsaZ5Gg/s2927/sc-desktop-system-2023.jpg" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="2097" data-original-width="2927" height="458" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjX04LAol8q85urwrYP2FqVYQRxiwS5pcT8jJuVoYPEQzpmDPm-7dSeAdcXEWTLPyrD-6doyKmfKixggW1SVlbjEq4JeJc05WU6USiFVco6ecRLoPBw5lMlvr6P44A0CbtMq00KD7I0_p4h7jURo0Yl9JHtUgMj_H76xvcC78yyUFAiCjUXLVCJsaZ5Gg/w640-h458/sc-desktop-system-2023.jpg" width="640" /></a></div><br /><span><a name='more'></a></span><p>...and what you see above is the result of this little project. So, what are we looking at !??!</p><p>A streamer based on RPi4 technology and a DAC based on the IMO still excellent performing<br />Khadas Toneboard 1.</p><p>To get more in detail:</p><p></p><ul style="text-align: left;"><li>the streamer is a Raspberry CM4 4GB module</li><li>on a <a href="https://www.waveshare.com/cm4-io-base-a.htm" target="_blank">WaveShare Mini Base Board (B)</a> CM4-IO-BASE-A </li><li>and a NVME M.2 256GB Samsung SSD</li><li>the <a href="https://dickson.industries/products/cm1" target="_blank">Dickson heatsink</a> keeps the CM4 below 40C°.</li><li>the DAC is a Khadas Toneboard 1</li><li>and there's also a iFi USB filter attached</li></ul><p></p><p>Inside the cases there's some powerrail filtering and buffering in place. Powering is done through the GPIO headers on both devices.</p><p>A slightly modified dual-DC-rail linear Allo Shanti feeds both devices.<br />The Shanti IMO still is one of the best 5V-DC supplies out there. </p><div><br /></div><div>Slightly modified !?!? That's what I did to the Shanti:</div><div><div class="separator" style="clear: both; text-align: center;"><br /></div></div><div class="separator" style="clear: both; text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgABuO6Nq7mfHLmtGftcXFBYRXxikI30RrPgqYAe6pFl6Ez7L0kFER8c6Unbr54C22sORoSy_43KRt-qKSsjcBsapGlk0vpd2yrHgnizrkbZ_LC4u7vexmfwcl9X_93iWXFNXZBAWdCLlJlWtOZwLh_oQHXEpmv153isFJjIs_yJ1PlYxrVTUaWuxyw8A/s2016/Allo_Shanti_Mods.jpg" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="1512" data-original-width="2016" height="300" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgABuO6Nq7mfHLmtGftcXFBYRXxikI30RrPgqYAe6pFl6Ez7L0kFER8c6Unbr54C22sORoSy_43KRt-qKSsjcBsapGlk0vpd2yrHgnizrkbZ_LC4u7vexmfwcl9X_93iWXFNXZBAWdCLlJlWtOZwLh_oQHXEpmv153isFJjIs_yJ1PlYxrVTUaWuxyw8A/w400-h300/Allo_Shanti_Mods.jpg" width="400" /></a></div><br /><div class="separator" style="clear: both; text-align: center;"><br /></div><p>As you can see I added switches to the Shanti to be able to turn both rails on and off separately.<br />Further I am able to avoid the Shanti supercap discharge over the connected devices. </p><p>There's a small drawback: Adding any switches, plugs, soldering joints etc. to a DC power rail will degrade its (audio) performance. This fact alone actually speaks for not adding switches.<br />However. No matter what you do, you'll end up with a compromise at one point. And, as you can also see, I didn't cut or replace the DC cables on the Shanti. My main system Shanti looks different.</p><p><br /></p><p>Let's talk a bit about software. </p><p>The streamer runs my own custom-build (from scratch) 64bit Raspberry Pi OS based streaming OS. <br />The audio streaming solution is based on a Logitechmediaserver and squeezelite. Both applications are running on the same system. </p><p>My own latest custom realtime kernel builds the core of the system.</p><p>The system optimizations and tweaks I've been applying follow the same strategy as outlined on this blog. Just have a look the audio streaming series, sKit, etc.. </p><p>However. The level of optimizations go further of what you learn in the Audio Streaming Series.</p><p>My solution can also act as Airplay client. I can feed Qobuz streams from the iPhone right into <br />my system. It doesn't sound as good as the LMS/squeezelite setup. However, it's closer than I expected it to be. On the first try I was shocked how close it was. The to me huge advantage for this Airplay setup is its convenience factor. With minor effort you can set up a pretty good sounding system in no time. And everybody with an Airplay capable device can stream audio to that device.</p><p>All that fun, at what price ???</p><p>The project took me a couple - partly challenging - hours of work. <br />Let's get a "couple of hours" defined. All in all the project ran over a period of three weeks. I'd estimate a 16 to 20 net hours effort for research, planning, sourcing, building, testing etc.</p><p>I spent roughly $/€ 40 for the two cases, $/€20 for some little parts, such as switches, LEDs, asf. and another $/€40 for tools, such as a special cone drill for the large holes.</p><div>The electronics, RPi, DAC, cabling, etc. I had lying around. And, to be honest, that's what actually makes the project feasible.</div><div><br /></div><div><br /></div><p>Bottom line, to wrap things up. All in all it's been a fun project. Something different compared to all the hacking. </p><p>There's a lot to think of. All these tiny details. ...where to drill the hole, how to drill it, how to avoid scratching these very sensitive surfaces, what parts, what switches, what LEDs, where to source it, asf, asf. There's a lot to do and a lot can go wrong. For sure, you'll learn a lot. And yep, there's some stuff you do more than once. ;) For sure you'll face a steep learning curve. And you also feel some kind of pressure (I did) if you intend to build something like that properly with an appealing touch to it at the first shot.</p><div>And last but not least... ...what about perceived sound quality. <br /><br /></div><div>The setup sounds really nice on my Adam speakers. Very clean and punchy. Lot's of energy. Lot of details. Nothing to complain about. Even though I have a few commercial DACs and well cased RPi's around, I kind of prefer this new solution. A good working and sounding setup all done by yourself is simply more enjoyable.</div><div><br /></div><div>Now there's one major caveat that comes with above solution. As you might have noticed I am using a CM4 on Waveshare board with NVME-m.2 SSD attached. This solution was prefered over any other RPi based solution. However. The Waveshare motherboard won't come with a seperate USB controller, like the RPi4. That's a problem. And I am not just talking about the lack of USB3.0. The PCI lane is occupied by the NVME adapter. The USB2 ports run in a OTG-driver mode. And that's not good, not at all. You can face nasty XRUNS on your DAC. Running a realtime kernel, like I do, makes things worse. It took me a major effort to get the XRUN issue somewhat under control up to 96kHz. All I am saying. Things can get tricky with such a setup!</div><div><br />My today's main audio system looks pretty similar to above desktop system. The main difference is the DAC. I am using an Allo Katana HAT DAC with class-A output stage. No need for USB. I have turned USB completely off. The main system performs - non-surprisingly - better then my new desktop setup. Beside that the main system still lacks a nice housing... ...the next winter is coming soon. ;)<br />And just to mention it: On the main system, the Shanti switches, you saw above, are replaced with remote controlled relays (Wifi-ESP8266). Why? The main audio system requires 6 power switches, a mix of AC and DC switches. It requires a certain timing for the power-on and another one for the power-off cycle. E.g. The Katana DAC should get some 20s settling time before the RPi starts up, after that I start the amp to avoid the turn-on thumps. That works really nice. By pushing a single shortcut on iOS - see below - the whole system starts up piece by piece. Finally, after more than a decade of DIY chaos the entire family can handle the stereo system. ;)</div><div><br /></div><div class="separator" style="clear: both; text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhBhYZSkr-ZsccTvVKVqon1-GvNpjupyEFkqNXI_mzdDSImaTSN3urxQJB6HAQaBznVfpJumvj--NyPDvqeZDiUD_DzZofps-7-sUIBamXqgt5DzRsVX8wc7aGZHBmp1UKqItXJW1hFQ3iclX5Ag6sg_Vb69FoDHP1p5Yje7MxVs3nWw7cCQDFWLFRkXw/s2778/audio-remote.png" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="2778" data-original-width="1284" height="320" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhBhYZSkr-ZsccTvVKVqon1-GvNpjupyEFkqNXI_mzdDSImaTSN3urxQJB6HAQaBznVfpJumvj--NyPDvqeZDiUD_DzZofps-7-sUIBamXqgt5DzRsVX8wc7aGZHBmp1UKqItXJW1hFQ3iclX5Ag6sg_Vb69FoDHP1p5Yje7MxVs3nWw7cCQDFWLFRkXw/s320/audio-remote.png" width="148" /></a></div><br /><p style="text-align: center;"><br /></p><p>Enjoy.</p><p><br /></p><p><br /></p><p><br /></p><p><br /></p><p><br /></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com7tag:blogger.com,1999:blog-1557068496303009409.post-62938774169272672782023-01-19T09:15:00.001+01:002023-01-19T09:15:42.457+01:00RPi kernel developments - 2023<p>Hi folks.</p><p>The foundation has updated its kernel tree a short while ago. It seems they plan to switch to kernel version 6.1 during the year. Usually the foundation goes for LTS (long term support) kernel versions.<br /><br />A lot of RPi related stuff had been merged into the so called mainline kernel tree of 6.1 over time. That's the master tree maintained by Linus Torvalds. The more stuff the foundation can transfer to mainline the less effort they and the developers will have to maintain two kernel trees. And the users should greatly benefit by getting much faster newer kernel versions and related features. Especially accessing performance relevant developments asap are of great value for embedded system like the RPi.</p><p>What's new on 6.1 ?</p><p>A lot.</p><span><a name='more'></a></span><p><br /></p><p>Especially the RPi graphic driver section is going to benefit. (not much of relevance for audio though)<br />But there can also be performance improvements expected on the kernel side. A lot has been<br />changed and improved on the internal kernel structure over time.</p><p>Multi-Generational Least-Recently-Used (aka MG-LRU) has been introduced. That's a promising<br />feature to boost the performance (google the details for more info). It's not enabled by default on<br />mainline, but the brave RPi folks have enabled it by default already now.</p><p>Beside that, most of the realtime kernel-patch code went into 6.1 mainline kernel by now. <br />They get very close to get rid of that huge separate realtime kernel patch. <br />That's a great development from "my" audio perspective. I am running rt-kernels since more than a decade by now. Unfortunately the feature is not officially accessible yet.<br /><br />Guess what !??! - I did some kernel hacking. Aaand... ....I got it going. </p><p>From what I see so far - it's working! <br />My guess. In the future we'll see a lot more realtime enabled audio kernels out there. <br />Again. It still requires kernel customization to get access to it, the feature won't come with a standard kernel.</p><p>The RPi foundation still maintains numerous RPi specific kernel customizations, that's why they have to maintain a separate kernel tree.</p><p>By now It seems they already did a good job merging their stuff with 6.1. I've been running 6.1.x for a week by now. So far I am not experiencing any issues.</p><p>The Raspbian folks are convinced it's stable enough to put it on the NEXT tree. <br />That's why on Raspbian based system you can (should) give 6.1.x a try by simply running:</p><p>rpi-update next</p><p>I couldn't find anything about piCorePlayer offering 6.1 yet. Probably it's not available. Usually it takes some time for the pCP folks to catch up. ( You need to push for it! ;) )</p><p>2023 just started. And great developments already on the horizon. <br /><br />Now we just need to get RPis back into the market. Just a couple of (6?) more months to suffer. There's hope.</p><p>Enjoy.</p><p><br /></p><p><br /></p><p><br /></p><p><br /></p><p><br /></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0tag:blogger.com,1999:blog-1557068496303009409.post-12188701637713696962022-10-23T15:08:00.000+02:002022-10-23T15:08:12.168+02:00RPI4 - All Lights OFF - Update<p>Not much going in RPi land these days. It's unbelievable that a whole industry branch gets dragged down currently.</p><p><br />However. Many of us still run and maintain RPis. I'd like to make you, the tweakers, aware of a recent change in the RPI firmware.</p><span><a name='more'></a></span><p><br /></p><p>It's about the RPI LEDs. The PWR LED in particular.</p><p>Since the very beginning I outlined how to turn off all LEDs on the RPi. </p><p>However. Since kernel 5.15.56 there's a not yet widely known change. <br />It now requires a slight change in the settings of /boot/config.txt to get<br />the Power LED turned off. </p><p>Because - as a matter of fact - since kernel 5.15.56 it simply stays on!</p><p>Below the config you have to apply to turn off ALL LEDs on your PIs from kernel 5.15.56 onwards!<br />The most current kernel is 5.15.75 or similar btw. </p><p>If you still run older kernels, this change will for sure hit you sooner or later. Just keep it in mind.<br /><br /></p><p>Here's the config.txt change:</p><p>###turn off onboard LEDs###################<br />dtoverlay=act-led<br />###disable the ACT LED<br />dtparam=act_led_trigger=none<br />dtparam=act_led_activelow=off <br />###disable the PWR LED<br />dtparam=pwr_led_trigger=default-on<br />dtparam=pwr_led_activelow=off<br />###turn off ethernet port LEDs<br />dtparam=eth_led0=4<br />dtparam=eth_led1=4<br />##########################################</p><p><br /></p><p>That's it. A pretty simple change. If you know it! Hope you'll manage.</p><p>Enjoy.</p><p><br /></p><p><br /></p><p><br /></p><p><br /></p><p><br /></p><p><br /></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0tag:blogger.com,1999:blog-1557068496303009409.post-74919795991821110742022-08-21T11:16:00.001+02:002023-08-04T11:30:51.401+02:00RPi Power Gadget - UcPi<!-- Google tag (gtag.js) -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-N3ZZ82CSQH"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'G-N3ZZ82CSQH');
</script>
<p>IanCanada recently released a new gadget called UcPi.</p><p>It's basically a supercap power buffer for the RPi. It follows a to me new approach - I just called it "HAB" -- Hardware Attached to Bottom. To accomplish it IanCanada makes use of spring contacts to connect to the soldered GPIO power pins at the bottom of the PI board.</p><p>Over at github you'll find the manual that also lays out this or that application:</p><p>https://github.com/iancanada/DocumentDownload/blob/master/UltraCapacitorPowerSupply/UcPi/UcPiManual.pdf</p><p>Audiophonics sells the product:</p><p>https://www.audiophonics.fr/en/single-board-computers-et-hats/ian-canada-ucpi-universal-ultracapacitor-power-supply-board-for-raspberry-pi-p-16717.html</p><p>I do use cap buffers to improve the RPi power rails for years with good results.<br />Therefore I consider the UcPi a promising gadget. However. @ $/€79 plus feet on top I consider the pricing a bit steep.</p><p>Further it seems to take a while to get the supercaps charged, which might make operation of the module slightly annoying, especially if you don't run an AlwaysOn system.</p><p>Beside that it's recommended to use a quality PS for charging. Another argument that weakens the UcPi proposition. Of course - the lack of a case is another bummer.</p><p>Anyhow. I do think for those who run quality RPi implementations with PS like Allo Shanti, iFi iPower or alike the UcPi should enhance the overall performance of the setup without being too complex.</p><p>If anybody out there have already tested the UcPi, please let us know.</p><p><br /></p><p>Enjoy.</p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com1tag:blogger.com,1999:blog-1557068496303009409.post-77604942432104970482022-04-19T11:57:00.007+02:002023-08-04T11:30:27.653+02:00soundcheck's Purifi amplifier<!-- Google tag (gtag.js) -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-N3ZZ82CSQH"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'G-N3ZZ82CSQH');
</script>
<p>Let's talk a bit about amplifiers. I owned numerous devices. Commercial and DIY. Jeff Rowland, Anaview Abletec, Hypex, T-amps, SMSL and the list goes on and on. I tried many of those ebay Class-D modules. I tried all kind of power supplies (connex/hypex/linears, ...), I tried several full digital amps. and I compared these to even more amps owned by audiophile friends of mine. </p><p>Oh dear. What a career. I swapped Class-D output transformers with Mundorf coils, replaced rectifiers, cabling, fuses connectors, power cables and the tweak list goes on and on.</p><p>And finally I got hooked up to a Purifi Audio Eigentakt 1ET400A Stereo Eval1 amp integrated into a Ghent Audio case.</p><p><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhPQQslULt2MfPDBuFH2Gub5XRdzBqDRS6p1k5NBxEvwKsMtkA8SPX3iCgRaFw0MaGxMgP298oyzGujEAslIAo8c6S8m7Dl0JqVABODsm4le1nYFKZSyyJeCwGE2mw5xMtDgBRru8S5nLYViSKnfaqiPngtT-DA9KYCB8ZKpJmeggWMOrCWDFdjcIXaCw/s400/SC-PURIFI-2-20210923_082850.jpg" style="margin-left: 1em; margin-right: 1em; text-align: center;"><img border="0" data-original-height="400" data-original-width="303" height="320" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEhPQQslULt2MfPDBuFH2Gub5XRdzBqDRS6p1k5NBxEvwKsMtkA8SPX3iCgRaFw0MaGxMgP298oyzGujEAslIAo8c6S8m7Dl0JqVABODsm4le1nYFKZSyyJeCwGE2mw5xMtDgBRru8S5nLYViSKnfaqiPngtT-DA9KYCB8ZKpJmeggWMOrCWDFdjcIXaCw/s320/SC-PURIFI-2-20210923_082850.jpg" width="242" /></a></p><p><br /></p><p>I am running this amp for more than a year by now. Puh. What a journey.</p><div><br /></div><span><a name='more'></a></span><p><br />Spoiler alert: This is the best amp I ever had - supplying me with almost "boringly" clean amplification. A so-called wire-with-gain. Everything that gets thrown at it will be displayed. I simply can't find any amp specific mannerism. The numerous measurements and reviews out there confirm it. </p><p>My old Hypex stuff as all my other (tweaked) amps performed worse. Period. My Abletec/Anaview AMS0100-2300 was IMO close (not close enough though). However. On performance/cost the Anaview was and IMO would still be the better choice! <br />Whatsoever. There was simply no better amp then the Puriri in my system yet.</p><p>The next question one the table: What is/was done with the preamp-section or volume control!?!? <br />I did have a Jeff Rowland Pre in the past. I sold it quite a while ago. I was running several passive pres. And the probably best solution was a Bent Audio Transformer potentiometer. </p><p>The best of all preamp solution to me though is to use "No PreAmp at all". Since years I am hooking up my amps right to the DACs and make sure the amp gain structure matches the DAC output voltage and my speaker sensitivity. Levels and impedances should also comply to specs btw! A preamp and its cabling will add flaws. Always. That's basic physics. </p><p>There's one caveat. At system turn-on or during DAC initialization turn-on pops might blow your speakers. Better look for a DAC that doesn't have that turn-on-pop problem. On well designed DACs the output stage gets muted at DAC init to avoid this behaviour.</p><p>However. You better make sure there's a proper system turn-on sequence in place. That lowers the risk substantially.</p><p>What's been the learning so far!?!? </p><p></p><ol style="text-align: left;"><li>quality amps play not such a big role - not causing huge differences<br /></li><li>I can get rid off preamps and HW VCs if <br /> 1. I get an amp that mates/matches the DAC output <br /> 2. the amp gain level matches my speakers sensitivity<br /> 3. I run a quality SW volume control<br /> 4. I just run a single source <br /> 5. I make sure that the chain won't blow my speakers</li><li>fully integrated solutions (DAC and AMP, powerDAC, full digital amps) are IMO still the best solution for the future. These amps are reducing DAC>amp interface related flaws even further. ;)</li></ol><div><br />Let's get a bit more specific about my setup. </div><div><br /></div><div>My latest amp is a Purifi Eigentakt Eval1 amp housed in a Ghent Audio case. I also use the Ghent Audio harness. PS is Hypex SMPS1200. Power cable is a ViaBlue X-40. I run the amp differential hooked up to Neutrik plugs + Voodoo solid core silver wires.</div><div class="separator" style="clear: both; text-align: center;"><br /></div><br /><p></p><div class="separator" style="clear: both; text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgVL7fqJcmEI2AbjORaAqp_OgtdwvnhRwzQJknjPAGjlYcPUmV77bJGRlYTUTzPqVk1kb3y2-i1LQ2ExHp3JL05n5an4iYrwD7EY2IdxkI5yhklds3ruuC6zCPPAcH7LUFKxJpOt_d0SuZvUDQsVByDXtI_crl2HqYGkgR6Pw-t9YopOpxiIR6Slo9w4g/s320/SC-PURIFI-1-20210923_082850.png" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="320" data-original-width="320" height="320" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgVL7fqJcmEI2AbjORaAqp_OgtdwvnhRwzQJknjPAGjlYcPUmV77bJGRlYTUTzPqVk1kb3y2-i1LQ2ExHp3JL05n5an4iYrwD7EY2IdxkI5yhklds3ruuC6zCPPAcH7LUFKxJpOt_d0SuZvUDQsVByDXtI_crl2HqYGkgR6Pw-t9YopOpxiIR6Slo9w4g/s1600/SC-PURIFI-1-20210923_082850.png" width="320" /></a></div><br /><div class="separator" style="clear: both; text-align: center;"><br /></div><br /><div><br /></div><div><br /></div><div>I did have to tweak the Ghent Audio case. I removed that IMO "awful" Ghent Audio power switch and connector section as well as the internal power cabling and attached the Viablue right to the Hypex SMPS. This noticeably did improve the perceived sound quality. I run a central mains switch at one place (power distribution). All my DC rails (DAC and RPi) are controlled via relays and wifi. That allows for a proper turn-on and turn-off sequence. The AC rails get a slight cleansing by using an iFi AC purifier. I also disconnected the Ghent LED board recently. I don't need it. I also have the speaker cables soldered to the Purifi board. There's nothing better and cheaper than a nice silver-tin soldering joint.</div><div><br /></div><div>The Purifi Eval 1 comes with the generic Purifi input buffer stage. The Purifi buffer stage allows for choosing different pre-gain levels. That's great. I run it at 0dB - no pre-gain - setting. <a href="https://www.audiosciencereview.com/forum/index.php?threads/review-and-measurements-of-purifi-1et400a-amplifier.7984/" target="_blank">Over at the ASR review</a> you can see the 0dB setting is the best performing setting. And IMO you can also hear it. To me it sounds cleaner. As I mentioned before, such a setting has to match your DAC and speaker sensitivity levels. YMMD.<br /><br /></div><div>Audiophonics offers a similar Purifi based amp. It comes with a custom input buffer though. I think the Audiophonics offering is a pretty fair commercial solution. Actually you can find several 3PP buffer offerings for the 1ET400A modules out there. I had a look at them and couldn't find any advantage over my generic Eval1 buffers - which essentially get (partially) bypassed @0dB anyhow. <br />Perhaps there are also better power supply options out there. To be honest. For me it's good enough the way it is for the time being. Optimizations on my upstream gear and front-end (DAC/RPi) still cause a lot more differences on perceived sound quality then the tweaking on the amp side. </div><div><br /></div><div>Now. The source.</div><div><br /></div><div>I was running my seriously tweaked Gustard A18 differentially straight into the Purifi amp. This was and IMO still is a great combo. Now I am doing it with the Allo Katana. It works nicely. The Katana base output voltage is a bit lower than the Gustard. This still works quite well. </div><div><br /></div><div><br /></div><div>Anything else to report?</div><div><br /></div><div>The amp is IMO quite expensive. I'd say: It's a rather good "audiophile" value for the money though. However. My Abletec Anaview ALC0180-2300 modules were also performing great and that at $180. That keeps bugging me. </div><div><br /></div><div>However. The Purifi amp is a great performer. For sure it's able to challenge the best stuff out there. <br /><br />I wouldn't call the Purifi my final amp though. Because I think a fully integrated DAC+amp or PowerDac/Full digital amp solution or a digital active speaker solution has the potential to beat a chain build from seperate devices. </div><div><br /></div><div>I also would like to see the volume control on the high power side. A voltage level controlled amp output. I think TI had a solution for it on it's full digital amps. EC-Designs is offering a 4W PowerDAC-S solution that would cover a lot of aspects on my wishlist. Price and power constraints make it a NoGo for me.</div><div><br /></div><div>Keep in mind to choose your amplifier based on your system topology and requirements. Making your buying-decision just by looking at the standalone (lab-) performance of such a device might hit you on the wrong foot... ...if you happen to run a 98db/W speaker you probably better have a look at e.g. the above mentioned EC Designs solution.</div><div><p></p></div><div>Enjoy.</div><div><br /></div><div><br /></div><div>PS:</div><div><div>...while still dreaming about affordable and highest quality full-digital amps... <br />What I don't understand is to see Peter Lyngdorf from Tact listed as co-founder in the Purifi team . Lyngdorf is one of the full-digital amp pioneers. Tact had a real good rep in the full-digital amp arena. And now he seems to support a full-analog Purifi amp again. Hmmh. That must have been a big step back for Lyngdorf. Perhaps - fingers crossed - one day Purifi will go the all-digital route. For sure they have the right team onboard for getting it done.</div><div><br /></div></div><div><br /></div><div> </div><div><br /></div><div><br /></div><div><br /></div><div><br /></div><div><br /></div><div><br /></div><div><br /></div><div><br /></div><div><br /></div><div><br /></div><div><br /></div>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com2tag:blogger.com,1999:blog-1557068496303009409.post-13307347299202911342022-04-10T12:12:00.012+02:002022-04-14T10:22:06.520+02:00Gustard V18 vs. Allo Katana - a 2022 assessment<p>For almost a year or so I am running an IMO excellent Gustard A18 DAC. The Gustard builds around the top of the line AKM AK4499 and Accusilicon clocks. The DAC received quite some nice <a href="https://soundnews.net/sources/dacs/gustard-a18-dac-review-outstanding-performance-and-value/" target="_blank">reviews</a> confirming my experience with it. Most reviewers saw that DAC challenging much pricier rivals. Until now I was never in doubt about the overall DAC performance...</p><span><a name='more'></a></span><p>Before Gustard I was running a tweaked Khadas Toneboard1. And before that plenty of tweaked RPI HATs. In between I listened to this or that (incl. high-end gear) to see where I am standing. </p><p>Over the last year I applied - I had to - plenty of tweaks to get this rather nice Gustard DAC setup working satisfactory. You might have seen it in my "my system" section.</p><p>What I did:</p><p>1. power cable swap<br />2. fuse swap<br />3. applying USB filters and isolators<br />4. using a new USB cable<br />5. applied lots of RPi OS optimizations<br />6. applied some RPi HW modifications<br />7. Changed the DAC default settings</p><p>Basically all the stuff you'd do with most DACs out there. All this lifted the DAC into a different arena.<br />Without all these tweaks applied I always had the feeling my "tweaked" Khadas was able to rival a stock Gustard. </p><p>After all I concluded that the Gustard A18 device -- with all tweaks applied -- is an excellent performer. I am running the DAC differential right into a Purifi - Eigentakt amp. These devices mate extremely well. Volume control is done via software. Bottom line. I love this setup. I enjoy(ed) every minute with it.</p><p><br /></p><p>However. Yep. Here it comes.</p><p><br /></p><p>While cleaning up my lab the other day I stepped over my Allo Katana HAT. It's been settling dust for quite some time (years). The Katana is no longer built/sold. It used to be an around $250 device.<br />As a matter of fact it mates well with the RPi4 family, despite its docs showing RPi3 support only!</p><p>The Katana comes with an excellent Sparkos-Labs discrete Class-A output-stage and requires multiple high quality rock solid 5V supplies - if "excellent audio performance" is among your targets. The Katana is built around a rather entry level ESS Sabre 9023 DAC - implemented the Allo way though. Allo tried to tweak most out this DAC. If you look at the 3 boards closely you'll see how much complexity Allo built into this tiny stack to accomplish their goal to create a great performing "class leading" HAT DAC. </p><p>During those early Katana days I considered the Katana a solid performer, offering (requiring) a bit too many "power options". I used it for testing and documentation purposes. I never really used the DAC in my systems.</p><p>Today, some years further down the road, probably bored by that well performing Gustard/Purifi system, I thought why not give the Katana a try once more. Just to see where I stand with my USB-DAC stuff. I didn't expect it to become a huge effort to get it hooked up. The task required a bit of reshuffling. I soldered differential cables to the DAC. Attached some power supplies. Had to reprogram my PI setup to cope with DAC specifics. T'was done in an hour or two.</p><p>What struck me first: No more annoying and expensive USB tweaks. What a relieve!</p><p>For now I have 3 5V supplies attached. 1 (iPower) for the USB-SSD, 1 (Shanti) for the PI4 and 1 (Shanti) for the DAC.</p><p>According to Allo it requires at least one more rail for great sound. And then I do not have an isolator in place. I read reviews claiming an isolator was making things worse with the Katana. That's not gonna be up high on my priority list. Anyhow. It looks there's further tweak potential out there.</p><p>I run the DAC with its linear filter and on-DAC volume control on my own RPi OS. Just to mention it.</p><p>Last night I finally powered Katana up for the first time in years. I gave it some time to settle and to warm up. Class-A loves stable conditions.</p><p>Allo recommends to power up the audio stack first! Luckily I have my DC rails attached to relays and Wifi (DIY home automation based on Arduino/Tasmota/iOS). <br />I can simply shout a "Hey Siri" "Stereo ON" and magically all devices (rails) turn on in the right sequence and timing.<br />(Ok.Ok. Not much magic here. Just technology) </p><p><br /></p><p>I then pressed the "Play" button on one of my fav tracks expecting nothing... </p><p>... and nothing... ...but clean and dynamic sound hit my face.</p><p>Oh boy. </p><p>I wasn't expecting that. I couldn't localize any obvious downside. Nice. It all sounded so familiar to what I've been listening to for a while. Hmmh. Similar - is the better wording. The low end, the highs, the depths and separation, the tone-colors, realism, naturalness, the impact, basically everything seemed to work fine. No booming bass, or sharp Ses. Nice.</p><div>After a bit of breaking-in and stabilizing, the DAC developed some more to me obvious (subtle) qualities. It stayed on for the rest of that enjoyable night.</div><p>I had no problems to conclude after 3 hours, that even at this early stage I am tempted to rate the Katana performance above my highly tweaked Gustard A18 setup. Wow. I wasn't expecting that. For sure this little project was well worth the 2h effort.</p><p>During the coming days/weeks I'll put some more effort into the project. If I am able to gain even more, unfortunately the Gustard will be gone for sure. </p><h2 style="text-align: left;">Aftermath</h2><p>Folks. I sometimes do get the feeling that I am spinning in circles. I do know I am not alone on that one. I simply realized after almost 4 years that I didn't make any progress on the DAC side. Actually. It's been the other way around. All new DACs, even after heavy tweaking, I had or tested didn't perform as well as the Katana in the setup I've been listening to last night. Puh.</p><p>Let's not talk about the time and money that's been wasted. That's kind of frustrating. But again. I've been enjoying that IMO rather flawless Gustard DAC for quite some time now. Fact is, the differences between Katana and Gustard for now are not night and day. I could live with either setup. But honestly. The Katana comes with a better feel to it - why ? - successful DIY Audio projects which surpass commercial high quality solutions are simply most satisfying.<br /><br />Another learning. It seems that the overall DAC evolution approaches some kind of stagnation state.<br />Which doesn't mean there's no space for improvement. What I mean. All these DAC still heavily respond to upstream improvements and changes. Stagnation exists because most DAC developments just follow that generic "standard specs" path, which <a href="https://soundcheck-audio.blogspot.com/2021/04/everything-that-matters.html" target="_blank">we all know</a> is not sufficient to explain the entire soundimage in front of us. </p><p>Some thoughts about Allo - since Allo just made it back into my main focus - I'd like to share. I was rather close to the Allo folks in the early days. Later on things developed into the wrong direction. However. I have to admit Allo did a great job during those days (let's not talk about the Katana callback disaster - Allo usually acted pretty customer focused ). Most of their products were performing above average and were priced quite fair. And for the most part they were listening to the customer base and DIY-Audio community given advise. The company IMO had huge potential. A pity that it's been pretty quiet around that company for quite some time. Let see if they recover from the plague (chip shortages). Let's hope they'll survive these challenges. There's IMO more to it I guess. Allo IMO should have stayed in the HAT business. By entering the USB segment with the Revolution DAC, they were facing a different and pretty rough and tough competition and lots of pressure. Another highly complex product, the RPI CM3 based Allo USB Bridge, must have been turned out to be a nightmare project for them. A major project. A major effort. Exactly at the time of the USB Bridge market launch the much better RPi4 showed up. Basically the CM3 and the Allo board could be considered immediately outdated at launch. Really bad luck. I'm also wondering if the newest kid-on-the-block, the Boss 2 project, sells as good as it measures. It's neither a real HAT nor a "Boss". Allo should have marketed it differently. Hifiberry shows how it works: "Stay focused, stay on your turf". Therefore I think, it was not just the ship shortages putting pressure on Allo.<br />Sorry for talking too much business here. I just hope that Allo finds back to its high quality HAT roots one day. The Katana simply proves to me now they were on the right track back then. I wish Allo good luck and a more focussed approach with future projects. ;)</p><p>Back to my little project. This little exercise just showed that I have another viable audio setup at hand (and best of all - there's still tweak potential). And instead of jumping on the next "greatest DAC ever" hype or offering I better head down to the basement and continue my DIY efforts. There might be even more hidden treasures down there. </p><p>Enjoy. Stay healthy. Peace.</p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com5tag:blogger.com,1999:blog-1557068496303009409.post-85629246232573461932022-02-04T11:39:00.000+01:002022-02-04T11:39:34.845+01:00Raspberry PI OS 64 bit officially released<p>The RPI foundation finally officially <a href=" https://www.raspberrypi.com/news/raspberry-pi-os-64-bit/" target="_blank">announced</a> the release of RPI OS 64.</p><span><a name='more'></a></span><p>It'll support</p><p></p><ul style="text-align: left;"><li>RPI Zero 2 W</li><li>RPI 3 family</li><li>RPI 4 family</li></ul><p></p><p>And can be downloaded over <a href="https://www.raspberrypi.com/software/operating-systems/" target="_blank">here</a>.</p><p>RPI OS 64's been on Beta for years. From what I read in the past they had to get the advanced GPU driver developed from scratch before going public. A major effort.</p><p>Anyhow. Running the Beta headless didn't require a GPU driver. That's why I am running a RPI OS 64-bit Beta system for quite some time. Never seen any glitches. It's rock solid Debian based stuff. </p><p>All the RPI audio distros out there do not have an excuse any longer for not providing you with a full 64 bit OS base. </p><p>Great developments.</p><p>Enjoy.</p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0tag:blogger.com,1999:blog-1557068496303009409.post-17945906962548697892022-01-18T17:53:00.001+01:002022-01-18T18:03:27.212+01:00audio@advise passes 1.000.000 site visits<p>Hey... </p><p>...just figured that the site visit counter of blogger passed the 1.000.000 for audio@vise!!!</p><p><br /></p><div class="separator" style="clear: both; text-align: center;"><a href="https://blogger.googleusercontent.com/img/a/AVvXsEhehDJInRM8isyCXw_UC4GxZ4t-VawWAndV04CyHImOJyHUSCsV4oo2327ZPPfZ4B-8jFLTZNXvySu4Pmdi-5Ky1Cx0R4pCY0pEwZ7GztTrgD3zC4Xcel-8dELpVFnwI-CxtqwCFeg005DvMgKTyzRehUJJthTZEouP2dCrtZR3V9ziUsiKyEUpoOaCEg=s267" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="123" data-original-width="267" height="123" src="https://blogger.googleusercontent.com/img/a/AVvXsEhehDJInRM8isyCXw_UC4GxZ4t-VawWAndV04CyHImOJyHUSCsV4oo2327ZPPfZ4B-8jFLTZNXvySu4Pmdi-5Ky1Cx0R4pCY0pEwZ7GztTrgD3zC4Xcel-8dELpVFnwI-CxtqwCFeg005DvMgKTyzRehUJJthTZEouP2dCrtZR3V9ziUsiKyEUpoOaCEg" width="267" /></a></div><div><br /></div>I thought it'd be worth mentioning. <span><a name='more'></a></span><div><br /></div><div>Many of you folks out there come by on a regular basis and try to and many successfully apply the advice that's been given. I really hope you enjoy it. Thx a lot for your trust and loyalty. </div><div><br /></div><div>If there are problems with the content or you'd have improvement proposals just send me note.</div><div><br /></div><div>Those of you who come by regularly did notice that there's not a lot of progress to be seen on the blog recently, basically since the introduction of sKit during summer 2021. And. Yep. That's right. </div><div><br /></div><div>That doesn't mean, that there's no progress to be made. I am continuously working on this and that. <br />On my own system I made some huge steps forward last year. </div><div><br /></div><div>Why not sharing???? It'd be a lot of work. It'd be a lot of cost on my side. @basically no return. </div><div>Things that I share are highly valuable for many audiophiles, but also for the more commercial minded </div><div>folks out there. The point is. I am not willing to further boost these commercial projects free-of-charge.</div><div><br /></div><div>Bottom line. My resources are limited. Which doesn't necessarily mean there won't be any progress in the future. Many of you know that I keep the project/blog going and up2date for years. Let see how things are developing.</div><div><br /></div><div>And once more, a very big thx to the very few people that have been "supporting" the project. </div><div><br /></div><div>All the best</div><div>\SC</div><div><br /></div><div><br /></div>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0tag:blogger.com,1999:blog-1557068496303009409.post-19641466168667446132021-12-16T14:13:00.000+01:002021-12-16T14:13:32.916+01:00sKit and pCP81 - regression test successful<p>On December 5th pCP 8.1 was released.</p><p>I just did a sKit regression test on my RPi4. I used the 64 bit version.</p><p>And I did a maiden installation.</p><p>It all went OK. Not a single hiccup! Great.</p><span><a name='more'></a></span><p><br /></p><p>And yes. I always suggest to go for the latest release asap.</p><p><br /></p><p>You can download the pCP81 64bit version <a href="https://docs.picoreplayer.org/releases/pcp810/" target="_blank">over @pCP</a>. </p><p><br /></p><p>Once more:</p><p>sKit is made for RPi4 running a 64bit OS. Because that's IMO the best performing base.</p><p>sKit gets tested only this way.</p><p>If you simply ignore these prerequisites and end up with problems later on (this happens quite often!),<br />please don't leave any comments or send me any mails asking for help! </p><p>You won't get it. </p><p>And you also won't get support if you do not provide the information I ask for on the sKit page.</p><p>Keep in mind. Every sloppy inquiry (e.g. a simple copy/paste of a printout without context ) on your side might cost me hours of digging. I simply can not provide that kind of service. </p><p><br /></p><p><br /></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com2tag:blogger.com,1999:blog-1557068496303009409.post-38176662668291294042021-11-17T11:43:00.005+01:002021-11-18T17:59:24.574+01:00Releasing sKit-custom-squeezelite.sh 1.4 for test<p>I just uploaded a seriously modified version 1.4 of sKit-custom-squeezelite.sh</p><p>(UPDATE Nov-12-21 - I uploaded 1.4.2-beta. It comes with a few minor corrections, slightly enhanced stability and traceability)</p><span><a name='more'></a></span><p>Background</p><p>I received several inquiries related to pCP extensions download issues. </p><p>Fun-fact. I actually experienced these issues recently too, while checking how everything is going. <br />And I was already looking into it.</p><p>The clunky behaviour of the pCP repositories is well known. Contacting the pCP folks (Paul) about it once more wasn't of any help. They simply seem to accept that behaviour - that was my impression at least.</p><p>Folks. </p><p>It took me more than 3 days for (hopefully) finding a way around it, rewriting and testing the code! </p><p>I was that annoyed that I thought to shutdown the project altogether and start over with Raspberry PI OS. I might do that anyhow. I consider Raspberry PI OS the far better base for RPi 4. But many of you know that.</p><p>Anyhow. I hope the sKit extension download is much more stable now.</p><p>With the updated tool you'll see exactly what's going on while running the custom squeezelite installation and you'll also see what feature fails. The tool will then try again.</p><p>I had to write my own functions for doing that. The pCP tools were of no use. </p><p>The new tool tries 5 times to get the whole stuff downloaded. In my case I ran into 3 tries max.<br />I never made it in 1 attempt!</p><p>You can simply install the update via sKit-manager.sh "update". <br />Than you can run sKit-custom-squeezelite.sh. </p><p>And. Better stay away from the 2nd repo (mirror). sourceforge is simply the worst. I never managed to get a download properly done.</p><p><br /></p><p>Please let me know how it works.</p><p>I btw made two recommendations to the pCP team.</p><p></p><ol style="text-align: left;"><li>Look for another - a more reliable - download server</li><li>Install the standard wget download tool instead of that rather slim thus pretty weak busybox-wget.</li></ol><p></p><p>Ah.</p><p>I added one more little feature: </p><p>You'll now find the "git-commit-id" as unique version identifier for the custom-squeezelite binary. <br />This way you can also look-up in the pCP browser which exact sKit-squeezelite-custom version is active.Over <a href="https://github.com/klslz/sKit" target="_blank">@github </a>you could lookup what has been changed by opening the respective commit-ids.</p><p><br /></p><p>THX & Enjoy</p><p><br /></p><p><br /></p><p><br /></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com5tag:blogger.com,1999:blog-1557068496303009409.post-19480339992468220292021-10-27T10:42:00.005+02:002021-10-27T10:42:57.161+02:00squeezelite - update - flac processing improvement<p> Hi there.</p><div>Just recently a small - potentially interesting - performance enhancement for flac processing had been introduced.</div><div>Those of you who run sKit and a (my) custom squeezelite binary should update the binary.</div><p>I did update my own squeezelite branch on github accordingly.</p><div>I havn't tested what the update does - performance and soundwise - though.</div><p>If you run the update and experience some improvement - please let me/us know.</p><p><br /></p><p>Enjoy.</p><p><br /></p><p><br /></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0tag:blogger.com,1999:blog-1557068496303009409.post-16252293176785942932021-08-12T10:33:00.217+02:002021-11-11T19:58:41.800+01:00The audiophile power of Youtube<p>Did you ever realize that you can easily discern sound quality differences on most of these numerous audiophile a/b device comparisons over @</p><p><br /></p><div class="separator" style="clear: both; text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEik9M4wAI7BMxX5kZUeRqU6zcybR6wvLImkHPBogEwe2Qdmzr_XIT0wXWXFkYBor0yHfBY2-LOZrLWmUOngJ5ByDP-puu1z_P9plEiBU1G-Tnb6Jv-_AmkavHJXPlJxgJvs2F_Q2X_esZrI/s502/youtube.png" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="211" data-original-width="502" height="84" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEik9M4wAI7BMxX5kZUeRqU6zcybR6wvLImkHPBogEwe2Qdmzr_XIT0wXWXFkYBor0yHfBY2-LOZrLWmUOngJ5ByDP-puu1z_P9plEiBU1G-Tnb6Jv-_AmkavHJXPlJxgJvs2F_Q2X_esZrI/w200-h84/youtube.png" width="200" /></a></div><br /><p><br /></p><span><a name='more'></a></span><p>Many of you are well aware about the endless discussions around audiophile listening skills (golden ears), double-blind-testing-with-nobody-would hear-a-difference results and of course our scientific<br />measurement friends who simply refuse to accept any sound difference if the (their) measurements won't show it. Measurements beyond any audible thresholds...</p><p>Golden Ears are usually pushed into defense caused by the lack of hard proof. </p><p>Now. Youtube is a place where I'd be very careful what stuff to trust. You'll find a lot of crap out there.</p><p>However. Over a Youtube you can also find numerous videos showing audio equipment, cable, fuse and whatever tweak comparisons that show clearly and obvious A/B sound differences. <br /><br />Would or should you accept such a video as proof of anything!?!? </p><div>If a video and its findings reflects exactly my own experience, I'd say. What the heck. It's not just me. (I know - it's not just me ;) ) Such a video is just confirming my and other peoples experiences in quite a generic way. Not just confirming. It's proving it.</div><p>The interesting thing is. These videos don't even need to be produced with top-of-the-class video/audio recording equipment. These videos still show more than obvious A/B differences. </p><p>Even on rather mediocre recordings and with entry level earplugs on my side (I use a $15 headset) hooked up to your smartphone, you can easily tell cable A from cable B, fuse A from fuse B, device A from B whatever that comparison is all about, apart. </p><p><a href="https://www.youtube.com/watch?v=EDmhcswZ-7I" target="_blank">Here's an example</a> (just one out of many)</p><p>You'd of course not be able to realize the absolute quality and sound performance of an audio system. </p><p>That'd be impossible. And that's not the point and intention.</p><p>The point is: </p><p>You can clearly and objectively identify "a change" on an audio system within these Youtube videos. <br />And that independent of what overall quality level that audio system under test exhibits. And - fun-fact - not just "a" change can be identified. IMO quite often you can easily say if the performance got better or worse. So what!?!? I do not know any audiophile, who doesn't want to improve the system. You can use all these videos simply as supporting material, a trigger to start your own investigations and projects or simply to shut up these ignorant folks out there. All that - to me - clearly underlines once more the fact that there's a huge (in audiophile terms) difference between test-tone measurements and related conclusions and real world music playback situations. </p><p>You basically need both, the measurement and the listening. <br />And you need serious listening. I do highly question many of these measurement-only prophets, if they are willing to run objective listening tests. What if they hear a difference they haven't measured?Something like that that would prove them wrong. Would prove their whole philosophy wrong. </p><p>There is a way out for them.</p><p>Simply consider these Youtube videos - peer reviews. Peer reviews of these partially misleading measurements and related conclusions. A real scientist would, would have to, start to try to explain what the obvious and proven difference (measurements vs. audition) is all about. Yep. Real scientists would feel challenged to get it all explained. </p><p><br /></p><p>Bottom line.</p><p>I do consider these Youtube videos quite useful in proving an audiophile point.<br />They clearly show that measurements alone won't tell the whole truth.<br />The videos also animate to look into and evolve your own system. Sometimes with minor efforts.<br />And yes, these videos also show what a difference a simple fuse can make.</p><p>I think it's fair to conclude: Fellow Golden Ears - keep trusting your ears! </p><p>...and talk about it.</p><p>As engineer myself, I don't give up the hope that one day the measurements folks out there are able to tell us what the differences many people hear is all about.</p><p><br /></p><p>Enjoy.</p><div><p>PS: This post (a slight variation) I posted initially <a href="https://www.audioasylum.com/cgi/vt.mpl?f=pcaudio&m=185058" target="_blank">over @ Audio Asylum</a> btw. . I thought it's worth to put it on the blog</p></div>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0tag:blogger.com,1999:blog-1557068496303009409.post-17844332331798290532021-08-07T18:29:00.001+02:002021-08-09T14:16:22.239+02:00introducing sKit 1.5<p>Hi there.</p><p>I uploaded a new version (1.5.) of sKit.</p><span><a name='more'></a></span><p>I changed</p><ul style="text-align: left;"><li>the initial installation procedure</li><li>the "license accept" feature has only to be done once from now on</li><li>some minor stuff</li><li>some text editing</li><li>change of git repo name</li><li>made sure that it is pCP8.0 compatible<br /><br /></li></ul><p>Nothing earth shattering.</p>Running the "update" via sKit-manager.sh should work. 1.5 is tested on pCP 8.0. <br />Before you update sKit update pCP. As usual, make a backup first.<p>On major pCP release changes, you should consider to build a pCP<br />from scratch instead of using the insitu upgrade. That's also what the pCP <br />folks recommend. </p><p>Please let me know if something is missing or not working.</p><p>Enjoy.</p><p><br /></p><p><br /></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0tag:blogger.com,1999:blog-1557068496303009409.post-21598330807147774542021-07-27T22:05:00.011+02:002021-07-28T11:16:53.821+02:00Full Digital Amp News<p>In 2012 I wrote an article about <a href="https://soundcheck-audio.blogspot.com/2012/08/ever-heard-about-full-digital-amps.html" target="_blank">Full Digital Amps</a>. I just updated it a little. That's one part of the news.</p><p>During those days I was trying and tweaking several of these FDAs. I really liked the technology and the idea behind it. And they sounded extremely clean and dynamic already back then. IMO cleaner and more dynamic then most DAC-amp combos I had listened to. In the article I describe what I didn't like and why I left that idea alone. Just to mention it: It was mainly due to the mediocre digital pre-stages of these amps. These amps also need top notch USB receivers, filters, clocks, DSP/SRC, power supplies, regulation, layouts, isolation etc. During those days that's IMO been a major weak spot. </p><p>I still kept watching for developments. Not much was happening over the years.</p><span><a name='more'></a></span><p><br /></p><p>It recently changed when SMSL introduced the <a href="https://www.smsl-audio.com/portal/product/detail/id/747.html" target="_blank">VMV A2</a>. </p><p>IMO a very interesting product. <a href="https://www.audiosciencereview.com/forum/index.php?threads/smsl-vmv-a2-review-stereo-amplifier-dac.24640/" target="_blank">ASR</a> basically confirms its excellent specs. Many of the recent SMSL DACs received pretty good reviews too. SMSL seems to intend to position themselves further up in the higher quality audio gear market.</p><p>...which would explain a rather steep (probably still reasonable price/performance ratio) price tag of 1k for that amp. I am not saying (at this point) it's not worth it.</p><p>I recently got myself two new toys (amp and DAC). If I would have known earlier I for sure would have tried the VMV A2 first. There are IMO quite some arguments for having a closer look at the FDA technology - if that's done right.</p><p>If I'd get a chance to at least try this device, I'd go for it without hesitation. You might want to have a closer look at that one too. However. I do predict further developments. Buying one now. Hmmh.<br />Better not. </p><p>I also learned no long ago that Peter Lyngdorf, beside some other amplifier hot-shots, is a Co-founder of <a href="https://purifi-audio.com/about/" target="_blank">Purifi</a> (I'm running such an amp for a while). The Purifi amp is considered a top notch device. I'd sign that. What has this to do with FDAs?<br />What I find really strange here is that the Purifi amp is just a Class-D amp. (Bruno Putzeys Hypex inheritance? ;) )<br />Especially Lyngdorf (TaCT) did build excellent first generation FDAs. Then there's Lars Risbo at the table, who according to the Purifi page, pioneered PCM-PWM conversion technology - marketed later by TI. Pretty solid FDA competence I'd say. With all that FDA competence at hand I am not sure why Purifi didn't go for a FDA solution. Perhaps they'll do one day. I hope so. </p><p>Competition and community pressure is the best fuel for evolving technologies. And SMSL hopefully started a new round in the game around that IMO evolutionary technology. If you read the thread over at ASR, people do find a hair in the soup ( is this a gemain term? ;) ) . For the time being I'll keep watching (since I really like what I have in front of me right now). There's no pressure on my side.</p><p>Enjoy.</p><p><br /></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0tag:blogger.com,1999:blog-1557068496303009409.post-72786011911322198202021-04-30T19:40:00.012+02:002023-11-25T12:05:08.495+01:00Everything that matters...<p><b><span style="font-size: medium;"></span></b></p><blockquote style="border: none; margin: 0px 0px 0px 40px; padding: 0px; text-align: justify;"><br /></blockquote><blockquote style="border: none; margin: 0 0 0 40px; padding: 0px;"><p style="text-align: center;"><b> <span style="font-size: large;">"Not everything that matters gets measured and not everything that gets measured matters"</span></b></p></blockquote><p> </p><div class="separator" style="clear: both; text-align: center;"><a href="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEidAA413ROM6ViyEgyH4VurBp7oea3OVW8Wa2_-zGHJq2GrNe0skfF3hXhIDfQhaHYHFCP8Zw8_C7Txb6UP5Q7p_-_4Z9hOUVDS2cGslWZ9rXVLWSJkOhGnpzVqwgFvBZd-6WwVIAAhqdVw/s864/einstein.png" style="margin-left: 1em; margin-right: 1em;"><img border="0" data-original-height="747" data-original-width="864" src="https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEidAA413ROM6ViyEgyH4VurBp7oea3OVW8Wa2_-zGHJq2GrNe0skfF3hXhIDfQhaHYHFCP8Zw8_C7Txb6UP5Q7p_-_4Z9hOUVDS2cGslWZ9rXVLWSJkOhGnpzVqwgFvBZd-6WwVIAAhqdVw/s320/einstein.png" width="320" /></a></div><br /><div><br /></div><div><br /></div><div>Nope. It wasn't him expressing it. Consider us kindred spirits. ;) Alberts original version reads</div><div><b style="font-size: large;"><br /></b></div><div style="text-align: center;"><b style="font-size: large;">"Not everything that can be counted counts and </b><b style="font-size: large;">not everything</b><span style="font-size: large;"> <br /></span><b style="font-size: large;">that counts can be counted" </b> </div><div style="text-align: center;"><br /></div><div style="text-align: center;"><br /></div><span><a name='more'></a></span><div><br /></div><div><br /></div><h1 style="text-align: left;">What's the point?</h1><div><br /></div><div>Since the beginning of audiophile times we all face a major ongoing dispute in the audiophile audio world. There's one group, let's simply call them "Wannabe-Scientists", claiming that if an audio device measures great, showing qualities (common specs) beyond the audible threshold, such a device is simply not to be questioned regarding its sound qualities and performance. Basically saying: If you can't measure it, it can't exist.</div><div><br /></div><div>And then there is the other group. Open-minded audio folks who simply listen and judge. Let's call them GoldenEars. That's the term Wannabe-Scientists usually use to belittle members of this group. People such as audio-gear manufacturers and designers, audio-measurements gear manufacturers, professional and hobbyist reviewers, engineers, die-hard audiophiles, audio enthusiasts and even our beloved wifes/partners. All these folks have one thing in common, they can tell A from B apart by simply using their ears. The vast majority of this group would confirm that not one device sounds the same, no matter how well it measures.</div><div><br /></div><div>By writing this article I intend to (1) point out the hypocrisy and ignorance of these Wannabe-Scientists and (2) encourage fellow people of the GoldenEar group not to let you get intimidated by these Wannabe-Scientists. Simply trust their own ears.</div><div><br /></div><div><br /></div><h1 style="text-align: left;">What triggered this article ???</h1><div><br /></div><div>The situation around Wannabe-Scientists and GoldenEars is not new. Both groups coexist side-by-side</div><div>since decades. I do know many manufacturers who simply consider it the most basic knowledge that measurements AND listening tests are a MUST to design audio equipment. I do also know quite a number of reviewers who would also confirm it - specs AND listening tests are a MUST. Actually the vast majority of interested audio folks will be confirming it. Most of these folks run the hobby for several decades based on reading (and understanding) the specs and listening to numerous systems and devices over the years. I actually consider the GoldenEar audio folks the main reason why our audio systems still evolve and never sounded better than today. </div><div><br /></div><div>Now. All this is not new. But, why are these "Wannabe-Scientists" so ignorant? It almost feels they follow some kind of ideology.<br /><br />In 2021 I came across a Youtube video from 2015 called "<i>RMAF15: What The Specs Don’t Tell You… And Why</i>". It's a presentation on USAs largest audio show by Jonathan Novick from Audio Precision. Audio Precision is manufacturing the best audio measurement gear money can buy. Basically all serious audio manufacturers and professional reviewers will have an AP product at hand for the work to be done. (You can watch the video later on)</div><div><br /></div><div><br /></div><div class="separator" style="clear: both; text-align: center;"><iframe allowfullscreen="" class="BLOG_video_class" height="266" src="https://www.youtube.com/embed/2V6YN-mshmY" width="320" youtube-src-id="2V6YN-mshmY"></iframe></div><br /><div><br /></div><h1 style="text-align: left;"><b>What is the key message ????</b></h1><div><br /></div><div>To make a long story short and simple. Jonathan in, as a part of his presentation, clearly outlines that standard audio measurements as provided by Audio Precisions own, even its best-in-class measurement equipment the APx555@$28000, would not sufficiently describe the <b>audible <i>sound </i></b>of an audio device or system. </div><div><br /></div><div>Let that sink in for a while!</div><div><br /></div><div>To hear that message from such a party. That's gotta be a major bummer for our Wannabe-Scientists out there. </div><div><br /></div><div>Let's rephrase a bit what's been said. I try to get to the essence of the story.</div><div><br /></div><div>Widely used measurements, such as </div><div><ul style="text-align: left;"><li>total-harmonic-distortion and noise (THD+N), </li><li>signal-to-noise and distortion ratio (SINAD), </li><li>dynamic range,</li><li>linearity</li><li>jitter, </li><li>asf., </li></ul>using specific test-signals (incl. multi-tone signals), would not sufficiently mirror the sound signature of a highly complex music signal on the same system. </div><div><br /></div><div>According to Audio Precision, there are simply no measurements that would properly reflect the underlying real music complexity. ( And I'd like to add to that - the real world audio system complexity.)</div><div><br /></div><div>However. All these audio measurements as provided by their devices would sufficiently prove that all the basics are right. These kind of measurements are useful and usually sufficient to make sure that production quality gets maintained. </div><div><br /></div><div>Great. </div><div><br /></div><div>Great to get such a message from a manufacturer who sets the baseline for large parts of the audio industry and the audio market, the audio sellers and users, the reviewers, you name it.</div><div><br /></div><h1 style="text-align: left;"><b>What now???</b></h1><div><b><br /></b></div><div>The Audio Precision message proves the GoldenEar fraction right. Ok. We always knew we were right. And that message puts these Wannabe Scientists to shame. It basically confirms to me that beside running a standard set of measurements, it'll require listening tests to get the full picture about a devices overall performance. </div><div><br /></div><div>It's obviously a known fact in the industry that all measurements that'd be required to define to 100% the overall performance of an audio device or system are not identified to date.</div><div><b><br /></b></div><div>How can we cope with this situation. The message implies that we need </div><div><ol style="text-align: left;"><li>standard measurements AND</li><li>listening tests</li></ol>to describe the sound signature of a system. That's the situation we're actually in - since decades.</div><div><br /></div><div>I am not aware of a scientific study or paper that would analyze the topic. No idea why science never really "properly" looked into this particular subject. </div><div><br /></div><div>Talking about science (or perhaps better Wannabe Scientists). There's a popular place called <a href="https://www.audiosciencereview.com/forum/index.php" target="_blank">Audio Science Review</a> founded by Amir Majidimehr, an electrical engineer who used to work for Microsoft. </div><div><br /></div><div>I really like this place. Not that many places like that on the net. Its name implies scientific reviews to begin with. How can a serious audio lover stay away from such a promise ??? </div><div><br /></div><div>As expected, just by using highest end measurement gear the site founder claims to provide scientific</div><div>reviews. Hmmh. And no. Listening tests are not scientific according to that site. Listening tests are subjective. As a consequence listening tests are simply left out to make things "scientifically" easier.</div><div>Amir also judges devices, he even recommends devices based on just measurements. Hmmh. </div><div>And not to get anybody confused out there, people who start talking about listening tests will get a pretty tough time over there. I know, it sounds awful. It actually IMO is.</div><div><br /></div><div>Why than liking this place? Right. I said that. I like ASR for its measurements of all kind of audio gear. Amir uses the top of the line measurement gear from Audio Precision. A statement! Amir IMO produces reliable measurements. Presents them nicely and also presents list rankings. For me this is simply a good starting point when looking into new gear. </div><div><br /></div><div>What you can also see is that manufacturers respond rather quickly to his findings. I'd go that far that a certain number manufacturers seem to be using the ASR review charts and rankings to define their design goals for beating the competition. Not a bad thing. We as users love it.</div><div><br /></div><div>But that has another weird side effect. If you look at the ASR rankings you'll notice that the number of devices on top of the list is continuously growing. You can consider is a saturation effect. It simply can't get any better. Because the measurement gear limitations simply prohibit from showing even better results. According to ASR a lot of that gear under test runs nowadays in areas far beyond audible thresholds. One could wonder why running the tests at all???</div><div><br /></div><div>Amir's place is also a nice spot to get in touch with Chinese audio gear manufacturers, such as SMSL, Topping, Gustard, you name it. It's one of the very few places on the net where you can get in touch with them inside the forum. </div><div> </div><div>There's a huge problem with this place - Audio Science Review. It's name already refers to the problem. The site name implies a scientific approach. What's presented are basic audio measurements.</div><div>And people who talk about listening experiences are attacked, ignored or even expelled. </div><div><br /></div><div>We know by now, ignoring listening experiences or tests - which are a must - can't lead to reliable reviews or conclusions about a device or system. Such an approach would never fulfill the most basic scientific demands of being called scientific. </div><div><br /></div><div>People need to be aware that the reviews and even recommendations over at Audio Science Review</div><div>can be quite misleading, because listening tests are not part of the story.</div><div><br /></div><div><div>Don't get me wrong. This is no Audio Science Review bashing. Getting your own baseline right by looking at the ASR measurements is absolutely OK. If you simply ignore Amirs conclusions, recommendations, demystifying videos asf and you trust your own ears - it'll do. </div><div><br /></div><div>One more thing that's bugging me. If you've watched the Audio Precision video from earlier, you might have noticed the ASR founder Amir also being present at that event. He basically was being told by his own measurement gear provider that the measurements he creates are not sufficient to explain what we all actually hear and experience. </div><div><br /></div><div>Then, why is he still pushing his "scientific" storyline!?!? He didn't listen, forgot about it or he actually didn't want to listen - didn't want to hear it. Not good.</div><div><br /></div><div>It's not just ASR though. You'll find people all over the net trying to shut you up about listening impressions. </div><div><br /></div><div>Don't let you intimidate you! Trust your own ears... ...and share it. You'll be surprised how many people</div><div>will hear the same good or bad thing you experienced by yourself.</div><div><br /></div></div><div><br /></div><div>Back to that Audio Precision RMAF presentation which was called "What specs don't tell you...".</div><div><br /></div><div>There seem to be some ideas floating around about what's missing and where to look for it. But even now in 2023 - 8 years later - I haven't seen any progress on clearly explaining what's missing. And at that point I do hope that real "science" steps in at one point.</div><div> </div><div><br /></div><div><br /></div><h1 style="text-align: left;">Bottom line</h1><div><div>"Not everything that matters gets measured and not everything that gets measured matters". </div><div><br /></div><div>Buying your next audio device simply based on measurements or pseudo-scientific reviews might wrong-foot you.</div><div><br /></div><div>Make sure you also run listening tests on your own and make sure you'll be able to return the device under test. </div><div><br /></div><div>Your listening skills will improve over time. The more you hear the better you're able to judge a device. </div><div>You learn what to "hear" for. It's difficult in the beginning to judge what's right or what's wrong.</div></div><div><br /></div><div>Audio is about finding the best compromise that meets your personal taste. </div><div><br /></div><div>Sharing your findings with others will grow your confidence. Been there done that. ;) </div><div><br /></div><div>Perhaps one day we see a sincere scientific approach that explains properly the gap of what we measure today and what we actually hear. Basically describing the audio experience to almost 100%. I am sure training an AI for the task would be an interesting development.</div><div><br /></div><div>The big Q. Does the industry really want a close to 100% definition of sound quality?? Do they want a set of measurements that explains close to 100% of what we hear !?!?</div><div><br /></div><div>All differentiation would be gone in no time. And a huge industry would face a major problem.</div><div><br /></div><div>For us audio folks times would be great. Audio gear would become a commodity @ low prices and excellent and predictable performance. </div><div><br /></div><div>We finally would be able to listen to music! Nah. That'd be boring. We'd get our hobby eliminated.</div><div><br /></div><div>Enjoy.</div>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com3tag:blogger.com,1999:blog-1557068496303009409.post-41293820936040323632021-04-27T14:31:00.002+02:002021-04-27T14:31:37.225+02:00introducing sKit 1.4 (beta)<p>Hi folks.</p><p>I worked a little on sKit to make it more convenient and stable.</p><p>* most of the tweaks get automatically installed now<br /> no real need to dig into the Audio Streaming series anymore. </p><p>* sKit-manager remove function - removes everything from now on<br /> there shouldn't be any sKit trace left anymore once that is done.</p><p>What's basically left for to you do is</p><p></p><ol style="text-align: left;"><li>you install a fresh pCP image,</li><li>increase it to at least 200MB, </li><li>configure your audio interface via WEB-UI and</li><li>install sKit</li></ol><p></p><p> At that point you should have a well tuned system up'n running.</p><p>On 1.4 I added some changes to the affinity tweak. I now assign squeezelite to<br />CPU 1 and 2 without isolating these CPUs anymore.</p><p>The documentation also got updated. And of course this or that error got corrected.</p><p>sKit 1.4 needs a fresh pCP install as base. Using sKit-manager update won't work. Sorry for that.</p><p>sKit 1.4 is still beta. I'll wait for 2 weeks for feedback before officially releasing sKit 1.4. I count on your feedback.</p><p><br /></p><p>Enjoy.</p><p><br /></p><p><br /></p><p><br /></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com9tag:blogger.com,1999:blog-1557068496303009409.post-71174273370336794322021-02-16T18:15:00.003+01:002021-02-16T18:15:34.185+01:00introducing sKit 1.1<p> I've added some interesting new features to sKit (1.1).</p><span><a name='more'></a></span><p><br /></p><p>1. sKit-check.sh</p><p>sKit-check.sh checks your installation against the mods and suggestions made in the Audio Streaming series. You'll get a quick overview about your situation.</p><p>I made it because some people might get confused by all the mods to be applied. Even I loose frequently control about what is enabled and disabled.</p><p>I hope it's much better now.</p><p><br /></p><p>2. sKit-restore.sh</p><p>pCP just recently introduced a nice online-backup feature. I though why backup only!?!? And wrote a online-restore tool. To me it's very convenient.</p><p><br /></p><p>3. Error corrections and fine tuning</p><p><br /></p><p><br /></p><p>4. How to update to sKit 1.1.</p><p><br /></p><p>As promised there's no new installation required. However. You need to run the sKit-manager</p><p>from github because I had to change the repo-urls.</p><p>You run the known commands:</p><p><b>url="https://raw.githubusercontent.com/klslz/tuningkitpcp/master/sKit-manager.sh"</b></p><p><b>f=/tmp/sKit-manager.sh; wget -q "$url" -O $f;chmod u+x $f;$f</b></p><p>from a pCP terminal.</p><p>And then you choose "update" inside the sKi-manager.</p><p>Once the update is done you can try the new tools. I documented them over at </p><h1 style="text-align: left;"><a href="https://soundcheck-audio.blogspot.com/p/the-skit-pcp.html" target="_blank">sKit</a></h1><p>That'd do.</p><p><br /></p><p>Enjoy.</p><p><br /></p><p><br /></p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com1tag:blogger.com,1999:blog-1557068496303009409.post-38621781935663072372021-02-14T15:12:00.001+01:002021-02-14T15:12:46.521+01:00introducing soundcheck's tuning kit - pCP (sKit) 1.0<p> I'd like to introduce the first release of <b><i>soundcheck's tuning kit - pCP (sKit) </i></b>.</p><p>sKit provides a small and potentially growing number of little tools that'll help to setup piCorePlayer 7.0+ and Rpi4 and 3 in line with the advanced tuning recommendations outlined in my <b><i>Audio Streaming Series</i></b>.</p><p>The underlying idea is to take quite a basic operating system like piCorePlayer and tune it that way that it'll provide you with the best possible performance -- challenging most other (audiophile) OSes and streaming solutions out there.<br /><br />This is not just another of these meanwhile numerous "audiophile" operating system out there though. In principle these do nothing else then applying this or that tweak you'll also find on this blog for years. The idea is to provide you with a tuning kit instead. You take it and apply it to that standard operating system installation yourself. And sKit supports you on that. </p><p>You'll put yourself in the driver seat. And you control the pace.</p><p>sKit already goes slightly beyond the earlier outlined Audio Streaming Series scope. You should still go through the Audio Streaming Series step-by-step though. sKit doesn't cover all settings yet. At the relevant spots inside the Audio Streaming Series I refer to sKit.</p><p><br /></p><h1 style="text-align: left;"><a href="https://soundcheck-audio.blogspot.com/p/the-skit-pcp.html" target="_blank">sKit 1.0</a></h1><p><br /></p><p>I hope you'll enjoy the results.</p><p>SC</p>soundcheck (aka Klaus)http://www.blogger.com/profile/01547136694650184460noreply@blogger.com0