Moode Forum
[SOLVED] Very distorted when NOT in cover view - Printable Version

+- Moode Forum (https://moodeaudio.org/forum)
+-- Forum: moOde audio player (https://moodeaudio.org/forum/forumdisplay.php?fid=3)
+--- Forum: Support (https://moodeaudio.org/forum/forumdisplay.php?fid=7)
+--- Thread: [SOLVED] Very distorted when NOT in cover view (/showthread.php?tid=4952)

Pages: 1 2

RE: Very distorted when NOT in cover view - Reddo84 - 07-12-2022

(07-11-2022, 10:03 PM)DRONE7 Wrote:
Quote:Is the RasPi 3+ inside the USBridge Sig underpowered for the task?

Should be ok.... I note a post re Pi3 and increasing the swapfile size.
No idea if this is of any help but here is the link..

https://www.diyaudio.com/community/threads/moode-audio-player-for-raspberry-pi.271811/post-7061368

Thanks for the suggestion, I tried it this morning. Unfortunately, it didn't help with the problem at hand.

RE: Very distorted when NOT in cover view - TheOldPresbyope - 07-12-2022

(07-11-2022, 07:42 PM)Reddo84 Wrote: So I've disconnected the USBridge Sig from my system completely and pieced it back in - bit by bit - checking for distortions after each step (after rebooting). For a DAC, I used a Dragonfly Red instead of my regular miniDSP 2x4 HD. Everything went fine, no distortions.

It's a fact of life that digital computers generate prodigious amounts of digital noise.[1] Circuit designers try to minimize the impact of this noise (in order to meet regulatory requirements if nothing else) but it can leak across various parts of the system (aka EMI or, crudely, "cross talk"). The Allo claim is that the USB port on its USBridge is especially "quiet" in this regard. It certainly seems there's not enough leakage to affect the Dragonfly, but then I haven't had this problem with any regular RPi and several different USB DACs either.

Quote:The last thing I changed back to normal was the DAC ... et voilá: With the miniDSP 2x4 HD, the distortions are back full force when not in CoverView. When in CoverView, everything is fine.

OK, so this suggests there is leakage to the miniDSP which is causing havoc. The input circuitry to the Dragonfly and the miniDSP differs and the internal processing of the reconstructed digital signals differs so I can't comment on why this is occurring.

The hard bit is knowing how the leakage is occurring so it can be mitigated. Could be through poor electrical grounding, poor power supply filtering, poor cable shielding, yada yada yada. One could write a book about this. Indeed, many such books have already been written. There are all sorts of tricks to EMI mitigation but I have no idea which trick(s) will apply in your case.

Quote:Another observation: When I boot up the USBridge Sig and only then connect the screens HDMI cable to the device, the resolution is off (too low), but there's less distortions.

This reinforces in my mind that the interference is being generated by the Javascript being processed in the CPU and the results then processed in the GPU-related subsystem since lower resolution means fewer pixels are being flipped per interval which means lower bitrate which means lower frequency.

Quote:...

Is there more load on the RaspberryPi when:
  • not in CoverView?

Do you see more things changing on the screen in one view? That implies more processing (in Javascript).

Quote:
  • running a miniDSP 2x4 HD instead of a Dragonfly Red DAC?

See above discussion.

Quote:
  • the resolution of the screen is higher? 

Yup. See above.  More bits implies more processing

Quote:Is the RasPi 3+ inside the USBridge Sig underpowered for the task? I contacted Allo some days ago, still no reaction. I'm trying to get my hands on a regular Raspberry 4 to test this hypothesis. 

I can't wait to hear if the miniDSP is susceptible to this noise when coupled to a stock RPi.

Regards,
Kent

[1] Ca 1972 I was using the Maniac III computer at the University of Chicago to cross-compile programs for the PDP-11 I used to control experiments and collect data in my lab.We wizards who stayed up late would put a transistor radio (remember those?) next to the console and try to compose Maniac programs which would cause recognizable if noisy tunes to play on the radio when the programs ran. The chain printer could be used to add a raspy counterpoint.

RE: Very distorted when NOT in cover view - Tim Curtis - 07-12-2022

For moOde local display always have the display connected first before booting the Pi otherwise the display's resolution may not be detected correctly by X11. It prolly defaults to 640x480 or something like that.

RE: Very distorted when NOT in cover view - Reddo84 - 07-15-2022

(07-12-2022, 09:17 PM)TheOldPresbyope Wrote: I can't wait to hear if the miniDSP is susceptible to this noise when coupled to a stock RPi.

That could take some time, I'm not paying top dollar for a RPi just to try this out, so I'm looking for a good offer for a used one. 


In the meantime, I have pulled apart my system completely and tried some more things, with headphones connected to the DACs (with 8.1.2):
  • When the miniDSP 2x4 HD was just connected to the USBridge Signature the problem persists anyway.
  • When the USBridge Sig was connected to a Steinberg UR12 Audiointerface DAC, the distortions sounded different, but were very present nonetheless. Please note that the Steinberg is powered through USB, so no power supply issues there.
  • The distortions from the Steinberg didn't react to the USBridge Sig being powered by its stock power supply, another 5V power supply, my custom battery setup or a 5V powerbank.
  • The distortions from the miniDSP didn't react to the USBridge Sig being powered as described above. Also powering the miniDSP straight from a 12V motorcycle battery didn't change anything.
  • Grounding the USBridge Sig through its dedicated grounding connection didn't change anything.
  • Trying different USB-cables from the USBridge Sig to the DACs didn't change anything.
  • Starting a track stored locally on the USBridge Sig and then disconnecting the Ethernet-cable didn't change anything. The distortions were still present.
So ...
  • I pulled apart my system and made sure all the cables were at the most possible distance from one another. I think cable interference can safely be ruled out.
  • Even powering the whole setup - the USBridge Sig as well as the DAC - straight from batteries didn't change anything. I think power supply can safely be ruled out.
  • Trying different cables didn't change anything. I think faulty cables can safely be ruled out.


Quote:This reinforces in my mind that the interference is being generated by the Javascript being processed in the CPU and the results then processed in the GPU-related subsystem

I think so, too. Is there any way to mend this? I am still convinced that it is the circular track-progress indicator that is causing the problem. Sometimes, it freezes for a couple of seconds and when it comes back to life again, this is always accompanied by a burst of distortions. 

Another thought: The Audioquest Dragonfly Red - the only DAC so far not showing this problem - is a second gen smartphone DAC, specifically engineered to not drain the smartphones battery as fast as its predecessor. Maybe the Dragonfly Red doesn't put as much strain onto the USBridge Sig ... what could be learned here?

RE: Very distorted when NOT in cover view - TheOldPresbyope - 07-15-2022

@Reddo84

The distortion may show up when a specific bit of Javascript code is executed in the Chromium browser engine, but that's just a symptom. The heart of the problem is the leakage of parasitic signals into the external DAC; e.g., it's a hardware issue.

There's an offhand chance that this leakage is occurring through what's called "common mode" noise current. The usual hack for solving this in the case of radio frequency noise is to pass the interconnecting cable (the USB cable in this instance) through a ferrite core noise suppressor. Indeed, some cables come with one already in place. In severe cases, the cable may be looped through the core more than once. These devices are cheap so you might try one to see if it helps mitigate the problem. They are often seen on power cables as well, although your experiments so far suggest this isn't the entry point. No harm in trying anyway.

As I said, the input circuitry and signal processing differs in the different DACs you are using so it's not so surprising that you see different responses to the hypothesized leakage. The other day I looked up the miniDSP 2x4HD and found it uses the XMOS Xcore200 microcontroller to handle the USB interface and the Analog Devices ADSP21489 400MHz Sharc microcontroller to handle DSP so there's plenty of opportunities for mischief. No idea about DragonFly and Steinberg other than to guess they have their own solutions.

I'm running out of suggestions. I'm disappointed you haven't received useful advice directly from Allo. Perhaps there are other USBridge Signature owners among the moOde forum members who can chime in?

Regards,
Kent

RE: Very distorted when NOT in cover view - Reddo84 - 12-08-2022

Problem solved, case closed. It was the Allo USBridge Signature! I finally got a RPi 4 today and voilà - all issues are gone. Also, I think the stock RPi sounds actually better then my Allo USBridge Signature!

Spending $$$ on the Allo streaming transport - wasted! So, so many hours of trying this, trying that - wasted! All the time I spent with the utterly USELESS Allo support - wasted! I'm so mad right now, I can't even ...