Topic: Hi-Res files with IdPi on Ras Pi 4B (Read 3930 times)

ASenna1 · « **on:** October 17, 2019, 12:51:43 am »

I'm currently using Volumio on a Ras Pi 3 (Ethernet) as a DLNA client, connecting my JRiver library to a Denon PMA-2500 via USB.

With this combo, 44.1/16 FLAC playback is flawless, but I do notice occasional drop-outs with higher resolution (above 44.1KHz) and higher bit-rate (24-bits) files.

Does anyone have experience with the IdPi-Ras Pi 4 combo in a similar setup with high-res files? Would switching to IdPi-Ras Pi 4 combo solve the drop-out problem I hear with my Volumio-Ras Pi 3 config?

MikeyFresh · « **Reply #1 on:** October 17, 2019, 06:32:49 pm »

I have no experience with that particular Denon unit, however I have used Pi3B, Pi3B+, and Pi4B with both Volumio, and Moode, and PCM up to 352.8kHz as well as DSD up to DSD128 plays back with zero drop out.

There can be many reasons why you'd hear drop outs, but the RPi version at least including RPi3B and newer should not make any difference, they all work well in terms of playback connected to a JRiver server.

In my experience you should use the server called Generic DLNA, don't use "24 Bit Audiophile" or whatever the heck it's called, and make sure it isn't defaulting to a server called L16, I actually delete that one because the program seems to want to always default to it.

ASenna1 · « **Reply #2 on:** October 17, 2019, 07:37:08 pm »

Interesting on the Generic DNLA recommendation because I am now using the 'Audiophile 24-Bit DAC' option. I'm going to make the switch tonight and take a listen.

Thanks for the suggestion!

MikeyFresh · « **Reply #3 on:** October 17, 2019, 08:44:45 pm »

Quote from: ASenna1 on October 17, 2019, 07:37:08 pm

Interesting on the Generic DNLA recommendation because I am now using the 'Audiophile 24-Bit DAC' option. I'm going to make the switch tonight and take a listen.

Thanks for the suggestion!

I bet that will solve it, the DAC in the PMA-2500 is 32-bit capable, not that there is any 32-bit source material, however there is no reason why the bit depth needs to be limited to 24-bit, which is what I think that "Audiophile" server actually does, that's my guess anyway.
Check the rest of the configuration too, I set everything to "no conversion" and "original", disable the software volume control(s), and then there are absolutely no glitches. Even the low powered Pi Zero W has no issue with either Volumio or Moode as the Renderer.

JimH · « **Reply #4 on:** October 18, 2019, 02:01:04 am »

Quote from: MikeyFresh on October 17, 2019, 08:44:45 pm

... there is no reason why the bit depth needs to be limited to 24-bit, which is what I think that "Audiophile" server actually does, that's my guess anyway.

It has nothing to do with the media bitdepth. It's just how the DAC receives data.

ASenna1 · « **Reply #5 on:** October 18, 2019, 03:20:28 am »

Thanks MikeyFresh because the random dropouts I was noticing lately had been driving me ca-razy! I listened to a series of 48/24 and 96/24 files after following your recommendations and everything sounded clean again.

Thanks again!

(Any reason to look into Moode as a renderer?)

ASenna1 · « **Reply #6 on:** October 22, 2019, 11:58:54 pm »

I've been suspicious that it's the USB performance of RPIs that contributes most to dropouts that are especially apparent in hi-res audio files.

https://github.com/raspberrypi/linux/issues/2215

The question for me still is, will RPI4 USB improvements be able to provide dropout-free performance at all sample rates/depths?

JimH · « **Reply #7 on:** October 23, 2019, 06:41:22 am »

Thd only way to know is to try it.

ASenna1 · « **Reply #8 on:** October 23, 2019, 11:47:55 am »

That I know.

MikeyFresh · « **Reply #9 on:** October 23, 2019, 10:58:38 pm »

It's not the RPi hardware, even the lower spec'd Pi Zero W has exactly zero problems in playing hi-rez content, I'm talkin' 24-bit 192 kHz PCM, or DSD128, no issues using either Moode or Volumio as the player software on the renderer/endpoint side.
Ditto 32-bit which is only in the end just 24-bit with zero padding as there is no actual 32-bit content.
Neither of the above are the issue with your playback dropouts.
Is your Denon fully updated as far as firmware goes?

ASenna1 · « **Reply #10 on:** October 24, 2019, 12:41:29 pm »

The PMA-2500NE does not contain firmware. CPUs create noise and the Denon is a very pure integrated amp that's CPU-free. With my Lenovo T430 in place of my RPI 3B, everything high res plays senselessly and that's with the T430 connected to my LAN in WiFi.

I also have CA DACMagic Plus (24bit) USB-connected to a W10 desktop machine that drives (balanced audio) two near-fields and I've never heard a single dropout. That machine is connected to the library server through my LAN.

I disagree that 32-bit does not have an impact. "Zero padding" is irrelevant because it doesn't matter that bits 25 through 32 are 0 or 1, it's still more data. See this calculator > http://www.theaudioarchive.com/TAA_Resources_File_Size.htm

And see this info from my Moode RPI. > https://photos.app.goo.gl/xrmFyo24iSn2Qppu8

There's no shortage of examples of RPI USB limitations. Whether the 4B improves on that is a reasonable question.

MikeyFresh · « **Reply #11 on:** October 24, 2019, 01:26:59 pm »

I think the point I was making is that even the low-spec hardware Pi Zero W has no issue running either Moode or Volumio as a DLNA/UPnP endpoint renderer. I have used them both with about 6 different USB DACs, no issues whatsoever with the dropouts you are describing, PCM up to 352.8kHz, and DSD128.

Good luck.

ASenna1 · « **Reply #12 on:** October 24, 2019, 02:35:58 pm »

The question is still the same and that's RPI high-speed compatibility.

Please note this from raspberrypi.org..

3. Esoteric USB sound cards
Expensive "audiophile" sound cards typically use far more bandwidth than is necessary to stream audio playback. Reliable operation with 96kHz/192kHz DACs is not guaranteed.
As a workaround, forcing the output stream to be CD quality (44.1kHz/48kHz 16-bit) will reduce the stream bandwidth to reliable levels.

https://www.raspberrypi.org/documentation/hardware/raspberrypi/usb/README.md

>>> "operation with 96kHz/192kHz DACs is not guaranteed."<<<

That's exactly the issue I've been experiencing and precisely the point I've been trying to make.

The Q about RPI4 USB performance remains.

ASenna1 · « **Reply #13 on:** October 25, 2019, 12:42:30 pm »

Based on the USB documentation provided by raspberrypi.org and my experience with my equipment, it would be helpful to everyone if JRiver provided the same guidelines with their IdPi product announcements.

syndromeofadown · « **Reply #14 on:** October 25, 2019, 03:59:06 pm »

Quote

The Q about RPI4 USB performance remains.

I have been using Pi's for a few years from version 3 to 3+ to 4 with MC. I have a stereo that has a Pi inside running some variant of Volumio that went from 2 to 3 to 3+. I have never had dropouts. I do however have networks involved and hats to output via optical. I can test a Pi4 connected to Modi Multibit DAC via USB playing 24-196 audio if that's what you are looking for. How frequent are the dropouts you experience?

ASenna1 · « **Reply #15 on:** October 25, 2019, 08:13:12 pm »

The dropouts are infrequent, 2 to 4 in a 192/24 album, about the same if not a bit more from a SACD. And that's what's obvious.

Thanks for the offer to test, but you would need to connect to a 32-bit DAC to mirror the USB path of my Denon PMA-2500NE. The Modi Multibit DAC is 24-bit.

I use optical for connection to a rarely used CD player and optical may bypass the RPI's USB limitations, but for a variety of reasons it's not the path I'd want to go for the MC library connection.

bob · « **Reply #16 on:** October 25, 2019, 09:11:39 pm »

Quote from: ASenna1 on October 17, 2019, 07:37:08 pm

Interesting on the Generic DNLA recommendation because I am now using the 'Audiophile 24-Bit DAC' option. I'm going to make the switch tonight and take a listen.

Thanks for the suggestion!

Is there any reason to not use "Original Format"?

ASenna1 · « **Reply #17 on:** October 25, 2019, 09:45:58 pm »

Generic DLNA/Original Format is what I've tested with.

ASenna1 · « **Reply #18 on:** November 02, 2019, 08:10:44 pm »

Problem solved with a nuc7i3bnh.

Maybe the RPI 4B solves previous USB limitations, but until proven

otherwise, if you've spent decent money on a DAC or integrated amp, I don't think it's worth the money (and maybe A LOT of time) gambling on a RPI.

TECH TIP: If you go the W10/NUC/Headless route, don't use Remote Desktop to enable WASAPI in MC. Use Team Viewer or an equivalent.

(UPDATE: Open Remote Desktop (Local machine) > Show Options > Local Resources > Remote Audio Settings > Play on Remote Computer)

MikeyFresh · « **Reply #19 on:** November 02, 2019, 10:10:41 pm »

Quote from: ASenna1 on November 02, 2019, 08:10:44 pm

Problem solved with a nuc7i3bnh. Maybe the RPI 4B solves previous USB limitations, but until proven otherwise, if you've spent decent money on a DAC or integrated amp, I don't think it's worth the money (and maybe A LOT of time) gambling on a RPI.

TECH TIP: If you go the W10/NUC/Headless route, don't use Remote Desktop to enable WASAPI in MC. Use Team Viewer or an equivalent.

I'll say this one last time and leave it at that, I don't want others considering the use of Raspberry Pi to be misled, so I'll detail the number of different USB DACs and RPi models that I have found are completely compatible with each other for flawless hi-rez audio (up to PCM 352.8kHz and DSD128 fully tested):

RPi 3B, RPi 3B+, RPi Zero W, RPi4B are all working without any glitches or any other issues as UPnP Renderer/Endpoints running either Moode or Volumio playback software with the following USB DACs:

Oppo HA-2, Resonessence Labs Herus, LH Labs GO100, GO450, GO720, GO1000, and Pulse SFi, Topping D10, iFi iDAC2, and nano iDSD, ADL Stride, and Simaudio MOON 280D (12 total DACs or 9 different models).

The above is with RPi 3B (JRMC22), RPi 3B+ (JRMC24), and RPi 4B (JRMC25) as Id Pi DLNA servers.

The only limitation at all is inherent to the ADL Stride's UAC1 specification, which means it tops out at PCM 24/96, and no DSD. Thats it, the rest do PCM 352.8kHz and DSD128 even on the low powered RPi Zero W, with either Moode or Volumio playback software, and an Id Pi DLNA server.

ASenna1 · « **Reply #20 on:** November 02, 2019, 11:38:10 pm »

"Misled"?

Until proven otherwise with the RPI4B, my experience suggests following www.raspberrypi.org recommendations.

3. Esoteric USB sound cards
Expensive "audiophile" sound cards typically use far more bandwidth than is necessary to stream audio playback. Reliable operation with 96kHz/192kHz DACs is not guaranteed.

As a workaround, forcing the output stream to be CD quality (44.1kHz/48kHz 16-bit) will reduce the stream bandwidth to reliable levels.

https://www.raspberrypi.org/documentation/hardware/raspberrypi/usb/README.md

ASenna1 · « **Reply #21 on:** November 03, 2019, 11:05:53 am »

Also from www.raspberrypi.org

The USB host port inside the Pi is an On-The-Go (OTG) host as the application processor powering the Pi, BCM2835, was originally intended to be used in the mobile market: i.e. as the single USB port on a phone for connection to a PC, or to a single device. In essence, the OTG hardware is simpler than the equivalent hardware on a PC.

OTG in general supports communication to all types of USB device, but to provide an adequate level of functionality for most of the USB devices that one might plug into a Pi, the system software has to do more work.

duncan4791 · « **Reply #22 on:** November 03, 2019, 02:05:27 pm »

If you are referring to the Raspberry Pi 4B then you have misread the article at https://www.raspberrypi.org/documentation/hardware/raspberrypi/usb/README.md.
Review the section on the Pi 4B quoted below:

Quote

Overview for the Pi 4
For the Pi 4, a fully-featured host controller drives the downstream USB ports. Downstream USB is provided by a Via Labs VL805 chip - that supports two USB 2.0 ports and two USB 3.0 ports. This is connected to the BCM2711 SoC using a PCIe link, which is extremely fast. Therefore, the Pi 4 does not have the speed constraints of previous models, which means very fast datarates, especially when using the USB 3.0 ports.

It also means that almost all of the known issues with USB on previous models are no longer present.

The OTG hardware present on previous models of Pi is still available and it has moved to a single connection on the USB-C port. The OTG hardware is intended to be used in device-only mode on Pi 4.

Throughout the article it states "not Pi 4" on most of the subsections. Your last two posts refer to sections about the Pi 3 or earlier which all have the USB/Ethernet on the OTG buss, not the Pi 4.

I have tested USB3.0 hard disk raw read speeds up top approx. 350MB. There may be compatibility issues with some USB 2.0 DACs with the USB implementation or problems in the ALSA setup used by JRiver MC?, USB port bandwidth is not the problem.

ASenna1 · « **Reply #23 on:** November 03, 2019, 03:56:55 pm »

"Mis-read"?

Which part of "Until proven otherwise with the RPI4B, my experience suggests following www.raspberrypi.org recommendations." is not clear?

"It also means that almost all of the known issues with USB on previous models are no longer present."

Almost all"? Then allow me to repeat: "Until proven otherwise with the RPI4B..."

"Your last two posts refer to sections about the Pi 3 or earlier which all have the USB/Ethernet on the OTG buss"

Have you, duncan4791, read this thread from the beginning?

JimH · « **Reply #24 on:** November 03, 2019, 04:01:05 pm »

Quote from: ASenna1 on November 03, 2019, 03:56:55 pm

"Mis-read"? Which part of "Until proven otherwise with the RPI4B, my experience suggests following www.raspberrypi.org recommendations." is not clear?

"It also means that almost all of the known issues with USB on previous models are no longer present."

Almost all"? Then allow me to repeat: "Until proven otherwise with the RPI4B..."

"Your last two posts refer to sections about the Pi 3 or earlier which all have the USB/Ethernet on the OTG buss"

Have you, duncan4791, read this thread from the beginning?

Please tone it down a little. The bolds are unnecessary, among other things.

I haven't read everything you've said, but you're taking a strongly negative position on the RPi's capabilities, and some of it is just wrong.

MikeyFresh · « **Reply #25 on:** November 03, 2019, 05:45:20 pm »

Quote from: JimH on November 03, 2019, 04:01:05 pm

Please tone it down a little. The bolds are unnecessary, among other things.

I haven't read everything you've said, but you're taking a strongly negative position on the RPi's capabilities, and some of it is just wrong.

Thank you for that, I wholeheartedly agree with your assessment Jim.

My note above listed very specifically the different versions of the RPi I have and use either as Id Pi servers, or as UPnP endpoint renderers in conjunction with either Moode, or Volumio player software.

I additionally listed all of the different USB DACs I have successfully used in this manner. I own all of those DACs, it wasn't just some quick test where I failed to come across some sort of glitch or limitation, they have all been used extensively at one time or another over the course of 4 years.

So that's my data point, and I don't care about assertions to the contrary gleaned form the RPi forum, because I already know what works and what doesn't. The above combinations all work exactly as I described them.

Jim, you are correct, and I'm sure you also know through actual testing, the RPi is a very capable performer for the specific task of audio server or renderer when used with any number of USB DACs. Hi-rez tracks? No problem.

Do some DACs have firmware that for whatever stupid reason doesn't work well with Linux? Yes, there are many such reports out there. Odd but true.

ASenna1 · « **Reply #26 on:** November 03, 2019, 07:08:10 pm »

"I haven't read everything you've said, but you're taking a strongly negative position on the RPi's capabilities, and some of it is just wrong."

"Strongly negative"? Oh really? What exactly have I posted that is "just wrong"?

ASenna1 · « **Reply #27 on:** November 03, 2019, 07:15:00 pm »

"I don't care about assertions to the contrary gleaned form the RPi forum, because I already know what works and what doesn't. The above combinations all work exactly as I described them."

"Assertions to the contrary gleaned form the RPi forum"

https://www.raspberrypi.org/documentation/hardware/raspberrypi/usb/README.md

That's https://www.raspberrypi.org > documentation > hardware > raspberrypi > usb > README.md

That ain't no forum.

INTERACT FORUM

Author Topic: Hi-Res files with IdPi on Ras Pi 4B (Read 3930 times)