I have a Fiio BTR5 USB DAC that can support hi-res playback. However, Plexamp on the Mac is not aware of the DAC and thus doesn’t play tracks in hi-res format through the DAC. The audio program Audirvana does recognize the DAC and plays FLAC files in their full hi-res glory. However, Audirvana doesn’t have any Plex integration, so I don’t really want to use it.
1 Like
For it to make any sense, Plexamp would have to support “exclusive mode” and bitperfect audio like Audirvana or Tidal. Otherwise, everything would just get re-sampled to macOS native 48kHz before being passed to the BTR5 via USB.
Ironically, this is how “hi-res” Amazon Music works - since the app in Chromium based, it can’t do exclusive mode or send bitperfect audio to a USB DAC. So even though you may be paying for lossless/hi-res, 44.1kHz or 96kHz tracks just gett re-sampled to 48kHz in the software before passing to the DAC.
In order to actually get lossless audio to a USB DAC, Plex would have to create custom drivers for each operating system that bypass the native OS audio drivers.
You can do this in MacOS, Windows (via ASIO), and Android (via USB audio API), but it takes a lot of resources to do it.
You are better off sticking with Audirvana if you want hi-res output and then firing up Plexamp when you want to listen and don’t care that it’s downmixed.
I’ve asked about this before and unfortunately audiophiles like us are too few and far between for them to dedicate to resources to let us listen to hi-res audio in Plex
Taking this a step further, even though Plex and Plex amp support playing hi-res FLAC audio across the spectrum, you can’t actually listen to hi-res music using any of the Plex apps as a result of the lack of audio drivers.
Ultimately, there is no benefit to playing lossless (44.1kHz) or hi-res (88.1kHz, 96kHz, 192kHz, etc) FLAC in the Plex ecosystem without the driver support.
You might as well just listen to lossy tracks from iTunes at 256 AAC or Spotify at 320 Ogg Vorbis since they are output natively at 48kHz.
You can’t state that in this absoluteness.
First of all, lossless 44.1 kHz material is played losslessly. (Unless you take exception to the volume change to achieve loudness leveling. Which is done at 32bit float-precision, so any claim that this influences the sound quality in a negative way is highly doubtful.)
And there is a way to play high-res audio losslessly as well:
Jun 22, 2020 10:06:22.060 [0x00001684] INFO - BASS: Device opened and sample rate is 96000 (preferred was 44100), latency is 70ms (minimum buffer: 10ms).
...
Jun 22, 2020 10:06:23.897 [0x00001be0] INFO - BASS: Opened stream 826050 in 76 ms (paused: 1) with gain -3.0 dB, max 0 Kbps (picked flac, 2771 Kbps, actual 2771 Kbps/sec).
Jun 22, 2020 10:06:23.897 [0x00001be0] INFO - BASS: Created a gapless source stream for 826050 (channel: 0x80000015) with sample rate of 96000 and 2 channels (paused: 1)
This is on Windows. The audio interface has been set to be the standard device and the sample rate of 96kHz has been set as the default sample rate for it.
Which means all audio is upsampled to 96kHz during playback, unless it is already in 96kHz.
But that doesn’t matter much to me.
The upsampling process from the default 44.1kHz (CD) to 96 kHz produces artifacts so low, that they don’t matter to me. And I’m using a high-quality DAC and studio monitors.
Don’t forget that enabling bit-perfect playback would mean
- no sweet fades
- no loudness levelling
- short interruptions every time the sample rate of the interface has to be switched
2 Likes
I suppose you could argue that it’s not important because most people can’t hear an audible difference between 256 AAC and lossless FLAC. But for those that buy lossless music, it would be nice to hear it at 44.1, or a multiple thereof.
So what’s happening is Plex is outputting the audio at whatever the native bitrate is, and then Windows Direct Sound is downsampling it to 48kHz before sending it to the DAC. The only way around this is to utilize ASIO or Wasapi drivers to the DAC directly.
Which DAC are you using? If it has a screen it will show you what bitrate it is receiving. Ideally you want it to be bitperfect, but if not, then at least upsample to a multiple of the original track. In the case of redbook this would be 44.1 and you’d want to upsample to 88.2, not 96kHz.
Technically, all of these introduce distortion, which again defeats the purpose of lossless audio (with the exception of clicks on bitrate change which can be compensated for instead the DAC itself).
Overall, I suppose I would have hoped this community would have been more open to a new perspective than having moderators act dismissively condescending, or say that the opinion of an audiophile doesn’t matter.
I may be in the minority, but that doesn’t mean I am incorrect or the my opinion doesn’t count.
Plexamp does not convert to AAC.
In a local network, it doesn’t convert at all, unless it cannot understand the file format (e.g. DTS).
In a remote connection, it doesn’t convert if the bitrate setting for remote connections is high enough for the file. i.e. if you set i to ‘Maximum’, it will fetch the original flac or alac file from the server. No lossy compression applied.
No. Windows Direct Sound is using the default sample rate of the interface, which in my case is 96 kHz.
RME Fireface UC. Yes, I can monitor the sample rate. I can even verify the bit depth of the data which arrive at the interface.
While that is true in theory, I have compared the results and _up_sampling is much less prone to distortion. Particularly if the difference in sample rates is rather large. (On the other hand, converting 44.1 to 48 kHz or vice versa is definitely something you want to avoid, no argument about that)
I have set it to 96 kHz because I also use video and the default sample rate there is 48 kHz.
Plus, the majority of high-res audio is available in 96kHz, not 88.2 kHz, at least that is my experience so far.
1 Like
Right, I was being facetious by pointing out that Windows DS is so terrible, that if you can’t bypass it via ASIO, you might as well just listen to lossy 256 AAC 
Are you connected via USB or SPDIF? If SPDIF, try USB as that’s how most users will connect. Even still, that’s only Windows and each of: Windows, macOS, iOS, and Android output natively at 48kHz which will resample 44.1kHz to 48kHz.
Ideally you want to stick to upsampling in multiples, so 48 > 96 > 192.
Ultimately, all I’m saying is that if the goal is “best-in-class” audio (like Audirvana, Roon, Foobar2k, JRiver, etc) then allowing the user to listen in bitperfect CD quality lossless 16/44.1 should be the goal, no?
(*sidenote: most audiophiles turn off loudness normalization and then just try and stick to recordings that aren’t brick walled. The downside of loudness normalization is that it reduces the dynamics and perceived spacial instrument separation of the recording and most are willing to tolerate a +/-5 dB swing in SPL)
I am not using SPDIF. Which is also usually limited to 48 kHz.
My DAC is connected by USB. It can work at any sample rate I choose. I decided that I like its sound at 96kHz better than at 192kHz. So I picked 96kHz.
RME has a pretty good USB interface, but for the most part SPDIF will deliver cleaner sound without all that digital noise that can plague USB.
The only limitation of SPDIF is that it caps out at 24/192, which is as high as anyone really needs (with the exception of those DSD fans out there).
Getting back to the original point though, why not allow the user an option to listen to each song at it’s native bit/sample rate rather than forcing every track to resample?
Sorry, nothing is plaguing USB anymore. The technical issues with audio transport, which existed in USB v1.1 have been solved since years.
(We are talking the era of “Windows XP” here.)
Plus, as I said above, most implementations of SPDIF are capped at 48kHz/20 bit (sometimes even only 16 bits).
If you’re looking for a reason why audio quality is restricted to 48 kHz when using SPDIF, you should look at the specifications of the SPDIF ports of all involved components.
We’ll definitely look into better support for this over time. As you’ve sort of pointed out, fully supporting bit-perfect can be super hairy, but it’s my own personal hope that we can find a happy medium ground which makes the majority of people happy (happier?)
4 Likes
Totally agree that a happy medium can be reached. I’m not such a purist that I’m seeking ‘bit perfect’ playback for 100% of my tracks. I’m in the Apple ecosystem, and the best compromise that I’m looking for is playback using the best native OS bit depth and sampling rate available with an attached DAC (without custom Plexamp USB drivers).
For example, on my iPhone 11 Pro with a dongle the Fiio BTR5 DAC and the Onkyo HF player re-samples my 24-bit 96 KHz FLAC up to PCM 192 KHz. Plexamp 3.1.1 on my iPhone DOWNSAMPLES the same FLAC track to 44.1 KHz and I’d guess 16-bit, but my DAC doesn’t display bit depth.
And just like the Onkyo HF player, please display the the negotiated OUTPUT bit depth and sample rate so we have a visual indicator what’s going on.
I’m not asking for custom USB audio drivers or 100% bit perfect playback 100% of the time. I like cross fades, etc. Just best utilize the OS driver and USB DAC to not compromise on sound quality when possible.
Totally agree with that. Aim for optimal fidelity, but within reasonable constraints.
Any chance of having gapless playback in PMP in the future? I’m running it headless on an RPi3 with a Topping E30 and love having hi-res bit-perfect PCM and DoP. The only downside is the breaks between tracks.
Thanks.
If this app is to be taken seriously it needs to be able to do bit perfect on all major platforms. Android is a little bit complicated but it’s doable look at UAPP. There is no excuse for not having it on Mac or Linux. It seems to be available on windows .
I just got my SMSL M400 DAC setup with my Mac, and I’d SO love for PlexAmp to use it to it’s fullest extent (native sampling rate and bit depth, passing MQA, etc.). I’m testing out Audirvana and it sounds great, but as we know it has no Plex integration. The holy grail for me would be an ecosystem of PMS + and Hi-Res/MQA capable player (PlexAmp) but that doesn’t seem to exist in our universe. Exceptionally disappointing.
1 Like
Just one more note for PlexAmp wish list: allow system VST audio plug-ins. I currently use FabFilter Pro Q3 for headphone EQ and it’s amazing! Audirvana can use such plugins.
So until PlexAmp is more audiophile friendly, I have a workaround that seems to work OK. It doesn’t solve the MQA problem, but I’m not a big fan of MQA tracks anyway. But on the Mac you can get SoundSource 5 and directly route PlexAmp through to your USB DAC. You can also plug-in FX engines, such as Pro-Q 3 (which is a professional Parametric EQ engine). To my ears, this setup sounds the same as using Audirvana. Audirvana is great, but the lack of Plex integration sucks. Since my ‘hack’ for DAC use sounds the same to me, I’ll stick with this until PlexAmp gets DAC and FX engine support.
4 Likes
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.
