PDA

View Full Version : BitPerfect



steff
2017-10-28, 16:29
Having played with Rpi3 for month with pure MPD, now I am using pCP.

I would like to know what kind of setting I need to set in order to achieve BitPerfect playback.
Let me say, avoid resampling, avoid software volume control, etc.

In particular, my external USB DAC is capable of hardware volume control, but I am not sure how to use DAC volume control and not something "software" (decimation by Rpi3, etc.)

Thanks a lot.

SuperQ
2017-10-29, 00:23
Set the volume to 100, that's all you need.

But, remember, software volume control is good as long as your gain chain is ok. It's done in 24 bits, so any rounding errors are several orders of magnitude below the noise floor introduced in even the best amplifier chains.

steff
2017-10-29, 16:26
Set the volume to 100, that's all you need.

But, remember, software volume control is good as long as your gain chain is ok. It's done in 24 bits, so any rounding errors are several orders of magnitude below the noise floor introduced in even the best amplifier chains.

Agree, but my question is:
since my DAC is capable to control the volume and the documentation says that in case the player is capable of controlling the volume by delegate the DAC... then it is better.
Looking aroud, seems that letting the player modify the volume it is called "software volume control", while controlling the volume by tellig to do it to the DAC is called "hardware volume control".

My DAC come with a driver for Windows that install a little app in the Windows bar and so, for example with JRiver, I disable the software volume control and control the volume with this little app that communicate to the DAC itself.

Is there any way with pCP to do that? I mean disble the software volume control and enable the hardware one?

paul-
2017-10-29, 16:31
No, There are no specific USB drivers to do any controls on the DAC. Those would need to be made by the device manufacturer.

steff
2017-10-29, 17:17
No, There are no specific USB drivers to do any controls on the DAC. Those would need to be made by the device manufacturer.

So, what happens when I select "hardware mixer" for example in MPD instead of software mixer?

Maybe I am confused, sorry for that.

iPhone
2017-10-29, 17:25
Agree, but my question is:
since my DAC is capable to control the volume and the documentation says that in case the player is capable of controlling the volume by delegate the DAC... then it is better.
Looking aroud, seems that letting the player modify the volume it is called "software volume control", while controlling the volume by tellig to do it to the DAC is called "hardware volume control".

My DAC come with a driver for Windows that install a little app in the Windows bar and so, for example with JRiver, I disable the software volume control and control the volume with this little app that communicate to the DAC itself.

Is there any way with pCP to do that? I mean disble the software volume control and enable the hardware one?

Doesn't matter what one calls it, the only point that matters is "Where" the volume control is actually happening:

If it is being done in the Digital Domain, the stream is no longer Bit Perfect.
If it happens in the analog section, then the data has been decoded, the DAC has done its job, and the analog level is being adjusted which doesn't affect the Bit Stream because all that was already completed. Almost all DACs and Digital Output devices that can control Volume do so in the digital domain all at the cost of Bit Perfect. Off the top of my head, I have only seen one DAC that had volume control that didn't affect the bit stream because it was actually a DAC/Pre-amp so the volume control on the remote controlled the analog pre-amp's various inputs as the output of the device went directly to the amplifier.

steff
2017-10-30, 17:09
Doesn't matter what one calls it, the only point that matters is "Where" the volume control is actually happening:

If it is being done in the Digital Domain, the stream is no longer Bit Perfect.


Got it!
Clear explanation.

So both pCP and my little DAC (M2Tech) control volume in digital domain and so, if volume is less than 100%, no more bitperfect.

Maybe several ways are available to control volume in digital domain, with different quality. Need to understand if it is better to let pCP to control volume or the DAC.

For sure the other DAC I have ( Teac ) manage volume in analog domain, after the signal has been decoded.

Greg Erskine
2017-10-31, 05:56
hi steff,

Generally, I think LMS, rather than end device, handles the volume.

But, to confuse things a little more, the ALSA volume settings on pCP needs to be looked at. It is usually set to 100% (+4dB) or 0dB?

Next, some audio cards add +6dB by default. Some drivers allow adjustment.

regards
Greg

steff
2017-10-31, 16:41
When I was student at the university, I designed and build a digital volume control by using a 8 bit DAC.
It was a cheap Burr Brown chip and the idea was: since in the DAC there were a sort of resistors array properly scaled and powered by an external reference voltage, I replaced this one with the audio line signal. Of course the DAC was not clocked.

Today one of my DAC ( Teac ) seems to use a very similar idea: the rotary volume control has 256 steps and the encoder controls the gain of the output drivers.

I am not sure if decimating a digital signal is worst than using analog solutions, but a digital control can be designed even with an analogue approach.

Some months later, during the final exam of digital electronics, we had to design and program all the above with a Xilinx, but we are talking about 1993, more or less

Unfortunately I am no more into this stuff... so sorry having to deal with budget and economics :-(

Mnyb
2017-11-01, 11:16
You can dig out some old treads about this too . But here it goes...

Back when I thought this was very important Iíve used some test to indirectly verify that the transfer really was bitperfect .

There are signals thatís usually are packed/disguised in regular pcm namely hdcd & DTS or ac3 ( Dolby digital ) thatís how for example a DVD player sends ac3 over spdiff or toslink to a ht-reciver .
The thing is this must be bitperfect or the encoding gets lost and you get gibberish.

So a great test is to get some ac3 encoded wav fileís, dial in stuff to what you think are zero gain, 100% bitperfect and then conect the digital out to a home theatre processor or reciever and listen you get 5.1 sound or noise .

On one of the side topics here , iím More or less convinced that a good well dithered digital volume control is actually performs better than any analog implementation a good digital systems noise if of the same character as good old analog noise , s/n ratio is the same if achieved by analog or digital means ( forget al those staircase graphs in sterephile thatís not itís actually working ). Itís also an issue of implementation of the rest of the system .
The whole gain structure must be reasonable so that full scale signal is just as loud as you ever want to listen . Then quieter volume are well quieter :) the volume is so low that any possible degradation is inaudible anyway .

And then you have your 24 bit dithered digila system with an s/n ratio of say 144dB and then you play some old 70ís rock album 70dB s/n ratio thatís less than 13 bits .... there are ofcourse better records , but what I mean that in practice the system noise produced by a digital volume adjustment is far below the noise in your actual recordings you listen to , even if you bougth 24 bit or DSD album , what most of the data actually is are random noise from the recording .

steff
2017-11-02, 14:48
Talking about "bitperfect", I would like to ask a question to you.

If we take different devices labeled as "bitperfect", I suppose that there should not be any difference among them.
I mean... no difference in the bitstream delivered...
is this sentence correct?

Since I am playing around this matters since years, some days ago i got an OTG USB cable and connected my DACs to a Sony Experia Z1 Android smartphone where I installed "USB Audio Player PRO".
Everything is played from a NAS and it works well even with my 352 kHz \ 24 bit FLAC with no problems.

The question is: what could be the difference between an Android smartphone running USB Audio Player Pro and pCP and a PC running Jriver or any other high-end multi k$$$ streamer?
If bitperfect is bitperfect... there shuld be no difference at all (speaking about bitstream and not other features), or not?

thanks.

Mnyb
2017-11-02, 21:17
Talking about "bitperfect", I would like to ask a question to you.

If we take different devices labeled as "bitperfect", I suppose that there should not be any difference among them.
I mean... no difference in the bitstream delivered...
is this sentence correct?

Since I am playing around this matters since years, some days ago i got an OTG USB cable and connected my DACs to a Sony Experia Z1 Android smartphone where I installed "USB Audio Player PRO".
Everything is played from a NAS and it works well even with my 352 kHz \ 24 bit FLAC with no problems.

The question is: what could be the difference between an Android smartphone running USB Audio Player Pro and pCP and a PC running Jriver or any other high-end multi k$$$ streamer?
If bitperfect is bitperfect... there shuld be no difference at all (speaking about bitstream and not other features), or not?

thanks.

Correct they will deliver the exact same bit stream :) they will also sound the same unless one of the sources is seriusly broken or the DAC is very shoddily built . Now devout audiophiles starts to talk about ďjitterĒ etc as a factor .....

Jriver is decent software and does not claim any magic features , but there are other fake software like amarra and audio optimiser , that claims to optimise your pc for audio playback , beware

steff
2017-11-03, 01:23
Regarding jitter, since almost all DACs today are internally reclocked, it should not be a matter of the source\streamer.

Correct?

Mnyb
2017-11-03, 03:55
Regarding jitter, since almost all DACs today are internally reclocked, it should not be a matter of the source\streamer.

Correct?

Not really , and the levels of jitter claimed in most data for almost anything is still 1000' of timesbelow whats considered audible ..
So even if source jitter do affect even reclocked in a very minute ways , you can measure I think , think about it as a fillter , but still it does not get audible even if can show the effect .

Thats totally eliminated by for example asynchronus USB or ethernet direct to the streamer (also async comunication ) .

Then you have electrical noise etc thats alos claimed to do "harm" .

I'm no longer a believer in this myself so i'm in agrrement :) really ( i has been a devout believer in all kind of audiphool nonsense ) .

But its really down to cluthing at straws arguments to justife an imaged difference "heard" trough uncontrolled non blind testing aka your are not even wrong as no real information cuts above the noise :)