HydrogenAudio

CD-R and Audio Hardware => Audio Hardware => Topic started by: db579 on 2012-07-31 00:21:49

Title: If the output is digital does the player make any difference to the so
Post by: db579 on 2012-07-31 00:21:49
I'm wondering whether if you had two media playing pieces of hardware outputting digital music would the quality of the device make any difference? For example say you had a cheap-to-mid range android phone or mp3 player playing a flac file and outputting over HDMI to a receiver, and a higher spec digital media player like say the Olive 06HD playing the same flac file output over HDMI to a receiver would you hear/should there theoretically even be any difference in audio quality?

Thanks
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-07-31 01:05:24
I'd say that if you stay in the digital domain and don't do any digital signal processing and jitter doesn't matter then the quality of the device doesn't matter either.
Title: If the output is digital does the player make any difference to the so
Post by: db579 on 2012-07-31 01:13:16
Wouldn't jitter occur in the DAC process if at all rather than the output? I assume I'm wrong but why? Would this affect the audible sound?
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-07-31 01:18:34
I think with HDMI it's entirely in the 'hands' of the receiver.
Title: If the output is digital does the player make any difference to the so
Post by: db579 on 2012-07-31 01:50:28
Same thing for digital optical?
Title: If the output is digital does the player make any difference to the so
Post by: itisljar on 2012-07-31 07:33:53
I think it's the same with all modern devices and all modern digital connections, coax, optical, hdmi. Receiving device does buffering and re-clocking the incoming signal, so jitter is practically non-existent.
Title: If the output is digital does the player make any difference to the so
Post by: probedb on 2012-07-31 08:15:16
Plus from reading around here, if you can actually hear jitter something is very seriously wrong with the device.
Title: If the output is digital does the player make any difference to the so
Post by: db579 on 2012-07-31 13:56:06
So with that in mind if you wanted portable audiophile sound quality, all you'd need is any flac or wav player (quality not particularly important) with a digital out, good headphones and a good portable DAC? Does such a DAC exist? How do they compare to AV receivers?
Title: If the output is digital does the player make any difference to the so
Post by: Nessuno on 2012-07-31 14:09:58
if you wanted portable audiophile sound quality, all you'd need is any flac or wav player

No need  for lossless codec, actually, if SQ is the target. High bitrate (i.e. transparent) lossy will suffice (and this could sometimes require lower hardware specs for the player).
Title: If the output is digital does the player make any difference to the so
Post by: Kohlrabi on 2012-07-31 14:14:38
So with that in mind if you wanted portable audiophile sound quality, all you'd need is any flac or wav player (quality not particularly important) with a digital out, good headphones and a good portable DAC? Does such a DAC exist? How do they compare to AV receivers?
"Audiophile" audio quality is more a question of skewed perception than hardware. I found most of the things you mention pretty irrelevant for portable audio. Well encoded lossy audio on hardware like the Sansa Clip+ with good headphones is good enough from my experience. Why would you need digital out in a portable device?
Title: If the output is digital does the player make any difference to the so
Post by: skamp on 2012-07-31 14:21:06
So with that in mind if you wanted portable audiophile sound quality


Just get a Sansa Clip+ or an iPod (and Rockbox it).
Title: If the output is digital does the player make any difference to the so
Post by: greynol on 2012-07-31 15:30:30
a good portable DAC? Does such a DAC exist?
Yes, and it can be found in the portable media player.

Quote
How do they compare to AV receivers?
Short of any objective listening tests proving otherwise, they are every bit as good.

I don't see how adding rockbox to an iPod changes its intrinsic sound quality. While someone may be inclined to talk about the snazzy eq or some other feature, the topic is about hardware; and since the topic is about hardware, it will be moved to the appropriate forum.
Title: If the output is digital does the player make any difference to the so
Post by: db579 on 2012-07-31 15:36:26
I'm asking more out of theoretical interest than anything else. I'm just wondering whether it is possible to get the same sound quality out of an entirely portable system (assuming you are sat in the same conditions obviously) as a top end separates system (I'm talking about exact sound output as well as just 'what you can hear').

Since I am assuming that no portable player in production at the moment has a DAC built in that is as good as top of the range separates DACs (happy to be corrected if this is wrong?) I was wondering whether something like this exists. Sort of like a top end separates system but portable? (I understand why in 99.9% of cases this would be pointless and you wouldn't hear a difference I'm just curious)

Thanks guys

edit. Apologies if this was the wrong forum
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-07-31 15:39:50
I think that portable players can easily sound better than the headphone out on your receiver. Many receivers have resistors in series with the headphone jack raising the output impedance up to several hundred ohms. This will cause audible frequency response deviations up to several dB.
The best DAC won't matter if the output impedance causes, lets say, a narrow +5 dB peak around 100 Hz.
Title: If the output is digital does the player make any difference to the so
Post by: greynol on 2012-07-31 15:43:27
The answer is yes, it is possible to get the same quality out of a portable player.  To determine whether any particular device achieves the same quality as any particular non-portable device will require an objective listening test. How else can someone assess how something actually sounds (as opposed to how it measures)?
Title: If the output is digital does the player make any difference to the so
Post by: db579 on 2012-07-31 15:48:24
Interesting responses and not what I expected, thanks. Has anyone done any such tests here/does anyone have a measured data on this?
Title: If the output is digital does the player make any difference to the so
Post by: skamp on 2012-07-31 16:46:22
The O2/ODAC (http://nwavguy.blogspot.fr/2012/04/odac-released.html) is a portable DAC + headphone amp that rivals top of the line components in measurements, and is claimed by its designer to be audibly transparent in double blind tests. According to fellow HA member DanBa, you can use it with a Samsung Galaxy S III (or a laptop or any other portable source that works with regular USB Audio devices). I guess that meets you criteria of portability and top notch quality.

Whether you could tell the difference between such a rig and a simple Clip+ or iPod or other portable players that measure less but still pretty well, only you can answer that.

Disclaimer: I love both my O2/ODAC and my iPod Classic. I don't know that I could tell the difference between the two, but then again I use them in different situations and with different headphones.
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-07-31 16:54:42
To determine whether any particular device achieves the same quality as any particular non-portable device will require an objective listening test. How else can someone assess how something actually sounds (as opposed to how it measures)?

High output impedance causing a bass 'boost' will sound bassier like the measurements show.    If the differences were fractions of a dB then I'd also say a listening test is required to see whether there's an audible difference or not, but not if the difference are as big as several dBs.

db579: Here (http://www.hifi-forum.de/index.php?action=browseT&forum_id=211&thread=338) are some portable players tested with sensitive, low-impedance (therefore even more susceptible to high output impedance) IEMs. It's german but you should be able to read the graphs nonetheless. Even some portable players show a relatively high output impedance compared to 16 or 32 ohm headphones.
Title: If the output is digital does the player make any difference to the so
Post by: db579 on 2012-07-31 18:01:36
Thanks guys, really appreciate the detailed and helpful responses on here! Fortunately I speak german so will have a proper read through this!

Skamp - Am I right in thinking the O2+ODAC requires a power socket so can't really be used on the move?

xnor - He seems to suggest that the response of the DACs he tests is better when given an optical digital input than when given a USB digital input. This thread had given me the impression that shouldn't be the case?
Title: If the output is digital does the player make any difference to the so
Post by: DVDdoug on 2012-07-31 19:12:37
Quote
xnor - He seems to suggest that the response of the DACs he tests is better when given an optical digital input than when given a USB digital input. This thread had given me the impression that shouldn't be the case?
It shouldn't make a difference unless different digital data is coming-in the different inputs, or unless "something else" is different.    If the "numbers" are the same, it shouldn't matter how the numbers get into the DAC. 

Digital audio data is numbers...  This message is coming to you as digital data...  It doesn't matter if the data is sent over wires, or fiber-optics, or the distance the message travels, or whatever...  As long as the data doesn't get corrupted somehow, you are reading exactly the letters I typed.    If there are any "typos", it's probably not a problem with the digital transmission...      Digital data storage & transmission is super-reliable.  If we can send data around the world without corruption, you can safely send data across the room to your DAC...

Yes, you might run into a case where the driver is sending different data (such as a different sample rate) over the different digital interfaces.   
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-07-31 20:07:10
It doesn't matter if the data is sent over wires, or fiber-optics, or the distance the message travels, or whatever...  As long as the data doesn't get corrupted somehow, you are reading exactly the letters I typed.    If there are any "typos", it's probably not a problem with the digital transmission...      Digital data storage & transmission is super-reliable.  If we can send data around the world without corruption, you can safely send data across the room to your DAC...


But it does matter. Very few USB devices are isolated whereas S/PDIF typically is. Actually, many devices are also powered by the USB port so that noise in the computer's PSU also affects the device. Yeah, if 'properly implemented' such devices should reject that noise and you shouldn't hear any difference.

Another problem with USB DACs is DPC latency. I've heard many people report problems with 'streaming' audio over USB, like random glitches due to latency spikes. Again, if the drivers for you computer's hardware are 'properly implemented' you shouldn't have any problems.

The default transfer mode for streaming audio over USB is isochronous, which means that for example every millisecond your computer has to send audio data. Data loss is possible. One DPC latency spike and you might experience a glitch because the data won't arrive in time.

I don't know if the same is true for HDMI. Old receivers had horrible jitter problems but I guess this problem was fixed one or two years ago.
Title: If the output is digital does the player make any difference to the so
Post by: greynol on 2012-07-31 20:08:12
Identical numbers are fine and dandy, but they don't account for the possibility that the connection might introduce noise such as (but not limited to) a ground loop.
Title: If the output is digital does the player make any difference to the so
Post by: mzil on 2012-07-31 20:41:54
...Digital audio data is numbers...  This message is coming to you as digital data...  It doesn't matter if the data is sent over wires, or fiber-optics, or the distance the message travels, or whatever...

But what does matter is if you are sending the original, raw digital file vs. some manipulated version of it.

When using a USB cord I can perfectly replicate a digital file from one location to another in just seconds, true, however when sending a signal over HDMI, from say my DVD or Blu-ray player (or Toslink for audio, I suspect) I'm not sending the original raw data file found on the optical disc (or storage drive) but rather a streaming version of that file in "real time" (a one minute song takes one minute to transfer), and the conversion from the original, raw file to the streaming version may vary device to device, so to answer the OP's question, I think the technically correct answer is:

"Yes, the original source device, when playing the exact same digital file, may have a slightly different output from another device streaming a digital version of the same song over HDMI."

I make no claims there would be an audible difference in real world use, however in reading reviews of DVD players and Blu-ray players using HDMI out, there seems to be machine measurable differences in, for example, (video) dynamic range. This tells me that HDMI outs, at least when it comes to video playback devices, aren't "all the same and bit for bit replications of the stored file".
Title: If the output is digital does the player make any difference to the so
Post by: db579 on 2012-07-31 21:14:07
Very few USB devices are isolated whereas S/PDIF typically is. Actually, many devices are also powered by the USB port so that noise in the computer's PSU also affects the device.


So I think I've read some new android phones, the galaxy nexus for example, can output audio over mhl via the micro usb port at the same time as being charged by it. Any way of knowing how well isolated this might be or how good the sound quality would be short of buying them all and trying it?
Title: If the output is digital does the player make any difference to the so
Post by: yourlord on 2012-07-31 22:33:17
All things being equal, the delivered data should be identical regardless of whether it was sent over a USB cable, fiber, or via radio waves in the atmosphere so long as it's sent via a reliable link or protocol.

If you are powering a DAC/amp using the USB port, then other factors can come into play such as noise on the power from the USB port. But, that's not an issue of the data being delivered incorrectly. The noise on the power rail could affect the analog output stages past the DAC, but it won't do anything to the data being fed to the DAC..

2 separate devices, one feeding a known data set at a sufficient bit-rate for real-time presentation to the other, should perform according to the specs of the receiving device.. Always.

There is no "streaming version of a file". There is only data. It can be packetized and delivered as needed, but it's still the same data at transmitter and receiver. Any data differences seen in video or audio delivered via HDMI would have to come down to differences in the decoders on the source devices since HDMI normally carries uncompressed audio and video. What is delivered to the receiver is exactly what the decoder produced at the source. 

When you "stream" audio from an internet radio station they are simply encoding a pcm data stream and then sending you the resulting data as it becomes available. The delivered encoded data will always be identical to the source if delivered over a reliable link (TCP). Different decoders may produce different results from the same encoded data due to differences/bugs in implementations, but that stream delivered to a million devices that all use the exact same decoder will all get the exact same resulting decompressed data.

Obviously once the digital data has been transformed to analog, all bets are off and all kinds of things can then happen to the data. I'm just trying to drive home that data delivered via a reliable link will not contain errors. Encoders can screw up, decoders can screw up, but the data transmitted across a $3 HDMI cable, $5000 HDMI cable, or a $2 USB cable will all be bit identical.
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-07-31 23:21:33
[...] the delivered data should be identical [...] so long as it's sent via a reliable link or protocol.

For most DACs with an USB receiver this premise is not true.

Unlike TCP, there is no retry if delivery of data failed, no guarantee of delivery. In fact, isochronous transfer is unidirectional.

(Btw, when talking about USB audio streaming I don't mean listening to an internet radio station..)
Title: If the output is digital does the player make any difference to the so
Post by: mzil on 2012-08-01 00:16:40
The signal the HDMI output sends gets transfered pristinely and without degradation, however that signal is not a bit for bit identical copy of the file on the disc/hard drive, it is a converted version of that file, which to the best of my knowledge can not be reassembled on the receiving end to make an exact copy of the file from the original source. [Even if there weren't copy protection schemes in place to prevent it and pretending for the moment that there were stand alone recorders with HDMI in]. HDMI streaming is not the same as a USB file transfer.

I'm not arguing that the difference from device to device is audible, however. I'm just saying it's not a bit for bit accurate copy of the original file and may vary from device to device, which was the OP's question.

"Since all of the content will be transferred digitally to your display, this should result in perfect picture quality with no issues since "bits are bits", right?  Unfortunately the reality is far more complex than this, and there are many factors that can come into play."

That quote is from this magazine (http://www.hometheaterhifi.com/secrets-blu-ray-player-hdmi-benchmark/secrets-blu-ray-player-hdmi-benchmark/blu-ray-player-hdmi-benchmark-introduction.html) and the article shows how there are many changes to the HDMI signal outputs between, for example, an OPPO and a Sony machine (that fails in some regards compared to the OPPO), using a very pricey Quantum Data HDMI Analyzer, however they unfortunately dont address the audio over the HDMI.

I guess everyone here (but me) thinks the audio is simply an unmodified S/PDIF signal and interchangeable between devices, however I'm not so sure. Machines modify their HDMI out video signals according to that magazine, so how can we be so sure they don't also manipulate their HDMI out audio signals?
Title: If the output is digital does the player make any difference to the so
Post by: yourlord on 2012-08-01 03:33:06
The signal the HDMI output sends gets transfered pristinely and without degradation, however that signal is not a bit for bit identical copy of the file on the disc/hard drive, it is a converted version of that file, which to the best of my knowledge can not be reassembled on the receiving end to make an exact copy of the file from the original source. [Even if there weren't copy protection schemes in place to prevent it and pretending for the moment that there were stand alone recorders with HDMI in]. HDMI streaming is not the same as a USB file transfer.


You are correct. The file on the disc is compressed using a lossy codec. I'm not arguing that the data transmitted has any resemblance to the file on the disc. I'm saying the data transmitted, which is the decompressed output from the decoder in the player, will arrive perfectly intact at the destination.

If the decoder in the player has a bug or other issue in it's implementation then that decoded output may differ from another player, but either way HDMI will faithfully deliver the decoder output to the receiver.
Title: If the output is digital does the player make any difference to the so
Post by: andy o on 2012-08-01 15:46:29
I don't know if the same is true for HDMI. Old receivers had horrible jitter problems but I guess this problem was fixed one or two years ago.

Could you elaborate on this, please?
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-01 16:37:44
For example the Yamaha RX-V3900 was measured to have 7660 ps of jitter.
Title: If the output is digital does the player make any difference to the so
Post by: mzil on 2012-08-01 17:07:11
^Do you have any evidence that level is audible to humans under controlled conditions?

All I know is this (http://amorgignitamorem.nl/Audio/Jitter/Detection%20threshold%20for%20distortions%20due%20to%20jitter%20on%20digital%20audio%2026_50.pdf) study which suggests < 250 ns (250,000 ps) is safe. The Yamaha easily qualifies.
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-01 17:17:13
Nope and that's why I didn't make that claim, but such amounts of jitter have a big measurable negative effect on SNR. I can imagine that it could be audible with highly dynamic recordings if you turn up the volume quite a bit.
Title: If the output is digital does the player make any difference to the so
Post by: mzil on 2012-08-01 17:31:22
If you are allowed to turn up the volume and listen to the noise floor, almost everything sounds different! Suddenly comparing a device with a noise floor 100 dB down becomes discernable from one that's 102 dB, with the right material and fast A/B switching, but in real world use it is completely meaningless and undetectable. [Not hard to believe since all of us would have no problem detecting a 2 dB change in audible hiss in real world conditions.]

"Poor" low level linearity down at those levels [when greatly amplified] also becomes audible, but similarly, in real world use, is completely meaningless and undetectable.
Title: If the output is digital does the player make any difference to the so
Post by: greynol on 2012-08-01 17:59:00
Do DACs matter?

Perhaps, but if you can't demonstrate audible differences through double-blind testing then for practical purposes they don't, at least not as far as this forum is concerned.
Title: If the output is digital does the player make any difference to the so
Post by: Roseval on 2012-08-01 18:13:40
Unlike TCP, there is no retry if delivery of data failed, no guarantee of delivery. In fact, isochronous transfer is unidirectional.


The first part is correct.
As an isochronous stream is a soft real time stream, a retry will always be too late.
However USB in isochronous mode does reserve the appropriate bandwidth.
So delivery is guaranteed downstream but delivering of the right bits is not

The second part is not correct.
One can combine isochronous (that’s the way USB audio works) with 3 different type of synchronization.
In case of synchronous or adaptive synchronization, the protocol is used unidirectional (the DAC adapts its clock speed to the incoming data stream).
In case of asynchronous synchronization a feedback loop is established controlling the amount of data send by the PC.
This allows the DAC to run with a fixed clock.
In this case the protocol is used bi-directional.

http://thewelltemperedcomputer.com/KB/USB.html (http://thewelltemperedcomputer.com/KB/USB.html)
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-01 18:34:43
As an isochronous stream is a soft real time stream, a retry will always be too late.

It probably would be too late, but typically there are no retries (see adaptive below).

Quote
However USB in isochronous mode does reserve the appropriate bandwidth.
So delivery is guaranteed downstream but delivering of the right bits is not

If frames need to arrive every ms but do not it doesn't matter how much bandwidth is reserved. Also, most glitches occur because data doesn't arrive in time (DPC latency on Windows), not because of flipped bits.

The second part is not correct.
Quote
One can combine isochronous (that’s the way USB audio works) with 3 different type of synchronization.

Theoretically, yes, but most USB audio devices function in "isochronous adaptive" mode.

Devices using "isochronous transfer with asynchronous synchronization" are typically simply called 'asynchronous'. I guess I should have been more pedantic with the terms I used before. Noted.
Title: If the output is digital does the player make any difference to the so
Post by: greynol on 2012-08-01 18:40:54
I guess I should have been more pedantic with the terms I used before.

...as should I have when talking about double-blind testing trumping measurements as a means to demonstrate audible differences.  Noted.
Title: If the output is digital does the player make any difference to the so
Post by: Roseval on 2012-08-01 18:50:20
Quote
If frames need to arrive every ms but do not it doesn't matter how much bandwidth is reserved. Also, most glitches occur because data doesn't arrive in time (DPC latency on Windows), not because of flipped bits.


It works slightly different.
The frames always arrive in time, that’s the way USB works.
The bandwidth is reserved on the hub to avoid interference of other USB devices.
But this is downstream not upstream.
The frames will always arrive in time but you might get glitches due to an insufficient amount of data in the frame e.g. due to DPC latency upstream

BTW: a lot of new USB DACs support async: http://thewelltemperedcomputer.com/HW/USB_DAC_Async.htm (http://thewelltemperedcomputer.com/HW/USB_DAC_Async.htm)
Title: If the output is digital does the player make any difference to the so
Post by: pdq on 2012-08-01 19:51:23
Nope and that's why I didn't make that claim, but such amounts of jitter have a big measurable negative effect on SNR. I can imagine that it could be audible with highly dynamic recordings if you turn up the volume quite a bit.

On the contrary, jitter noise is proportional to the amplitude of the signal, so in very quiet passages the jitter noise also becomes much smaller. Jitter has no effect whatever on silence.
Title: If the output is digital does the player make any difference to the so
Post by: andy o on 2012-08-01 20:34:13
For example the Yamaha RX-V3900 was measured to have 7660 ps of jitter.

One problem with one receiver doesn't jibe with this statement though: "Old receivers had horrible jitter problems but I guess this problem was fixed one or two years ago".

(If that was even a problem, as others are saying.)
Title: If the output is digital does the player make any difference to the so
Post by: greynol on 2012-08-01 20:38:13
How about the word horrible?

EDIT: Just saw your edit:
(If that was even a problem, as others are saying.)
Title: If the output is digital does the player make any difference to the so
Post by: Arnold B. Krueger on 2012-08-02 18:35:17
So with that in mind if you wanted portable audiophile sound quality, all you'd need is any flac or wav player (quality not particularly important) with a digital out, good headphones and a good portable DAC? Does such a DAC exist? How do they compare to AV receivers?


I often use a Behringer UCA 202 ($30) with older laptops that have problematical audio interfaces. Its headphone amp leaves something to be desired with some headphones. The FIIO E7 seems to have a better headphone amp at a far higher cost, particularly in the area of providing a consistent low source impedance.

In general good portable equipment such as the Sansa Clip/Fuze perform about as well as a good AVR, other than power output.  The headphone jack on a Clip/Fuze may outperform some AVR headphone jacks, particularly in the area of providing a consistent low source impedance.
Title: If the output is digital does the player make any difference to the so
Post by: Arnold B. Krueger on 2012-08-02 18:41:25
Quote
If frames need to arrive every ms but do not it doesn't matter how much bandwidth is reserved. Also, most glitches occur because data doesn't arrive in time (DPC latency on Windows), not because of flipped bits.


It works slightly different.
The frames always arrive in time, that’s the way USB works.
The bandwidth is reserved on the hub to avoid interference of other USB devices.
But this is downstream not upstream.
The frames will always arrive in time but you might get glitches due to an insufficient amount of data in the frame e.g. due to DPC latency upstream

BTW: a lot of new USB DACs support async: http://thewelltemperedcomputer.com/HW/USB_DAC_Async.htm (http://thewelltemperedcomputer.com/HW/USB_DAC_Async.htm)



As far as I can tell, the Asynch DAC issue is the latest in an long history of efforts by audio's high end to justify higher prices for what has largely become a commodity item.

I'd like to see someone make a Behringer UCA 202 (limited to line level loads) or Fiio E7 flunk an ABX test comparing it to some highly-reviewed (technically) asynch DAC.
Title: If the output is digital does the player make any difference to the so
Post by: Arnold B. Krueger on 2012-08-02 18:50:04
Nope and that's why I didn't make that claim, but such amounts of jitter have a big measurable negative effect on SNR. I can imagine that it could be audible with highly dynamic recordings if you turn up the volume quite a bit.


I don't know about SNR, but if it is really humongous, jitter can have an bad effect on THD+N. Not because it creates THD, but because the sidebands that jitter does create might be confused by the measuring technique with noise or harmonics.

As correctly stated by others, jitter has no effect on the noise floor again unless it becomes really huge and creates clicks and pops.

What's interesting is that as far as I can see,  none of the technical data shown at the well-tempered computer page linked out of this thread actually shows jitter.

BTW I have heard a lot of jitter in my day, most intentionally created to mimic what I see in test reports, but at levels high enough to actually hear. Jitter seems to have a characteristic, generalizible sound. In general the subjective comments by high end reviewers don't seem to describe what I hear.

Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-02 18:50:33
As I said, isochronous (adaptive) transfer can work perfectly fine.
Title: If the output is digital does the player make any difference to the so
Post by: Arnold B. Krueger on 2012-08-02 18:55:44
As I said, isochronous (adaptive) transfer can work perfectly fine.


I would like to see a real-world reliable test that shows it working any way but fine.

All we seem to have is these anecdotes.
Title: If the output is digital does the player make any difference to the so
Post by: greynol on 2012-08-02 18:58:21
...and poor subject/verb agreement, lol.
Title: If the output is digital does the player make any difference to the so
Post by: Roseval on 2012-08-02 19:05:26
A test by Jim Lesurf: http://www.audiomisc.co.uk/Linux/Sound3/TimeForChange.html (http://www.audiomisc.co.uk/Linux/Sound3/TimeForChange.html)
Title: If the output is digital does the player make any difference to the so
Post by: Arnold B. Krueger on 2012-08-02 19:26:16
A test by Jim Lesurf: http://www.audiomisc.co.uk/Linux/Sound3/TimeForChange.html (http://www.audiomisc.co.uk/Linux/Sound3/TimeForChange.html)


* no reliable listening test

* What do the spurious responses he measured mean given that they are given in parts per BILLION?

Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-02 21:21:40
All we seem to have is these anecdotes.


So you think a lot of people, including me (on my old notebook anyway), imagine audio glitches like VERY ANNOYING clicks or drop-outs with USB DACs? Screenshots of DPC latency measurements that clearly show that the hardware isn't able to stream audio glitch-free are anecdotes too?

Just because you haven't seen it yourself doesn't mean it's not real.
Title: If the output is digital does the player make any difference to the so
Post by: saratoga on 2012-08-02 21:28:32
All we seem to have is these anecdotes.


So you think a lot of people, including me (on my old notebook anyway), imagine audio glitches like VERY ANNOYING clicks or drop-outs with USB DACs?


That does not follow.  I don't think anyone here doubts that audio drop outs can happen with buggy hardware.  I've certainly seen data corruption from USB drivers for instance.  Doesn't mean I think the MSC spec is broken.

Thats the problem with anecdotal reports about glitches.  You don't know what caused it.  Driver bug?  Chipset errata?  Problem with the USB spec?  What conclusion can you safely draw when you don't know what happened?
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-02 21:38:56
What conclusion can you safely draw when you don't know what happened?


If a user that reported the problem reduces the DPC latency and the problems disappear then I draw the conclusion that most likely a too high DPC latency was the problem. Unfortunately, not all users can fix their hardware configuration / drivers so they end up returning the device and replacing it with an S/PDIF or Firewire or PCIe or async. USB interface which seem to make less problems. HDMI, as I said, I don't know.

On my old laptop, for example, disabling WLAN reduces the DPC latency considerably and gets rid of nasty spikes that caused drop-outs. Still not convinced?
Title: If the output is digital does the player make any difference to the so
Post by: saratoga on 2012-08-02 21:44:14
On my old laptop, for example, disabling WLAN reduces the DPC latency considerably and gets rid of nasty spikes that caused drop-outs. Still not convinced?


I'm absolutely convinced that you have broken hardware.  Beyond that, what else am I supposed to take away from the anecdote?
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-02 21:57:30
I'm absolutely convinced that you have broken hardware. Beyond that, what else am I supposed to take away from the anecdote?

Well that's just (being) stupid but I'm okay with that.
Title: If the output is digital does the player make any difference to the so
Post by: [JAZ] on 2012-08-02 22:13:13
@ xnor: That example does not apply.

I do have high DPC latencies sometimes in my laptop with Windows 7 and integrated realtek soundcard. And obviously, those spikes do cause clicks and glitches to the audio.

So talking about DPC latencies to say that USB audio fails is just attributing a problem A to an unrelated subject B.

In this regard, and if we consider "Laptop with windows Vista/7/8" a "player", then, ok, we can say that the player matters to get glitch-free bit-perfect  digital audio playback. Otherwise, i'm with the usual thinking that digital is digital.
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-02 22:35:09
I do have high DPC latencies sometimes in my laptop with Windows 7 and integrated realtek soundcard. And obviously, those spikes do cause clicks and glitches to the audio.

So talking about DPC latencies to say that USB audio fails is just attributing a problem A to an unrelated subject B.


My laptop also has onboard sound and it works perfectly fine without glitches even though the DPC latencies go consistently crazy with WLAN enabled. It's just the USB audio devices (have not tried async ones) I've had problems with, but only with high latencies. This is in no way unrelated.

What's unrelated is saying the hardware is broken when it's clearly a software (driver) problem.
Title: If the output is digital does the player make any difference to the so
Post by: Arnold B. Krueger on 2012-08-03 13:15:46
What conclusion can you safely draw when you don't know what happened?


If a user that reported the problem reduces the DPC latency and the problems disappear then I draw the conclusion that most likely a too high DPC latency was the problem. Unfortunately, not all users can fix their hardware configuration / drivers so they end up returning the device and replacing it with an S/PDIF or Firewire or PCIe or async. USB interface which seem to make less problems. HDMI, as I said, I don't know.

On my old laptop, for example, disabling WLAN reduces the DPC latency considerably and gets rid of nasty spikes that caused drop-outs. Still not convinced?


My old laptop and my new laptop run audio over USB just swell using a very basic interface - the Behringer UCA 202 with no parameter changes from defaults, and both or either WLAN or wired LAN active and in use.

There was a lot of discussion of asynch USB over at AVS a while back, and subtle SQ and jitter were the only issues raised. No advocate of Aysnch USB was able to proffer a DBT or any measurement that supported their claims.  In short, each and every problem was perceived and subtle which is to say, probably imaginary. I'm not saying that Asynch can't be a problem solver, I'm saying that it is an effective solution that is in search of a real-world problem that is actually causing someone audible problems. OK, it finds a few such people. Next!

I think that we all know that measuring a problem or nailing it in a DBT are both slam dunks when the kind of problems you describe are active. Since this isn't happening, the basic problem must be pretty rare.

What I take away is that the problems you mention are rare, and when they arise they are due to pre-existing technical difficulties that are both far more global within the laptop, and/or  also very much peculiar to the laptop sample where they are observed. Most people are interested in Asynch USB because they are led into it by people with commercial interests ar a desire to be opinion leaders, not because of any actual audible problems that they are experiencing.  In many cases you can buy a whole new really pretty good laptop (or several!)  for the price premium of the Asynch USB interface.

Also, it appears that adding Asynch mode to a USB audio interface may be just a matter of changing the firmware. It shouldn't be increasing the price of the equipment by 1,000% or 10,000% which it currently is doing.

The world is full of USB audio interfaces that served their owners very well without any asynch support at all. I can wait for it to become a standard feature, which might even take a few years.

Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-03 15:01:09
My old laptop and my new laptop run audio over USB just swell using a very basic interface - the Behringer UCA 202 with no parameter changes from defaults, and both or either WLAN or wired LAN active and in use.

My new computer also runs fine (average DPC latency of 60 us). So what?

Quote
I think that we all know that measuring a problem or nailing it in a DBT are both slam dunks when the kind of problems you describe are active. Since this isn't happening, the basic problem must be pretty rare.

Right, it's so rare that pro audio manufacturers have articles on this problem.
For example: avid/m-audio kb (http://avid.force.com/pkb/articles/en_US/Troubleshooting/en404491?retURL=%2Farticles%2Fen_US%2FFAQ%2Fen417151&popup=true)
or NI kb (http://www.native-instruments.com/knowledge/questions/847/Windows+7+Tuning+Tips+for+Audio+Processing)
or echo audio support (http://echoaudio.com/Support/FAQ.php#2)
or focusrite (http://www.focusrite.com/answerbase/en/article.php?id=1058) even has a video
...
or just search in their forums.

Yes, it absolutely MUST be a pretty rare problem.

Quote
What I take away is that the problems you mention are rare, and when they arise they are due to pre-existing technical difficulties that are both far more global within the laptop, and/or  also very much peculiar to the laptop sample where they are observed.

These problems are not limited to laptops. In fact, for example, certain nvidia driver versions had horrible latency problems. So do/did other devices' drivers.

Quote
Most people are interested in Asynch USB because they are led into it by people with commercial interests ar a desire to be opinion leaders, not because of any actual audible problems that they are experiencing.  In many cases you can buy a whole new really pretty good laptop (or several!)  for the price premium of the Asynch USB interface.

Also, it appears that adding Asynch mode to a USB audio interface may be just a matter of changing the firmware. It shouldn't be increasing the price of the equipment by 1,000% or 10,000% which it currently is doing.

I don't know about the firmware but I agree with the other points you make. I'm not advocating to buy an async interface either. As you've said it's probably cheaper to get a better laptop or PC that will run flawlessly with the interface.

All I'm saying is that even if the output is digital the player (a computer in this case) can make a difference under certain circumstances. These circumstances are certainly not the rule but those problems do exist and I wouldn't say they are pretty rare.
Title: If the output is digital does the player make any difference to the so
Post by: probedb on 2012-08-03 16:25:36
Right, it's so rare that pro audio manufacturers have articles on this problem.
For example: avid/m-audio kb (http://avid.force.com/pkb/articles/en_US/Troubleshooting/en404491?retURL=%2Farticles%2Fen_US%2FFAQ%2Fen417151&popup=true)
or NI kb (http://www.native-instruments.com/knowledge/questions/847/Windows+7+Tuning+Tips+for+Audio+Processing)
or echo audio support (http://echoaudio.com/Support/FAQ.php#2)
or focusrite (http://www.focusrite.com/answerbase/en/article.php?id=1058) even has a video
...
or just search in their forums.

Yes, it absolutely MUST be a pretty rare problem.


You mean those pro audio manufacturers that are trying to get you to buy their expensive gear?

Not really a good argument there.
Title: If the output is digital does the player make any difference to the so
Post by: Arnold B. Krueger on 2012-08-03 16:55:06
My old laptop and my new laptop run audio over USB just swell using a very basic interface - the Behringer UCA 202 with no parameter changes from defaults, and both or either WLAN or wired LAN active and in use.

My new computer also runs fine (average DPC latency of 60 us). So what?


The good news is that you then don't have a dog in this fight. ;-)

Quote
Quote
I think that we all know that measuring a problem or nailing it in a DBT are both slam dunks when the kind of problems you describe are active. Since this isn't happening, the basic problem must be pretty rare.

Right, it's so rare that pro audio manufacturers have articles on this problem.
For example: avid/m-audio kb (http://avid.force.com/pkb/articles/en_US/Troubleshooting/en404491?retURL=%2Farticles%2Fen_US%2FFAQ%2Fen417151&popup=true)


Mentions tics and pops which asynch I/O being not a perfect panacea, may not address. Does not recommend or even mention Asynch USB.


Quote
or NI kb (http://www.native-instruments.com/knowledge/questions/847/Windows+7+Tuning+Tips+for+Audio+Processing)


Ditto, no mention of asynch as a solution.

Quote
or echo audio support (http://echoaudio.com/Support/FAQ.php#2)


Ditto, no mention of asynch as a solution.

Quote
or focusrite (http://www.focusrite.com/answerbase/en/article.php?id=1058) even has a video
...


Ditto, no mention of asynch as a solution.

Quote
or just search in their forums.

Yes, it absolutely MUST be a pretty rare problem.


What is the relevant "it". Is it clicks and pops (which will always be with us) or clicks and pops or just poo SQ that is going to be fixed with aynch USB.

My comments about rare were in the context of a discussion of asynch USB. If you try to make them far more general than I obviously intended you can go anywhere you want.


If it was really common people would be lining up to buy and no market stimulation would be required.
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-03 16:55:11
You mean those pro audio manufacturers that are trying to get you to buy their expensive gear?

Not really a good argument there.


This makes absolutely no sense whatsoever. 
Title: If the output is digital does the player make any difference to the so
Post by: Arnold B. Krueger on 2012-08-03 16:59:47
You mean those pro audio manufacturers that are trying to get you to buy their expensive gear?

Not really a good argument there.


This makes absolutely no sense whatsoever. 



I agree because I know of no traditional pro vendor audio interface who is active with asynch USB interfaces.  By active I mean formally announced or delivered products.
Title: If the output is digital does the player make any difference to the so
Post by: yourlord on 2012-08-03 17:17:52
All I'm saying is that even if the output is digital the player (a computer in this case) can make a difference under certain circumstances.


That's never been questioned.

One truism in the computer world is "garbage in, garbage out".. You pipe garbage or delayed data into a USB cable at one end, it's going to arrive at the other the same way.
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-03 17:19:09
Mentions tics and pops which asynch I/O being not a perfect panacea, may not address.

How do you know that?

Quote
Does not recommend or even mention Asynch USB.

Right, any manufacturer should recommend that if their users have problems with their own hardware to buy a new interface with async. I/O. Preferably from a different manufacturer. Makes sense.

later you even posted:
I agree because I know of no traditional pro vendor audio interface who is active with asynch USB interfaces.  By active I mean formally announced or delivered products.


Those articles mention DPC latency as being one source of the problem. You said those problems must be pretty rare. I disagreed and posted those links. That's all there is to it.


Quote
What is the relevant "it".

I was talking about DPC latency problems, which should be blatantly clear by now.

Quote
My comments about rare were in the context of a discussion of asynch USB.

What? So what rare problems in the context of async USB are you talking about then?
To me it seems you're trying to avoid the DPC latency issue now after you've seen that even manufacturers acknowledge it's a problem.


Quote
If it was really common people would be lining up to buy and no market stimulation would be required.

I see, 'pretty rare' turns into 'common'..
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-03 17:28:16
That's never been questioned.

One truism in the computer world is "garbage in, garbage out".. You pipe garbage or delayed data into a USB cable at one end, it's going to arrive at the other the same way.

Read the thread again.
Title: If the output is digital does the player make any difference to the so
Post by: Arnold B. Krueger on 2012-08-03 22:26:43
Mentions tics and pops which asynch I/O being not a perfect panacea, may not address.

How do you know that?


Simple. The tics and pops come from data that is lost inside the PC. We've had these as long as we have had computer interfaces that were entirely internal to the PC, and long before USB or even SPDIF.

Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-03 22:38:40
Simple. The tics and pops come from data that is lost inside the PC.

I don't understand what you mean by that. Do you have any details on such issues? I guess now we're really talking about pretty rare events.
Title: If the output is digital does the player make any difference to the so
Post by: Roseval on 2012-08-04 09:43:27
It is very simple.
USB Audio is a isochronous stream, the track is decoded and stored in a buffer in memory. It is spooled from memory to the internal USB hub and then from the hub to the USB DAC. The internal hub reserves sufficient bandwidth for the audio stream.

There are 2 type of errors possible.
DPC latency problems simply means the PC is too late in feeding the internal USB hub so it can’t maintain the needed quasi real time stream.
Obvious DPC is a upstream problem (before the internal hub) and therefore the method used to sync the audio device has no relevance.
No USB (audio) device is able to control what is happening upstream.

Sometimes there are CRC errors. Here the bits got mangled between the hub and the audio device. Compared with DPC they are a bit rare but they can happen. As the UAC1/2 protocol doesn’t have a retry like all bulk mode device, this might yield clicks, pops, etc as well.

You won’t find many pro audio interfaces using USB audio (UAC)
The reason is probably that up to 2009 there was only UAC1 and this standard is limited to 2* 24/96.
If you want to do a multi channel recording with low latency, this won’t do.
Most pro USB interfaces (e.g. RME) use bulk mode. As a consequence they do have to write their own device drivers at the PC side.
By design bulk mode is async.
Title: If the output is digital does the player make any difference to the so
Post by: db579 on 2012-08-04 16:29:50
Thanks for that Roseval really clear helpful explanation.  Do you know whether the same applies to MHL-USB? Is that isynchronous?
Title: If the output is digital does the player make any difference to the so
Post by: Roseval on 2012-08-04 21:22:34
Thanks

There are a couple of exiting developments going on in the Android world
One is the implementation of UAC2 (USB Audio Class 2)
However this is the straight “old” Linux stuff.
There are a couple of new developments going on and I must admit I’m totally at loss where it is about technically and sound quality wise
http://thewelltemperedcomputer.com/SW/Andr...Android_USB.htm (http://thewelltemperedcomputer.com/SW/Android/Android_USB.htm)

Title: If the output is digital does the player make any difference to the so
Post by: Arnold B. Krueger on 2012-08-04 23:43:50
Simple. The tics and pops come from data that is lost inside the PC.

I don't understand what you mean by that.


There's a lot more places to lose data in a computer than just a USB link. On paper, many look like they should be abundantly fast and never lose a sample. But in the real world, stuff breaks and other tasks interfere. Data gets lost and you have a click or a pop. I'm not going to teach a computer architecture class, but there's where the answers lie.
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-05 00:36:44
On paper, many look like they should be abundantly fast and never lose a sample. But in the real world, stuff breaks and other tasks interfere.

Like ISRs/DPCs?  You don't have to try to teach anything here, just tell me what you mean.

Quote
Data gets lost and you have a click or a pop.

Do you mean data loss for example due to a memory failure? No interface can deal with that so it's of no relevance to the topic. If something breaks it of course makes a difference..
Title: If the output is digital does the player make any difference to the so
Post by: saratoga on 2012-08-05 00:55:29
I'm absolutely convinced that you have broken hardware. Beyond that, what else am I supposed to take away from the anecdote?

Well that's just (being) stupid but I'm okay with that.


What a tactful way to avoid answering the question. 

I don't know about the firmware but I agree with the other points you make. I'm not advocating to buy an async interface either. As you've said it's probably cheaper to get a better laptop or PC that will run flawlessly with the interface.


No, its cheaper just figure out what broken device you have that is blocking inside an ISR (!!!) and throw it out a window.
Title: If the output is digital does the player make any difference to the so
Post by: db579 on 2012-08-05 15:02:44
@roseval So am I right in thinking this would be implemented through software on new android devices rather than hardware (the USB port is already there right...) so would older models be able to use it?
Title: If the output is digital does the player make any difference to the so
Post by: Roseval on 2012-08-05 17:14:59
The USB DACs being UAC1 or UAC2 compliant (use the native mode drivers as supplied by the OS for USB Audio) will work when UAC1/2 is implemented in Android.
Android is based on Linux and ALSA does have the drivers.
Obvious this is a matter of software.
A couple of mobiles already do: http://www.audiocircle.com/index.php?topic=103447.0 (http://www.audiocircle.com/index.php?topic=103447.0)

The architecture as proposed by Google (the device, in this case the DAC) provides the power, not the host (e.g. a mobile running Andoid) needs different  hardware.
Title: If the output is digital does the player make any difference to the so
Post by: phofman on 2012-08-05 21:03:59
No, its cheaper just figure out what broken device you have that is blocking inside an ISR (!!!) and throw it out a window.


Actually, it would be a crappy driver to be thrown out. The device itself is innocent
Title: If the output is digital does the player make any difference to the so
Post by: [JAZ] on 2012-08-05 21:40:35
Actually, it would be a crappy driver to be thrown out. The device itself is innocent


That is not always the case 
Remember the Creative Soundblaster Live! , it had a design that was using the PCI bus in excess and caused problems with some chipsets. ( I've quickly searched for a link about it and found this: http://www.tomshardware.co.uk/forum/37211-...opping-crakling (http://www.tomshardware.co.uk/forum/37211-10-live-popping-crakling) . Guess i could find a better source with a better searching)
Title: If the output is digital does the player make any difference to the so
Post by: xnor on 2012-08-05 23:39:53
That is not always the case

Yeah but if you can fix all the glitches by reducing the DPC latency then the sound card is not broken, regardless if somebody is absolutely convinced of that. As absurd as it sounds, DPC latency can be (temporarily) reduced even by just restarting/disabling certain Windows services.

saratoga, you don't know what laptop I have, what sound card I use or anything else yet you're absolutely convinced my hardware is broken. I'm sorry if I offended you but imho it's exactly what I said (http://www.hydrogenaudio.org/forums/index.php?s=&showtopic=96272&view=findpost&p=804120) it is. Same goes for your "throw it out a window" statement.
Title: If the output is digital does the player make any difference to the so
Post by: Arnold B. Krueger on 2012-08-06 08:14:43
On paper, many look like they should be abundantly fast and never lose a sample. But in the real world, stuff breaks and other tasks interfere.

Like ISRs/DPCs?  You don't have to try to teach anything here, just tell me what you mean.


Like ISRs, DPCs, and every data bus, register, and byte of memory inside the PC.

I don't have to diagnose the problems, all I have to know is that the problems still happen.

It is a simple fact that people still have tics and pops in PCs that don't route audio through USB. Therefore it is abundantly clear to anybody with an ounce of brains that you can systematically fix USB all you want, and those kinds of problems will still exist, and there will be tics and pops.
Title: If the output is digital does the player make any difference to the so
Post by: phofman on 2012-08-06 17:43:55
If you raise playback buffer/playback latency, will it cure the pops and clicks? The clicks in most cases are simple late delivery of samples to RAM for the soundcard to read via DMA. If you keep the playback process more ahead of the reading pointer, you get more time headroom for the playback chain.
Title: If the output is digital does the player make any difference to the so
Post by: [JAZ] on 2012-08-06 19:00:34
phofman:  Yes and no.

With windows Vista and upwards, microsoft changed the way the audio stack worked inside the OS.
With windows XP, I can almost halt the PC if I run an audio thread at high priority and uses all the cpu (ok, one core), but with Vista, you get audio skips.
They moved the audio stack out of the kernel, and that way it lost part of the "privileges" that it had until then.

When defining the buffer in an application, you're defining the buffers that it gives to the driver. With ASIO (and real drivers, not API emulators like ASIO4All), the same buffers are then sent to the soundcard, but with other APIs, that's not necessarily the case.


Now, don't forget that the meaning of multithreading is, in fact, slice tasks of different threads/programs in time fractions determined by the Operating System, and then run each one sequentially, determined by task priority and other order mechanisms.

And here is where the DPC latency plays a role. Basically, the problem is that once the soundcard driver needs audio, it schedules a callback that will provide this data, and this acts as a queue. (separated from user programs and with higher priority).
The latency of the DPC queue is determined by the different drivers that run, and the code it executes. It is expected that these calls are fast, but when they are not, the scheduled callback might not return in time to avoid audio skips.

(Better explained in http://www.thesycon.de/deu/latency_check.shtml#background (http://www.thesycon.de/deu/latency_check.shtml#background) )

Title: If the output is digital does the player make any difference to the so
Post by: phofman on 2012-08-06 20:19:42


When defining the buffer in an application, you're defining the buffers that it gives to the driver. With ASIO (and real drivers, not API emulators like ASIO4All), the same buffers are then sent to the soundcard, but with other APIs, that's not necessarily the case.


I am talking about the direct DMA buffer, not a buffer which gets redefined by the mixing layer.


Now, don't forget that the meaning of multithreading is, in fact, slice tasks of different threads/programs in time fractions determined by the Operating System, and then run each one sequentially, determined by task priority and other order mechanisms.

And here is where the DPC latency plays a role. Basically, the problem is that once the soundcard driver needs audio, it schedules a callback that will provide this data, and this acts as a queue. (separated from user programs and with higher priority).
The latency of the DPC queue is determined by the different drivers that run, and the code it executes. It is expected that these calls are fast, but when they are not, the scheduled callback might not return in time to avoid audio skips.


DPC is in single milliseconds at worst, while soundcards easily handle DMA buffers of hundreds of millisecs, some even seconds. The trick (unless low latency required) is to raise the buffer sufficiently.

(Better explained in http://www.thesycon.de/deu/latency_check.shtml#background (http://www.thesycon.de/deu/latency_check.shtml#background) )


Well, that document does not mention the DMA buffer at all. Raising the buffer fixes any DPC problems (unless a really buggy driver hogs the CPU), provided one does not need to keep latency low for recording/authoring work. That is a situation very different to pure playback scenario.

I am aware of the fact that raising the buffer is often not an option on windows. That is why I use linux for audio. But the hardware itself is OK, fault of the rigid software stack. I have not heard of the SB Live problem in linux - every PCI card communicates over PCI constantly, reading the sound samples via DMA. IMO the cause is a too low DMA buffer chosen by the proprietary driver, causing underruns for slower PCI hardware.
Title: If the output is digital does the player make any difference to the so
Post by: [JAZ] on 2012-08-06 22:08:56
DPC is in single milliseconds at worst, while soundcards easily handle DMA buffers of hundreds of millisecs, some even seconds. The trick (unless low latency required) is to raise the buffer sufficiently.


Tested latency monitor (http://www.resplendence.com/latencymon) this afternoon:
Code: [Select]
REPORTED DPCs
____________________________________________________________________________________________________
_____
Highest DPC routine execution time (µs):              19488,501337
Driver with highest DPC routine execution time:       ACPI.sys - Controlador ACPI para NT, Microsoft Corporation

Highest reported total DPC routine time (%):          2,344699
Driver with highest DPC total execution time:         ACPI.sys - Controlador ACPI para NT, Microsoft Corporation

Total time spent in DPCs (%)                          3,438667

DPC count (execution time <250 µs):                   381482
DPC count (execution time 250-500 µs):                0
DPC count (execution time 500-999 µs):                1334
DPC count (execution time 1000-1999 µs):              6675
DPC count (execution time 2000-3999 µs):              887
DPC count (execution time >=4000 µs):                 0


Yes, that says the maximum has been 19 milliseconds. Just playing a song with directsound set at 5 buffers of 1576 samples (176ms of audio latency, 44Khz).
And obviously, i had audio skips.

I can get as low as 59ms with directsound and 25ms with WASAPI (And even less with ASIO and Asio4all). It really depends if the system wants to lag or not.

I was going to ask why you talk about DMA buffers, but I guess this sentence of yours is the key "I am aware of the fact that raising the buffer is often not an option on windows. That is why I use linux for audio."

So, now that we're on linux... What problems have you experienced with USB Audio over it? It's just so that we can conclude that Windows is more prone to the problem, and as such, it's audio stack, and not the USB protocol.
Title: If the output is digital does the player make any difference to the so
Post by: Roseval on 2012-08-06 22:43:29
Maybe this link is of use: http://www.rme-audio.de/en_products_firefa...s/en_uc_general (http://www.rme-audio.de/en_products_fireface_uc_systems.php?page=content/products/en_uc_general)
Title: If the output is digital does the player make any difference to the so
Post by: phofman on 2012-08-07 11:46:01
Quote
Yes, that says the maximum has been 19 milliseconds.


Well, I do not know what that number means, since the output also says:

Code: [Select]
DPC count (execution time >=4000 µs):                 0


It could as well be a total time spent in DPCs of ACPI.sys during the test period. No source code available to find out, but you can try to ask the developer.

Quote
Just playing a song with directsound set at 5 buffers of 1576 samples (176ms of audio latency, 44Khz).
And obviously, i had audio skips.


I do not know if these are DMA buffers the soundcard is using, or some intermediate buffer for the windows audio subsystem which allocates a different size buffer for the actual transfer to the soundcard. My knowledge of windows audio is almost zero

Quote
I can get as low as 59ms with directsound and 25ms with WASAPI (And even less with ASIO and Asio4all). It really depends if the system wants to lag or not.


For pure playback the figures do not have to be so low. E.g. MPD uses 500 ms buffer by default.

Quote
I was going to ask why you talk about DMA buffers...


Because that is the buffer which counts when dealing with audio nondelivery to the soundcard. For PCI/e devices (not USB) the ASIO buffer is the direct DMA buffer, buffer allocated by linux alsa for raw hw:X device (e.g. used by MPD playing to hw:X device) is the direct DMA buffer too. This is to show that the DMA buffer is accessible, it does not have to be something hidden deep in the driver layer. I have no idea about other windows sound systems. In fact there are developers of some commercial closed-source audio players (I will not name, it is googleable  ) who have no idea about the DMA transfer and believe the CPU sends data directly to the soundcard.

USB devices work a bit differently, the audio buffer is used by the USB driver to prepare/mix final USB frames for all current USB streams to the actual DMA buffer for the USB controller to read.


Quote
So, now that we're on linux... What problems have you experienced with USB Audio over it? It's just so that we can conclude that Windows is more prone to the problem, and as such, it's audio stack, and not the USB protocol.


In the end every OS communicates with the hardware in basically the same manner, unless one features a bug in the driver. Linux offers access to even the lowest software layers which means an advanced user can tweak the params at will. I have tried to torture the USB audio stack and it proved to be surprisingly resilient - http://www.diyaudio.com/forums/pc-based/93...tml#post1719044 (http://www.diyaudio.com/forums/pc-based/93315-linux-audio-way-go-92.html#post1719044)