Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: More misinformation (Read 107101 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Re: More misinformation

Reply #25
When playing FLAC 16/44.1 through DirectSound (e.g. WMP 12) and sound card shared is set to 24/48, Windows has to resample on the fly to the target format, the same with other sounds. WASAPI exclusive is used only in Foobar/Winyl etc.


Re: More misinformation

Reply #26
If you really don't have analog volume control I would upgrade to something better.

I dont know what you think resampling does but its not right.

I certainly hope there's no analog volume control anywhere near my toslink-connected DAC, not before the signal hits my preamp anyway, once it is already in the analog domain. That would be a completely unnecessary D->A->D step, and the same goes for USB/Firewire/Thunderbolt-connected DACs.

With a 24bit digital volume control, you can attenuate a 16bit signal by 48dB using simple bit decimation with no loss of dynamic range at all. That's a lot of attenuation, and I would be surprised if most analogue volume controls could attenuate that much without losing any dynamic range.

Re: More misinformation

Reply #27
If you really don't have analog volume control I would upgrade to something better.

I dont know what you think resampling does but its not right.

With a 24bit digital volume control, you can attenuate a 16bit signal by 48dB using simple bit decimation with no loss of dynamic range at all.

No you cannot.

Re: More misinformation

Reply #28
The discussion seems to be sometimes kind of weird here.

Maybe it is a topic for another forum, but I am also concerned about the fact, considering the sample rates, which has been discussed (and proved by those that have neccessary equipment) in the past here https://hydrogenaud.io/index.php/topic,104547.0.html. E.g. that some common cards perform better at 48/96 kHz than 44.1 (CD). For example, one of my computers has audigy SE (P17V) and when i set the card native rate at the creative control panel to 44.1 the audible noise going to headphones slighlty increases, I do not know the cause  of this behaviour. So I keep it at 48 kHz at those control panel settings (does not prevent using WASAPI 44.1 mode though). Also on this card, when I use the tests downloadable here http://www.audiocheck.net/audiotests_aliasing.php, the aliasing effects, I can do ABX playback test which is asked for here and clearly get the difference between aliasing on 44100 and 48000 kHz file - post below.

foo_abx 2.0.2 report
foobar2000 v1.3.9
2016-02-22 08:58:25

File A: audiocheck.net_aliasingcheck_44100.wav
SHA1: a93746596ba6e363635bf4911682f9e529997b07
File B: audiocheck.net_aliasingcheck_48000.wav
SHA1: 7ef5277406e1a32df872060dfeddd8ee9af7fcc4

Output:
WASAPI (push) : Reproduktory (SB Audigy), 24-bit
Crossfading: NO

08:58:25 : Test started.
08:58:50 : 01/01
08:58:54 : 02/02
08:58:59 : 03/03
08:59:04 : 04/04
08:59:13 : 05/05
08:59:17 : 06/06
08:59:20 : 07/07
08:59:24 : 08/08
08:59:27 : 09/09
08:59:33 : 10/10
08:59:38 : 11/11
08:59:55 : 12/12
09:00:00 : 13/13
09:00:03 : 14/14
09:00:06 : 15/15
09:00:10 : 16/16
09:00:10 : Test finished.

 ----------
Total: 16/16
Probability that you were guessing: 0.0%

 -- signature --
1f22ce7883f9cbcb0e83bc296111193c1d8ef0a2

So maybe it would be beneficial to discuss actual problems of users than trying to talk "yes/no,can/cannot" style.


Re: More misinformation

Reply #29
Did you look at that thread you linked?  The old audigy cards used very poor quality resampling, which is why you get aliasing.

Re: More misinformation

Reply #30
If you really don't have analog volume control I would upgrade to something better.

I dont know what you think resampling does but its not right.

With a 24bit digital volume control, you can attenuate a 16bit signal by 48dB using simple bit decimation with no loss of dynamic range at all.

No you cannot.

Well, that was an informative post.

Perhaps I should have written "no audible loss of dynamic range, provided a DAC with reasonable SNR is used". Of course, I might be mistaken, but a reasonable digital volume control will still be no worse than a normal analog volume control.

But just in case I'm wrong, please explain why you consider digital volume controls unacceptable.

E: https://hydrogenaud.io/index.php/topic,103887.0.html

Re: More misinformation

Reply #31
Did you look at that thread you linked?  The old audigy cards used very poor quality resampling, which is why you get aliasing.

This was I think the case of old Live! with EMU10k. Audigy SE has C0106 processor and should support 44.1 natively, but (as asked here) I get audible artifacts when using those WAVs. No significant audible problems when using 48 kHz mode through WMP 12 and shared mode 24/48, or Foobar WASAPI with SoX resampling to 48 kHz.

I am just trying to explain, that various users and card can have various optimal settings and there are some examples when 44.1 does not work properly and 48/96 kHz does, not mentioning the more room 48 kHz/96kHz provides. Still, it is probably best to use native 44.1 when playing standard CD FLAC without any resampling provided if sound card supports 44.1 sample rate well (without artifacts).

Re: More misinformation

Reply #32
And by the way, if a card (like the tested Audigy SE) performs better at 48/96 than at 44.1, is it better (theoretically) to (software) resample to 48 kHz or 96 kHz? Or the results are the same ( or inaudible difference) regardless of the target rate ?

Jan

Re: More misinformation

Reply #33
If you really don't have analog volume control I would upgrade to something better.

I dont know what you think resampling does but its not right.

With a 24bit digital volume control, you can attenuate a 16bit signal by 48dB using simple bit decimation with no loss of dynamic range at all.

No you cannot.

Well, that was an informative post.



96 +48 = 144db so no you absolutely cannot get a DAC that can do that. Thermal noise exists.  Sarcasm aside do I really need to state this?

Edit: follow-up question, what does this have to do with anything?

Re: More misinformation

Reply #34
And by the way, if a card (like the tested Audigy SE) performs better at 48/96 than at 44.1, is it better (theoretically) to (software) resample to 48 kHz or 96 kHz? Or the results are the same ( or inaudible difference) regardless of the target rate ?

Jan

From what you described above, your card is resampling and also probably not a great choice. You can use a better quality resampler which will prevent aliasing, or you could just use something better.

Do you have onboard audio? If it works just use that.

Re: More misinformation

Reply #35
Thank you. On other computer I have onboard Realtek ALC662 and there are no audible resampling artifacts. On this computer I would probably use SoX for foobar wasapi and/or Windows resampling to 48 kHz shared, and probably will move to different card in the future (as onboard audio is very bad on this PC).

Still I think that from those issues many hot and long debates arose, and moving to better common standard for audio, like 24/48 or 24/96 would be the best solution for general use.

Jan

Re: More misinformation

Reply #36
96 +48 = 144db so no you absolutely cannot get a DAC that can do that. Thermal noise exists.  Sarcasm aside do I really need to state this?

Edit: follow-up question, what does this have to do with anything?

Did I run over your dog or something?

Well, fair enough, I didn't take the SNR into account. But a DAC with a very reasonable SNR of 110-115dB is easily affordable, and will provide plenty of attenuation with no audible loss of quality or dynamic range, when you take the noise floor of your listening environment into account. Good enough to match an analog volume control, which is what you were arguing against.

As long as you're not wildly boosting the signal after digital attenuation, there is no issue.


Re: More misinformation

Reply #38
Bad designs will always exist.  Doubling the data rate guarantees nothing, except wasted bandwidth and the reinforced notion that it's necessary.

True on the first point. But doubling the data rate to the DAC allows you to utilize any SNR your DAC may have above 96dB.

And that's if the Windows/OSX/Linux software volume control is doing simple bit decimation, which may not be the case.


Re: More misinformation

Reply #40
Thank you. On other computer I have onboard Realtek ALC662 and there are no audible resampling artifacts. On this computer I would probably use SoX for foobar wasapi and/or Windows resampling to 48 kHz shared, and probably will move to different card in the future (as onboard audio is very bad on this PC).

Still I think that from those issues many hot and long debates arose, and moving to better common standard for audio, like 24/48 or 24/96 would be the best solution for general use.

Jan


I don't see how changing audio standards would make some ancient, low quality device better.  Nor do I see why we should care.  In the decade since that part was released these issues have long since been solved.  Maybe this would have been a concern in the 1990s, but not today.

Re: More misinformation

Reply #41
96 +48 = 144db so no you absolutely cannot get a DAC that can do that. Thermal noise exists.  Sarcasm aside do I really need to state this?

Edit: follow-up question, what does this have to do with anything?

Did I run over your dog or something?

Well, fair enough, I didn't take the SNR into account. But a DAC with a very reasonable SNR of 110-115dB is easily affordable, and will provide plenty of attenuation with no audible loss of quality or dynamic range, when you take the noise floor of your listening environment into account.

You replied to my post stating that analog volume control not digital attenuation with high dynamic range was the preferred solution in the overwhelming majority of pcs to tell me that you didn't understand how volume control works.  I literally have no idea what your point is.

Quote
Good enough to match an analog volume control, which is what you were arguing against.

I never did whatever you're referring to. Use digital volume control if you like, but do not think it is a sound argument for distributing audio in 24 bit, because it's not.

Re: More misinformation

Reply #42
That's not what I'm arguing, I'm simply arguing that your statement "If you really don't have analog volume control I would upgrade to something better" is disingenuous, as digital volume control is perfectly fine and comparable in quality to analog volume control, in most cases.

Distributing music in 24bit is pointless, but using a 24bit digital volume control for 16bit content is not.

Re: More misinformation

Reply #43
Distributing music in 24 bit is not pointless. Actually it would be great for music industry to standardize on "final" format for consumer audio at 24/48 or 24/96 kHz. Then many discussions would disappear and every consumer could experience  (buy) the maximum reasonable quality at standard prices. At todays storage capacities and after DVD invention it would be logical (and probably final) digital audio evolution.

Re: More misinformation

Reply #44
if a card (like the tested Audigy SE) performs better at 48/96 than at 44.1, is it better (theoretically) to (software) resample to 48 kHz or 96 kHz?
From Wikipedia:
Quote
Creative Labs advertised the Audigy as a 24-bit sound card, a controversial marketing claim for a product that did not support end-to-end playback of 24-bit/96 kHz audio streams. The Audigy and Live shared a similar architectural limitation: the audio transport (DMA engine) was fixed to 16-bit sample precision at 48 kHz. So despite its 24-bit/96 kHz high-resolution DACs, the Audigy's DSP could only process 16-bit/48 kHz audio sources.

Also: https://hydrogenaud.io/index.php/topic,50729.msg454469.html#msg454469

Re: More misinformation

Reply #45
Yes, but I have Audigy SE which is NOT based on EMU processors, rather on (simpler) CA0106 chip. And I dont do any effects.

Re: More misinformation

Reply #46
That's not what I'm arguing, I'm simply arguing that your statement "If you really don't have analog volume control I would upgrade to something better" is disingenuous, as digital volume control is perfectly fine and comparable in quality to analog volume control, in most cases.

He was arguing that he needed 24 bit source material because of Windows volume control.  This is:

1) wrong.
2) a good indication that you probably need an analog volume knob somewhere if this is actually becoming a problem.

Digital volume control can work fine, but its harder to do well then just adjusting gain, which is why few systems lack analog volume somewhere in the chain.

Re: More misinformation

Reply #47
Then many discussions would disappear and every consumer could experience  (buy) the maximum reasonable quality at standard prices.
Please present your evidence that 44/16 is not reasonable quality.

Re: More misinformation

Reply #48
I was arguing that one of the reasons for 24 bit is also the digital volume control. I know that doing digital volume control at 16 bit is possible and introduces no or very small (inaudible) distortions.

Still after all those discussions I clearly see that consumers would highly benefit from 24/48 kHz, if that would not be agreed the 24/96 standard. When I could buy my albums at those rates, I would not discuss the things I am discussing now. CD 16/44.1 is hifi, but is too close to limits and especially at computers age slowly sees its limits and borders.