do you use bit-perfect audio?

D

dioxholster

Gawd
Joined
Aug 11, 2010
Messages
801
Or is it bit-matched? Sound cards have an option to do bit-matched playback, and im guessing wasapi or asio have to be utilized by the media player in order for that to work. Is it worth it over the normal mode with sound resampled, remixed or whatever happens. I figured out how to get wasapi in media player, which i think pybasses windows mixes, dont know if its any better. Anyway, lets say i dont enable bit-match and use the soundcard's EQ, wont the sound still "enter" the card cleaner with wasapi and thus better even when the card itself doesnt output it bit-perfect? Also I read in the thread below that all volumes like the sound card's and software must be max for bit-perfect to work, is that true? I use x-fi, and im a fan of crystilizer, so im wondering if bit-matched is a better alternative especially for DVDs and FLAC.

http://www.head-fi.org/t/312005/auzen-prelude-which-mode-do-you-use/30
 
The resampler in Windows 7 is apparently of high enough quality to merit setting your sound card to the highest bit depth/sample rate upon playback.

WASAPI exclusive mode, but only in the pull configuration (used in WMP iirc) is confirmed to be bit perfect, but this is useless because you won't be able to hear the difference. It also doesn't matter what you set the bit depth or sample rate to in exclusive mode, because the application will take direct control of these settings for the session anyway.

Also, you won't be able to hear the difference between bit-perfect and not. Repeated ABXing have confirmed that it is basically impossible; bit-perfect is much more useful for recording, or if you have a very, very high quality setup.
 
Im not sure if i completely understand the technical terms you used, but bascially you saying that it makes zero difference whether windows mixes it or not. Was wasapi made just for recording? Judging directly between two incremental audio settings is tough, I always view audio quality as a subconscious effect so for stuff like this, I just follow the math; if its supposed to sound better than surely it does. If I follow my ears then everything sounds great as long as it aint distorted.
 
dioxholster -

Your argument is valid. I would agree with you that a lot of audio wizardry you hear about is all related to a subconscious/placebo effect. Also, if a placebo does make you enjoy the audio more, why not, right? In which case, using WASAPI or ASIO should make your output bit-perfect. They weren't "created" for recording, per se, but they work very well with recording, where being accurate is a lot more important.

Anyway, on your original point about using EQ and using WASAPI. I'm not sure exactly how your card works, but I'd assume "bit-match" means that the card isn't trying to do any funky resampling. I would enable this all the time. And yes, your volumes should be all maximized except for your amplifier, if you have one. Also, yes, it would "enter" the card cleaner, but isn't enjoying the sound the way you like to hear it more important than whether the 0s and 1s are a perfect match?
 
I personally use Foobar + ASIO and have Audio Creation and bit-matched enabled on my X-Fi when I solely listen to music. Though apparently Vista/Windows 7 doesnt really affect soundquality when volume is maxed out, but I still use it for the peace of mind much like I prefer to use FLACs over high bitrate MP3s even though I cant really hear the difference with reliable results.
 
Again for the average user who just wants to listen to music (regardless if its 96/24 or 128k MP3's) will not be able to hear the difference. Bypassing the Windows mixer (at least the windows 7 one) has not shown to produce an actual audible difference in blind testing.

Convincing yourself that WASAPI or ASIO will make your music sound better is just you buying into something with zero empirical evidence to back it up like spending insane amounts of money on supposed "better" cables when they have been shown time and time again to be utterly pointless. I challenge you to try out the ABX mode in Foobar and tell me there is a difference , you won't be able to hear the difference.
 
I used to be very concerned about this stuff. However even on my higher end speakers, I could not tell the difference.
 
Godmachine said:
Convincing yourself that WASAPI or ASIO will make your music sound better is just you buying into something with zero empirical evidence to back it up like spending insane amounts of money on supposed "better" cables when they have been shown time and time again to be utterly pointless. I challenge you to try out the ABX mode in Foobar and tell me there is a difference , you won't be able to hear the difference.
Click to expand...

The problem with this argument is that you don't listen to music in a double-blind situation. If the OP believes that WASAPI/ASIO makes his music sound better, and he is psychologically fooled into hearing so, let him do it. It's free, anyway.
 
Perpetuating placebo effects like that just leads to a lot of problems in the long run tho.
 
I agree with Impulse, and if you want to find out once and for all whether or not DirectSound alters your music just do this:

1.) Record a track using DirectSound,

2.) Record the same track using WASAPI.

3.) Download audio DiffMaker.

4.) Subtract the DS recording from the WASAPI recording and see if you hear anything. If yes, DS is NOT transparent, and if no, then yes it is transparent.

5.) It's really about as simple as that AFAIK.

But honestly, I use WASAPI too. Since I haven't gone to the trouble of doing what I just described, by using WASAPI, I know the Windows audio stack is not altering my sound. Being 100% sure is better than debating it.

Impulse said:
Perpetuating placebo effects like that just leads to a lot of problems in the long run tho.
Click to expand...

I agree, it's better to accept the truth for what it is and go from there, than try to do whatever makes you feel better. What if you are deciding between using vinyl or Compact Disc, or deciding between CD and SACD, or deciding between MP3 and lossless, and one option is significantly more convenient than the other? Which one would you like more, to accept the truth and take the easier option, or take the more intense/difficult option because it "feels" better?

EDIT let me explain why I chose that example. Compact Disc is better than vinyl because it is more prevalent, easier to play/digitize, has superior sound quality, and does not degrade every time you listen to it. CD be better than SACD because studies have shown that after 550 trials, the listeners could not differentiate between a 24/96 and a 16/44 recording, they were guessing with 49.8% accuracy. Finally, there is nothing wrong with MP3 because odds are, if you use ABX comparator and compare FLAC to MP3 in an acceptable bitrate like V2 VBR or more, the quality will usually be identical in most cases (until you buy $1,000 headphones).
 
Last edited:
XacTactX said:
I agree with Impulse, and if you want to find out once and for all whether or not DirectSound alters your music just do this:

1.) Record a track using DirectSound,

2.) Record the same track using WASAPI.

3.) Download audio DiffMaker.

4.) Subtract the DS recording from the WASAPI recording and see if you hear anything. If yes, DS is NOT transparent, and if no, then yes it is transparent.

5.) It's really about as simple as that AFAIK.

But honestly, I use WASAPI too. Since I haven't gone to the trouble of doing what I just described, by using WASAPI, I know the Windows audio stack is not altering my sound. Being 100% sure is better than debating it.



I agree, it's better to accept the truth for what it is and go from there, than try to do whatever makes you feel better. What if you are deciding between using vinyl or Compact Disc, or deciding between CD and SACD, or deciding between MP3 and lossless, and one option is significantly more convenient than the other? Which one would you like more, to accept the truth and take the easier option, or take the more intense/difficult option because it "feels" better?

EDIT let me explain why I chose that example. Compact Disc is better than vinyl because it is more prevalent, easier to play/digitize, has superior sound quality, and does not degrade every time you listen to it. CD be better than SACD because studies have shown that after 550 trials, the listeners could not differentiate between a 24/96 and a 16/44 recording, they were guessing with 49.8% accuracy. Finally, there is nothing wrong with MP3 because odds are, if you use ABX comparator and compare FLAC to MP3 in an acceptable bitrate like V2 VBR or more, the quality will usually be identical in most cases (until you buy $1,000 headphones).
Click to expand...

Some good points but headphones are the only part of the equation that truly differs in sound from one another. Lower bit rate recordings (I'm talking 128k MP3) can also be easy to pick out from say FLAC. The only difference I hear with 96/24 recordings versus 44/16 is very slightly more detail and 96/24 sounds a bit "louder" but in a blind test I could likely be prove to not even know the difference. I like thick and heavy bass but with extreme detail , I think detail to me is the most important feature next to proper non-overwhelming bass.

But the vast majority of bullshit in cables , specialized balanced amps and DAC's makes it even more unlikely you'll hear a difference between the windows mixer and a pure ASIO run to the DAC. I honestly feel sorry for people that spend thousands upon thousands on there dac's and amp's for something that a blind test can prove they won't be able to tell a difference with compared to cheaper solutions.

My best advice to anyone that wants to improve upon what they hear , buy better headphones. Sink ALL of your money into great headphones and find somewhere you can audition them at or buy from someplace like say Amazon with an excellent no bullshit return policy that will allow you to try different headphones until you find the best you can afford at that price point.
 
Hell, even if you already invested in one good pair of HPs, you can always buy a second good pair of headphones and it'll have a bigger impact on your experience than most other options, no model excels at everything or can satisfy every usage case.
 
Impulse said:
Hell, even if you already invested in one good pair of HPs, you can always buy a second good pair of headphones and it'll have a bigger impact on your experience than most other options, no model excels at everything or can satisfy every usage case.
Click to expand...


*glances at the ridiculous collection of headphones*

Yeah, thats the way it pretty much goes. :D
 
I don't worry about it at all. In part for simple convenience, if everything is up-sampled to 96kHz I don't have to fuck with it. I have music at 44.1k, 48k, and 96k. So no matter what I'm playing, it all gets converted and life is good. No changing the sample rate all the time.

Another part is because my receiver proceeds to throw a ton of math at the signal it gets, so there's no such thing as "bit perfect" in my system anyhow. When the digital audio hits my receiver a couple of Analog Devices processors go to work on it applying a FIR filter (well six of them actually, one per speaker) to correct for the imperfections in my room. That then gets sent on to the DACs.

I'm not after bit-perfect, I'm after something that sound neutral and correct. Resampling doesn't hurt that so it gets to happen.
 
I use WASAPI in exclusive mode on my media player because why not use it if it's available?

Works great and I have the peace of mind that the bits are just going right from point A to point B.
 
SirMaster said:
I use WASAPI in exclusive mode on my media player because why not use it if it's available?

Works great and I have the peace of mind that the bits are just going right from point A to point B.
Click to expand...

Well if it personally assures you then that's fine but in reality it makes no difference. You are not missing out if you don't use it.

In fact I've found a VST plugin for Foobar2000 far more useful than WASAPI or ASIO and its called : http://www.toneboosters.com/tb-isone/

Isone one is fantastic for helping you tune your cans to there best strengths and also provides a natural sounding crossfeed. I love it.
 
I use Foobar2k with the WASAPI plugin through an X-fi -> Schiit Bifrost. If you're going to use your computer as a transport... might as well do it right.
 
If you like Crystallizer, you're mucking up the audio already so the impact of bit-matched audio is likely to be slim in your perceived playback. I personally prefer ASIO and rate locking, so everything is bit-perfect. I don't even believe in equalizer DSP being in the source chain; that should be done later, before your output just like compression. And while SSRC is pretty impressive, and can alter the noise floor of lower quality audio, I'd much rather listen to that lower quality audio as-is, so there is no chirping or audible crunchy sounding artifacts in rare circumstances. I also only believe in noise shaping and dithering only for 16-bit sources -- that is if the DAC or codec measures better at a native 24-bit sample rate versus just playing back the original audio (I'm looking at you, Cirrus Logic.)

WASAPI Exclusive is nice, but it locks out the mixer from other applications/games playing back audio. ASIO is capable of keeping the locked rate, and playing back other sounds at the same time. If your rate is locked at 44kHz, and a 48kHz sound plays in the background from a game, sure it's going to stretch the sound a little, but at least it's playing. WASAPI (non-exlusive) is hit or miss with this, and can cause whatever application you're using to error out, having to stop and replay the audio again when the device is available (from an application in the background requesting exclusive.) ASIO also has better multi-channel support across a wide range of devices, though if you're listening to music on headphones or a 2.1 system, it's all down mixed anyway.

Volume of course is another problem from manufacturer to manufacturer in this case, some interfaces send hotter signals. That's why we have signal generators to match inputs -- but some information can be lost from a digital pot being reduced from 100% to match some analog device in the chain. I keep most analog pots at 50% gain going to an amplifier to prevent clipping, but even doing -2dB on a digital volume control loses data. It's best to measure via null comparison in this case, to see if any of the audible data has been lost and adjust until there are no differences. So many devices differ, even the same DAC/ADC (or codec) implemented differently can throw things off.

If you aren't recording, or tracking music as a hobby (or professionally) -- do whatever sounds best TO YOU, and tell us audio purists (and audiophiles, even though they're evil) to shove it. I know a few people that listen to music with CMSS3D on, and think it sounds great. Audio engineers of the world are fully aware that you're going to have an equalizer in a V-shape and you think Bose speakers and Beats by Dre are the dogs balls. Any mastering engineer worth their salt went out to the car and checked that it would still make their foot tap before sending it to be pressed ;)
 
Farkle said:
If you like Crystallizer, you're mucking up the audio already so the impact of bit-matched audio is likely to be slim in your perceived playback. I personally prefer ASIO and rate locking, so everything is bit-perfect. I don't even believe in equalizer DSP being in the source chain; that should be done later, before your output just like compression. And while SSRC is pretty impressive, and can alter the noise floor of lower quality audio, I'd much rather listen to that lower quality audio as-is, so there is no chirping or audible crunchy sounding artifacts in rare circumstances. I also only believe in noise shaping and dithering only for 16-bit sources -- that is if the DAC or codec measures better at a native 24-bit sample rate versus just playing back the original audio (I'm looking at you, Cirrus Logic.)

WASAPI Exclusive is nice, but it locks out the mixer from other applications/games playing back audio. ASIO is capable of keeping the locked rate, and playing back other sounds at the same time. If your rate is locked at 44kHz, and a 48kHz sound plays in the background from a game, sure it's going to stretch the sound a little, but at least it's playing. WASAPI (non-exlusive) is hit or miss with this, and can cause whatever application you're using to error out, having to stop and replay the audio again when the device is available (from an application in the background requesting exclusive.) ASIO also has better multi-channel support across a wide range of devices, though if you're listening to music on headphones or a 2.1 system, it's all down mixed anyway.

Volume of course is another problem from manufacturer to manufacturer in this case, some interfaces send hotter signals. That's why we have signal generators to match inputs -- but some information can be lost from a digital pot being reduced from 100% to match some analog device in the chain. I keep most analog pots at 50% gain going to an amplifier to prevent clipping, but even doing -2dB on a digital volume control loses data. It's best to measure via null comparison in this case, to see if any of the audible data has been lost and adjust until there are no differences. So many devices differ, even the same DAC/ADC (or codec) implemented differently can throw things off.

If you aren't recording, or tracking music as a hobby (or professionally) -- do whatever sounds best TO YOU, and tell us audio purists (and audiophiles, even though they're evil) to shove it. I know a few people that listen to music with CMSS3D on, and think it sounds great. Audio engineers of the world are fully aware that you're going to have an equalizer in a V-shape and you think Bose speakers and Beats by Dre are the dogs balls. Any mastering engineer worth their salt went out to the car and checked that it would still make their foot tap before sending it to be pressed ;)
Click to expand...

+1 well said
 
I've tried WASAPI, ASIO, Kernel Streaming, and another one. I forget which.

All of them give me problems when I play something that's really high bit rate or perfect rip, eg I have some PCM 1411kbps files from a playstation 1 game that had no copy protection, and they stutter all the time. Tried all different bit outputs, buffer length, and all that jazz. None of them had any edge over the other.
 
LinkinMcOwnage said:
I've tried WASAPI, ASIO, Kernel Streaming, and another one. I forget which.

All of them give me problems when I play something that's really high bit rate or perfect rip, eg I have some PCM 1411kbps files from a playstation 1 game that had no copy protection, and they stutter all the time. Tried all different bit outputs, buffer length, and all that jazz. None of them had any edge over the other.
Click to expand...

Are you using foobar?
 
You must log in or register to reply here.
Back
Top