Native Sample Rate of Games

[Blackstone]

Does anyone know what the native audio sample rate of most games is. For example, in Battlefield Bad Company 2, is it best to have my audio set for 44.1 or 96 audio?

 

[DeathPrincess]

They are supposed to be 16bit PCM with the audio work done from 32bit floating point 48kh/z audio (which is perhaps a weird as there isn't much difference in sound quality when working on sound between 24 and 32bit, but personally more difference between 48/96/192kh/z sound, especially as native blu ray is 192-kHz 32-bit stereo or 96kh/z 32bit up to 7.1 though most audio equipment doesn't go that high).

There is a big article about the sound in that game here.

 

[rigurat]

Most games (especially old games) seem to be very low. Try browsing the install directories of the games and if they happen to be readable files, you might see them to be low resolution, low bitrate files.

I used to do that back in the old days (pre 2003), most games at that time were usually 22 khz wav files like Starcraft, and Warcraft 3 used 112 kbps mp3 files (once you extract them).

Times have probably changed since then but still it seems audio isn't top priority in gaming teritory since graphics seems to be taking the spotlight.

 

[Sycraft]

Worrying about bit matched playback, for games in particular, isn't worth it. Thus want you want to do is run you audio at a high rate that our playback devices support (48kHz or 96kHz would be good choices) and let Windows handle the resampling. Windows 7 has a high quality resampling engine and will take care of everything for you. Then just don't worry about it.

However if you obsess over it, most games are 44.1kHz or 48kHz.

Do note that unless you are in Audio Creation mode, your X-Fi will resample anything Windows doesn't. That is one of its selling points. It resamples all audio so there isn't a need to mess with shit.

My advice is to set it to 96kHz and call it good. Everything will get upsampled and work well.

 

[SirMaster]

Worrying about bit matched playback, for games in particular, isn't worth it. Thus want you want to do is run you audio at a high rate that our playback devices support (48kHz or 96kHz would be good choices) and let Windows handle the resampling. Windows 7 has a high quality resampling engine and will take care of everything for you. Then just don't worry about it.

However if you obsess over it, most games are 44.1kHz or 48kHz.

Do note that unless you are in Audio Creation mode, your X-Fi will resample anything Windows doesn't. That is one of its selling points. It resamples all audio so there isn't a need to mess with shit.

My advice is to set it to 96kHz and call it good. Everything will get upsampled and work well.

Except a real-time re-sampling process with consumer grade equipment (or software) most definitely adds in unwanted artifacts into the sound. The audio re-sampler in Windows and in Creative's drivers were written with speed in mind, not quality.

You are much better off outputting the audio as it was mastered. 16-bit 44.1kHz and in some cases perhaps 48kHz.

 

[Sycraft]

Except a real-time re-sampling process with consumer grade equipment (or software) most definitely adds in unwanted artifacts into the sound. The audio re-sampler in Windows and in Creative's drivers were written with speed in mind, not quality.

You are much better off outputting the audio as it was mastered. 16-bit 44.1kHz and in some cases perhaps 48kHz.

Says who, audiophiles that listen with their wallets, not ears? Resampling isn't voodoo, it is well understood and not hard to do right. Also, as I noted, the X-Fi is resampling anyhow. Unless you put it in Audio Creation mode, enable bit-matched playback, and have an application send it data using ASIO or WDM/KS it WILL re-sample all audio, that is just how it works.

This silly idea of audio purity for its own sake really needs to stop, particularly in a consumer environment. If you can hear the difference in a blind test, or show it on a scope (and it is within the limits of human hearing) then it is something to worry about. If not it is not.

Resampling seems to be some boogeyman for various audiophile types, probably because they don't understand it. However it is not hard to do right with the processing power we've got these days.

 

[Nicholars]

Says who, audiophiles that listen with their wallets, not ears? Resampling isn't voodoo, it is well understood and not hard to do right. Also, as I noted, the X-Fi is resampling anyhow. Unless you put it in Audio Creation mode, enable bit-matched playback, and have an application send it data using ASIO or WDM/KS it WILL re-sample all audio, that is just how it works.

This silly idea of audio purity for its own sake really needs to stop, particularly in a consumer environment. If you can hear the difference in a blind test, or show it on a scope (and it is within the limits of human hearing) then it is something to worry about. If not it is not.

Resampling seems to be some boogeyman for various audiophile types, probably because they don't understand it. However it is not hard to do right with the processing power we've got these days.

Resampling quite obviously sounds worse than bit perfect playback to be honest, I would recommend setting it to 44.1 or 48khz whatever games use the most and then using WASAPI for other things like music which will send the bitperfect signal to your DAC and switch the sampling rate automatically. Although the sound quality drop from resampling is probably less noticeable in games than in music so setting it to 96 or 192 probably wont be as obvious as it is in music.

 

[Nicholars]

I just realised this thread is 3 years old sorry for the necrobump.

 

[B00nie]

Does anyone know what the native audio sample rate of most games is. For example, in Battlefield Bad Company 2, is it best to have my audio set for 44.1 or 96 audio?

If it makes a difference you can just do a quick comparison which sounds better. If you can't hear the difference just leave it as it is and forget about it.

Edit: And thanks for the necrobump lol.

 

[Sycraft]

Resampling quite obviously sounds worse than bit perfect playback to be honest, I would recommend setting it to 44.1 or 48khz whatever games use the most and then using WASAPI for other things like music which will send the bitperfect signal to your DAC and switch the sampling rate automatically. Although the sound quality drop from resampling is probably less noticeable in games than in music so setting it to 96 or 192 probably wont be as obvious as it is in music.

Sigh. It really isn't obvious or indeed audible with proper resampling. If you wish to test, ok well have a listen to this and this and tell me which one is resampled. They were both made from 44.1kHz samples mixed at 44.1kHz in Sonar. However I took one in Sound Forge and upsampled it to 192kHz, then back down to 44.1kHz using iZotope SRC, the other I did not. They are different, but see if you can tell the difference audibly, not using a spectrum analyzer or the like. My bet? You can't.

If you want to do an empirically valid test grab an ABX testing utility like this one and run a test. See if in a blind test you can identify the difference.

The question with resampling is not if it is bit accurate, it isn't, but if it is audible, which it isn't when done well. It just doesn't matter much.

 

[cageymaru]

They sound completely different but I don't know which is supposed to be resampled. Wav #1 sounds kinda dull or less dynamic than #2. It's the driving beat that is so different between the two tracks. #1 sounds like it is blended together like an alphabet soup. #2 has definition in the sound and I can pick the instruments out better. I think track #1 has more bass than track #2. So I'm torn as to if you added in more bass with the resample or did it lose the bass and something boosted the treble to accentuate the guitar sounds.

With that said I might have my Zx set to resample to a higher bit rate. Regardless I like #2 even though #1 has the perception of more bass. Is #2 the original since it lacks the muddier bass track from #1? I'm getting older so let me know if my ears have died on me from all of the thousands of hours listening to 150+ db car stereo systems back in the day.

Lastly if you listen to them too much they start sounding the same. It's like my brain wants them to sound the same even though they sound like night and day to me on the first 25 listens. Funny shit. Makes me wish you had posted 3 tracks; the original, the original again, and the resampled for another test to see if we can tell them apart.

Oh yes, I'm listening to them on $55 Pioneer headphones after tax so I blame it on them if I failed.

 

[Sycraft]

My original post is below the line... however don't bother testing. I had a look and you probably can tell the difference... because I screwed up the resampling. Oops. The problem of doing things too fast and not checking your settings.

I will fix it and reupload. However this does provide an example of what I said "with proper resampling." You F it up, resampling can for sure change the sound.

Since I did mess up I'll tell you that #1 is resampled, #2 is original.

---------------------


The reason I didn't post 3 is a simple file diff would tell you what is what. If you want to test to see if you can hear a difference, that's what the ABX utility is for. You can use the one I linked, Foobar also has a good one. What it'll do is you set one of the tracks as A, the other as B. It then randomly picks one as X. You listen to A, B, and X and tell it which of A or B is X. Do it 15-20 times, see how you do. It'll let you know your accuracy and confidence level. then let your ears rest, and do it again. If you can get good accuracy and good confidence, then ok, you can hear a difference. If not, then it is all in your head .

 

[Sycraft]

Ok. Fixed my screwup. Now you have this one and this one. I verified they do, indeed, sound the same to me.

Also you can see the difference, if you like with this. This file is a difference of the two, I took the tracks, inverted one, and mixed them. Only the differences remain. If you look at it on a spectrum analyzer all you see is a tiny bit in the highest frequencies, as a result of the anti-aliasing filter inherent to resampling.

 

[B00nie]

Lastly if you listen to them too much they start sounding the same. It's like my brain wants them to sound the same even though they sound like night and day to me on the first 25 listens. Funny shit. Makes me wish you had posted 3 tracks; the original, the original again, and the resampled for another test to see if we can tell them apart.)

This is exactly why some snake-oil high-end vendors will feed you cock and bull stories about burn in times. They know that even if you think you bought expensive crap on your first listen, your brain will adjust during the burn in and convince that the components just needed burn in lol.

If it doesn't sound good fresh out of the box, it isn't. Simple as that.