Following my personal concerns and curiosity, I decided to do a quick test with a bunch of VSTs at 44.1k and 88.2k, to see which ones sound more or less the same and which don't.
Just a snare drum hit (not ideal for some effects, I know) copy pasted into separate tracks in Renoise, with a different plugin on each.
Tested these 16 plugins: FH132 (beta), Freeamp2, Freeamp3, GClip, Greasetube, M30 Reverb, MarvelEQ, MDA Stereo, Mo Verb, Modern Compressor, Omniverb, PLParEQ3 (this one doesn't work properly in some hosts, including Renoise), ReaEQ, Reverberate LE, Stereoizer, VFX.
For anyone interested, you can download the rendered tracks here:
http://www.mediafire.com/?0bbdsq994eot4d8
(EDIT: new link, 88k files have been converted to 44k for listening convenience)
The result? Some sound more or less the same to me (MarvelEQ, MDA Stereo, Modern Compressor, Omniverb, ReaEQ, Stereoizer and VFX), others present either subtle or drastic differences. I'd watch out for those that sound too different, if you're often working with various sample rates and need consistency.
Also, I think it would be great if developers let us know when a plugin has been tested to work at different sample rates (and at which ones) to avoid surprises. 44.1 is probably still the standard (we should do a poll some day), but CPUs are capable these days and using something else is not so uncommon.
VST Effects At Different Sample Rates
-
- KVRAF
- 2655 posts since 18 Mar, 2011 from Spain
It had a troyan virus in the download!
My whole computer restarts in loop. What have you done?
Just kidding, sorry.
Well, this kind of tests are always interesting and welcome.
I always working at 44,1 Khz at 16 bits. My old soundcard can 96 Khz at 24 bits. SoundBlaster Extigy.
My single core CPU is slow and old as hell, so i always thought 'More Khz = More CPU usage'
But i'm not sure, really.
Could anybody confirm this to me?
If in case this were not true, i will start making music with maximum quality, no doubt. A bit late lol.
EDIT:
Confirmed: More Khz = More CPU Usage. I was correct.
But obviously, this only happens when making a song in real time.
Once the track is done, change the output quality and cross your fingers for not resampling issues when render.
Cheers and have a good day.
My whole computer restarts in loop. What have you done?
Just kidding, sorry.
Well, this kind of tests are always interesting and welcome.
I always working at 44,1 Khz at 16 bits. My old soundcard can 96 Khz at 24 bits. SoundBlaster Extigy.
My single core CPU is slow and old as hell, so i always thought 'More Khz = More CPU usage'
But i'm not sure, really.
Could anybody confirm this to me?
If in case this were not true, i will start making music with maximum quality, no doubt. A bit late lol.
EDIT:
Confirmed: More Khz = More CPU Usage. I was correct.
But obviously, this only happens when making a song in real time.
Once the track is done, change the output quality and cross your fingers for not resampling issues when render.
Cheers and have a good day.
-
- KVRist
- 126 posts since 16 Jun, 2004
I have done a lot of testing with compressors using different sample rates and they differ largely most of them. Internal oversample in most plugins don't do the same positive difference that an increase in sample rate does. Better attack and release behaviour. At 192kHz most compressors sounded really smooth while some needed 384kHz too become usable to me. Regards.