We have a situation where we recorded tracks at 48khz. However, the final mix will be printed through an analog console. Is there any benefit - and if so, how much - to printing the mix through the console in 96khz instead of 48khz?

  • 'printed' or mixed? printed inferrs no changes will be made, ie the analog mixer is simply acting as a summing mixer... If you are actually mixing through an analogue mixer (ie using processing such as EQ, compression, delays, reverb etc) then thats a different scenario
    – user49
    Jan 28, 2014 at 23:19
  • @timprebble: I'm not entirely sure. The mix is worked on in Pro Tools, but the engineer said he will do the print through the board. I'm not sure if the board will pick up the effects from Pro Tools and do them analog, or if it just mixes the already-effected tracks from Pro Tools.
    – Claudiu
    Jan 28, 2014 at 23:39

3 Answers 3


The relevant question is what you're going to do with the mix yet. Distributing a final master at more than 48 kHz makes no sense whatsoever. The only reason it can be useful to record at higher rates is to avoid aliasing issues in any nonlinear effects plugins.

That certainly includes mastering compressors, so if your plan is to feed the digital mix to some such final stage then I would say it is indeed a good idea to sample your analogue mixer's output at 96 kHz, do the mastering on that file, and downsample to 44.1 / 48 kHz for consumer distribution. Since it's only stereo, neither hard-drive space nor CPU power should be an issue, so no matter how little aliasing if you can avoid it it's worth the effort.

Realistically, there will probably be no significant difference unless the mastering includes rather excessive limiting / soft-clipping. If you don't do any digital processing of the mix at all, you might as well record straight to the final target sample rate (note that your ADC will in this case do the equivalent downsampling).


Not really, the only benefit from resampling again at 96 on the analog output is that the output of any time-based effects (delay, reverb, chorus, etc) will be sampled at the higher quality. If you're simply adjusting levels in your mix, you'll receive no benefit. You would be able to capture slightly more data resolution by increasing the bit-depth, however. then you're capturing the minute analog amplitude changes.

  • It's already at 24-bit. There are definitely a good amount of effects going into the mix I think..
    – Claudiu
    Jan 27, 2014 at 23:42
  • 1
    Of course, that's all dependent on WHERE in the chain those time-based effects are occurring. If they happen prior to the initial conversion that feeds into the console, you're not going to be capturing any of those nuances you're referring to. Jan 27, 2014 at 23:58
  • @ShaunFarley: This is true, good point.
    – Claudiu
    Jan 28, 2014 at 0:34
  • What has time-based to do with it? Delay and reverb are perfectly linear, so for these 48 vs 96 kHz doesn't make any difference whatsoever. Chorus and other effects that include pitch-shifting do benefit from higher sample rates, but in practise the shifting range is usually less than a semitone, it won't matter either. The only type of effect where aliasing is really a big issue is fast compressors and direct nonlinear distortion. Jan 28, 2014 at 14:11
  • 1
    "Perfectly linear" in the mathematical sense. Such mappings can be reduced losslessly (via Hilbert space projection; e.g. for a reverb this is implemented by sinc-resampling the convolution kernels) to their equivalent for bandlimited signals, i.e. processing at 44.1 kHz gives exactly the same result as processing at 96 kHz and then downsampling. Jan 31, 2014 at 17:27

No. It would be a different story if you were changing bit depth (16 to 24), but you won't get any benefit by increasing the sample rate in this situation.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.