Sample Rate for Tracking

[NeilParticipant]

Hello everyone,
I'd like to revisit the much debated topic of which sample rate is best for tracking - referring to 48khz vs 96khz here, mainly. I.e., what do you use and what discernible benefits do you see/hear in 96k?
Looking forward to your thoughts,
Neil

[Keith FelsParticipant]

White Sea Studios did a great video on why he uses 192. Breaks out the math in the higher frequencies. I know my ears can’t hear 24khz but he used it as an example. At 48k samples you only get 2 samples per cycle of 24khz. That’s still only 4 samples per cycle of 12khz and the rest is averaged.

[NeilParticipant]

Hi Keith, many thanks for your reply. Do you happen to have the link for that video and what do you use yourself?

[NeilParticipant]

Found the video - thanks!

[Keith FelsParticipant]

I still use 48 because I haven’t had the chance to test anything else and it works for me. Somebody with younger ears can probably tell me I should move up.

[NeilParticipant]

I believe this is a rather well-founded approach & exposition (also extensively debunking the above video): https://people.xiph.org/~xiphmont/demo/neil-young.html

[NeilParticipant]

also good I think: http://www.ryanschwabe.com/blog/96k

[NeilParticipant]

http://www.lavryengineering.com/pdfs/lavry-white-paper-the_optimal_sample_rate_for_quality_audio.pdf

[NeilParticipant]

https://www.youtube.com/watch?v=cIQ9IXSUzuM

[Keith FelsParticipant]

For me what it comes down to is the music you put out for good or bad is the music people hear regardless of your technique or equipment or settings or even talents. If you want it to have a vinyl sound, tape sound or digital sound is up to the creator. Whether or not a setting would sound better is subjective and if that isn't how I created it then it isn't the sound I liked when I hit bounce to disc. If someone feels like they aren't getting the right sound out of a certain sample rate then try another and see if it works better. I'll probably not hear the difference between 44.1 or 48K unless I was specifically looking for it and was told which was which. I just use 48 cause it seems to be the current most common standard.

[NeilParticipant]

Of course the music (creative ideas, good recordings, production, mixing) should be by far the most important part of any discussion related to technology, equipment and the like. But: since we have the technology at hand and 96khz/24bit is nowadays a format widely and easily supported by most interfaces, computers and DAWs, it is necessary to decide if I want to use it or not (and stick to lower sample and bit rates). So I think the questions here are certainly relevant when you're serious about audio recording and processing, whether it's for your own music or that of other people.

I personally find 96khz/24bit to sound better - more open, more details, more presence, etc. One explanation of that seems to lie with the issue of anti-aliasing, which apparently works better at 96khz, because the low-pass filter can slope off much slower in the high end (which is theoretically not audible, but still seems to make a difference in the audio processing. Here's a pretty consistent explanation of that (which can also be found in other places): https://www.youtube.com/watch?v=FG9jemV1T7I

So for now I think I will stick to that format for both recording and mixing, and then also exporting 96/24 wav files, despite the additional disk space and processing power required (my CPU can handle it rather well).

I recently subcribed to the audio streaming service TIDAL which offers audio streaming like Spotify but in Hifi and even MASTER quality (usually 96/24). This is of course a major difference compared to Spotify 320kbit files, but I even think that there's discernible differences between Hifi (44.1/16) and Master quality. The listening pleasure is definitely huge. I recommend everyone to check out this streaming service, you can do 1 month for free (but beware, you might keep your subcription after that :).

[NeilParticipant]

Lastly, this statement by Lavry also seems pretty convincing to me, addressing the issue of "optimal sampling rate":

"At 60 KHz sampling rate, the contribution of AD and DA to any attenuation in the audible range is negligible.
Although 60 KHz would be closer to the ideal; given the existing standards, 88.2 KHz and 96 KHz are closest
to the optimal sample rate. At 96 KHz sampling rate the theoretical bandwidth is 48 KHz. In designing a real
world converter operating at 96 KHz, one ends up with a bandwidth of approximately 40 KHz."
(http://www.lavryengineering.com/pdfs/lavry-white-paper-the_optimal_sample_rate_for_quality_audio.pdf, p. 3)

[Author]

Posts