Record at 48kHz Mix down 96khz

[micv]

Any reason why you would or would not export mixdown to 96khz when all tracking and project is at 48Khz?

I've always used 48K but now that computers are more powerful and sound-card can easily do 96khz.  

Can anyone actually hear a different?

[Byron Dickens]

Because no delivery format is 96k. 

[Promidi]

19 hours ago, micv said:

Any reason why you would or would not export mixdown to 96khz when all tracking and project is at 48Khz?

None that I can think of.

[AB99]

I always mix down to 96K in case a format becomes popular later that can use it.

[martsave martin s]

you will benefit more if you record @96k or 88.2k

Edited January 20, 2020 by martsave martin s

[Gswitz]

Recording at higher sample rates gets you more headroom with equivalent fidelity.

Most vsts for audio will sound the same at 44.1 as at 96.

Using all available bits (loudest moment in the song hits 0), 44.1 can theoretically reproduce everything a human can hear up to around 108 decibels at 16 bit. At 24 bit 44.1, there is enough information to reproduce everything a human could hear if one could listen at 160 dB. Thunder is 120 dB. (This doesn't guarantee your stereo does a good job of reproducing.)

Now soft synths are different. Guitar amp simulators may be different. Those may audibly benefit from a higher sample rate depending on how they are written. 

To this end, you can choose to have cakewalk run any vst at a double rate without altering the project.

Bouncing a 44.1 Recording to 96 will likely have no detectable benefits imho.

Edited January 21, 2020 by Gswitz

[mettelus]

Depending on FX used and their accuracy, you might get a little more precision from a mixdown; but if simply converting a rendered file, you won't get anything new/better. As for ...

12 hours ago, Alan Bachman said:

I always mix down to 96K in case a format becomes popular later that can use it.

... technology has surpassed the capability of human hearing for a long time already. Unfortunately, people tend to focus on the technology side versus its application (the listener). Just because something could be recorded at 128-bit/384 KHz, doesn't mean a listener would know or care. 44.1 was chosen because it exceeds human hearing, and no one's ears have gotten any more capable. As we age, hearing deteriorates even without loudness damage (ironic that musicians tend to do this intentionally) ... for the longest time I forgot where I had seen this, but it was on an episode of "Brain Games"

Another interesting test is to use a brick wall EQ (or steep as you can get) as a HPF, set it way high initially (well over 10KHz), play a tune and see how low you need to drop the threshold before you can identify the song. It is more fun as a party trick but is an eye opener for where the crux of music is located.

[chuckebaby]

On ‎1‎/‎18‎/‎2020 at 11:02 PM, micv said:

Can anyone actually hear a different?

You will receive many different opinions from many different people and their personal experiences and uses.

A lot of good info has been covered here, including that in which VST plug ins can utilized at higher sample rates. To me, Sample rate will always be subjective to personal opinion and even though data doesn't lie (when explaining this theory) either does the human ear, which to the average user, cant always define.

[user 905133]

Hmmm.  Had to get out a scope to "hear" it.  ?  

Edited January 20, 2020 by User 905133
to add a picture showing confirmation that an 8K tone is there.

[sjoens]

As I Understand It:

You're much be better off reversing that formula. Record @96 and mix @48 or even 44. You're not gaining anything mixing at a higher rate than your recording. All you'd be doing is making a huge file for no reason. You won't hear any audio nuances much above 48 anyway.

It's not so much WhatYouHear but how much head room the original signal has to capture finer details. Higher resolution rates afford recorded sound that headroom. Also allows any effects/processing more headroom to work in. Once you've applied any processing to the audio, like "bouncing" it (destructive editing), you're safe to down sample to a playable resolution.

Since most players don't support over 48, and probably never will, it makes little sense to mix to it.

Edited January 21, 2020 by sjoens

[Gswitz]

Because of video, for some years i have been archiving 48k 24 bit files. My phone plays them fine, which i use to listen to mixes on my commute to my day job.

I haven't burned a cd for myself in 2 years.

I think that CbB bounces to 32 bit by default internally. I allow this. Internal bounces have never driven my io out of range.

I can surpass io limits when practicing a tune a couple of hours a day. The takes add up and require deleting from the project. I don't know why i record practice. I guess it is so as i develop as a player for the lead, i can then switch to develop my comping and have a better lead to work with and vice versa.

I don't think you can do internal bounces to a different sample rate than the project. You would have to be doing an export to change the sample rate. I always do an internal bounce prior to export. That way i can keep versions of the mixes as i work my way through them.

I end up with a stack of mixes to compare and choose from. I save scenes for each mix. When the artists swing by, they can get a 10k foot view on each version of the song to quickly assess the mixing choices and veto things they don't like. You can use exclusive solo to flip from version to version helping them hear differences in equally loud versions.

If you export your mix without an internal bounce, comparing different bounces would be hard.

Edited January 21, 2020 by Gswitz

[micv]

I realize that if you are doing ITB, it probably doesn't buy you anything.  However I routed my buses to external analog hardware and export the audio, thinking that it's like 'tracking' the final mix bus so let use the higher rate there. 

[Michal Ochedowski]

4 hours ago, sjoens said:

You're not gaining anything mixing at a higher rate than your recording. 

I disagree. Lets' not forget about saturation or distortion type plugins, that can add a lot of harmonic content. Sometimes they don't offer oversampling e.g. Wave Arts Tube Saturator Vintage. If micv was to use them during mixing at 48 kHz, some unwanted nastiness/aliasing would fold back into the signal. Mixing at 96 kHz will prevent a lot of aliasing from appearing below Nyquist frequency. Then when he mixes down from 96 kHz to 48 kHz all this unwanted stuff will be discarded.

I tested it many times myself. Even though aliasing doesn't seem to be totally gone at 96 kHz, it's level is significantly lower at this sample rate. I'm talking about a single track use. If you consider multiple tracks in a single project, this nasty inharmonic content becomes much more apparent. 

[msmcleod]

3 hours ago, Michal Ochedowski said:

I disagree. Lets' not forget about saturation or distortion type plugins, that can add a lot of harmonic content. Sometimes they don't offer oversampling e.g. Wave Arts Tube Saturator Vintage. If micv was to use them during mixing at 48 kHz, some unwanted nastiness/aliasing would fold back into the signal. Mixing at 96 kHz will prevent a lot of aliasing from appearing below Nyquist frequency. Then when he mixes down from 96 kHz to 48 kHz all this unwanted stuff will be discarded.

I tested it many times myself. Even though aliasing doesn't seem to be totally gone at 96 kHz, it's level is significantly lower at this sample rate. I'm talking about a single track use. If you consider multiple tracks in a single project, this nasty inharmonic content becomes much more apparent. 

Very true - this video explains it all:

 

[sjoens]

6 hours ago, Michal Ochedowski said:

I disagree. Lets' not forget about saturation or distortion type plugins, that can add a lot of harmonic content.

Right. Must have slipped my mind momentarily.

Edited January 21, 2020 by sjoens

[Gswitz]

6 hours ago, Michal Ochedowski said:

Lets' not forget about saturation or distortion type plugins, that can add a lot of harmonic content. Sometimes they don't offer oversampling e.g. Wave Arts Tube Saturator Vintage. If micv was to use them during mixing at 48 kHz, some unwanted nastiness/aliasing would fold back into the signal.

This is true of the bounce happened at a higher sample rate, but i don't think that is what the op is getting.

Project at 48...

Saturation plug

Export audio to 96...

...

This will result in the saturation being processed at 48 then the result converted to 96 with badness printed. Yes?

....

On the other hand, using cakewalk up-sampling to 96 for the plug and exporting audio to 48 avoids the badness of the plug.

...

No harm is done exporting to 96, but unless i misunderstand, i don't think there it's much potential benefit. Perhaps if he has an interface from the 2000s the higher rate might reduce issues related to the interface.

[micv]

1 hour ago, Gswitz said:

This is true of the bounce happened at a higher sample rate, but i don't think that is what the op is getting.

Project at 48...

Saturation plug

Export audio to 96...

...

This will result in the saturation being processed at 48 then the result converted to 96 with badness printed. Yes?

....

This is zero'ing in on what I don't yet understand.  The project is at 48k, when I export to96k, my soundcard got set to 96K, what rate is the project now?  I thought that the entire project would be up-converted to 96k, but if what it does is process at 48k and just simply up-convert the 'bounced' audio to 96k, then I can see that there is no real value.  If that's the case how would you actually mix that the higher rate?

 

[bitflipper]

Quote

If that's the case how would you actually mix that the higher rate?

You would have to convert every file in your project to 96kHz, import it into a new project and re-mix. Most likely to no benefit.

The only good reason to convert on export is if you'll be sending the files to a third party who has specifically requested 96k.

[mettelus]

6 hours ago, msmcleod said:

Very true - this video explains it all:

I still get chuckles watching that video, especially for his side comments, but that is definitely a recommended video for just about anyone to see at least once. His point about checking plugins for upsampling (to determine the "lazy" programmers) as a valid one. SONAR/CbB introduced 2x upsampling; but for lower sample rates, that may not be enough to prevent the "grains of sand" (i.e., crap), from showing up in the "hear-able range." His explanation is nicely done overall, but I am surprised at times that internal upsampling by plugins doesn't come to the fore more often. IMO, that is a good litmus test for a digital plugin to be "valuable" or not, especially those that can cause aliasing issues.

You also mentioned this when you posted this video previously, but Melda has 16x upsampling capability in all of their plugins. I am not aware of any other vendor who has gone to that level of detail with plugin design (is there another vendor that does this?).

[Michal Ochedowski]

35 minutes ago, mettelus said:

You also mentioned this when you posted this video previously, but Melda has 16x upsampling capability in all of their plugins. I am not aware of any other vendor who has gone to that level of detail with plugin design (is there another vendor that does this?).

It's true that Melda Production has gone much farther with oversampling capability than any other vendor. However it's usually recommended to use the lowest amount of upsampling until aliasing is gone, at least to save some CPU power. I remember checking Voxengo Tube Amp and comparing all modes. When I got to 8x, which is the maximum, all nasty stuff was gone. Bare in mind that it wasn't a low level signal during my test. I went totally overboard to see what this little gem was capable of. Tube Amp is free if anyone is interested to see for themselves.