Recording 96k, mixing 44.1, bouncing 96

[papiel]

With my native G5+Logic system I can't really mix 96&88khz stuff, but I was wondering whether it will make sense to downsample the tracks to 44.1khz, work on the mix, then replace the files to the original 96khz files and do an offline bounce.
Did anyone ever tried it?

Thanks,
Papiel

[u b k]

going from 96 to 44.1 will sound like ass. going from 88.2 to 44.1 will sound like 44.1. why bother?

track and mix in 48k, it's so much better than 44.1 and is easy on the cpu.


gregoire
del ubik

[phelbin]

I'm told that most commercial projects are being recorded at 48 these days. 96/88.2 does sound significantly better, but the industry has a few years to catch up. And eventually everything is going right back down to 16 bit 44.1KHz.

But to answer your original question. If it was recorded at 96, convert it to 48 instead of 44.1. Dividing by 2 is a lot easier equation for the computer, fewer 1's and 0's will get lost. Then keeping it at 48 until mastering will be just fine.

[AdamJay]

i work at 48khz, for the reasons Ubik mentioned and also it just seems to have the best quality/resource consumption ratio.
sounds great but without the huge cpu hit.

i dunno if its the filtering at 48khz or what, but it sounds damn good.

[Nu-tra]

Which G5? What is it in logic that is crapping out? Is it the cpu or the audio meter? Are you freezing your tracks? Are you using the firewire 800 ports for an external drive for audio?

[papiel]

Thanks guys.
But I'm not sure you understood my question.
Let's say I'm recording at 88.2khz. I have my stuff recorded at top resolution, but I can't mix it in Logic due to both hardisk and CPU limits (dual 1.8 G5, internal SATA drives).
So I downsample everything to 44.1khz. Work on the mix, complete the mix. Now, instead of just bouncing it all (ITB mix) to 44.1khz, I remove the converted files from my project and insert the original 88.2khz instead. All plugins, levels, pans, sends, everything remains the same, just the audio files are the 88.2 originals.
Now I bounce the mix to a 88.2khz file and continue to mastering... (I of course need to change the soing sample rate to 88.2khz as well)

The real question is, if I had enough cpu/disk power to mix with the 88.2khz files all the way through, would I make the same mix as I do with the described method?

Thanks,
p

[papiel]

Why can't I see the messages in this thread?

[papiel]

OK, now I can

[ISedlacek]

Originally Posted by u b i k

going from 96 to 44.1 will sound like ass. going from 88.2 to 44.1 will sound like 44.1. why bother?

track and mix in 48k, it's so much better than 44.1 and is easy on the cpu.


gregoire
del ubik

Not true - there is basically NO DIFFERENCE between 96-44 and 88-44 conversion (unless you use your calculator for that :-)

If you use highest quality SRC (like Voxengo r8BrainPro), you will preserve most of the superior 96k sound quality even in 44k.

48k sounds almost the same as 44k. The same cannot be said about 96k. That´s already quite a difference ...

See also here: ******//gearslutz.com/board/showthread.php3?t=52593

Downsampling to 44k and then upsampling and then again downsampling is probably not very good idea ... (if I understood Mr.Papiel well)

[Apemandan]

Originally Posted by ISedlacek

Not true - there is basically NO DIFFERENCE between 96-44 and 88-44 conversion (unless you use your calculator for that :-)

If you use highest quality SRC (like Voxengo r8BrainPro), you will preserve most of the superior 96k sound quality even in 44k.

48k sounds almost the same as 44k. The same cannot be said about 96k. That´s already quite a difference ...

See also here: ******//gearslutz.com/board/showthread.php3?t=52593

Downsampling to 44k and then upsampling and then again downsampling is probably not very good idea ... (if I understood Mr.Papiel well)

I totally agree, we track to 96K and then downsample to 44.1 using a Weiss Gambit SRC, the difference between that and tracking in 44.1 is vast!

[papiel]

I'm not talking about downsampling then upsampling. Please read my 2nd post. It explains my question.

[Rush909]

Originally Posted by papiel

I'm not talking about downsampling then upsampling. Please read my 2nd post. It explains my question.

HAHAHAH... I don't think anyone is really reading your question... they are just eager to argue their case for 44k/96k!!! hahaha

ok.. I'll take a crack at it... I only use 44k but I understand what you want to do... the only thing I would be concerned about is that you will be making mix decisions based on downsampled 96K samples... so you will apply EQ etc... based on downsampled files at 44K and then once you are happy with the mix, swapping those files to 96K files might not preserve the mix 100%...

since I have not tried 96K yet, I am not sure what you are gaining (is there more high end, clarity etc...) so I am not sure how your EQ/processing decisions will translate once you switch the files back from 44K... it might be worth a try, maybe try it with a rough mix, do your experiment and then see if the mix holds up... if it does then go for it, if it sounds like frankenstien then you should maybe just downsample to 48K and mix and bounce there...

hope this helps...

r.

[papiel]

Thanks Rush909, finally.
I did an experiment, but didn't have 96khz to begin with. I took a mix I'm working on, which is 44.1khz, converted the files to 96khz, bounced the mix to 96khz and it sounded like shit, probably because going from 44.1khz to 96khz on the first place did a lot of degradation. I need to test this with "real" 96khz files.

[Apemandan]

Originally Posted by Rush909

since I have not tried 96K yet, I am not sure what you are gaining (is there more high end, clarity etc...) so I am not sure how your EQ/processing decisions will translate once you switch the files back from 44K...

I'm not sure how this will work...will you run your plug-ins at 88/96kHz? If not then will 88/96k audio just drop back into a 44.1 session? Half the beauty of running sessions at these rates is to get the best from your plugins as well as captured audio.

[illacov]

How about is you recorded on REALLY good converters at 44.1 khz 24 or 32 bit float and then mixed in 44.1? And then mixed down or however your daw is working to a higher sample rate? I'm already satisfied that plug ins like 192 khz than 44.1 it seems to improve their sound quality.

Try it that way.

Just make sure your converters are at least on say EMU 1212m levels then you can record at 44.1 and get very good results along with the upsample for the mixdown.

Peace
Illumination