Achieving the best audio quality

3tuxedo

Senior Member
Apr 2, 2011
393
0
16
Hey guys, so lately i've been trying to understand the things that make engineering tick, and I have a few questions about audio quality. When I export something from Cubase, it seems like the quality is lost, it seems almost "crystalized" for those of you that know what I mean, it loses any brilliance it had. I never really understood sample rates very intricately, so could someone explain how to get the best possible quality when bouncing to a .wav? And also, what do you fellas us to convert your .wavs to mp3s at the best quality? Thanks.
 
Assuming you export at the same sample rate you work at, there shouldn't be a difference. Some plugins work a little bit differently at higher sample rates (for instance, one of the delays I use has more feedback at 96000 than lower rates), so your mix can change if you work at a different rate than you export at.

Otherwise, make sure you're listening to them at the same volume?

If we're getting into small procedural stuff that probably doesn't make an audible difference, I work at 96/24, export at 96/24, and then downsample with r8brain. That way, it's only being filtered at the Nyquist of 44100 once, instead of every time a plugin oversamples.

As far as MP3s go, you can't go wrong with LAME.
 
Chances are either there's some sound enhancer effect in your media player, or you're just hearing things because you don't have the Cubase screen in front of you.
 
or you're just hearing things because you don't have the Cubase screen in front of you.

Not saying something couldn't be fuck'd up in your setup but this is something that should not be dismissed lightly.. God knows how many times I've finished the mix, exported it, played it back, and went: "Jeebuz Christ how did I miss that!?!?"

Screen is the enemy. Turn the fucker off every now and then.
 
Not saying something couldn't be fuck'd up in your setup but this is something that should not be dismissed lightly.. God knows how many times I've finished the mix, exported it, played it back, and went: "Jeebuz Christ how did I miss that!?!?"

Screen is the enemy. Turn the fucker off every now and then.

+1 on this!

I once spent a good deal of time trying to figure out why my tracks didn't sound the same in Reaper & Windows Media Player, only to find out that I had accidentally enabled the equalizer in Media Player. So that kind of thing can happen, too.

I generally mix at 44.1/24 bit, and keep it that way until the final mixdown to CD (which I usually do directly from REAPER). Always dither when reducing the bit depth from 24 to 16. If you need to perform a sample rate conversion down to 44.1, use the best re-sampling algorithm that your DAW allows, even if it takes awhile longer. I don't know a ton about MP3 encoding, but again I've always had good results doing it directly from REAPER. Every time I've run into a problem, it's been the result of some setting in Windows Media Player or the aforementioned "screen effect."
 
Never heard of the screen getting in the way, but that totally makes sense. I also will look at windows media player and make sure the equalizer isn't on. Could someone chime in why they record/mix in a specific sample rate?
 
There's nothing wrong with recording at 44.1 kHz, and there's actually disadvantages to using extremely high sampling rates.

Don't lose any sleep over it - if the mixes aren't sounding right then there's probably other factors at play.
 
You could also try importing the bounced WAV/AIFF alongside your the rest of your session (as long as there's no master buss processing) and invert the signal polarity - as long as they're both sync'd (and I mean down to the sample level) they should cancel.
 
Simply record at 44.1/24 Bit - and when you export your track -> 44.1/16 Bit.

If you´re exporting an audio file from your DAW that you want to use later on in your mix, always go from 24 bit to 16 bit and everything is fine bro.

Never really got into the whole mp3 thing. Alot of quality gets away.
 
Simply record at 44.1/24 Bit - and when you export your track -> 44.1/16 Bit.

If you´re exporting an audio file from your DAW that you want to use later on in your mix, always go from 24 bit to 16 bit and everything is fine bro.

Never really got into the whole mp3 thing. Alot of quality gets away.

You mean from 24 bit to 24 bit.
 
Native PT10 24bit 44.1 for recording, mixing and exporting, mastering I have it dithered to 16bit for YT, Physical, and for MP3 just put in the master and bounce out as MP3 320kpps, never convert elsewhere.

I'm not comforable with recording above 48 without compromising my buffer size which I have on 128 at 44.1. My friend has a similar PC spec and he's on 96 and just so his CPU is OK has to record at 1028 and mixes at 512, which is a bit high for recording though the audio is heavily edited anyway, but for 32bit plugs in PT10 and 64 in 11 it just gets stupid delay.

I'm constantly told how 96 is sooooo much better, not really sure myself. Is their a relevant difference?

Can someone explaine LAME too? Why is it better than say, putting in your 16 or 24 bit wav master into Pro Tools as a stereo track, and exporting as MP3. Or even exporting as MP3 straight out of PT?

I know most wav-mp3 conversion sites are sucky but surely the difference between Pro Tools and Lame, aren't worlds apart?
 
I'm constantly told how 96 is sooooo much better, not really sure myself. Is their a relevant difference?

This linked paper gets into it a bit, but there's nothing really close to night and day when recording at say 88.2 as opposed to 48 or 44.1. If your cpu can take the hit, 88.2/96 are good but not going to make or break anything.

lavry-white-paper-the_optimal_sample_rate_for_quality_audio(2).pdf
http://www.sendspace.com/pro/dl/gkksh5


Can someone explain LAME too? Why is it better than say, putting in your 16 or 24 bit wav master into Pro Tools as a stereo track, and exporting as MP3. Or even exporting as MP3 straight out of PT?

I know most wav-mp3 conversion sites are sucky but surely the difference between Pro Tools and Lame, aren't worlds apart?

Lame got it's good reputation for it's conversion sounding better when dealing with low kbps, like 128 or 192, ...but once you're converting at 256/320 (which everyone should be doing these days if you have to go the mp3 route) even converting with the latest iTunes which uses the Fraunhofer codec is a contender and just as good ime as most codecs/converters out there, including RBrain and Sox etc.

I'd skip 192 and especially 128 altogether and code 320kbps - non vbr when possible.
 
I do a lot of 88.2 but on big rock sessions I find 44.1 to be fine TBH. I do a fair amount of classical and those guys demand high sample rates but when I've gotten Hir-res rock to mix I'll often dumb it down. No one has ever noticed the difference. YMMV.
This is pure conjecture, but I feel like many converters have a sweet spot where they sound best and you have to figure it out. The plugin thing pointed out above is also big. It's subtle of course and psychosomatics are undoubtedly big.
 
48khz/24bit - because it's supported by iphone/ipod and Once I export just convert to apple lossless. Will covert to mp3 for web/soundcloud/etc. I know CD is still a major car format, but personal players support it so why not just do it. Sure technically we may not need the headroom and lower noisefloor, but on the off chance I do - I'm already there. Anything to take a step of dither, or samplebit reduction the better imho.