audible differences from 48khz to 96khz?

[lastounce]

I'm wondering what the real-world difference between 96k and 48k is when recording ultimately for a CD. Is it worth spending the extra cash on 96k converters? can you hear the difference?

[Bluemoon]

Simple answer - yes. It sounds like you've taken a cloth off the mic, everything sounds like it has a much more space. I don't know what other guys think, but I find it easier to mix, especially where you have multiple intruments of a similar ilk, e.g. two acoustic guitars, a brass section, strings of any kind, overheads, etc.

[JFreak]

you also have to remember that you cannot use adat inputs over 48kHz - so you will limit yourself to maxx 10 i/o simultaneously. if it's enought, it's enough. also, all plugins don't work in higer resolutions, though nowadays almost all plugins do. you will also need to use double hard drive space and all processing takes twice the cpu power, so you might take a hit on your track count, too.

higher resolution is always nice, but if the end product is a CD, i'd still use a 24bit/44.1khz session.

if you decide to try higher resolution knowing you will have a CD as an end product, consider choosing a 88.2kHz session rather than 96kHz - that way you will have a mathemathically easier job to go down to 44.1kHz

[FunkNotPunk]

In my opinion you simply have to determine what is more important to you. Being able to mix in realtime, OR getting the perceived benefit of a little extra sound quality.

If you use 96, then you must do one of two things, a) get a fast enough system to mix everything you want in real time with no compromises (which I have yet to really see on a native only system) or b) know ahead of time that you will be using a lot of Audiosuite plugs and be sure of your mixing choices before hitting "process."

Personally, I hate having to make too many mixing decisions outside of the context of the mix.

When I am doing a small project like an acoustic duo or something like that, I will you 24/96. I can hear enough of a difference to make it worth the CPU hit......and I know my computer is strong enough to handle the smaller track count.

If I am going to do a full band with live drums, multiple guitar layers, loops, layered vocals...so on and so forth, I will do one of two things. What I have done 85% of the time is do the whole project in 24/48 (and sometimes even 16/44.1) because it is a bigger plus in my mind to be able to mix the whole project in real time than to sacrifice listening to stuff together for my perceived better sound quality.

Something else I have experimented with on a couple of songs in order to compare the two resolutions is to track the whole song in 24/96, and do all of my editing/comping and whatnot at this resolution...........then dither the tracks to 24/48 and mix the the entire song that way. My system is powerful enough to have a no compromise realtime ability so I can mix without ever have to bounce tracks pre-maturely or use audiosutes. Then when I am completely done with the mix......which sometimes takes a few weeks....I have opened up the 24/96 version and matched everys single effects and mix setting (this is very time consuming).

I audiosuite all of my EQs and Compressions on individual tracks, while leaving delays and reverbs for bounce to disk. When I am positive I have matched every setting from the mix I am ALREADY happy with at 24/48 (again, this is time consuming)............then I bounced it all down to hear the results.

First off, I do not recomend mixing this way because it takes a lot of time and nearly made my brain explode.............however, the end result was that the 96 mix had a more airy and open sound to it and the whole thing sounded just a little bit sweeter and tighter............however, was it a night and day difference?

Not at all, after a few weeks of letting it all bake and not listening to either mix, I went to my 24/96k mix first. It sounded great..........then I listened to my 24/48K.............it sounded damn good too.

So I suppose my conclusion was that if I had the computer horsepower to do 24/96 in realtime, I probably would..........but not being able to do so IN NO WAY stops me from having great sounding mixes.

p.s.............I have tried this same experiment on both my PTLE system and my PC based Cakewalk Sonar system (although on different songs) with pretty much the same results.

[PhiDelta308]

The higher sampling rates are nice, and I feel that I can hear some difference. I still have flashbacks of that thread on the DUC last year that turned into a very technical discussion of the physics and math involved in the process (fourrier and nyquist stuff). The general feeling that I got from that discussion, and some follow up research, is that the higher sampling rates do not "render" the audio any more accurately. The higher rates allow for the low pass filters to be much higher, thus moving alias frequencies above the range of our hearing. It's the getting rid of that gunk that makes the audio sound so much better.

That all being said, I take too much of a performance hit on my setup to use the higher rates on large projects. I do however work at 24 bit all the time, and I think that has an even greater impact on the quality of the sound.

Try different combinations of bit depth and sampling rates, and see what works for you. If your blessed with a fast enough machine, by all means, work at 88.2 or even 96...but you may find that 44.1 @ 24 bit is a substantial improvement over 44.1 @ 16 bit and you don't take the performance hit.

-Chris

[stackatrack]

Most people I know doing HD TDM recording are doing 24 bit /48k.
By the time you've squashed, squirmed, processed, reprocessed your mixes...I doubt you'd ever hear the difference between any of the sample freqs above 44.1.
Of course in a perfect world who'd want to use a limiter anyway ?
There used to be a time, not that long ago, when mastering engineers painstakenly tried to figure out how to increase the dynamic range of their projects without frying out the cutter head on a lathe. Seems a lot of folk want to use a finalizer and truely finalize any possibility of dynamics.
The great recordings had incredible dynamics.
Go back
Listen
Peace

[Bluemoon]

Totally agree with all that's said above about processing, horsepower, the ability to do great things in 48 (or 44, or a Boss BR532), but it does sound better. This isn't up for debate, only whether it's worth it, and it does make things more open and airier. Talking to a guy from one of the big London studios (in the Pub naturellement), who's only this year moved to PT from Analog tape, but he won't work under 96k.

However, one point that has come out here is 24 bit recording. You simply must do this, and there's no physics involved just a bit of maths. Audio has 3 control bits, so if you record at 16 you're actually only recording 13bits. At 24 you are recording 21bits, so for a 50% increase in storage you're getting 61.5% in quality, and 61.5% in quality my grandmother could hear the difference. You're also 61.5% less likely to clip in my opinion, as you have acres of headroom.

[tele_player]

Audio has 3 control bits, so if you record at 16 you're actually only recording 13bits. At 24 you are recording 21bits, so for a 50% increase in storage you're getting 61.5% in quality, and 61.5% in quality my grandmother could hear the difference. You're also 61.5% less likely to clip in my opinion, as you have acres of headroom.

Control bits? You're making this up, right?