Converting 16 Bit to 24?
- Thread starter Bassman
- Start date
I'd make 24-bit copies of the original 16-bit files and import 'em into a new 24-bit project. Converting 16 bit to 24 bit should be trivial and you'll have the extra bits to work with while mixing tracks together. That will help a lot with the sound if you have more than a couple of tracks.
Hold off going back down to 16-bits 'til the very end, after mixing and mastering.
Hold off going back down to 16-bits 'til the very end, after mixing and mastering.
I'm pretty sure that with Cakewalk you can have tracks with different bit depths in one project. Your tracks aren't going to sound any better converted to 24 bit, they might even sound worse actually. i think the best corse of action is to do... NOTHING! just record your new tracks at 24. The old 16 bit tracks and new 24bit tracks will all sit on the same project I believe (not too familar with Cake so correct me if I'm wrong) When you mix down, mix down in 24... you shouldn't have any problems
-jhe
-jhe
sonusman
Banned
James HE said:
Not true James, at least the part of them sounding "worse". I agree though that they won't sound any better per se.
What WILL sound better is your plug-in's working with 24 bit files. Basically, a plug-in applies a algorythm to the original sound. The original sound has to be re-created because of the calculations the plug-in does. Your bit depth will increase. But, if you only have 16 bit files to work with, and play back at 16 bits, the extra bit depth is either "truncated", meaning the extra bits the file now has must be eliminated, with no interest if that may SOUND good or not, or they can be "re-dithered" back to the original bit length. Dithering simply add's a noise at the threshold of the sound to noise ratio spec of the bit depth you are working with, but you also lose a couple or few dB of sound to noise ratio BECAUSE of the dither, but this is considered a good trade off over truncating the word because the noise will hide the sudden elimination of sound.
Anyway, that is all technical talk about dithering and what not and not the subject at hand here DIRECTLY. But you need the background to understand what is going on when you apply DSP (which could be something as simple as changing the volume of a track in a software mixer. The .wav file must be re-calculated to create the new sound that is lower in volume....you get the idea here....then the bit depth is longer, thus you need to truncate, or re-dither is to the original bit depth it started out at). Blah blah blah.....
Now, the nice thing about 24 bit word length's is that you have the potential for much more signal to noise ratio, thus, when you apply DSP to a .wav file and the bit depth increase and get addressed by truncating or re-dithering, the truncate or re-dither happens AT A VOLUME THAT YOUR EAR CANNOT DETECT! Thus, you hear a far more natural sound and the result of the DSP is much smoother. Do this make sense? It is sort of complicated, but I have done my best to simplify the advantage (and those that may have a little more insight to the actual workings of all this, please cut me some slack for oversimplifying some of this.....thanks....
)So, yes, you SHOULD work with higher bit depths IF you plan to do ANY DSP to the original file. Even though the resolution of the file will not change just because it is now a higher word length, your file will sound much better after DSP is applied to it.
If you only have a couple tracks though in the project, and you don't anticipate having to do much DSP on them, if any, then the 24 bit increase will not mean as much, and as pglewis stated earlier, the more tracks you add, the more advantage you get with working in the 24 bit realm.
Be aware too that longer bit depths mean a bigger load on your processor too. You may not be able to play back as many tracks, and you will certainly not be able to apply as many real time plug-in's to the tracks before your processor is overloaded.
But if the original tracks sound very good, and you don't need a gazillion tracks, you should be okay.
Hope this helped clear the air a bit for you. Good luck!
Ed
All audio files in a CPA 9 project have to be the same bit width. No mix-n-match.
Converting a 16-bit file to 24 bits will sound exactly the same as the original 16 bit file. You won't directly benefit from the extra bit-depth, but it won't sound worse, either. The lower 8 bits will just get stuffed with zeroes since there isn't any dynamic info available below 16 bits.
SoNuSmAn (hee hee) covered a lot of the whys for doing the conversion anyway. If two 16-bit tracks are mixed together, there is math going on in your software. That math is going to yeild a result that is bigger than 16-bits (multiply a couple of 4 digit numbers together and you can't store the result in a 4 digit number... same deal). This is where rounding errors will creep in. Mix tracks together, change volume levels on tracks, use plug-ins... all these things contribute to rounding errors and a resulting loss of quality.
Now 24-bit files are going to have these rounding errors too, but the dynamic range of 24-bit audio is such that the errors become insignificant. So converting the original 16-bit files to 24-bits before running 'em through mixing software makes sense if you have more than a couple of tracks and you can take the performance hit that comes with moving up to 24 bits.
Converting a 16-bit file to 24 bits will sound exactly the same as the original 16 bit file. You won't directly benefit from the extra bit-depth, but it won't sound worse, either. The lower 8 bits will just get stuffed with zeroes since there isn't any dynamic info available below 16 bits.
SoNuSmAn (hee hee) covered a lot of the whys for doing the conversion anyway. If two 16-bit tracks are mixed together, there is math going on in your software. That math is going to yeild a result that is bigger than 16-bits (multiply a couple of 4 digit numbers together and you can't store the result in a 4 digit number... same deal). This is where rounding errors will creep in. Mix tracks together, change volume levels on tracks, use plug-ins... all these things contribute to rounding errors and a resulting loss of quality.
Now 24-bit files are going to have these rounding errors too, but the dynamic range of 24-bit audio is such that the errors become insignificant. So converting the original 16-bit files to 24-bits before running 'em through mixing software makes sense if you have more than a couple of tracks and you can take the performance hit that comes with moving up to 24 bits.
you are right Ed, they aren't going to sound worse... (I was just trying to frighten him a bit... 
) but am I wrong about keeping the original files at 16 bit, and just doing the mix with the 24 bit tracks at 24 bit?
The DSP isn't actually going to "sound" better, it's just that after the DSP, the additional word length will be dithered down to 24 insted of 16, thus you'd have less bits truncated after the dsp... in which case the DSP will sound better.
BUT!... if he takes my advise he's not going to truncate back down 16 bit's is he? he will still be taking those 16 bit files, adding DSP, then mixing down to 24. what is the point having those 16 bit files at 24? It really isn't going to make any difference at all. IT will however hogg is CPU, slow up his hard drive... etc.. ect... 16 bits + 8 useless bits isn't going to sound better than 16 bits! you still got 16 bits!
I could be wrong. I could misunderstand what happens when you take a track or a number of tracks in a multitrack program and mix them down to another track. Logically, to me it seems the DSP applied to a 16 bit track but "mixed" as a 24 bit file would sound the same as a 16bit file converted to 24 bits "mixed" as a 24 bit file.
Someone please correct me if I'm wrong!
but really what the hell happens exactly when you mix down a project? I've assumed (until now that Ed's got me thinking about this) that...
a) all the DSP and what not would simply be summed and go to the "master" buss and the file is written from there... or is it that..
b) all of the DSP on each track is applied to each track and the sum of that goes to the "master" and gets written. or...
c) something else?
if it's a, I don't see any need of (what's the word, not resampling, not dithering) er.. changing a few 16 bit files to 24 bit in a project.
if it's b, then yes. change them to 24
if it's c, god help me!
SO now I've changed my mind, unless someone can say for sure exactly what happens when you mix down a project in a multitrack program, Save them as 24 bit files to be sure.
Damn I have a headache after this! sorry guys...
-jhe
) but am I wrong about keeping the original files at 16 bit, and just doing the mix with the 24 bit tracks at 24 bit?The DSP isn't actually going to "sound" better, it's just that after the DSP, the additional word length will be dithered down to 24 insted of 16, thus you'd have less bits truncated after the dsp... in which case the DSP will sound better.
BUT!... if he takes my advise he's not going to truncate back down 16 bit's is he? he will still be taking those 16 bit files, adding DSP, then mixing down to 24. what is the point having those 16 bit files at 24? It really isn't going to make any difference at all. IT will however hogg is CPU, slow up his hard drive... etc.. ect... 16 bits + 8 useless bits isn't going to sound better than 16 bits! you still got 16 bits!
I could be wrong. I could misunderstand what happens when you take a track or a number of tracks in a multitrack program and mix them down to another track. Logically, to me it seems the DSP applied to a 16 bit track but "mixed" as a 24 bit file would sound the same as a 16bit file converted to 24 bits "mixed" as a 24 bit file.
Someone please correct me if I'm wrong!
but really what the hell happens exactly when you mix down a project? I've assumed (until now that Ed's got me thinking about this
) that...a) all the DSP and what not would simply be summed and go to the "master" buss and the file is written from there... or is it that..
b) all of the DSP on each track is applied to each track and the sum of that goes to the "master" and gets written. or...
c) something else?
if it's a, I don't see any need of (what's the word, not resampling, not dithering) er.. changing a few 16 bit files to 24 bit in a project.
if it's b, then yes. change them to 24
if it's c, god help me!
SO now I've changed my mind, unless someone can say for sure exactly what happens when you mix down a project in a multitrack program, Save them as 24 bit files to be sure.
Damn I have a headache after this!
sorry guys... -jhe
Damn Ed and pglewis got back here befor I finished that one...
why do you guys always make me put my foot in my mouth?
So no mismatch in Cakewalk in bit depth? yeah never mind..... convert them to 24...
I'd still like a response to my reply though If anyone can make head or tails of it!
-jhe
why do you guys always make me put my foot in my mouth?
So no mismatch in Cakewalk in bit depth? yeah never mind..... convert them to 24...
I'd still like a response to my reply though If anyone can make head or tails of it!
-jhe
Wow....... Well, I have my answer, I certainly appreciate the detail, I kind of already knew the pro-cons and other small details of using 24 bit over 16bit. I actually do everything at 24 now anyway when it is a new project, I guess I was more concerened what would happen to the original 16 bit tracks if I did a "convert audio" in cakewalk to 24bits to match the new tracks.
So convert it is, I will do it tonight after and report if there are any problems
Oh and Ed, this is the tune you recently reviewed for me. Its funny because I hadn't listened to this tune through my new monitors since I bought them. The problems you heard were that much more apparent then when I listened to it, say in the car.. Makes me feel like Im getting my monies worth...
Thanks everyone...
So convert it is, I will do it tonight after and report if there are any problems
Oh and Ed, this is the tune you recently reviewed for me. Its funny because I hadn't listened to this tune through my new monitors since I bought them. The problems you heard were that much more apparent then when I listened to it, say in the car.. Makes me feel like Im getting my monies worth...
Thanks everyone...