24 bit recording.........Why ?

  • Thread starter Thread starter Torpid-x
  • Start date Start date
Torpid-x

Torpid-x

Member
Hey all

I read his somewhere and I don't know if it's true,(I think it was a Home Recording.com article) if you record in a higher than 16 bit 44.100kbs resolution, you will have to convert to 16 bit to put it on CD. Also said that you'll be glad you recorded in 16 bit in the first place. My questions Are: is there really an audible difference???, What type of recording would benefit the most from 24 bit (i.e. Guitar, Voice, MIDI, etc.).
My opinion (not having recorded in 24 bit yet) is, 16 bit 44.1 is accurate, if they standardize CD's to 24 bit or higher, this would cause the consumer to have to buy new CD players to read the new format to achieve minimal audible benefits.

You cut my hair, you die!!!!!
Rich
 
Here's a little background on digital recording. Each sample is a "snapshot" of your sound at a particular moment in time. At 44.1kHz, there are 44,100 samples per second. The bit depth is how many bits are used to represent each sample.

With 16 bits, you have 2^16 or 65536 different values you can hold. You can think of it as 0 being silence and 65535 being the loudest. With 24 bits, you have 2^24 or well over 16 million different values you can hold.

The real value of the extra 8 bits is at the "bottom of the bit-bucket" near and into the noise floor. With quieter sounds you're using fewer bits to represent the sample (the upper bits are just set to 0). With 6-bits, you only have 64 different values to work with. The smallest relative volume change you can make will be 1/64th. The quieter the sound, the fewer bits you work with and the more coarse the volume adjustments will be. Listen closely to slow 16-bit fades and you'll hear this.

With the extra resolution of 24-bits, you'll have more bits to work with near the nasty noise floor and thus you'll be able to make much more accurate adjustments to the volume there, if necessary. This adds up a lot when multi-tracking. Panning changes the volume levels in each channel, so it stands to reason that your stereo field can be much more defined as well.

And now the short answer: YES! Use as high a resolution as you can afford on your source tracks and while mixing. Shit flows downhill. Clobber the sound too much while tracking and mixing and nothing you do in mastering will be able to fix it.
 
Thanks for the info pgl

I understand what you mean, I'm glad someone could explain so clearly the benefits are. You probably saved me from remaking a lot of tracks. I'm new at this and I don't have anything serious recorded yet.
One more question. what do you do about mixdown, just convert the finished mix?

Rich
 
Yes, after you've done all your processing at the highest bitdepth and resolution possible, convert everything down to 16-bit/44.1 kHz so you can burn it to CD.

Unless, of course, you're planning to ship your mix out to a mastering lab, in which case leave it at the higher resolution - they can handle it and know what to do with it, but check with them first.

[This message has been edited by dobro (edited 07-02-2000).]
 
I'll let you in on a little secret: everything I've recorded up to this point has been 16-bit. It's all I can afford to use until I get a new drive. But understanding the weaknesses of 16-bit recording has helped me keep my mixes a lot cleaner lately.

You do know to use dither if and when you go down to 16/44.1, right? Make sure you don't just lop those extra bits off.
 
Actually pglewis, depending upon how far down your noise floor is on the finale mix, dithering may be a mute point. 16 bit D/A converters only have about 92-96 db sound to noise ratio. If your mix already has noise at that level or above, it would dither itself really. I just caught that little tidbit on www.digido.com, and some of my experiments have supported this theory. It is where you dynamic range AND s/n ratio is better then 92db that you have to start thinking seriously about dithering. Dithering is much more important if you are dealing with 16 bit files too because you do lose some bits from DSP, so then extended work length would start to create noise above this level.

Anyway.......

Ed
 
I forgot to say that pglewis offered an excellent essay above (sort of essay, sort of short, which is what made it great) about how 24 bit is better for DSP.

Well said friend.

Ed
 
hey everybody,

thanks to the discussons here and those digido.com articles, i'm fully convinced of the usefullness and rationale of 24-bit recording with the highest sampleing rate possible. unfourtunately i only semi understand the technical stuff(but i'm getting better)...so here's my question

My Dman 2044 has 20 bit A/D but will only write a 16 bit file to disk, i think it truncates or dithers on the card.

So when mixing and editing in cakewalk the wordlenths expand (a la digido.com article). Would my mixing and editing benefit from saveing the files as 24 bit files despite the fact that my audio is 16-bit. I know the audio won't magicly recover bits, but will any changes made to the audio be more accurate?

hope that made sense,

rhoadz
 
Rhoadz: I'm not sure in that situation... I'm sure someone can probably field that one. I'm still learning a lot of this stuff myself. I have a bit-twiddling background, so I understand the theory but you should know that I'm not experienced in the "real world" with 24-bit recording. My ears have yet to pass judgement there. I can tell you with confidence that bit-width makes a huge difference with DSP. I'd focus on avoiding crappy DSP before sweating too much over 32-bit/192kHz recording.
 
You must log in or register to reply here.
Back
Top