Author Topic: 24bit to 16bit (Read 7389 times)

creekfreak · « **on:** August 28, 2004, 12:01:31 PM »

Ok, learning the hard way here. What extra step do I need to take when going from 24bit down to 16bit? I ran my 24bit recording hot, so when I went down to 16bit I was clipping left and right since I had no headroom..what is the best way to fix this?

F.O.Bean · « **Reply #1 on:** August 28, 2004, 05:20:59 PM »

hmmm, you mean if you run hot in 24 bit and dither down to 16 bit it is hotter

creekfreak · « **Reply #2 on:** August 28, 2004, 07:00:52 PM »

when I dithered it down I got a ton of pops and clicks, levels jumping through the roof. I saw a converation on this board that if you plan to convert to 16bit that not to run the levels too hot,give your self some headroom.

dklein · « **Reply #3 on:** August 28, 2004, 07:21:42 PM »

That sounds weird - were you clipping in 24 bit? The translation from 24 bit to 16 bit shouldn't change the levels, just the accuracy. The only way it should top out in 16 bit is if it did the same in 24 bit. Unless you did some post processing that pushed it beyond full scale. You could drop the levels by 0.2dB in the 24 bit (maybe use something like L2 so you can do that and dither to 16 at the same time).

caymanreview · « **Reply #4 on:** August 28, 2004, 08:32:53 PM »

i run mine as hot as possible w/out clipping

dither to 16bit with wavelab 5 using the uv22hr and have never experienced a problem like this

what are you using to dither?

creekfreak · « **Reply #5 on:** August 28, 2004, 10:51:24 PM »

that might be the problem, I am just resampling, not dithering

caymanreview · « **Reply #6 on:** August 28, 2004, 11:03:32 PM »

i dont understand how resampling to 44.1 or whatever you are resampling to would cause clipping though

creekfreak · « **Reply #7 on:** August 29, 2004, 08:42:44 AM »

I am trying to find the thread, Tim was the one that had a bunch of info on this. I was not clipping in 24bit, the 24bit sounds great, but when I resample to 16bit it just becomes filled with pops and clicks, and watching the wave form, its now clipping....Tim you out there? I think you mentioned that this will happen due to the lack of headroom that 16bit has over 24bit

caymanreview · « **Reply #8 on:** August 29, 2004, 09:09:58 AM »

hmmm, ive never heard anything like that

ive always ran my levels fairly aggressive lately in 24/48 and 24/96 and have had zero issues with this

actually just resampled/dithered/tracked umphreys from friday night and ran my levels as hot as possible, and have zero issues like that with the recording

what are you using to dither/resample?

wavelab???

creekfreak · « **Reply #9 on:** August 29, 2004, 12:10:22 PM »

I an using a program called goldwave, and I think the fact that I am not dithering is the problem. When I do a straight up resample with no dither going from 24--16bit (which I just tested) the wave form before and after looks jacked....need to see if godwave will dither also.

dmonterisi · « **Reply #10 on:** August 29, 2004, 12:11:24 PM »

it sounds like the resampling method in that program is jacked...do you have wavelab?

creekfreak · « **Reply #11 on:** August 29, 2004, 12:18:06 PM »

No, I do not, but I can get it. I emailed the goldwave people to see if their resampling is just a flat chop, or if it actually uses a dithering method, my guess is not.

Tim · « **Reply #12 on:** August 29, 2004, 01:17:02 PM »

Quote from: dmonterisi on August 29, 2004, 12:11:24 PM

it sounds like the resampling method in that program is jacked...do you have wavelab?

what Damon said.

I was talking strictly about recording. Once you've laid it down changing from 24 down to 16 shouldn't be causing you these problems.

Sorry if I confused you!

MattD · « **Reply #13 on:** August 30, 2004, 11:18:27 AM »

Vocab time:

Resampling = changing the sample rate (96 kHz to 44.1 kHz)
Dither = changing the bit depth (24-bit to 16-bit)

I'm having a hard time following what you're using to try to accomplish whatever it is you're doing.

creekfreak · « **Reply #14 on:** August 30, 2004, 11:49:38 AM »

I am trying to take a 24/96 recording down to 16/44.1, and I think I found out the problem. When going from 24 to 16bit I don't think the program I am using has all that good of a dither, if any that works well.

caymanreview · « **Reply #15 on:** August 30, 2004, 11:51:51 AM »

if you need a wavelab 5.0 "demo" pm me

BC · « **Reply #16 on:** August 30, 2004, 11:52:22 AM »

just FYI:

NEITHER resampling or dithering should cause any changes in the levels or pops/clicks in the recording.
I think something is wrong with the software you are using.

Take care,
Ben

creekfreak · « **Reply #17 on:** August 30, 2004, 12:12:09 PM »

something is wrong.....never had this issue before, but then again, never did 24bit before either

rustoleum · « **Reply #18 on:** August 30, 2004, 01:59:21 PM »

It sounds like you are simply chopping off 8 bits to move the file from 24 bits to 16... the end result would be a rather harsh recording in some parts since the low-level detail is simply cut off... When dithering, those low-level sounds are preserved by essentially pushing pieces of them into the 16 bit world. Dither your 24 bit recording (as oppossed to just saving it to a 16 bit file, which it sounds like is what you're doing) and I think all will be better.

creekfreak · « **Reply #19 on:** August 30, 2004, 02:26:43 PM »

Quote from: rustoleum on August 30, 2004, 01:59:21 PM

It sounds like you are simply chopping off 8 bits to move the file from 24 bits to 16... the end result would be a rather harsh recording in some parts since the low-level detail is simply cut off... When dithering, those low-level sounds are preserved by essentially pushing pieces of them into the 16 bit world. Dither your 24 bit recording (as oppossed to just saving it to a 16 bit file, which it sounds like is what you're doing) and I think all will be better.

I think you nailed the problem, that is exactly what happened.

sygdwm · « **Reply #20 on:** August 30, 2004, 03:58:58 PM »

fwiw. tim also told me to run pussy levels w/ 24 bit to leave headroom. crank the the shit out of your levels at 16 bit. i think.

Tim · « **Reply #21 on:** August 30, 2004, 04:05:35 PM »

where's that thread?

I think the confusion about where to set 24 bit levels was cleared up a bit.

16bit you need to run hot. 24bit I was always told to run a bit lower...

Tim · « **Reply #22 on:** August 30, 2004, 04:07:30 PM »

here's the thread...

http://www.taperssection.com/yabbse/index.php?topic=19485.90

start around page 7 and you'll see the relevant info on levels, posted by folks who are much smarter than I

willfoy · « **Reply #23 on:** August 30, 2004, 04:15:30 PM »

But maybe I'm getting the wrong message from the thread, but that is you can lower your levels for greater headroom, but not that you should lower your levels at 24 bit...

At any rate, after spending most of the last 18 months of my taping life in 24bit land, the pops and clicks are probably based on improper dithering. I use WaveLab to do mine, but Goldwave or SoundForge should do the job. CDWave's current incarnation does it as well, but I'm not pleased with the results in current production releases... Perhaps they'll improve over time.

rustoleum · « **Reply #24 on:** August 30, 2004, 04:28:27 PM »

Simply put, the digital world allows 6 db of dynamic information per bit. So, a 16 bit recording allows 96 db of info, while a 24 bit recording allows 144 db. Sooooo... when folks say you don't need to push the levels in 24 bit recording, this is simply because even if you're peaking out at -18 db you're still capturing 126 db of info which is still a higher resolution than a perfectly set 16 bit recording.

You should still shoot to set levels perfectly as the more detail in the recording the better, but if it's a question of setting levels higher and potentially clipping the recording vs setting levels conservatively I'd go with a conservative set.

creekfreak · « **Reply #25 on:** August 30, 2004, 04:39:32 PM »

Quote from: willfoy on August 30, 2004, 04:15:30 PM

But maybe I'm getting the wrong message from the thread, but that is you can lower your levels for greater headroom, but not that you should lower your levels at 24 bit...

At any rate, after spending most of the last 18 months of my taping life in 24bit land, the pops and clicks are probably based on improper dithering. I use WaveLab to do mine, but Goldwave or SoundForge should do the job. CDWave's current incarnation does it as well, but I'm not pleased with the results in current production releases... Perhaps they'll improve over time.

I think that was my problem, I was just saving the file back to 16bit, not applying any dither, will try that later tonight..

BC · « **Reply #26 on:** August 31, 2004, 04:50:35 PM »

Quote from: rustoleum on August 30, 2004, 04:28:27 PM

Simply put, the digital world allows 6 db of dynamic information per bit. So, a 16 bit recording allows 96 db of info, while a 24 bit recording allows 144 db. Sooooo... when folks say you don't need to push the levels in 24 bit recording, this is simply because even if you're peaking out at -18 db you're still capturing 126 db of info which is still a higher resolution than a perfectly set 16 bit recording.

Just FYI: no piece of recording gear comes even close to the theoretical 144 dB dynamic range of 24 bit PCM audio. For example I think the V3 is spec'd around 113 dB. But even with this, if your levels peak at -17dB (113-17=96) at 24 bit, you are getting the same resolution as recording at 16 bit hitting 0. High resolution digital rocks!!!

dklein · « **Reply #27 on:** August 31, 2004, 08:42:15 PM »

Quote from: MattD on August 30, 2004, 11:18:27 AM

Vocab time:

Resampling = changing the sample rate (96 kHz to 44.1 kHz)
Dither = changing the bit depth (24-bit to 16-bit)

I'm having a hard time following what you're using to try to accomplish whatever it is you're doing.

Since we're trying to be precise, dither does not = changing the bit depth
Changing the bit depth is just that. Others call it word length reduction or whatever. It is simply throwing away the least significant bits, kind of like rounding off numbers. You see 13.345679 on the calculator and you say 13.3 (except for that little rounding rule we use). Same thing but in binary.

Dither is a mathematical technique used to capture detail that would not normally be captured at the reduced word length. When you truncate from 24 to 16 bits, you're basically chopping off the last 8 bits. On their own, these are only used to capture the very quietest sounds. During a peak, they capture the finest info about the event, like the last decimal place in the number example above. You can't chop the 8 bits on the other end of the 24 bit word because they cover all of the musical peaks so you chop the fine detail.

Dither involves adding some noise to raise those little bits up to the point where they will have an effect within the 16 most significant digits of the 24 bit recording. Then you chop off the last 8 and you have some of that detail you wouldn't have had without the dither. Noise shaping is about how you choose to apply the dither (out of scope!)

That's the theory on it. With respect to levels, I try for peaks around -6 to leave room for the unexpected. There's absolutely no point in going over.

Quote from: rustoleum on August 30, 2004, 04:28:27 PM

Simply put, the digital world allows 6 db of dynamic information per bit. So, a 16 bit recording allows 96 db of info, while a 24 bit recording allows 144 db. Sooooo... when folks say you don't need to push the levels in 24 bit recording, this is simply because even if you're peaking out at -18 db you're still capturing 126 db of info which is still a higher resolution than a perfectly set 16 bit recording.

True, but 24 bit recording benefits aren't limited to the additional dynamic range in the sense that you can record louder sounds or anything like that. It's more like using a fine grain film or a higher image density in a digital photo. You get a more detailed account of each sampling event.

That said, I have yet to prove its worth to myself. I've been recording in 24 bit for over a year and just kind of do it on faith. I can do 24 bit playback but I'm not sure that live shows are really the way to showcase the sonic capabilities of 24 bit. Right now I can only playback 24 bit through an emagic A62 - I really don't know how good the d>a is. Got something else coming soon but I guess I need to take a show and run A/B on the 16 and 24 bit through the same d>a. hmmm...maybe I'll get on that.
<edited for spelling>

Electric Cowgirl · « **Reply #28 on:** August 31, 2004, 10:29:38 PM »

Creek-- Didn't see what software you are using, but here's some info just for the hell of it.

If you are using Wavelabs open the masters window (view->specialized windows->masters section) apply the dither then click render in the mastering window. This will give you a new file that is 16bit, and allow you to keep your 24bit master with one click. Just in case you or anyone reading didn't know

creekfreak · « **Reply #29 on:** September 01, 2004, 09:53:16 AM »

thanks Becky, just got wave lab (thanks caymanreview!!!), will be trying it out soon

jk labs · « **Reply #30 on:** September 01, 2004, 11:07:13 AM »

Just a quick comment.

Brute force conversion from 24 bits down to 16 by just chopping off the 8 lowest bits, should not result in your problems: the 16 most significant bits would remain the same.

But if the program you are using is applying dither to the trunacation process it could attempt
to add a little bit of "noise" to an already full scale 16 bit sample. Half the time that "noise"
is positive and the full scale sample "overflows".

I think, with some reservations due to a hasty read :-), that the problem here is how your
program deals with adding a small value to a full scale sample. Ideally it should remain full scale
(no change). It sounds as if your program has a diffrent policy on what to do with overflows.

Any chance we could see a display of the samples?

creekfreak · « **Reply #31 on:** September 04, 2004, 05:04:51 PM »

used wavelab, dithered it down to 16bit/44.1, no issues this time, just starting using it, but wavlab is damn cool

caymanreview · « **Reply #32 on:** September 04, 2004, 05:16:42 PM »

creekfreak · « **Reply #33 on:** September 04, 2004, 09:18:01 PM »

yes, very

..thanks again...+T

Marc Nutter · « **Reply #34 on:** September 08, 2004, 08:40:42 AM »

Quote from: rustoleum on August 30, 2004, 04:28:27 PM

Simply put, the digital world allows 6 db of dynamic information per bit. So, a 16 bit recording allows 96 db of info, while a 24 bit recording allows 144 db. Sooooo... when folks say you don't need to push the levels in 24 bit recording, this is simply because even if you're peaking out at -18 db you're still capturing 126 db of info which is still a higher resolution than a perfectly set 16 bit recording.

You should still shoot to set levels perfectly as the more detail in the recording the better, but if it's a question of setting levels higher and potentially clipping the recording vs setting levels conservatively I'd go with a conservative set.

Perfectly said.
Oh course, since many of our recordings are being transferred to 16-bit without level adjustments, only dithering, peaks around -6db for what most of us record are still quite nice.

Still, I'm with you. It's better to be a bit low than the clip.

Marc