why bother with 24 bit

[Nick The Man]

whats the point of recording in 24 bit? when you put it on a CD it becomes 16 bit anyways so why waste space on your computer and i think its hogging soem of my processor speed too.

Same with sample rates higher than 44.1. Why bother? when its on a CD its reduced to 44.1. the human ear can only hear up to 20khz anyways.

i must be missing something?

the reason i bring this all up is because ive noticed that when i record at 24 bit my computer seems like a slug and i get many pops. if someone thinks they can help me, ill go into better detail on what my computer actually does.

 

[cusebassman]

Well, it would seem that recording at the highest possible quality means that when the day FINALLY comes that we move to a higher quality medium than CD, that the recordings made years ago will still be up to the quality of what is then available.

Let's say a new format comes out that reproduces music at 24+ bit, at 96kHz samplerate... it will be more noticeable when material is released that was recorded at 16 than the stuff that would be getting produced in the current time period. Of course, the layman probably won't notice the difference. I record to DAT tape decks at 16-bit, and the sound comin out sounds pretty much the same as what's going in (if you do a blind test between running my mic & pre through the DAT and simply running it straight to the analog board and listening back), so I partly agree with you.... but, time marches on, and unless your machine is fairly underpowered, I think most of what gets built these days can handle recording at 24-bit.

Not to say that your machine being a little older is bad - computers are damn expensive. And, in the long run, does it matter to you personally that you record at 16-bit? If it sounds decent, then why worry I obviously don't, although I would like to upgrade to an HD24 hard disk machine, which in fact records at 24-bit... *hmm*

In my mind, this all sums up to forward compatibility / quality, and that's about it, since most of the music we record either gets dumped to CD or mp3 anyway (for the time being).

 

[boogle]

As far as bit depth goes, 1) It gives you larger dynamic range, taking you further away from the nasty digital noise floor. 2) that larger dynamic range also allows for your processing to have more "room" to work. It is the equivalent to having a large canvas to paint on that you then have to shrink down, verses trying to do your entire painting on a small canvas. As far a sample rate goes I have run into a lot of problems down sampling form 48k to 44.1 or 96K to 44.1. Those are really hard on a computer, but if you can get 88.2 the downsampling is simply dividing everything by 2. Very easy for a computer to do. Even though we can only hear up to 20K (usually less) the higher the sampling rate the more accurately the higher frequencies will be captured. This helps prevent that square wave sound you can get from 44.1.

Your sampling rate is basically how many pictures your converter takes per second of your audio file, the more pictures it gets the better the resolution.

 

[kessel]

The sampling rate given by the nyquist theorem must be at least twice the highest frequency that you want to capture. So 20 KHz means a minimum sampling rate of 40 KHz. However, the minimum sampling frequency is not usually desirable since aliasing can still occur. I don't notice much audible difference between 44.1 KHz and 48 KHz though. As far as bit depth there is an audible difference between 16 bit and 24 bit because of the large difference between the possible combinations. 16 bit yields 65,536 combinations. 24 bit yields 16,777,216 combinations. Like boogle said, its best to record at higher bit depths and then dither than record at lower bit depths in the first place.

 

[Robert D]

No argument from me on the sample rate, but the extra bits are well worth it. It's not that you can't make a good recording with 16 bits, lot's of good recordings were done that way. But it's much harder to achieve and maintain good dynamic range and stay well above the noise floor with 16 bits. So unless you're really careful, all the way through the process, one side or the other is likely to be compromised.

 

[mrface2112]

well, for starters, 24bit sounds better to my ears than 16, and the longer i can put off having to listen to 16bit, the better (and the happier i am).

as stated, recording in 24bit affords you more headroom due to the increased bit depth. that's the biggest "win".

and depending on what sampling rate you use, it could also mean that your projects will be able to be remixed for the next change in format.

i try to avoid resampling and i know all of my stuff will end up on cd, so i record at 24/44.1. just stick a good dithering plug on your master bus when you bounce to disk and be done with it.

it's totally possible your machine can't handle 24bit. or it's possible that you just need to change your buffer sizes. either one could cause pops and crackles. or maybe it's the cereal you're eating.


cheers,
wade

 

[HangDawg]

but if you can get 88.2 the downsampling is simply dividing everything by 2.

I don't believe that is true. It was explained here before and it's not that simple.

 

[boogle]

I will dig around the board and try to find some info on how the decimator actually downsamples.

Another really important point that has not been brought up is the filter used in the d/a conversion.

Strong arguments have been made that the sample rate is not as important as the filter samplers must use to eliminate aliasing. With a great filter I do not believe many if any people can tell a diff between 96 and 44.1 (several blind test have been done to support this).

The advantage to 96 is you do not need near as steep of a filter as you do at 44.1.

No one is arguing that their is a noticeable difference between 16 and 24 bit. It is not huge, and tons of great recordings have come out at 16 bit, but it is noticeable to a trained or at least exercised ear. As far a sampling rate goes we are talking much smaller differences here. Personally I track at 44.1. No down sampling and I am pleased with my results (or at least my converters results)

 

[NL5]

I don't believe that is true. It was explained here before and it's not that simple.

It's absolutely not true.

 

[Reggie]

As far as bit depth goes, 1) This helps prevent that square wave sound you can get from 44.1.

Your sampling rate is basically how many pictures your converter takes per second of your audio file, the more pictures it gets the better the resolution.

Ha.
The square wave sound happens when people clip over 0dbFS; not due to sample rate. And the more samples per second = the faster (higher) frequencies you can capture; not necessarily better resolution of a sound.

On the 16 bit thing, I wouldn't worry too much about it. Whatever your computer can handle without hindering you. It won't be a deal breaker as far as getting a decent mix, especially if you are mixing in a 32-bit float program or something. And again, especially if you are just doing rawk recordings for the kids.

And one more thing, you don't really get more headroom with 24-bit; you actually get more footroom. You can maintain resolution of lower volume sounds and so therefore it is not as critical to record things so close to the limit -- zero db. Depending of course on the dynamic range of your converters' analog path as well.

 

[beezelbubba]

I bought my Audiophile 24/96 to launder some drug money!
I was perfectly happy with my AC97!

 

[snow lizard]

the reason i bring this all up is because ive noticed that when i record at 24 bit my computer seems like a slug and i get many pops. if someone thinks they can help me, ill go into better detail on what my computer actually does.

Sometimes the popping and stuff can be helped by adjusting buffer and cache sizes to help accomodate the task of recording. There should be settings in your recording software as well as your operating system that you can tweak. Your manuals should help to explain this. Default OS settings aren't aimed at smooth audio recording.


sl

 

[Robert D]

And one more thing, you don't really get more headroom with 24-bit; you actually get more footroom. You can maintain resolution of lower volume sounds and so therefore it is not as critical to record things so close to the limit -- zero db. Depending of course on the dynamic range of your converters' analog path as well.

I think that depends on how hot you record. If you hit the meters near 0 like people used to do with 16 bit, then yes, you've gained no headroom, just footroom. But if you record down at -18 dbfs, then you've gained headroom.

 

[Nick The Man]

well, for starters, 24bit sounds better to my ears than 16, and the longer i can put off having to listen to 16bit, the better (and the happier i am).

as stated, recording in 24bit affords you more headroom due to the increased bit depth. that's the biggest "win".

and depending on what sampling rate you use, it could also mean that your projects will be able to be remixed for the next change in format.

i try to avoid resampling and i know all of my stuff will end up on cd, so i record at 24/44.1. just stick a good dithering plug on your master bus when you bounce to disk and be done with it.

it's totally possible your machine can't handle 24bit. or it's possible that you just need to change your buffer sizes. either one could cause pops and crackles. or maybe it's the cereal you're eating.


cheers,
wade

hmmm buffer sizes .. this is something i havent come across yet how do i change these?? and what are they?

 

[SouthSIDE Glen]

I think that depends on how hot you record. If you hit the meters near 0 like people used to do with 16 bit, then yes, you've gained no headroom, just footroom. But if you record down at -18 dbfs, then you've gained headroom.

Simply put, 24 bits offers 48dB more digital dynamic range to work with than 16 bit does, and - unlike 16 bit - actually gives you enough room to guarantee recording your full analog signal from noise floor to peak without having to worry about clipping.

We may have to crop our sonic picture down to 16 bit at the end, but we need the extra space during tracking and mixing to be able to work with the entire image and decide just what gets into the crop and what doesn't.

G.

 

[Nick The Man]

ha .. figured it out, i was running 72 plugins at the same time ...


i had 23 tracks and each track had 3 plugings on it.. my defults got messed up.. thanks for your help.. i feel like im havinga blonde moment or something.

new problem that ive run into .. i dont think its anything to do with bit rate and such like that but here it is.

when i pan things there is a problem, its loud and normal in the center. but then i pan it 100% left and it is all in the left but its really quiet... and then like 50% left its like centered but just quieter .. i can figure it out.. any ideas?

 

[danny.guitar]

ha .. figured it out, i was running 72 plugins at the same time ...

My computer would explode with that many.

If yours was just being "sluggish" then I wish I had your computer

 

[Nick The Man]

My computer would explode with that many.

If yours was just being "sluggish" then I wish I had your computer

well i just built it its all custom i spent about $800... i guess i put it to the real test the past few days.. but really they were just simple plugins..

every track had a compressor, graphic EQ, and a noise gate.

 

[Farview]

1. Those are really hard on a computer, but if you can get 88.2 the downsampling is simply dividing everything by 2. Very easy for a computer to do.

I wish this myth would die already. This isn't true. Any time something is resampled, it is upsampled into the MHz range and then downsampled. You've got to remember that your converters are actually sampling a lot faster than the sample rate (oversampling). That gets paired down to the sample rate that gets stored on the computer. There is a theory that super high sample rates (192k) are actually less accurate because each sample is based on less information.



Anyway, back to the original question. As has been said, 24 bit keeps you away from the noise floor. This gets you quieter recordings (noise-wise) and allows you to record quieter. That keeps you from overtaxing your preamps and accidently clipping your converters.
another thing is how it all gets mixed together. When you mix, you are adding the noise floor of all the tracks together as well. 32 tracks with a noise floor of -96 is going to be noisier than the same number of tracks with a noise floor of -148db.
One more thing, every time you process something there will be rounding errors. Using 24 bit files will push the rounding errors farther off into the background.

Of course none of this makes any difference if your computer can't handle recording 24 bit files. You will have to record how ever you can until you can upgrade.

Thinking about this, I had a computer 10 years ago that could handle 24 bit files. There might be a configuration problem with yours (or your trying to record 24 tracks at once) because it really doesn't take that much power by todays standards. (or your still running a windows 95 machine)

[edit]I took a long time to type this, I see you figured out the computer problem.

 

[danny.guitar]

I've had this computer for about 7 years. It handles 24-bit/44.1 fine unless I put some big plugins on the tracks (GlaceVerb, Waves RVerb are both CPU fiends for mine). Maybe not for yours though.

I can seem to run about as many 'normal' EQs, compressors, etc. as I want.

If you're getting clicks/pops, make sure you're using the ASIO drivers in your software, assuming your interface comes with ASIO drivers (it should). If it doesn't, check out www.asio4all.com.

If that doesn't work then tweak your buffer settings like other people said and close down any other programs that are running.