Why 24bit?
- Thread starter Mofoman
- Start date
Monkey Mouse
Member
24 bit gives a lot more headroom and a lot higher (in theory) signal to noise ratio. It only takes up about 50% more space than 16 bit, but can really make a vocal or instrument shine through superior quality.
hello, i got kinda similar question, so i'll post it here, hope ya'll dont mind.
is there a point of saving my reason renderings at 96/24, if my refills/soundfonts quality is only 44/16? will it actually improve my sound? i dont think it does, yea probably malstörm and subtractor will sound better, but everything else? what do ya think?
thanks
is there a point of saving my reason renderings at 96/24, if my refills/soundfonts quality is only 44/16? will it actually improve my sound? i dont think it does, yea probably malstörm and subtractor will sound better, but everything else? what do ya think?
thanks
Lord Thathidge
Spectral Disco Pimp
wavez said:
Most professionals will answer quite simply that it is always best to work at the highest quality available even if you are intending to dither to 16bit for cd audio purposes later on.
This is because the after the stage at which you are - the export stage - most engineers will be processing the file further during the mastering stage and it is better to apply mastering effects to as higher resolution file as possible.
Bit depth and sample frequency determine the resolution of an audio file.
At 96KHz, snapshots of an original source signal are taken 96000 times every second. (Kilohertz = "cycles per second"). This is the 96KHz part. If you load a file into soundforge or adobe audition and then use the horizontal zoom to zoom as far into the file as possible you eventually get to an image of the audio file with little squares or blobs periodically placed on it...these blobs are the individual samples. The 24bit relates to the individual amplitudes of each of those samples.
24 bit means that you have 2 to the power 24 (roughly 16 million) individual amplitude levels, allowing each sample to be placed as near to its original amplitude level as possible. At least, nearer than if you were to work in 16 bit (2 to the power 16) which would only give you 32000 individual amplitude levels to work with...less chance of an accurate sample amplitude placement.
I'm not even sure if I'm making any sense any more...I've just finished a progress report for my final year degree project and i'm seeing double...hope I haven't blinded you with science...if in doubt...and if your computers processing power and HD space allow, always work with the best quality audio available.
...to more specifically answer your question, no, the rendered version won't sound superior or indeed any different at all, but your rendered version will have a better (if not at all audible) resolution than if you export at 44/16.
Think of a digital photograph.
The original digital photograph is displayed using 16 million colours.
If you then dither that image so that it is now displayed using a pallette of only 32000 colours - do you think it would appear different to the naked eye? Unlikely...
...same principle...
Last edited:
Tyree D.
Active member
Lord Thathidge said:
you explained this so well! But for this last paragraph, the answer is maybe. Super Nintendo could display 32000 colors while the PlayStation One had 16.7 million. Or if y'all remember 15 years ago PCs displayed 32000 colors. For some full motion graphics (not all), I could see the difference. But the average person won't, which is most important.
Bit Depth:
24bit is preferable to 16bit because of the improved dynamics of your song. the higher the bitrate the more accurate your resolution.
It's like having two sets of volume controls, one that goes
1,2,3,4 (2 bit)
and another that goes
1,2,3,4,5,6,7,8 (3 bit)
If you could only have all your sounds at one of those volume levels it is preferable to have it recorded at 3 bit because then you have 8 volume levels to choose from. More bits means more levels to choose from.
16bit is used by CD, but in this case it is relatively easy to scale down from 24bit to 16bit, and so your recordings are futureproof.
Sample Rate:
The sample rate is how many values are taken every second. There has to be at least twice as many samples per second as the frequency you want to record.
We're interested in sound so lets take a look at the human hearing range. We can hear roughly between 20Hz and 20kHz. So for us 44.1kHz is a nice sampling rate, it allows us to hear everything we can. For those interested in the theory behind this look up Nyquist/Shannon sampling theorem.
So why would we be interested in more samples than necessary? Essentially because you get more information about your sound. Also your recordings sound better as the waves being reproduced are smoother (draw a sine wave using only 5 data points and straight lines between them and then do the same with 10 data points - you'll see that you get a much nicer curve).
It turns out that 96kHz is not the best sampling rate available at the moment for general CD quality sound. CD's, as we all know, use 44.1kHz. If we want to improve the sound it makes more sense to double the sample rate - 88.2kHz. This is because we don't want lots of signal processing manipulating our recordings. Dropping from 88.2kHz down to 44.1kHz is easy. Dropping down from 96kHz to 44.1kHz is much more complicated and requires more maths. There should be as little processing here as possible.
However I remember DVD-Audio can use either 96kHz or 48kHz depending on how many chanels you want to use. So if you are recording for use on a DVD using the 96kHz sample rate is best.
At the end of the day if what you are using can cope with it more is best, but make sure you keep one eye on your desired medium and keep the sample rate the same (or double) the rate you'll need on the medium. Remember: you can always lower the quality at the end, you can never improve the quality without re-recording everything.
24bit is preferable to 16bit because of the improved dynamics of your song. the higher the bitrate the more accurate your resolution.
It's like having two sets of volume controls, one that goes
1,2,3,4 (2 bit)
and another that goes
1,2,3,4,5,6,7,8 (3 bit)
If you could only have all your sounds at one of those volume levels it is preferable to have it recorded at 3 bit because then you have 8 volume levels to choose from. More bits means more levels to choose from.
16bit is used by CD, but in this case it is relatively easy to scale down from 24bit to 16bit, and so your recordings are futureproof.
Sample Rate:
The sample rate is how many values are taken every second. There has to be at least twice as many samples per second as the frequency you want to record.
We're interested in sound so lets take a look at the human hearing range. We can hear roughly between 20Hz and 20kHz. So for us 44.1kHz is a nice sampling rate, it allows us to hear everything we can. For those interested in the theory behind this look up Nyquist/Shannon sampling theorem.
So why would we be interested in more samples than necessary? Essentially because you get more information about your sound. Also your recordings sound better as the waves being reproduced are smoother (draw a sine wave using only 5 data points and straight lines between them and then do the same with 10 data points - you'll see that you get a much nicer curve).
It turns out that 96kHz is not the best sampling rate available at the moment for general CD quality sound. CD's, as we all know, use 44.1kHz. If we want to improve the sound it makes more sense to double the sample rate - 88.2kHz. This is because we don't want lots of signal processing manipulating our recordings. Dropping from 88.2kHz down to 44.1kHz is easy. Dropping down from 96kHz to 44.1kHz is much more complicated and requires more maths. There should be as little processing here as possible.
However I remember DVD-Audio can use either 96kHz or 48kHz depending on how many chanels you want to use. So if you are recording for use on a DVD using the 96kHz sample rate is best.
At the end of the day if what you are using can cope with it more is best, but make sure you keep one eye on your desired medium and keep the sample rate the same (or double) the rate you'll need on the medium. Remember: you can always lower the quality at the end, you can never improve the quality without re-recording everything.
N
neilwight
Guest
look at the sticky at the top of the mixing and mastering thread.
it gives instructions on how to work with 24bit and you should all take a look, absorb what it says and then apply it to your working. your music will thank you.
theres much more to 24bit than the added dynamic range by itself. this huge increase allows you to work with headroom while maintaining huge amounts of quality. it helps the summing in the DAW, helps you mix with clean signals, keep dynamic range in the material itself and also keeps you away from clipping anywhere.
it all adds up to a far better output than working with 16bit which forces you to keep right up against 0dbfs since anything below is eating into quality that will go onto the final CD.
working inside a DAW is not the same as working in the analogue realm, different rules apply and you need to learn to work differently to exploit its potential. the argument that final cd audio is 16 so you are wasting HDD space working above this is so simplistic thats its a non argument at all.
this is a simple summary, check the sticky to get a better idea.
it gives instructions on how to work with 24bit and you should all take a look, absorb what it says and then apply it to your working. your music will thank you.
theres much more to 24bit than the added dynamic range by itself. this huge increase allows you to work with headroom while maintaining huge amounts of quality. it helps the summing in the DAW, helps you mix with clean signals, keep dynamic range in the material itself and also keeps you away from clipping anywhere.
it all adds up to a far better output than working with 16bit which forces you to keep right up against 0dbfs since anything below is eating into quality that will go onto the final CD.
working inside a DAW is not the same as working in the analogue realm, different rules apply and you need to learn to work differently to exploit its potential. the argument that final cd audio is 16 so you are wasting HDD space working above this is so simplistic thats its a non argument at all.
this is a simple summary, check the sticky to get a better idea.
thanks, Lord Thathidge! perfect post!
word! logical! *punching my face for not knowing this!
----------------
one more question tho
so i use cool edit.
now after i 'mastered' my track, i'll go to 'convert sample type', to lower my bitrate. it has a option to click dither on/off. so, on or off? :] and if on, then with what settings would be the best?
thanks
Lord Thathidge said:
word! logical! *punching my face for not knowing this!
----------------
one more question tho
so i use cool edit.
now after i 'mastered' my track, i'll go to 'convert sample type', to lower my bitrate. it has a option to click dither on/off. so, on or off? :] and if on, then with what settings would be the best?
thanks
Lord Thathidge
Spectral Disco Pimp
wavez said:
Lord Thathidge
Spectral Disco Pimp
Dither needs to be on.
When converting to 16bit from a higher bit depth, if you don't dither, then distortion can occur and the dynamic range can be affected detrimentally.
Dithering basically adds random noise to the 24bit audio file before it is converted - noise is broadband (it contains frequencies from the full spectral range) and can be filtered so as to fall mostly in areas of the spectrum where it is least audible.
A little added noise is more aurally acceptable than actual distortion.
As for settings, I don't use cool edit (I favour soundforge personally) so I don't know what the choice is.
If you can give an idea as to the choices then it may help. I assume when dithering there is the choice of dither "types" and "shapes"...perhaps if you list them we could be more specific.
Also, check out some of the material available for readin at the site below.
http://theprojectstudiohandbook.com/directory.htm
Good luck
When converting to 16bit from a higher bit depth, if you don't dither, then distortion can occur and the dynamic range can be affected detrimentally.
Dithering basically adds random noise to the 24bit audio file before it is converted - noise is broadband (it contains frequencies from the full spectral range) and can be filtered so as to fall mostly in areas of the spectrum where it is least audible.
A little added noise is more aurally acceptable than actual distortion.
As for settings, I don't use cool edit (I favour soundforge personally) so I don't know what the choice is.
If you can give an idea as to the choices then it may help. I assume when dithering there is the choice of dither "types" and "shapes"...perhaps if you list them we could be more specific.
Also, check out some of the material available for readin at the site below.
http://theprojectstudiohandbook.com/directory.htm
Good luck
Last edited:
thanks man!
heres a pic of the cool edit's dither settings.
http://com.msn.ee/wavez/dither.jpg?2
and thanks for that link!
heres a pic of the cool edit's dither settings.
http://com.msn.ee/wavez/dither.jpg?2
and thanks for that link!
Last edited:
Lord Thathidge
Spectral Disco Pimp
couldn't get the link to work but these links might help a little.
http://en.wikipedia.org/wiki/Dithering
http://en.wikipedia.org/wiki/Noise_shaping
Also, try searching in cool edit's manual regarding any potential settings. It may recommend specific settings for certain projects.
http://en.wikipedia.org/wiki/Dithering
http://en.wikipedia.org/wiki/Noise_shaping
Also, try searching in cool edit's manual regarding any potential settings. It may recommend specific settings for certain projects.
M
Mr.Alias Pain
Guest
kinda off topic but would i hear any audible improvements if i changed my sample rate to 48 khz in FL?
Lord Thathidge
Spectral Disco Pimp
Mr.Alias Pain said:
Not if you're currently working at 44.1, but if your sound card supports files at this setting then it is good practice to work at as higher frequency as possible.
M
Mr.Alias Pain
Guest
ok so what if i changed my setting in FL to 48 khz, but all of my samples are in 44.1? would i still notice any difference? would i get more headroom in such?
and BTW my sound card is an emu 1212m.
thnks
and BTW my sound card is an emu 1212m.
thnks
M
Mr.Alias Pain
Guest
questions been answered.... thanks