Ableton Forum

Hi

i’ve read a lot about bit depth but am still confused about the facts. I will do my best to explain the way i understand how it works.

The following questions sum up the the way i've perceived all the info on the subject but you see they are still posed as questions.

Please help me to understand the advantages of going Deeper n how it works.

16 bit is cd quality if you convert this to 24 or 32 bit you will have no improvements on the sound quality?

You will however add zeros to the end of the file 234100 or 23410000 making the file itself larger and easier to manage by effects processors and the like?

If you start with 16 bit and pass that signal thru a sound processor and save the processed wave in 24 or 32 bit the new file will not have zeros added to the end but fill the space with new data providing Deeper sound that will have sonic improvements as well as improved processing abilities?

Forgive me if this is obvious i'm only into tech house the other tech stuff is out of my depth for now.

Thanks

J

you are 100% correct

64 bit music is indeed 4x better than old fashioned 16 bit music

However, we should all be looking forward to 128 bit music, which is just around the corner

The dynamic range (amplitude) of the content in a 16bit file saved out at 24bit, will not gain any extra dynamic range. However, The new file will have the ability to contain content with the higher dynamic range that 24bit provides. So if you want, you can use an expander to increase the dynamic range of the 16bit material once it's placed in a 24bit workspace/file.

16bit has 96dB of dynamic range.
24bit has 144dB of dynamic range.

96 dB dynamic range is like hearing a mouse squeak from 15 feet and a jumbo jet from 20 feet in the same time segment

Thanks globalgoon & shuutobi

@ shuutobi i will have to read up on the expanders but lets say i have a 16 bit track and i put a lil compression on it, then export as a 24 bit file.

will that new file will have the 24 bit dynamics?

where as a converted file 16 to 24 bit would only have the dynamics of 16 bit with the size of 24 bit?

does the quality of those dynamics depends on the quality of soundcard?

i'm running an acer aspire 5738zg 32 bit 7 installed with an apc 40

i bought a fastrack pro and took it back because if i ran loops i had distortion on the loop points tried everything with the guy at the shop and we couldn't get it to stop.

i will mainly dj with the apc but as i learn production skills i will mix in some live to my sets

what sound card will be good for my live setup?

Thanks Again

J

soundcloud.com/injoy

The way I look at it (and this is very oversimplified), is that the higher the bit depth, the lower the noise floor. So I record everything in 24 bit to allow for plenty of headroom without having to worry about the parts being too quiet, because they still sit well above the noise floor.

Then after everything is boosted in volume at the mastering stage, it's ok to go down to 16 bit CD quality because it's all loud enough that we don't need such a low noise floor.

As I said, this is hugely oversimplified but this logic has served me well so far without having to get into the serious migraine stuff of bit depth...

When you look at a waveform, if you zoom in at the max, you'll see it's in fact segmented. Think the bit depth as the minimal step in the Y axis, while the sample rate is the minimal step in the X axis.

Thanks chapelier fou & Cezband

both posts shed more light on the subject.

J

chapelier fou wrote:When you look at a waveform, if you zoom in at the max, you'll see it's in fact segmented. Think the bit depth as the minimal step in the Y axis, while the sample rate is the minimal step in the X axis.

This actually explained a lot to me Thanks!

chapelier fou wrote:When you look at a waveform, if you zoom in at the max, you'll see it's in fact segmented. Think the bit depth as the minimal step in the Y axis, while the sample rate is the minimal step in the X axis.

it more or less true but it a dangerous explanation. you get a wrong picture from the sample rate with this!

it's hard to understand when you don't get the math behind that, but it's very important to know that you can recreate a signal perfect up to half of the sampling frequency (due to the not perfect filter a but less... but this is only a side note).

you absolutely don't gain more precision for a signal which has no content abouve 10 khz with 96 khz sampling frequency compared to 44 khz sampling frequency (note: 10 khz is just a example not the actual highest value for 44 khz sampling frequency). you just can represent higher frequencies with a higher sampling rate.

ps: writen by phone so don't mind spelling....

chapelier fou wrote:When you look at a waveform, if you zoom in at the max, you'll see it's in fact segmented. Think the bit depth as the minimal step in the Y axis, while the sample rate is the minimal step in the X axis.

that's a different view but a pretty awful explanation.


when analog is turned into digital the digital representation has a range of numbers it can use to describe the analog waveform.

think in terms of base 10 first (our regular daily way of counting, 0, 1, 2, 3, 4...100, 1,000). if we have 1 digit to work with we can describe the analog waveform with number from 0 to 9. if we have two digits we can use 0 to 99.

the more digits the larger the spread between those numbers, this is known as 'dynamic range.' the higher the dynamic range the smoother the result is. it also means that the largest signal can be that much further from the noise in the lower numbers due to rounding errors, dither and plain old hiss.


then apply that to binary. if you have an 8 bit system you have 2^8 numbers to work with, if you have 16 you have 2^16 numbers to work with etc.

more intersting ar the other effects of a higher bit depth that are not so easy o explain like..why does a mix soundmore open and airy on a higher bit depth ?

dynamic range allows a lower noise floor, noise suffocates a mix, high SNR lets a mix soar. hence gain staging makes a big difference.

but that's in the hands of the user.

my $0.02.

a mix without noise is like a girl without hair

reminds me of the scene in 24 Hour Party People where Martin Hannett is recording the air on a hillside.