recording thru 32 bit AD/DA

audio school
Posts: 4
Joined: Fri Feb 13, 2004 11:33 am

recording thru 32 bit AD/DA

Post by bsd » Fri Feb 13, 2004 12:22 pm

According to Steinberg's Cubase manual, recording and processing at 32 bits, 'virtually eliminates' any digital distortion. What are your experiences with 32 bit recording vs. 24 bit recording? What AD/DA's will do 32 bit recording well at both 48hz and 96hz sampling rates.
In the tape-op interviews that I have read so far, those who use digital processing and recording, do it at 24 bits. [Aside from the fact that 32 bit recording uses a lot more disk space than 24 bit: Why? Why not 32 bit? ]Does anyone one know of any interviews or articles where someone has experience with 32 bit recording & processing?
Right now, I have cheaper 16 bit AD/DA hardware and a computer that is quite short of processing and storage capacity. I intend to replace this system with a well researched system, but I would like to know how I can proceed with what I got now. If I have any more than a couple of tracks at a time in an open project, my computer crashes. I want to keep all of my tracks separate until its time to mix them. Are there any tricks I can do to get around this 'wall.'
Ben Drozdowicz
Note that I am only able to go online about once a week; usually Saturdays.

gettin' sounds
Posts: 103
Joined: Wed Oct 29, 2003 1:18 pm
Location: calgary alberta canada

Re: recording thru 32 bit AD/DA

Post by toddjal » Sun Feb 15, 2004 4:23 am

the first solution would be to post your questions in the right topic.

second, there is no such thing as a 32 bit ADC or DAC. infact there is no true 24 bit either. apparently all those 24 bit converters are actually 23 point something. the only way to achieve true 24, i was told, was at absolute zero temperatures which is -256 degrees celcius i think. what programs like cubase are doing is oversampling at 32 bit so that when any calculation is made, for something like a fade out, the calculations won't have to round up or down in as large increments. thus a closer approximation of the actual curve is achieved. the output is then dithered down internally. it's called "32 bit floating".


Who is online

Users browsing this forum: No registered users and 1 guest