48KHz vs. 44KHz

[dynomike]

Others have said it and I'll reiterate:

If your target is 44.1, record at 44.1. The sonic difference of recording at 48khz is nullified by the math involved in downsampling.

The main reason 48khz was chosen as a standard for DATs, etc, is because the tape could vary in speed slightly and the frequency response (up to 20khz at least) would still be maintained. Not as much room for error with 44.1. Now that few of use DAT, its not such a big deal.

[BradG]

The main reason 48khz was chosen as a standard for DATs, etc, is because the tape could vary in speed slightly and the frequency response (up to 20khz at least) would still be maintained. Not as much room for error with 44.1.

I could be wrong, but I'm almost certain that the 48KHz standard was set up during the SCMS wars to prevent direct CD copying. This is why the very earliest DAT machines to hit the states only recorded at 48KHz.

[misterock]

So is it true, per this thread, that 48 is a waste of time? I record 48 24 wav on my g4 001.

Should I record att 44.1? I hope not!

[Mr. Dipity]

The main reason 48khz was chosen as a standard for DATs, etc, is because the tape could vary in speed slightly and the frequency response (up to 20khz at least) would still be maintained. Not as much room for error with 44.1.

I could be wrong, but I'm almost certain that the 48KHz standard was set up during the SCMS wars to prevent direct CD copying. This is why the very earliest DAT machines to hit the states only recorded at 48KHz.

48khz was set up as a telecom standard. It's a multiple of 8khz, so they were able to band together 8khz phone trunk TDM streams (no, not the Protools TDM, but the underlying concept to it) to create it without wasting bits. Same with 32khz, 16khz. From telecom, it went to broadcast, and that's how it ended up on the earliest DATs.

Whether or not copy protection was an issue in the decision, I do not know, but I doubt it - all the gear it was supposed to interface with (as far as there was digital gear back then) was working at 48khz.

[Mr. Dipity]

hey folks,

thanks for all the informative replies. i think i will stick to generating, recording, processing and mixing at 32bit floating point 44KHz, as i often do a lot of digital post processing.

gregg.

You will not notice an audible difference if you convert from 24 khz before and after destructive processing, since your original signals are at 24khz anyway. The only reason to use 32 bit is when you are creating and modifying that signal - the dithering and modifying that goes on when you apply an effect destructively. Furthermore, 24 bit used properly contains more dynamic range than you can hear.

You will notice a difference in terms of the amount of processing power, and disk space that 32khz file require (25% more for each, give or take the vagaries of your DSP processing as far as the processing power is concerned).

My vote is 24 bit, converting to 32 bit when you hit sound forge or whatever your destructive editor is.

[high tek]

ive been recording at 44.1khz 24bit since i got my motu 828, and only recently (week ago) did a friend of mine that came over insisted that i do the 48 thing.

i have also heard from numerous sources (net, zines) that it is the thing to do (48)

however since it doesnt make any difference sonicaly (this is what its all about right?) as far as i can tell, ive decided to stick to 44.1

[Mr. Dipity]

ive been recording at 44.1khz 24bit since i got my motu 828, and only recently (week ago) did a friend of mine that came over insisted that i do the 48 thing.

i have also heard from numerous sources (net, zines) that it is the thing to do (48)

Please link to them, so we can shoot them down properly.

[misterock]

I think you guys are nutz:

48+48=96
96+96=192

44.1+44.1=88.2
88.2+88.2=176.4


Duhhhhhhhh....

Don't be a dum dum!

[high tek]

Please link to them, so we can shoot them down properly.

ill get back to you on this one.
(might take some time to find)