At 3:45 AM +0000 5/11/09, justinasia wrote:
> > I run a professional recording/mastering studio & one of the banes of
>> my existence is people bringing in files for mixing or mastering which
>> have been recorded at 48k. I can convert quickly, but at lower
>> quality.
>
>Hi Scott
>Thanks for the input. When you convert 48 to 44.1 quickly, you say
>it is at lower quality. Will the resulting 44.1 file be lower
>quality than if you had recorded directly at 44.1 in the first place?
>
>Also as yet I have not understood, is there any noticeable advantage
>for us to use even 96kHz, over 44.1kHz? IN particular if we will
>anyway be finally converting the files to 44.1kHz CD format?
>
>For example I read about aliasing distortions which apparently give
>96kHz an advantage. I can't say I understand what that really means,
>but am wondering, is that relevant or noticeable for our nature
>recordings?
>
>Justin
One can argue that many years from now, folks will be able to uncover
details in our recordings that are imperceptible to us now. A very
good long-term storage plan is implicit in this goal and the daily
routines involved can impact recording format decisions. I record
many hours of controlled material and everything gets saved as
16K/48Hz. 24/48Hz requires 50% more of my time and materials to
store and 24/96Hz takes 300% more. With time and money in mind, I
decided to conduct blind A/B listening tests before deciding on the
recording format.
A blind A/B test of field samples takes about an hour to conduct.
Record the tests in the field with fixed mics, fixed levels and
preferably low sound levels which are a little more demanding. Record
5-10 minutes of each unless the conditions are extremely consistent.
After the files are on your computer, without listening to them,
select a 10 second extract of least activity (thinnest amplitude)
from each, make a text note of the file type you are saving. Using
"Save-As" and enter key stroke garbage for a name and use Control-C
copy/paste this garbage string next to the corresponding data type
(don't stop to read it! ;-) ). Repeat this for all of your test
files. Place all of the 10 second file extracts into an iTunes
playlist and use your best monitoring equipment as you try to order
them in terms of quality. Use "shuffle" and run through them 2-3
times before you compare your results to the facts.
Another logical perceptual test is to up-convert all of these
original files to 24/96Hz, add some identical gain and EQ and output
a sample "master" file for each. You can compare these by ear if you
are curious. Finally, output a 44.1/16K file from these masters and
compare these.
I chose 48KHz because I use the sounds in video projects too. For
CD's I keep everything at 48Khz and resample the output the final
mix, not at any other step. Another factor to check is whether your
computer can handle a number of 24/96 files with automation,
plugs-ins etc-- especially if you do surround work. Mine will play
many more 16/48 original tracks than 24/96's which allows me to do
all of my EQ an dynamic changes in the same step that I output the 24
bit mix master.
We can speculate or we can test our perceptions. When it comes to
"quality," the later can tell one something while the former just
keeps one wondering,... :-) Rob D.
--
|