25 Nov
2015
25 Nov
'15
2:56 p.m.
On Wed, 25 Nov 2015 07:34:46 -0500, Paul Davis wrote:
We sample at typically 40kHz plus.
Sometimes guitars, vocals and synth can be recorded without loss, when
using 32 KHz 12 bit, but it also could lead to strange results and most
of the times there's still audible loss at 44,1 kHz 16 bit.
48 KHz 16 bit (or maybe 24 bit) is good enough to never get audible
loss. Getting audible loss at 48 KHz 16 bit happens often, but this
isn't caused by the sample rate and bit depth, a cause could be the used
hardware or software, some hardware might sound better at higher sample
rates, but again then it's the hardware, not the sample rate.
To work with audio data we should use 32-bit floating point.
To care if, when and how to use what kind of dither might be worth
reading.
If digital audio processing nowadays is able to simulate analog gear is
another issue and has nothing to do with this thread ;), but for a
professional audio recording without loss you need 48 kHz only.
44,1 kHz is not enough, more then 48 kHz is not needed.
--
http://www.grundgesetz-gratis.de/