bits & khz

[draven]

If you all ready recorded @ 44.1, SRCing to 48k wouldn't do anything but add in more bits and muck up the audio a bit (not really noticable tho).

[stoneinapond]

Bezo,

As Draven has stated, the question of 48/44.1 can and has been endlessly debated.

Virtually everyone agrees that 24-bit capture is superior to 16-bit, even if the project will eventually end up as 16-bit. The only proviso here is that it is possible to record with a truly great 16-bit A to D converter and get better results than using a really lousy 24-bit one.

On the sampling rate issue there is no such consensus. The problem is whether the gains you get in recording with the higher sampling rate are lost when you sample rate convert. The results here will also depend on the quality of the SRC implementation itself.

I try to avoid SRC myself, but that is simply a personal preference.

If you want to read more about it, I would go to Google and type in “sample rate conversion.”

Happy reading.

[Kickin.da.speaker]

Plus, recording at 48KHz means you add a conversion in the process, from 48 to 44.1. I've never heard any good 48 to 44.1 conversion. I suggest you work at 24 bits 44.1KHz unless you really know what you're doing.

[Duardo]

Of course. A 48k/24b dithered down and converted to 44.1/16 will sound better (yeah, yeah, everyone can argue this). I always record 48/24 unless it's a rough project and I need to save the HD space.

<font size="2" face="Verdana, Arial">I'll argue that...in a best-case scenario, all you'll get recording at 48 khz and converting to 44.1 kHz is the same thing you'd have had you recorded at 44.1 khz in the first place. In most instances, though, it will actually sound worse, as most SRC algorithms degrade the audio to a certain extent.

If you have an older converter that really sounds better at 48 khz than 44.1, then by all means record at 48 khz...but really all you're gaining is an extra tenth of an octave of information that's above the range of human hearing anyhow...

-Duardo

[Park Seward]

Originally posted by draven:
Of course. A 48k/24b dithered down and converted to 44.1/16 will sound better (yeah, yeah, everyone can argue this).

<font size="2" face="Verdana, Arial">If the low pass filter and the frequency response at 44.1k and 48k is the same, why would it sound better?

COnverting from 48 to 44.1 is no simple matter.

For a start, look at Multirate Digital Signal Processing by Crochiere and Rabiner (see FAQ section 1.1).

Almost any technique for producing good digital low-pass filters will be adaptable to sample-rate conversion. 44.1:48 and vice-versa is pretty hairy, though, because the lowest whole-number ratio is 147:160. To do all that in one go would require a FIR with thousands of coefficients, of which only 1/147th or 1/160th are used for each sample--the real problem is memory, not CPU for most DSP chips. You could chain several interpolators and decimators, as suggested by factoring the ratio into 3*7*7:2*2*2*2*2*5. This adds complexity, but reduces the number of coefficients required by a considerable amount.

[Bezo]

I found out that CD quality is 16 bit, 44.1khz. I'm current working on a recording using 24 bit, 44.1khz, that will eventually be bounced to CD. It seems useless to record w/24 bit. Can/Should I change the bit rate of the track to cut back on usage?

[pk_hat]

A very short and non-technical answer:

You will get better results by recording at 24 bit, even when dithering down to 16 bits for CD playback.

[Bezo]

Cool. Thanks.

Any benefit to upping the khz?

[draven]

Of course. A 48k/24b dithered down and converted to 44.1/16 will sound better (yeah, yeah, everyone can argue this). I always record 48/24 unless it's a rough project and I need to save the HD space.

[Bezo]

OK. I don't use many effects/plug-ins. And I doubt this, or any of my recordings will elcipse 15 tracks, so I'm sure I can spare the space.

Can I change the khz on the current track?