Digital Audio Questions sorta OT

[Burkeville]

Hello,

I have a sound card that is 24 bit with a SPDIF input . I record in 24 bit.

I have been connecting the soundcard as slave to a Apogee ad 500e(AES OUT) that is 16 bit and making it master( the sound card slaves to the Apogee clock).

This is through an AES to SPDIF converter (Apogee Master is AES out and the Sound Card Slave is SPDIF in). Things do sound better this way.

What is happening to the last 8 bits? How does this slave situation work with different bit depths? Are my files being recorded with 8 zeros at the end?

Thank you to those more knowledgable than I.

Best,
kb

[Cary B. Cornett]

The fact that your converter is 16 bit means that that's what you are getting-- 16 bits. The remaining "less significant" bits are padded as "0". That said, a GOOD 16 bit converter is preferable to a LOUSY 24 bit converter. So, just because it advertises "more bits" does not automatically make it better.

A few years ago I benefited from the fact that most people go for the "hype" when the Midiman "Flying Cow" converter was upgraded as a product from 20 to 24 bits. As a result, I got a NIB 20 bit Flying Cow for 10% of its original retail price. Yep... a $500 converter for $50! It works just fine, and is still the converter I monitor through (and sometimes record with) in my studio.

"Truncation" is what happens when a file with more "bits" of sample depth has to be output in a format with fewer bits, and dithering is not used in the process. Truncating from 24 down to 16 bits is not necessarily a big deal (the sonic difference would be subtle, and not noticed at all by many), but truncating down to, say, 8 bits will be sonically obvious compared to a properly dithered bit reduction. A 16 bit data stream fed into a 24 bit system is not be "truncated" at all.

The "slave situation" has nothing at all to do with the bit resolution of your setup. It refers strictly to the fact that the internal word clock oscillator of your Apogee converter is used as the system's master clock. If you are using only one converter box in your system, it is usually best to let the converter be your Master word clock reference/source. This is because the conversion process is where a very stable clock is most important, and your computer and sound card digital in are not at all affected by clock jitter (variation in clock pulse timing) unless it is really severe. Explaining why THAT is gets rather more involved... :eek:

[Sean McCoy]

And there is no advantage to recording 24 bit files from a 16 bit converter output. There are still some potential processing advantages to setting your edl bit depth to 24 even if you're using 16 bit audio files, but recording your files at 24 bits is just a waste of disc space.