recording at 24 bit 48k?

Farview's talking about the actual digital integers in that post, in which the crossing point is literally zero and the sound cycles into the positive and negative around it. If you create a completely silent file and look at it in a hex editor, it will be completely filled with zeros after the header. Zero db is +-32767 (in a 16 bit file). I think what you're suggesting makes sense when discussing the amplitude of a signal (and i think everybody pretty much speaks in db in that case anyway) but that's not what he's referring to in your quote.
 
That's just semantics at that point. Both of you explained exactly what I mean, just from a different angle.

But the point I was trying to make is that the system can not resolve anything in that noise floor range. So the lower the noise floor, the more you can resolve, or the more resolution you have. Therefore, 24 bits has more resolution that 16 bits, assuming that the acoustic and analog electronic noise floor is lower than the digital noise floor that a 16 bit file would have.
 
Jay, I totally agree but I wouldnt have said,

The thing is, you get down to the noise floor on every cycle.

It seems to imply that there is something problematic about getting down to the noise floor on every cycle. But every wave gets "down to the noise floor on every cycle". It's not the issue. If it was we could never record anything without noise intruding objectionably. We know that's not true.
The frequency of audible waves is not such that we "hear the noise" when the wave crosses from negative to positive and vice versa and momentarily "goes quiet", allowing us to better hear the noise floor. It doesnt work like that.
So I dont see how it helps to speak about the wave cycling from positive to negative and vice versa around the noise floor. From a listening point of view it's irrelevent.

Tim
 
with converters equivalent to 20 bits but any other part of the chain, including the aircon noise in the building, is at say 12 bits, then 12 bits is the maximum dynamic range you can resolve. In this case, recording at 16 bits would sound no worse. You would lose nothing.

Exactly, this is what many people overlook. And 12 bits is about right for most rooms. Especially home studios.

--Ethan
 
Thanks for this discussion. I've always recorded at 24/44.1 due to processor power, but always felt a little guilty. I'm putting that behind me now and focusing on the music.
 
I've recorded at 44/24 for one project I'm working on and 48/24 for another and to be honest I don't hear a grave difference between the two. There is a tad more "air" and that's about it and its completely what I expected since there are more harmonics happening up there somewhere that I can't really hear but I "feel".

In the end record what works best for you. I will tell you that the most important factor in whatever sample rate you record at is how you record. If you record a crappy sounding guitar without proper mic placement and such its going to sound crappy no matter what the sample rate is. Get it right at the source and it doesn't matter the sample rate. If you don't take the time to get the right tone and mic placements no sample rate, eq, compression, plugin, ANYTHING will help you.
 
There are other important reasons for higher sampling rates than frequency response relating to Nyquist–Shannon. I was pleased to see someone earlier mention higher sampling rates and apparently thinking ahead for posterity, as I often find myself alone in thinking like an archivist in these discussions. We should always be looking beyond our current state of the art with a view to preserving the best possible recording we can for new and better ways of capturing sound.

I approach recording quite differently than most people here because I heavily integrate analog in my studio. 24/48 is my standard. There's not so much detectable difference between 20-bit and 24-bit as there is between 16-bit and 20-bit. For audio in music recording 20-bit is more than adequate in most cases, but 24-bit is there and has been standard for some time now, so why not? That's about it when it comes to bit-depth. But I was perfectly happy when 20-bit or even 18-bit was the bee’s knees.

With sampling rate I recommend 48k regardless, but that's partly because I avoid D/D conversion, preferring resampling from 48k to 44.1 through the analog interface instead. In my experience and IMO, digital conversion from one sampling rate to another up or down does more damage to digital audio than does resampling the analog signal through a decent ADC. There’s an abundance of evils in converting back and forth between sampling rates within the digital realm and I believe this is the area more than any other where digital audio deteriorates and gets it’s bad reputation.

A typical process for me would be to track to analog multitrack, transfer those tracks to digital 1-to-1 @ 24/48 and then mix to ¼” half-track reel-to-reel analog. From there I master to either an HHB CDR-850 stand-alone CD recorder or back through my DAW interface using CD Architect. Depending on the project I may skip mixing to half-track analog, but everything else remains about the same. I use few plugs… very often none. I have a wall of outboard gear so I realize my methods might not be practical for everyone. But I’ll keep doing it this way for the foreseeable future. I’ve found the difference between my process as described above and doing everything ITB is a night and day, heaven and hell difference.

So anyway, all things being equal, 24/48 is the sweet spot for me when working in the early stages of the digital process. And of course 16/44.1 for the end product because that’s what we’ve got at this point, though things are changing (for the worse unfortunately) rather quickly. Even so I find converting from 16/44.1 to mp3 is a bit better fidelity if we can use the term fidelity when talking about these lossy compressed formats. Others are certain it’s better to convert to mp3 from higher bit-depth and sampling rates and that’s fine for them, but I’ll keep doing it this way until something comes along that really shakes things up technology wise. Perhaps if I were tracking initially to digital rather than analog I could get more excited about 96k, but at present it gives me no advantage.
 
Back
Top