sleepy said:
...I like many other people, don't like doing conversions. Dithering helps a bit when making 24 bit to 16 bit conversions but I'd rather work with 44.1khz straight rather than take 96khz and make it 44.1khz after and lose a bit of detail.
The thing is:
most people don't hear a difference between 44.1 and 96kHz recordings, unless they have incredibly good hearing. By contrast, the difference between 16bit and 24bit is
huge!
Rule over thumb:
the highest frequency that can be reproduced is half of the sampling frequency (the so-called Nyquist Frequency, I believe): 44.1k=22KHz, 48k=24KHz, 96k=48kHz. Now, most people can't hear anything beyond 20kHz (woman, btw, have a slightly higher hearing range). So true, 44.1Khz is already beyond the range of human hearing -- but the point is: With a sample rate of 44,100 samples per second the highest frequency that you can record is 22,050 hz, because you need at least 2 samples to reconstruct the waveform. So while at 44.1kHz you will have 2 data points representing a sine wave, you will have at least 4 sample points at 96kHz. In the first case you have a triangular wave -- which doesn't sound much like a sine wave (= "digital harshness") -- while its much more like a sine wave at 96kHz. Or the other way round: at 96kHz you get twice as many sample points in a recorded sound file, which will be much smoother. Even if you down-dither to a CD sample rate of 44.1 kHz, the sound wave will be smoother due to the interpolation between 4 data points instead of two. The downside, however, is that you need much more disk space to store the files at higher sample rates. That'll drain your CPU power, and the performance of your computer goes dramatically down! The resource factor is 3.2653 here. So if you were getting 36 tracks of audio before, now you're getting 11. WOW! (BTW: 88.2->44.1 is a cleaner conversion than 96->44.1 coz its exactly a factor of 2.)
Lets got back to the bit rate:
At 16bit you have a range of 0-65,000 to store values for volume, frequency, etc etc...
At 24bits, that range jumps up to 0-16.7million! Now THAT's a difference which is
clearly audible!
Here's an analogy from digital video:
Audio bit depth = number of pixels on your tv screen
Sampling frequency = frame rate.
As long as you are operating beyond the frame rate of the human eye (or frequency range of the ear), you're fine.
To make a long story short:
24 bits over 16 gives you a greater benefit over the sampling frequency. A good compromise is therefore 24/44.1 recording for CD production, while 24/96 is a must for DVD/surround recordings.
Sorry if that had absolutly nothing to do with Cubase/ProTools
Back to Pro Tools: ProTools LE is useless because of the track limitation. I think LE is nothing but a joke. ProTools HD, on the other hand, clearly sounds better than Cubase simply due to the better converters in the ProTools interfaces, which are top notch. However, unless you are operating a pro studio and need ProTools to attract customers (who
think a studio is only "pro" because you have ProTools), I'd get
Logic *LOL