Η πρότασή μου, στον αγαπητό vangelisorfas, που ξεκίνησε το post, ήταν να δουλέψει default όλα τα project 48/24, ούτε 96, ούτε 192. Συμφωνώ ότι δεν υπάρχει λόγος να ανεβούμε πολύ ψηλά κι αν είναι να ανεβούμε τα 88 είναι υπέρ αρκετά.
Από την άλλη, αφού τα 44.1 είναι "υπέρ αρκετά" αρκεί να είναι 24 Bit, τότε γιατί να μην είναι και 32 bit floating ή ακόμη καλύτερα γιατί όχι να μην δουλεύουμε στο νέο sonar που έχει true 64 bit επεξεργασία και όχι floating? Ο λόγος του post ήταν η αύξηση της δυναμικής περιοχής για να προσεγγίσουμε αυτά του εμπορίου και όχι τα 96 ή τα 5 δις Khz. Από τα συμπεράσματα στην αναζήτηση του θέματος στο internet, υπάρχει το εξείς αποτέλεσμα : Με βάση την ανάλυση Nyquist και την αναπαράσταση fourier δεν υπάρχει θεωριτικά λόγος για τίποτε παραπάνω από το ακουστικό φάσμα των 22050, αλλά σε σχέση με την πραγματική εφαρμογή και λειτουργία των DAC, υπάρχει μεγάλο όφελος στην λειτουργία σε μεγαλύτερες συχνότητες και έχουμε καλύτερο ηχητικό αποτέλεσμα στην ψηφιακή αναπαραγωγή. Μερικλα αποσπάσματα :
These brick wall filters are what make 44.1 khz signals lack in fidelity what 96 khz can provide, the "brick wall" filters are not "brick walls" but rather progressive filters that induce a lot of distortion, like phase shifts that can go up to 1000 degrees (unless they are really expensive)... At 96 khz, the brick walls are set at 48 khz (not 22.05 khz) and their effects (distortion) are NOT audible, hence why it's better (hence, cheaper).
At 96 khz there is much bandwidth for plug ins to work in, if they're poorly engineered, it may make a difference. The real achievable difference for having high sample rates would be re-pitching or non-linear processes like compression (and I mean really heavy one, like 20:1). 96 khz will sound more analog (actually it feels more than it sounds), but its a very subtle effect and only shown off with very good mixes always done and preserving the full 48khz audio bandwidth.
-------------------------------------------------
Most music is recorded at 44.1, which means the brick wall filter is at 22.05KHz. While most people over the age of 30 can't hear above 17KHz, *some* people have accurate hearing that go well into 23+KHz territory. Some people have complained about "high frequency ringing" from low sampling rates of 44.1KHz, which is why the original movie formats of DTS and DD were set at 48KHz, or the filter extended to 24KHz (nearly inaudible).While most people can't hear much above 20KHz, there are still higher frequency harmonics that can be felt. 96+ KHz sampling rates are of course, the answer for high fidelity audio.
-----------------------------------------------
Higher sample rates provide smoother curves at high frequencies. There can be a total loss of certain frequencies if the sample rate is timed at mid phase. There are also harmonic interactions between frequencies and if those frequencies aren't there, subtle changes occur in the content. Samples in their raw form produce a saw type wave that has a frequency of its own. These are partially corrected by smoothing software and hardware that fills in the gaps between samples. Higher sample rates have smaller gaps between samples.So, although 44.1KHz can accurately sample fundamental (that is sine-wave) frequencies as high as 22.05KHz, a complex wave-shape at 22.05KHz will require a much higher sampling rate if you want an accurate representation. 96KHz and 192KHz sampling (as used in SACD and DVD-A discs) can often sound better because of this. Nobody can hear a 40KHz note, but the notes you do hear (especially the high notes, 8Khz and up) will be more accurately represented. Depending on the nature of the source material and the quality of your playback equipment, the differences can often be easy to hear.
-------------------------------------------
Finally, the concept of higher sample rates creating more "points" between samples and therefore a smoother sound makes intuitive sense but is actually wrong. As Michael Fraser explained, the beauty of sampling is that, say, an 18kHz tone will sound identical at 44.1kHz and 192kHz. Or a 3.5kHz tone at an 8kHz sampling frequency. All the information necessary to reconstruct the tones perfectly is in the lower sample rate recordings.
Huh? ... This is very wrong. In your example, one cycle of the 18khz waveform would be defined by about 2.45 points on a graph. Less than 3 points to define one complete cycle of the waveform!
Sampling it at 192khz would allow the waveform to be defined, ie shaped, with 10.66 points.
Draw me this on a graph and tell me which represents the original waveform the most accurately? You can't tell me that a waveform defined by 3 points for the entire cycle is going to sound the same as the original. All those sharp angles and straight lines mean lots of distortion.
-----------------------------------------------
use the highest sampling rate you can, it can't hurt...maybe you can't hear the difference, but maybe you can feel it...if analog is a solid line, digital is a dotted line, but the more dots you can cram in, the tighter they get to each other and the closer you get to a solid line. Remember that it takes years of critical listening to pick up on the the kind of nuances which we are talking about, but the bottom line is more information equals more detail and more detail will always be better
----------------------------------------
Consider that at 44.1k, a 440Hz wave is measured about 100 times per wave cycle. That might seem like plenty of measurements, but then consider that a 4400Hz wave is measured only about 10 times per cycle, and an 9kHz wave is measured only about 5 times per cycle.
Yes, recording at 88.2kHz would extend the highest frequency you could record up an extra octave (from 22k up to 44k) but it does a few extra things too. It records everything else more accurately (twice as often) and it reduces the severity of the brick wall anti-aliasing high-frequency roll off that causes so much strange behavior at the top end of the spectrum.So it's not just about recording frequencies we can't hear, but about recording what we do hear, better.
--------------------------------------
My Wikipedia fueled explanation of Sample Rate and Bit Depth:
The Sample Rate, which is that second number you mention (96kHz and 192kHz) defines the number of samples per second (or per other unit) taken from a continuous signal to make a discrete signal. For time-domain signals, it can be measured in hertz (Hz). When recording, your interface/computer is actually taking small pictures of your signal called samples, every second. It is generally understood that the more pictures taken per second will result in a better overall image. Therefore, the higher the sample rate, the better characterization of your overall sound.
Αυτά....