[Audman71]

Hello,

I was working on a song when I had the idea of switching the all-MIDI production from 44.1 khz sample rate and 16 bit depth in the project and the file to 64 bit depth and 96 khz just to see what what would happen, if there would be any noticeable difference.

I have tried comparing prerecorded files of the same content before (not with what I'm currently working on), with limited differences between the recordings (In a DAW I panned the highest quality recording to one side and the "standard" CD quality recording to the other). I noticed some audible differences, but nothing I would notice without A/Bing the recordings.

However, this time in my production there were major differences, most notably the piano. I have two examples I made that are separate from my song:

Here is the first example, rendered at a typical 16 bit depth with 44,100 hz sample rate:




Here is the second example, rendered with a 64 bit depth at 96,000 hz:





Whatever differences you may or may not spot, you should notice the release of the piano is much shorter in the second example. It's like there's extra compression or something in the first one, but the only things I changed were the bit depth and sample rate; I didn't touch the instrument at all.

Is it just me, or is there a much higher dynamic range with 96 khz than with 44.1 khz? I'm not sure if bit depth made an audible difference here, as I tried it with the same bit depth both times in a different test as well. Any ideas?

 

[Roseval]

In case of PCM audio:

The maximum frequency possible is the halve of the sample rate.
The bit depth is the dynamic range. 16 gives you 16*6= 96 dB, 24 gives you 24*6=144, etc.
64 is a value no DAC can resolve but is used by audio editors to keep the quantization error as low as possible.


Obvious the 2 tracks are different.
The second sounds like the damper pedal is used.
Hard to believe this has anything to do with bit depth and/or sample rate.

 

[Audman71]

Indeed, it is hard to believe.

I forgot to mention the piano in the recording is a physically modeled virtual instrument, not a sampled instrument or a real piano, so the sample rate and bit depth don't affect it the same way as with audio.

I did not use a damper pedal. I drew the notes in a "piano roll" style editor, with the exact same settings for each note. There are no FX other than what is in the virtual instrument's environment. Like I said, I didn't touch the instrument or any settings other than the bit depth and sample rate.

 

[MindsMirror]

It might be that some parameters in your production are defined in number of samples rather than in time. For instance if the rate of decay of your MIDI piano were defined as x number of samples, that would explain why the higher sample rate gives a faster decay.

 

[Audman71]

Interesting. I will have to check this, because that would definitely explain it.