Better sound quality and depth

When the Pianoteq sample rate differs to the device sample rate, what does the sample rate conversion?  I haven't tried it yet, but if the quality is still excellent with the differing sample rates, that's great.

Greg.

Nice. I'm hearing noticeably less bite and dynamics (and volume?) in the acoustics, too, but these steps do create a very playable instrument that only takes up around 10-15% of the cpu.

I'm not sure I understand what happens. The first step raises the sound card\chip sample rate and the second lowers the program's sample rate, so the effect is to decrease the quality of the source material (with the 2nd step, on the Performance tab) and resample it, in a sense, at a higher rate than by default (with the 1st step, on the Devices tab)?

Somehow, I would expect the sound to degrade much more, given the deep cut in the Performance sampling rate.

"If the device rate was 44100,  Pianoteq's rate would be 11025, because Pianoteq appears to restrict it's internal rate to "integral submultiples" of the device rate."

Skip,

I'm getting confused--I've always just kept both sample rates at 44100. I thought this was the default, but I may have changed the settings at some point and forgotten...

It is the default, and that's how I run mine too.

What I meant was:
Given that the idea posed in this thead is to try to save a *lot* of CPU cycles by reducing Pianoteq's sample rate a lot,  rather than dropping it to 11025Hz, we can get back a *bit* more fidelity, consuming only slightly more CPU cycles,  by increasing it to 12000Hz. But to do that, we have to *also* change the device rate to 48000Hz - 12000Hz simply isn't listed as an option if the device rate is 44100Hz.  If we restricted the device rate to 44100Hz, that limits us to 11025Hz and 22050Hz, with no intermediate settings - it's a big jump.

Greg.

That's ok pom-pom man. 

Greg.

But it sounds better than that.

The mp3 is 44.1 kHz, so here's also some resampling done to further change the sound. You should have made a 48 kHz mp3 as well.

Also, you could have explored more dynamic levels in that improvisation, so that we can hear what is exactly lost.

I'm still using my PTQ3 on 48k/48k.

Actually, the highest note on the piano is around 4300 Hz and low notes don't produce harmonics much higher than 7000 Hz, so it makes sense, though I haven't tried it myself.

Yep, 48000Hz / 48000Hz is best! I've just been comparing and comparing and yes, the only way you're actually gonna get the best out of pianoteq is to use at least 48000Hz / 48000Hz. But for those who don't mind loosing a little presence and brilliance - especially noticed in the way some notes at certain velocities at 48000Hz / 48000Hz sound like kissess which is lost at any lower sample rate - go ahead and save on CPU!!!

Actually, this is a bit empirical for me. I just looked at some of my spectrum analysis graphs and never got any significant values above 7 or 8 KHz. (See my post in the "Two more pop-rock Yamaha videos" thread.)

8KHz is sometimes given when searching the web for this information but with no justification.

Still I managed to find the following reference under Google books which seems trustworthy and seems to agree.

http://books.google.ca/books?id=RUDTFBb...mp;f=false

By the way, my value for the highest note is off and closer to C#8 probably due to octave stretching.