KVR Audio

zeoy · Post by **zeoy** » Tue Feb 15, 2005 11:57 am

I was messing around the other day with this beast and I noticed a huge difference in sound when using 96K. More specifically the highs were way more pronounced (in pleasant way though) like a lowpass filter was bypassed. Is it normal? Similar experiences?

I am using Tracktion on Win XP with an audiophile 2496.

boin · Post by **boin** » Tue Feb 15, 2005 12:27 pm

Never tried PolyIblit in 96K but this is exactly what I would expect. The basic idea for using higher sampling rates is to allow better representations of high audio frequencies/transients. The final result obviously also depends on synth design, anti-alias filters, etc.

zeoy · Post by **zeoy** » Tue Feb 15, 2005 12:39 pm

I was expecting differences too, but I thought they would be slight ones. We are talking about huge differences here. Probably this has to do with the anti-alias filters as you say. I 'll check Asynth also to compare. Since Asynth has oversampling options to avoid aliasing I am expecting slighter changes. We 'll see.

boin · Post by **boin** » Tue Feb 15, 2005 1:10 pm

Using Pentagon with oversampling makes some patches to sound so different that I renamed them as different patches...

I'll give PolyIblit with higher rates a try when I got time.

zeoy · Post by **zeoy** » Tue Feb 15, 2005 1:13 pm

boin wrote:Using Pentagon with oversampling makes some patches to sound so different that I renamed them as different patches...

That's what I am talking about

Funkybot · Post by **Funkybot** » Tue Feb 15, 2005 1:18 pm

Too bad Polyiblit does't have Oversampling built in. Maybe someone should run it by Andreas. On the flip side of that though, I'm sure a few patches at 96k would just kill my CPU, but damn it is a nice sounding synth. I think my dream shareware/indie synth would be Polyiblit with ASynth's filters.

zeoy · Post by **zeoy** » Tue Feb 15, 2005 1:22 pm

Funkybot wrote: I think my dream shareware/indie synth would be Polyiblit with ASynth's filters.

Couldn't agree more

kritikon · Post by **kritikon** » Tue Feb 15, 2005 1:39 pm

Can't say I've tried at 96KHz - high sampling rates aren't worth flogging my CPU for IMO. But I thought the whole thing with PolyIblit was that it used band-limited oscillators, so I wouldn't have thought there were any higher frequencies in there that would be any better represented by using 96KHz?

zeoy · Post by **zeoy** » Tue Feb 15, 2005 1:44 pm

I never work @96K. I did it just from curiosity, but believe me some patches sounded totally different, like removing a LP filter. I am wondering if this is a problem with my setup.

stefancrs · Post by **stefancrs** » Tue Feb 15, 2005 1:48 pm

My guess is that it's just the filters that change, and if so, because of faulty coeffecient calculations.
There's no "anti-aliasing" going on in polyIblit, and afaik, the oscillators always outputs all harmonics < 22.05KHz when running in 44.1KHz. If it starts outputting all harmonics up to nyquist at sampling rates above 44.1KHz, the oscillators will most likely not sound different at all to anyone.

3*s · Post by **3*s** » Tue Feb 15, 2005 1:52 pm

This is most likely a bug. As stated earlier it might be attributed to code not designed to be run at 96khz.

zeoy · Post by **zeoy** » Tue Feb 15, 2005 1:57 pm

@stefan

Seems reasonable stefan. The extra highs were not in the 22-48K area definately

@3*s

Yeah, maybe. I ll' try to record an example to post in this thread later

stefancrs · Post by **stefancrs** » Tue Feb 15, 2005 2:05 pm

zeoy wrote:@stefan

Seems reasonable stefan. The extra highs were not in the 22-48K area definately

@3*s

Yeah, maybe. I ll' try to record an example to post in this thread later

Make a patch where the filters are fully open and then compare that patch at different sample rates!

punkrockdude · Post by **punkrockdude** » Tue Feb 15, 2005 2:11 pm

zeoy wrote:I was messing around the other day with this beast and I noticed a huge difference in sound when using 96K. More specifically the highs were way more pronounced (in pleasant way though) like a lowpass filter was bypassed. Is it normal? Similar experiences?

I have hade the same experience with IK Multimedia's Amplitube. I noticed a dramatic difference in 96kHz compared to 44kHz/48kHz. It happened when I flipped through the sample rates while playing.

Best Regards
Rickard Gerthsson

zeoy · Post by **zeoy** » Tue Feb 15, 2005 2:45 pm

stefancrs wrote:
zeoy wrote:@stefan

Seems reasonable stefan. The extra highs were not in the 22-48K area definately

@3*s

Yeah, maybe. I ll' try to record an example to post in this thread later
Make a patch where the filters are fully open and then compare that patch at different sample rates!

OK stefan, that's the way to see if it's a filters' bug

Polyblit sounding totally different at 96K