I just tried out 96KHz...

Shayne White · Post by **Shayne White** » Thu Dec 18, 2003 8:40 pm

...and I was blown away! I loaded in some of my custom-drawn single-cycle waveforms into the STS, and I couldn't believe and clean and pure the high-end sounded! Way better than 44.1! In fact, even bad "analog" samples sounded pretty good, almost like a VA synth! VA synths sounded a lot better, too -- more punchy and clear. I never knew how good it sounded before.

I just wish it wouldn't take up so much DSP and hard disk space, and that I wouldn't have to downsample everything to 44.1 in the end. I don't think my system is ready to move up yet, but it's something to look forward to.

Shayne

jtek · Post by **jtek** » Thu Dec 18, 2003 9:43 pm

I've also tried running my sfp at 96khz,and I wasn't exactly blown away, but it did add a 'slight edge' to my synths. Hard to describe...

On a related note...I own reaktor 4, and have tried running the internal samplerate higher than 44.1khz, and it does seem to improve the sound quality.

anyone else have experience with this?

Nestor · Post by **Nestor** » Thu Dec 18, 2003 11:21 pm

I think I'm going to try one day... it seems interesing. But I don't really see a strong point there, as I have heard the difference and it is not substancial once you have put it down to the normal rate, which is 44.1

ChrisWerner · Post by **ChrisWerner** » Fri Dec 19, 2003 6:53 am

Hi Shanye,
the difference between 44,1 to 96kHz is very clear to hear. Also the difference between 44,1 and 48kHz.
But, have you every tried to produce a small track in 96 and 44,1 and then dither down the ready 96 kHz track to 44,1 and compare it with the other 44,1 track?

I haven´t tried it yet.

mgranger · Post by **mgranger** » Fri Dec 19, 2003 7:25 am

I thought I read somewhere that 88.2 Khz was the best, As it doesn't hav to alter the signal much to dither it to 44.1 - I only has to halve the samples.

Could be wrong

ChrisWerner · Post by **ChrisWerner** » Fri Dec 19, 2003 8:45 am

Ah, I think that good dither programs or routines don´t care about 96 or 88,2 kHz.
But anyway 88,2 is not an option for SFP.

I´ll make this 96 and 44,1 test today.

ChrisWerner · Post by **ChrisWerner** » Fri Dec 19, 2003 11:38 am

Test is here:
http://www.planetz.com/forums/viewtopic ... forum=17&0

astroman · Post by **astroman** » Fri Dec 19, 2003 12:05 pm

I haven't yet listened to your track Chris, but I agree that there can be strange effects. Unfortunately the Pulsar One hasn't enough power to work with sophisticated devices at that rate, so I could only try things like the UKnow007.

Some stages of the instruments seem to react differently, with something like a shift in frequency you describe, some don't show a change at all.
The problem is to distinguish between a real processing at 96k which is very likely to produce a better result (at least on the 'master' machine) and a simple doubling of the samples.
The latter is really done on some soundcards at 96k, at least according to reviews in SOS. This also is supposed to sound different, but it can mislead the interpretion due to prejudice, since one expects an improvement.

At least in old CW devices I don't think there's an actual processing done.
To my ears the difference was either too small or the sound impression could be adjusted back to an identical result by moving a controller.
It was not like the difference between Minimax and Minimod or between the 2 versions of the ProOne.

my 2 cents, Tom

decimator · Post by **decimator** » Fri Dec 19, 2003 4:15 pm

I tried also ...
Fascinating would say Mr Spock !

It's truly a case by case enhancement / defacement like going under ( 44.1 )

alfonso · Post by **alfonso** » Fri Dec 19, 2003 4:56 pm

Sometimes I made this fantasy: If I had an old super pro stereo analog tape machine I could work at 96k and then mix directly on tape...no resampling, no tape emulation...I'm sure It would kick ears!

Shayne White · Post by **Shayne White** » Fri Dec 19, 2003 5:20 pm

I just made an interesting discovery. I recorded a couple of short licks of me playing one of my custom wavetables in the STS, and recorded it at 96KHz into Sonar. Then I exported it and loaded it into Goldwave, and, surprsingly enough, it sounded perfect when Pulsar was set to 44.1! Then I converted it to a 44.1 file, and it souded awful. So I reverted back to the 96 file, played it back through Pulsar at 44.1, and recorded it into Sonar at 44.1. It still sounded good! So -- why would that be?

My opinion is that aliasing problems only occur when there is no lowpass filter in place keep the frequencies within range, as happens in samplers. That's probably why MiniMax sounds so good -- they probably implemented something like that. Any further comments, anyone?

Shayne

astroman · Post by **astroman** » Fri Dec 19, 2003 8:37 pm

the Minimax filter does operate on a higher internal samplerate according to a statement from CW. I don't remember exactly where I read it, probably in a mag review.

They once suggested to not run Minimax at 96k, for whatever reasons.
Maybe the unmatching samplerates might cause trouble on the external input, but that's just my imagination, not a proven fact.

cheers, Tom