u-he

u-he-michael · Post by **u-he-michael** » Mon Sep 08, 2014 5:50 pm

Blipper wrote:I found rev2709 of Satin here:
http://www.u-he.com/downloads/Latest%20 ... 709Win.zip

The reason why we didn't post this revision was :
1. we didn't change stuff in the Satin-code
2. we didn't test it at all

In general, i do not recommend to search, download, post and use revisions,
which were not meant for our users. It may cause a lot of confusion.

However, if you found a major difference / bug, please send examples and/or steps how to reproduce it.

Thanks.
Cheers,
Michael

Blipper · Post by **Blipper** » Mon Sep 08, 2014 6:51 pm

Here is a link to audio files produced using Satin in the middle of the processing chain. The processing was the same for both other than the Satin version change. Satin is in the REAPER project using 176.4kHz. I use proprietary sample rate conversion to 44.1kHz. I listen using a 24-bit external DAC box (Lynx Hilo), and I use a tool that allows me to quick switch back and forth between the two audio files to do A/B subjective testing. I close my eyes during A/B testing to prevent visual bias and distractions.
https://mega.co.nz/#!nkxAmCJY!V1ZGIBiZh ... WvpOfZrU88

As I mentioned previously, this processing used the studio mode of Satin. The compander mix knob was set to 100%. I agree with Sascha that display quantization has nothing to do with audio quantization, so the previous issue I brought up about delay-mode display resolution is irrelevant to this audio quality issue.

Something to keep in mind is that changing compilers or math libraries can have an effect on quantization. Anything like this happen in the 132 revisions since 2577?

EdHarvey · Post by **EdHarvey** » Tue Sep 09, 2014 3:53 am

Another interesting XY editor bug in Zebra

The settings appear to work, but the 'bar' display is completely wrong...

Howard · Post by **Howard** » Tue Sep 09, 2014 5:36 am

Aha - appears to be bipolar, but shouldn't be...

Tested this, and noticed that the target tends to flip back from OSC4: Wave to OSC4: PhaseModDepth, the previous target in the list - which would be bipolar. Very strange bug...

EdHarvey · Post by **EdHarvey** » Tue Sep 09, 2014 6:32 am

Howard wrote:Tested this, and noticed that the target tends to flip back from OSC4: Wave to OSC4: PhaseModDepth, the previous target in the list - which would be bipolar.

That was going to be my next bug report...

Very strange bug...

It certainly gave me pause for a “WTF” moment, until I realised it wasn't my fault

sascha · Post by **sascha** » Tue Sep 09, 2014 9:31 am

Blipper wrote: https://mega.co.nz/#!nkxAmCJY!V1ZGIBiZh ... WvpOfZrU88

Downloading the zip doesn't seem to work. The page says 'starting' but then nothing happens. After a couple of mins, it says 'temporary error, retrying', but to no avail.

u-he-michael · Post by **u-he-michael** » Tue Sep 09, 2014 10:38 am

I downloaded the file yesterday and gave it to Sascha. He is currently having a listen

sascha · Post by **sascha** » Tue Sep 09, 2014 11:32 am

Blipper wrote:I close my eyes during A/B testing to prevent visual bias and distractions.
https://mega.co.nz/#!nkxAmCJY!V1ZGIBiZh ... WvpOfZrU88

Ok, the plot thickens... (drum roll):
...
...
...
...
...
Tadaaaaa:
http://www.u-he.com/sascha/_temp/Satin_ ... 9-2577.wav
This is the differential file of the two tracks. Enjoy

EvilDragon · Post by **EvilDragon** » Tue Sep 09, 2014 11:50 am

Haha. Pwnd!

Blipper · Post by **Blipper** » Tue Sep 09, 2014 1:52 pm

sascha wrote:Tadaaaaa:
http://www.u-he.com/sascha/_temp/Satin_ ... 9-2577.wav
This is the differential file of the two tracks. Enjoy

I see peaks of about -40dBFS in the difference file. Did you do an A/B listening test? As you well know, the ear is very sensitive to certain types of distortion/noise, and less sensitive to others.

sascha · Post by **sascha** » Tue Sep 09, 2014 2:10 pm

Sure, the peaks are around -40dBFS, but RMS is around -53dBFS. So, peak-to-noise ratio is ~13dB. What you hear is the noise masking, and sort of dithering to the transients, which is normal.
I re-did the test here with own percussive material, with rev 2577 and the latest side-by-side. In order to exclude noise shaping from the equation, hiss, asperity and w&f should be turned down completely. My results are that the difference signal has peaks at -72dBFS. It can't be -inf as there's still some random stuff going on deep down.
Therefore my conclusion still is: the DSP part behaves absolutely the same as before.

zvenx · Post by **zvenx** » Tue Sep 09, 2014 9:46 pm

zvenx wrote:Hi,
is the TDZ being uploaded or the one already uploaded corrupted?
been trying to download it for the past hour or two, but when I try to expand it, I am getting an error.
thanks
rsp

Technically this bug is with build 2570, but I can't seem to find that thread anymore.
Had noticed it a few days ago, but was waiting to see if a new ZHz would have been posted before I filed this......

some patches in ZebraHz, when selected show up a strange black box...... only very few patches, and the thing they seem to have in common is they are saved with the resonator parameters as the master fx.

rsp

Urs · Post by **Urs** » Wed Sep 10, 2014 7:29 am

zvenx wrote:Technically this bug is with build 2570, but I can't seem to find that thread anymore.
Had noticed it a few days ago, but was waiting to see if a new ZHz would have been posted before I filed this......

some patches in ZebraHz, when selected show up a strange black box...... only very few patches, and the thing they seem to have in common is they are saved with the resonator parameters as the master fx.

Ah, interesting! That is surely the same path displacement bug we've seen in XY Range controls and Diva Uhbie arrows. It'll be fixed with the next beta of ZHZ.

zvenx · Post by **zvenx** » Wed Sep 10, 2014 9:20 am

thanks Urs.
rsp

re*MD9@3!Tm^ · Post by **re*MD9@3!Tm^** » Wed Sep 10, 2014 9:22 am

Any chance for Zebra2 HD textures some day? Love the zooming feature, but the blurry fonts etc. are a bit distracting...

u-he

KVR Audio

Latest builds - Revision 2709