u-he

HiEnergy · Post by **HiEnergy** » Sat Nov 30, 2013 9:21 am

3ee: could be the lyrics for your next Top40 hit. People will sing along with that.

Urs · Post by **Urs** » Sat Nov 30, 2013 9:28 am

zvenx wrote:I assume there is a reason why you don't just send this version to the four ppl who the bug affects?

or they have it too?
rsp

They have it too. I haven't checked the support emails yet - will do so after coffee.

Hmmm, so far nothing in this thread seems to hint at *that* problem. I hope we didn't run into Heisenbergness - when the probes change the result. Would be a shame if we had to leave them all in there

geronimo · Post by **geronimo** » Sat Nov 30, 2013 9:41 am

An observation: with the SATIN first version , I could open 7 or 8 SATIN in LOGIC 9 and Digital Performer 8.
Now I can't exceed five instances SATIN without opening other plug-ins or Virtual Instrument !
This is a real Black Friday for me, literally .

Urs · Post by **Urs** » Sat Nov 30, 2013 9:48 am

geronimo wrote:An observation: with the SATIN first version , I could open 7 or 8 SATIN in LOGIC 9 and Digital Performer 8.
Now I can't exceed five instances SATIN without opening other plug-ins or Virtual Instrument !
This is a real Black Friday for me, literally .

Well, once the debugging tests are gone the performance will be much better

Bathrobe · Post by **Bathrobe** » Sat Nov 30, 2013 10:10 am

With the previuous Version i had a track go silent when satin vst3 64bit was added in osx 10.9.
With this version all seems fine.

edit: Just installed 2033 again and it did not mute a track but just muted the right channel.
With 2085 installed again everything seems fine.

Ameyah Audio · Post by **Ameyah Audio** » Sat Nov 30, 2013 10:29 am

HiEnergy wrote:3ee: could be the lyrics for your next Top40 hit. People will sing along with that.

geronimo wrote:An observation: with the SATIN first version , I could open 7 or 8 SATIN in LOGIC 9 and Digital Performer 8.
Now I can't exceed five instances SATIN without opening other plug-ins or Virtual Instrument !
This is a real Black Friday for me, literally .

i noticed also, tried on my notebook, the normal version 1 instance uses 44-48% CPU and now this version uses 100%. (well glad i have another desktop

)

so far nothing i will keep playing. anyway i hope someone will find this bug, that we can see some progress soon for Bazille. and of course see the finals over christmas for Diva, Satin and Zebra 2

Rusty Cadillac · Post by **Rusty Cadillac** » Sat Nov 30, 2013 10:50 am

As stated in the OP... this is not an RC.

"(this is a special beta version - not meant for production, but good for bug hunts!) "

And:

"Note: This version of Satin drains some serious CPU for debugging - if playback stutters, don't worry, it's only this build of Satin! CPU overloads are to be expected."

geronimo · Post by **geronimo** » Sat Nov 30, 2013 10:57 am

Rusty Cadillac wrote: "Note: This version of Satin drains some serious CPU for debugging - if playback stutters, don't worry, it's only this build of Satin! CPU overloads are to be expected."

Okay but since the first versions as well !

timbesamusca · Post by **timbesamusca** » Sat Nov 30, 2013 11:15 am

I tried it in Cubase 7.06 64 bit with the 64bit VST2 version, all worked well and no Bug Friday entry in the log as well.

Normally I use v1.1 (rev 1900) and I've never encountered the bug.

Urs · Post by **Urs** » Sat Nov 30, 2013 12:04 pm

We think that optimisation done for V1.1.1 (the one with "auto mute", faster gui loading and lower tape speeds) has either made the bug ore obvious than it was in version prior to that, or it is a newly introduced bug.

Aside from this bug we have indeed fixed a couple of issues that may have been present in 1.0 and 1.1 - there were a few bug reports with similar sympthoms.

All it is really is about getting the initial state right when the plug-in is created and before audio is processed. This is really complicated with Satin because there's like a hundred mini-modules in there that may all interact in various ways. Getting it to "not click on activation" and "not divide anything by zero" paired with an unpredictable call sequence from the host side is, well, a bit of a stunt.

#---

I think I'm going to add another set of checks today - instead of concentrating for illegal operations (divide by zero...) and random initialisations I'll add some checks focussed on "where does the signal go quiet?"

Maybe we can get a few Bug Fridays then (or shall I rename it to anything Saturday?)

- Urs

ducelare · Post by **ducelare** » Sat Nov 30, 2013 12:39 pm

ableton live 9.1, OSX 10.9, 64 bit. I had some mutes with rev 2033 (both AU and VST2), solved with rev 2085. Now looking forward to a less hungry version

Cheers!

tenfingers · Post by **tenfingers** » Sat Nov 30, 2013 1:38 pm

2085 beta (ML 10.8.5) Vst3 + 2 generally muting tracks on first instantiation - no Bug Friday though

MFXxx · Post by **MFXxx** » Sat Nov 30, 2013 1:39 pm

Urs wrote:
MFXxx wrote:Just sent it. Unfortunately where Machine2 is hanging on the loaded screen, soundcard is crashing and no crashlog produced.
Pssst... there might still be a Satin.log on your desktop

Hi urs tested 3 hosts in concession maschine being the last. Hopefully went the correct log. Out with the family at the moment so will check later over weekend.

Askeng · Post by **Askeng** » Sat Nov 30, 2013 1:44 pm

No problem here with Studio One x64, tested both satin.dll and satin(x64).dll
Loaded, removed and tweaked the plugin like crazy -- saved and closed about 30 times.

Urs · Post by **Urs** » Sat Nov 30, 2013 2:10 pm

tenfingers wrote:2085 beta (ML 10.8.5) Vst3 + 2 generally muting tracks on first instantiation - no Bug Friday though

Sounds good - I've started preparing a new version which checks that case too!

We might be able to draw a winner after all

u-he

KVR Audio

Bug Friday - win & suffer with u-he!