CWITEC

cerendir · Post by **cerendir** » Wed Oct 24, 2018 4:54 am

elcallio: all of them do.

ED: Thanks! That wasn't super-intuitive but I guess it makes sense now that I know it

cerendir · Post by **cerendir** » Wed Oct 24, 2018 6:51 am

OK, I added the samples path. Everything started loading without complaint except the icon (but that was an issue with the txprog files, right?), then I got this

And when everything was done, I got this

...upon witch REAPER promptly crashed. I saved the debugging minidump but it's close to 4 gigs in size, so...

elcallio · Post by **elcallio** » Wed Oct 24, 2018 8:58 am

cerendir wrote: ↑Wed Oct 24, 2018 6:51 am OK, I added the samples path. Everything started loading without complaint except the icon (but that was an issue with the txprog files, right?), then I got this

And when everything was done, I got this

...upon witch REAPER promptly crashed. I saved the debugging minidump but it's close to 4 gigs in size, so...

So I found why your programs did not pick up roots correctly. _Really_ lame error on my part. I suck.

But I cannot reproduce anything like the crash above. I am very curious (and worried) about what it might be.

cerendir · Post by **cerendir** » Wed Oct 24, 2018 9:07 am

You defintely do not suck

But you're just one guy and even with a group of beta testers it's impossible to account for each and every usage scenario (especially big and convoluted ones like mine).

It's possible this was just a glitch. I'm waiting for an update that takes forever to finish, then I'll fire REAPER up again and see if the same thing happens again.

elcallio · Post by **elcallio** » Wed Oct 24, 2018 9:39 am

cerendir wrote: ↑Wed Oct 24, 2018 9:07 am You defintely do not suck But you're just one guy and even with a group of beta testers it's impossible to account for each and every usage scenario (especially big and convoluted ones like mine).

It's possible this was just a glitch. I'm waiting for an update that takes forever to finish, then I'll fire REAPER up again and see if the same thing happens again.

Thanks for the encouragement. If you don't have any more info by tonight I will drop a 3.0.5 anyway, just to fix the root key stuff and some other things.

cerendir · Post by **cerendir** » Wed Oct 24, 2018 9:47 am

Nope, the project loaded up fine. Not sure what the heck happened last time.

RoadCrewWorker · Post by **RoadCrewWorker** » Wed Oct 24, 2018 10:04 am

I've also seen the Invalid<T> Vector subscript error pop up a few times in a row, though i can't remember exactly what triggered it (loading something maybe?). Neither the plugin nor the host crashed for me though.

And i'm very impressed by the speed of these bugfixes coming out. Just wish i had known about a beta thing earlier (i found out about the new version from the kvr frontpage) so i could've reported stuff quicker.

cerendir · Post by **cerendir** » Wed Oct 24, 2018 1:23 pm

Updated to 3.0.5. and the pitch issue appears to be fixed! Nice, thanks

However... (you knew there was going to be a "however", didn't you?) when playing some repeated notes on the ViolinEnsSpic patch, this happens. It just keeps popping up again indefinitely and I had to kill the REAPER process to make it stop.

Edit: and the warning pauses the audio process, I might add in case that's important.

elcallio · Post by **elcallio** » Wed Oct 24, 2018 2:26 pm

cerendir wrote: ↑Wed Oct 24, 2018 1:23 pm Updated to 3.0.5. and the pitch issue appears to be fixed! Nice, thanks

However... (you knew there was going to be a "however", didn't you?) when playing some repeated notes on the ViolinEnsSpic patch, this happens. It just keeps popping up again indefinitely and I had to kill the REAPER process to make it stop.

Edit: and the warning pauses the audio process, I might add in case that's important.

Wow. I don't understand how you could get that result from playing that program. It should not be possible. I certainly cannot repro it. Is the one you are playing in reaper modified from the one on disk (the txprog)? Is there anything else playing in TX?

I did a sneaky update of the current build (because it is midnight) just adding a few extra checks. Can you test the updated build (only build no bumped: 3.0.5 5045.1393) and see what it says for the same program?

cerendir · Post by **cerendir** » Wed Oct 24, 2018 9:54 pm

Haha. Did I win something? Seriously though, it doesn't appear to be modified. It's the same version of the patch that's on disk. Tried it again this morning after a reboot and the same thing happened. But after upgrading to the new build the problem appears to be solved!

cerendir · Post by **cerendir** » Wed Oct 24, 2018 10:26 pm

Just for kicks I decided to fire up my biggest orchestral template made with TX v2 (using around 11GB of samples). And I'm happy to report that it loads up just fine aside from a couple of icons. Loading big projects in v3 appears to be far slower though, but I'm guessing maybe it has to do some on-the-fly conversion of parameters so that is understandable. Aside from no round robins or mod wheel layer crossfades working anymore, I'd say it sounds and performs much like v2 which is very nice to see. I have gotten so used to programs and plugins getting increasingly bloated and demanding that I'm always pleasantly surprised when a new version performs equally to or better than the old one.

Edit: That is to say, it performs well until some dozen 4xRR instruments start playing with all RR notes sounding at the same time, at which point my RT CPU shoots through the roof

elcallio · Post by **elcallio** » Wed Oct 24, 2018 11:44 pm

Found the keyswitch problem. Will be fixed in next build.

As for loading speed, the biggest problem is probably that your orchestra library has quite a lot of relatively small samples that end up being RAM only eventually. The loading code is geared towards loading somewhat fewer large samples for DFD. I will look into getting the code path for small samples a little faster.

cerendir · Post by **cerendir** » Thu Oct 25, 2018 10:18 am

So v3 selects DFD or RAM intelligently, based on file size? If so, at what size does it switch to DFD? Just curious.

elcallio · Post by **elcallio** » Thu Oct 25, 2018 11:23 am

Yes, but there is a bug right now. I found it in your material. The good news is that with the fixes, the project you sent loads in under 20s and consumes under 600MB of RAM when loaded. Compared to +3500MB before.

cerendir · Post by **cerendir** » Thu Oct 25, 2018 12:00 pm

Whoa! Well that sounds awesome, let's see if my 7200 RPM sample HDD can live up to these modern features!

CWITEC

KVR Audio

TX16Wx v3 and backwards compatibility