@Another User said:

Could it be being caused by the low resolution?

Maestro2be

It _could_ have been related to 16 bit color depth, but not to the size.

 Hi Maestro2be,

to narrow down this issue, please perform these tests:

• Set the Venue Animation Speed to 100% (Options -> Preferences; requires restart) - does this change the never-ending impulse update?
• Have a look at the IR-update icon at the lower right corner of MIR - is it always active while you hear the "static sound"`?
• How much CPU load do you get in this situation?
• When you switch off MIR engine (right after starting MIR) and move an instrument icon around at the stage - how much CPU load do you get now?
• How much memory does MIR allocate on your machine?

And just to be sure - did you get any error message or hard crash? In this case a dump file would be useful... 

(When I tried to run MIR on an AMD machine the last time, it just crashed...)

Thank you, Florian

Florian,

Thanks for chiming in.  I will perform all of those tests and get back to you.

I do have a question though.  Where would I find the crash log for this because I think one or two times, it did in fact give me the "Windows 7" crash window which is new to me (but it did say MIR crashed and blah blah blah).  Where will that log be?

But so we are clear, it never crashed in the sense that it just disappeared and shut down on me.  It did however NOT like it when I tried to CLOSE it during the never-ending impulse loading distortion.

Maestro2be

 When you get the crash window, just open the task manager, right-click on "Vienna MIR.exe" in Processes tab and choose "Create Dump File". After the file is stored a pop up message will tell you the location of the file.

Just send me a message when you got a dump file, then I'll tell you how to upload it.

Florian

Ok Here we go with the updates (now that I am in front of the machine, everything else was from memory :).

1.  Windows 7 Enterprise information - 7600.win7_rtm.090713-1255

2.  I set Latency to 1024 in the RME configuration settings.  I left MIR set to "Def".

3.  Venue Animation Speed is set to 100% (it already was I didn't have to do anything).

4.  I was wrong earlier.  My monitor will in fact go WAY BEYOND 1024 x 768.  I set it to 1280 x 1024, 32bit color 75hz and MIR fits in the screen.  Wow is all of this text etc small LOL!

5.  Unfortunately (actually fortunately lol) I haven't been able to reproduce this problem but once.  I stopped the engine, and started it a few times in a row and it went bonkers and the never ending distortion started.  Then I gracefully stopped the engine one more time, waited a minute, started it up and let it get time to catch up and viola it worked fine.  I don't consider this a bug, it was me purposely trying to screw it up since I knew it couldn't possibly catch up to me stopping and starting the service so many times that quickly.

6.  My MIR Buffer multiplier is set to 2X at the moment, is that correct?  (I have 2 processors with dual cores each, total 4 cores).

7.  MIR Venue detail image level is set to 1X, is that good or should I set it to .5 or .25?

FIRST TEST - Ok after setting things to 1024 in RME and left MIR at "Def", it worked and I couldn't reproduce the problem until I did step 5 (force MIR to go bonkers by stopping and restarting the engine).  However, it eventually evened itself out after a minute.  I was hitting about 70% CPU here with 14 instruments going nonstop.

SECOND TEST - Set the latency to 256 samples in RME, Left MIR at "Def".  Was pushing 100% CPU the entire time.  No matter what I did, this one was in the red constantly.  This is not a good setting for my system lol.

THIRD TEST - Set the latency to 2048 samples.  System worked flawlessly, MIR was rock solid and never distorted once (I am actually in loop mode right now listening to it streamed from Logic over MOL, I am not using Remote desktop, I am physically walking from MAC to MIR machine).  The CPU is not hitting more then 30-40% while playback.  It is hitting about 60-65% when I start moving the instruments around and rotating them BUT it never distorts or stops and crashes.  After I STOP rotating the instrument or moving it, it returns to a very comfy 30% CPU.

So, Third test is where I am leaving it for now.  I assume it had a lot to do with the Latency settings mixed with low video resolution.  It seems fixed but I have one little bug now that maybe was there but didn't notice it.  I am getting stuck instruments and notes when the music stops etc.  Example would be, at the end of the piece, the violas are still playing.  I go STOP Logic from playback, and yet MIR is still playing notes for a few instruments.  I fix it by stopping the engine and restarting.

Other then that, it all seems to work flawlessly on my AMD machine :).  Do I get a cookie?

Are there are any tests or information you want from me with this AMD system since it sounds like yours don't work as stable as mine?

Maestro2be

Yes.  I made that apparent from the first day that I use an AMD processor :) and it was in my signature lol.  When I spoke to Christian we knew there was a possibility of a few issues but that he would see if he could get the programmers to help out.  And as you can see, there really wasn't any real incompatibility between AMD and MIR, it was just a few settings tweaks that needed to happen (at least in the case of my AMD setup).

I ended up using the system all night last night flawlessly.  It definitely uses a lot of resources and requires a fast memory bus.  That is definitely the weak chain in my link and I suspect in most setups.  I haven't pushed it to the max yet though so it will be interesting to see how many instruments I can get at once.  So far with the 14-15 instruments it was only at 30-40% CPU with the latency of 1024 or 2048.  If I can get a good 20-25 out of my setup, I will be very happy as I don't really use more then that on any of my projects at the moment.  I will test that tonight if possible.

I really appreciate the help Dietz :) in helping me get this to work.  I think I also figured out why the Remote Desktop was not acting properly.  On my MAC, I have a second monitor attached which is sitting on my digital piano so that I can watch what's going on while I record.  Well, the resolution on it for some reason was only set to 800x600 and it was affecting the remote desktop session (putting it in 800x600 mode).  That would definitely cause GUI issues with MIR.  I am going to test a higher resolution tonight and I bet it works.

Any chance you could answer the last few settings questions I had above please :)?  (the Venue detail and Buffer size).

Maestro2be

Yes, now I see it too - I think I just didn't _want_ to see an AMD as part of the equation. 😉

@cgernaey said:

[...] 

6.  My MIR Buffer multiplier is set to 2X at the moment, is that correct?  (I have 2 processors with dual cores each, total 4 cores).

7.  MIR Venue detail image level is set to 1X, is that good or should I set it to .5 or .25?

[...]

Larger buffer sizes will just allow MIR for more "time" to process IRs. Think of it as additional latency. In general, try to use buffer-settings as low as possible for a project you work on (this will of course depend a lot on its size).

Venue detail should be left at 1x (full resolution) under normal conditions. This parameter is here to help in special situations like Remote Desktop incompatibilities and the like. Animation Speed can be set as low as 20 or 30% for smooth zooms into the Venue Map. Raise it again if you encounter unforseen problems.

HTH, 

Thanks Dietz :).

For what it's worth, this application is simply unreal.  You guys did such an inbelievable job.  It's so advanced it's quite mind blowing but yet the learning curve was almost nothing (at least for a techie guy like myself).  I found it WAY easier to learn then the standard Space Designer, send to 28 different aux channels, then through 18 different layers to get different depths and other stuff...  It is just way to complicated that way (I know it works but to me it's to much of a time waste).

I am going to go ahead and dive in Dietz.  You guys have been amazing to me over the years and the products first class.  I have some open questions with Christian and Herb on this but I am going to go ahead and purchase it and let them think about it and get back to me.  I will be sending you guys my purchase email in a few days.

So be ready for more questions from me because I am going to stop with the old way of doing it as it's to cumbersome.  Going to change my workflow and setup into a MIR solution.  Just a few days of demoing it and I can already see how it will change my musical creation experience.  It will certainly speed up my song writing and I can't wait to see what advancements you all do with it.

Maestro2be

:-) Thanks for the positive feedback! Always good to hear.

@cgernaey said:

Are there are any tests or information you want from me with this AMD system since it sounds like yours don't work as stable as mine?

Hi Maestro2be,

good to know that MIR finally works on your machine!

Getting a few additional numbers concerning performance (how many instruments run without clicks at which buffer size with which CPU load?) would be great. Just to get a better feeling how Intel-optimized code works on AMD and how much benefit we get from better memory access on I7 machines.

Thank you, Florian