DiyLightAnimation
Software => Nutcracker: RGB Effects Builder => Topic started by: kgustafson on December 11, 2012, 11:27:36 am
-
Okay,
I just ran through the multi UserPattern generation for projects (local installation) and was able to generate a full song, copy each UserPattern without issue (there may be some minor problems with the last frame as I see some 'gap'ing) . I was able to load up each UserPattern without issue.
Here is how I did it:
- I ran Nutcracker and set up a project (Xmas Eve at Sarajevo) on my matrix (I am using a 16X25 smart string matrix this year on my bay window).
- I selected a series of effects that I had created earlier (mainly color washes, meteor and snowflake effects).
- I submitted the project with output for LSP.
- I saved each of the four generated XML files as is to the LSP Sequencer sub dir (under x86 Windows->Minleon->LSP Sequencer
- I renamed the first of the four XML files to UserPatterns.xml
- I started LSP Sequencer and configured it for my matrix and with my song with the same framerate (50ms)
- I highlighted the UserPattern-1 under Nutcracker-1 in the Patterns menu
- I selected the first cell in the sequence (top left most) and pasted it using the first two check boxes for the paste - Snap to nearest timings and Use clipboard timings
- The pattern pastes into the sequence for the first minute (the UserPatterns are broken up by minutes)
- I save sequence and quit LSP sequencer and open windows explorer on the subdir where the UserPatterns are kept
- I rename UserPatterns to UserPatterns_old1
- I rename the second generated UserPattern to UserPatterns
- Open LSP Sequencer
- Go to Patterns and highlight UserPatterns-2 under Nutcracker-1
- I highlight the top most cell in my sequence at 1 minute mark
- Paste and the next minute of sequence is pasted in
- Save and quit LSP
- Rename third gen'd UserPattern to UserPatterns
- Start LSP and load sequence
- Paste pattern Nutcracker-1->UserPattern-3 at the 2 minute mark
etc.
This seems to work.
Kurt
-
Does LSP get progressively slower doing the pastes? I've found it almost unusable trying to paste Nutcracker patterns after my LSP sequence gets more than a few patterns in it. It can take upwards of ten minutes to finish the paste.
-Mike
-
Does LSP get progressively slower doing the pastes? I've found it almost unusable trying to paste Nutcracker patterns after my LSP sequence gets more than a few patterns in it. It can take upwards of ten minutes to finish the paste.
-Mike
Yes,
This is why we broke up the big UserPatterns into smaller files (1 minute chunks). These are not exactly fast pastes, but they are tolerable.
Kurt
-
I hope everyone realizes I am using project and not effects here. I have yet to fix the effects code.
Kurt
-
I hope everyone realizes I am using project and not effects here. I have yet to fix the effects code.
Kurt
There is nothing to fix in the effects code. It generates one effect and then allows the users to pick how many of their effects they want into UserPatterns.xml. I have already suggested a person should not select more than 10. So users end up with 3oo mbyte UserPatterns.xml. The downside is that you need to pick each effect, you need to keep track of what are the next 10 effects .etc. Projects fixes trhis selection issue by putting all effects from a song into the output file.
In projects it takes all the effects in a song (30 to 35 in some songs) and builds one large UserPatterns.xml. This one large , > 1 gig file, is what you are fixing.
Now having LSP slow down while pasting, i dont know how to fix that.
Long term, we should not makde a UserPatterns.xml but a single msq file. No pasting needed, entire loads with a single command.
That will be next year ...
thanks
-
Okay. Thanks for the clarification Sean. I think I am running out of bugs to fix. I am adding a snowflake effect to my voice over for the show at the moment. I am also going to play with animated gifs, now that they are working (haven't worked with them since the problems were fixed).
I am also going to continue to decompose MSQ files. I am getting there slowly. Having a working key for LSP has been helpful.
Kurt
-
I might be a release or two behind - but my local install nor the network version is not generating a LOR .lms project properly. I can create effects OK but not the final project. The issue is that I have a 30 x 42 pixel tree but the resulting LMS file (not the animated GIF or the little dialog box you see that describes the output) ends up with 30 x 30 pixels.
User: budude
Model: MTREE-30S-42P
Either project that I have in my lists are not working. I can't retest at the moment as I'm at work but I've tried it several times. Again, creating effects with the same model and effects in the project works fine. I'll try it again later tonight.
-
I'm regenerating my local effects with consistent timing, and will be giving this a shot tonight. When using projects, does it matter how long the effect duration is in the effect definition?
-
I'm regenerating my local effects with consistent timing, and will be giving this a shot tonight. When using projects, does it matter how long the effect duration is in the effect definition?
Nope, Kurt overrides the sequence duration. He calculates exactly how many frames you need and launches the effect generator overriding the effects sequence duration.
he also allows you to use the same effect throughout a song even though each re-use could require a different seq duration.,
this is really cool in projects. So when i create new effects i just set the seq duration to like 2 seconds.
I am doing something like 30-60 p[roject generations per day.
thank you kurt for projects!
sean
-
Local install is leaving a bunch of .dat files in the C:\wamp\www\Nutcracker\effects\workspaces\529\ folder. I seem to remember these should clean themselves up.
Also, if a project is regenerated, it says no file found when it should have found it:
5 of 20 phrase(s) processed.
maxFrames=435,garland_gap=5
1 Generating BARBERPOLE No file found Generating workarea/Voltorb~MT30~BARBERPOLE~86.nc 25.329 secs
2 Generating BUTT_GREEN_BLUE No file found Generating workarea/Voltorb~MT30~BUTT_GREEN_BLUE~84.nc 55.33 secs
3 Generating MT_RD_GARL No file found Generating workarea/Voltorb~MT30~MT_RD_GARL~126.nc 255.397 secs
4 Generating BARBERPOLE_180 No file found Generating workarea/Voltorb~MT30~BARBERPOLE_180~82.nc 25.485 secs
5 Generating BUTTERFLY3 No file found Generating workarea/Voltorb~MT30~BUTTERFLY3~127.nc 51.18 secs
It seems to be using the user name instead of the user number (in my case 529). Shouldn't workarea be workspaces?
Oh, THANK YOU!
-
Local install is leaving a bunch of .dat files in the C:\wamp\www\Nutcracker\effects\workspaces\529\ folder. I seem to remember these should clean themselves up.
Also, if a project is regenerated, it says no file found when it should have found it:
5 of 20 phrase(s) processed.
maxFrames=435,garland_gap=5
1 Generating BARBERPOLE No file found Generating workarea/Voltorb~MT30~BARBERPOLE~86.nc 25.329 secs
2 Generating BUTT_GREEN_BLUE No file found Generating workarea/Voltorb~MT30~BUTT_GREEN_BLUE~84.nc 55.33 secs
3 Generating MT_RD_GARL No file found Generating workarea/Voltorb~MT30~MT_RD_GARL~126.nc 255.397 secs
4 Generating BARBERPOLE_180 No file found Generating workarea/Voltorb~MT30~BARBERPOLE_180~82.nc 25.485 secs
5 Generating BUTTERFLY3 No file found Generating workarea/Voltorb~MT30~BUTTERFLY3~127.nc 51.18 secs
It seems to be using the user name instead of the user number (in my case 529). Shouldn't workarea be workspaces?
Oh, THANK YOU!
Blame the differences on the author (aka me LOL). I use a subdir called workarea (not workspaces) under project. I used username vice member_ID for readability (and did not do Sean's better solution of putting a sub-dir of member_id under the workarea). In the next release, I think we should probably create an area under nutcracker called workspace and have it sub-divided by member_id and that all work from both effects and projects go under here. But that is for a later release.
If you are curious about the NC files they are gen'd this way: Username~Target~Effect~NumberOfFrames
If the file is not found, that means that a) This the first time for the effect to be generated or b) the number of frames changed -- meaning the phrase length was changed somehow or c) the program had an issue when generating the effect the first time and did not create an NC file the last go round.
Kurt
Kurt
-
I think I am running out of bugs to fix....
Kurt
Kurt,
Were you able to take a look at the "interval" issue I PM'ed you about? The one that causes the final effect cells in a LSP UserPatterns file to be abnormally long?
-Mike
-
I think I am running out of bugs to fix....
Kurt
Kurt,
Were you able to take a look at the "interval" issue I PM'ed you about? The one that causes the final effect cells in a LSP UserPatterns file to be abnormally long?
-Mike
Mike,
I know what is causing the issue, but I think this issue is appearing in effects, correct? (I don't think you are using projects, because I have not been able to re-create this issue in projects.) You are doing this generation under the 'effects' menu item, correct?
I know that there is a parameter set called MaxInterval that is set to 100000 that is probably causing the issue. Projects does not use this variable (I ripped the code to generate LSP from effects, but modified it to be used by projects). I will have to look at the effects generation tonight and see if I can't find (and fix) the culprit there.
Kurt
-
Kurt,
Yes, I'm using effects exclusively. Haven't ventured into projects at all.
-Mike
I think I am running out of bugs to fix....
Kurt
Kurt,
Were you able to take a look at the "interval" issue I PM'ed you about? The one that causes the final effect cells in a LSP UserPatterns file to be abnormally long?
-Mike
Mike,
I know what is causing the issue, but I think this issue is appearing in effects, correct? (I don't think you are using projects, because I have not been able to re-create this issue in projects.) You are doing this generation under the 'effects' menu item, correct?
I know that there is a parameter set called MaxInterval that is set to 100000 that is probably causing the issue. Projects does not use this variable (I ripped the code to generate LSP from effects, but modified it to be used by projects). I will have to look at the effects generation tonight and see if I can't find (and fix) the culprit there.
Kurt
-
Okay,
I think I found the issue and corrected it. Once sean uploads make_lsp-exec.php to meighan.net, it should correct the extra 100000 entry (the larger than normal cell). Please let me know if it is still 50ms off. This might be a logic error in how we are determining the frames and is pretty easy to correct if it is incorrect.
Kurt
-
Okay,
I think I found the issue and corrected it. Once sean uploads make_lsp-exec.php to meighan.net, it should correct the extra 100000 entry (the larger than normal cell). Please let me know if it is still 50ms off. This might be a logic error in how we are determining the frames and is pretty easy to correct if it is incorrect.
Kurt
i released make_lsp-exec.php
thanks kurt for looking into this
sean
-
I finished my 2nd song this morning using the new code, and it's working wonderfully! It takes FOREVER to paste 12,000+ channels into LSP, but it works, and the generated effects run MUCH smoother in the LSP scheduler than effects generated inside LSP. Thank you both for all of your hard work. You've exceeded my expectations! Merry Christmas!
-
I might be a release or two behind - but my local install nor the network version is not generating a LOR .lms project properly. I can create effects OK but not the final project. The issue is that I have a 30 x 42 pixel tree but the resulting LMS file (not the animated GIF or the little dialog box you see that describes the output) ends up with 30 x 30 pixels.
User: budude
Model: MTREE-30S-42P
Either project that I have in my lists are not working. I can't retest at the moment as I'm at work but I've tried it several times. Again, creating effects with the same model and effects in the project works fine. I'll try it again later tonight.
I just realized I kinda missed this message. Was target fixated on LSP. Looks like lms might need some TLC. I haven't played with the new code for animated GIF that Sean released yet. So this may have been fixed with the new release of animated gif code.
Looking at the generated master NC file, it all looks good as does the cooresponding LMS file. Realize that you have your target where it expects to have 45 smart strings of 42 pixels each. Your description says that you have 30 strings of 42 pixels each, but this is not what is programmed for the target. Perhaps this is what is skewing your results. I believe the 45 needs to be changed to a 30, but I will not touch it as I am not sure what you intended. BTW, I am using the meighan.net values and I think you are running locally, so your target may be correct (but I have no way to check it).
Kurt
-
Hi Kurt
Thanks for all of your work and I hope you are feeling better.
Rick R.
-
Hi Kurt
Thanks for all of your work and I hope you are feeling better.
Rick R.
I am fine now, thanks. My older son has it now. Like a 24 hour stomach thingy.
-
Hey Kurt, Have you had a chance to fix the model for start at bottom or top yet? Thanks for all of you efforts. As a Newbie here I've been using the copied effects into LSP and ran some sequences live, after my shows were over, last night and they came out awesome. I guess now I need to dive into projects. Is there a tutorial for projects?
-
Hey Kurt, Have you had a chance to fix the model for start at bottom or top yet? Thanks for all of you efforts. As a Newbie here I've been using the copied effects into LSP and ran some sequences live, after my shows were over, last night and they came out awesome. I guess now I need to dive into projects. Is there a tutorial for projects?
Not yet. I will have to collaborate with Sean on this fix as I am positive that the fix may be involved (I am almost certain that this parameter will affect every effect type.)
Glad that your sequences came out well! It is what we work on Nutcracker for.
Kurt
-
Gif timeline in projects no longer displays on local install. I can see the gifs in the gallery fine. Tried Chrome and IE.
-
Gif timeline in projects no longer displays on local install. I can see the gifs in the gallery fine. Tried Chrome and IE.
Voltorb,
Know problem (to me anyways). Noticed this about three days ago. I broke something and need to fix the timeline now (I am certain that it is related to me removing the effect id parameter, which I think timeline is still looking for.) Two steps forward, one step back. Great way to code! *grin*
Kurt
-
Not a serious issue, just wanted to get it on the list. Thanks again, this is GREAT stuff!
-
Okay,
Major rewrite (well not really re-write, just removal of all references of the failed effect_id calls). It appears to be working, however, I have not tested this thoroughly.
One thing to note, this new code uses a new value for the effects in project (effect_name vice effect_id) so your projects will all appear to blank out! I cannot help this happening as I am removing the effect_id (which is the value that the project was saving and the reason some of you were having issues with project since the effect generation did not guarantee that an effect_id was generated).
Kurt
-
Kurt,
You nabbed the large final cell issue. All cells in the pattern are now the same size. I tested the total duration issue and every effect is coming up 2 cells short (i.e. 100 msec short when I use 50 msec timing). Not a huge issue as I can probably generate the effects 100 msec longer than I actually need, as a workaround.
Thanks for all your work on this.
-Mike
Okay,
I think I found the issue and corrected it. Once sean uploads make_lsp-exec.php to meighan.net, it should correct the extra 100000 entry (the larger than normal cell). Please let me know if it is still 50ms off. This might be a logic error in how we are determining the frames and is pretty easy to correct if it is incorrect.
Kurt
-
Kurt,
You nabbed the large final cell issue. All cells in the pattern are now the same size. I tested the total duration issue and every effect is coming up 2 cells short (i.e. 100 msec short when I use 50 msec timing). Not a huge issue as I can probably generate the effects 100 msec longer than I actually need, as a workaround.
Thanks for all your work on this.
-Mike
Okay,
I think I found the issue and corrected it. Once sean uploads make_lsp-exec.php to meighan.net, it should correct the extra 100000 entry (the larger than normal cell). Please let me know if it is still 50ms off. This might be a logic error in how we are determining the frames and is pretty easy to correct if it is incorrect.
Kurt
Still not perfect though. I will work on this 2 cell difference. I am sure it something relatively easy to fix, but will take some thought and concentration. I won't call it completely fixed yet.
Kurt
-
I noticed that most, if not all, of my effects come up 2 cells short. I don't use projects, I am just generating effects for Vixen. I just always generate an extra .10 to .15 seconds of my effect so that it goes the all the way. If it goes too far, I just delete the extra. I had not brought it up yet as I could get around it, plus I have not narrowed it down as to which effects did it and exactly how many cells were missing. My timings are 50ms.
Thanks,
-Keith
-
I noticed that most, if not all, of my effects come up 2 cells short. I don't use projects, I am just generating effects for Vixen. I just always generate an extra .10 to .15 seconds of my effect so that it goes the all the way. If it goes too far, I just delete the extra. I had not brought it up yet as I could get around it, plus I have not narrowed it down as to which effects did it and exactly how many cells were missing. My timings are 50ms.
Thanks,
-Keith
Consistency is a good thing (even when in error). I am more convinced that our logic used to generate the length of things is obviously off by 2 cells. I will work on this as a low priority issue as we have good workarounds for it currently.
Kurt