DiyLightAnimation
Software => Nutcracker: RGB Effects Builder => Topic started by: smeighan on September 29, 2014,
-
Release 3.4.15. Now you can stitch multiple parts of a voice together. You can also merge multiple xlight sequences.
Here is a video showing multiple voices, from multiple files that shows xlights can now create singing faces
http://nutcracker123.com/nutcracker/images/thriller.mp4
XLIGHTS/NUTCRACKER RELEASE NOTES:
=================================
3.4.15 Sep 28, 2014
- ENH (Sean) merge_xml.awk will combine multiple xlights xml sequences into one.
- ENH (Sean) corofaces.awk awk now will stitch together multiple files for the same voice.
- DOC (Sean) New documentation, README,corofaces. You will see thsi when u install the
latest code
3.4.14 Sep 26, 2014
- ENH (Sean) scripts added that can take Papagayo voices and make xlights sequences from them
- ENH (Sean) Scripts are found in your installation directory. corofaces.awk, coroface.bat
- ENH (Sean) Installation of xlights now installs gawk.
So how to do Papagayo (This is interim method until Don Julien finished up Papagayo tab)
1) Take a wave file, example thriller.wav
2) Break it into 3 chunks using audacity.
Example
thriller1.wav
thriller2.wav
thriller3.wav
3) Use papagayo and mark up your voices.
save thyem as pgo files.
thriller1.pgo
thriller2.pgo
thriller3.pgo
4) Go into Papagayo and export each voice into a *.dat file
Example , there are three voices.
voice1a.dat
voice1b.dat
voice1c.dat
voice2a.dat
voice2b.dat
voice2c.dat
voice3a.dat
voice3b.dat
voice3c.dat
the "*a.dat, *b.dat,*c.dat indicate that these came from 3 different pgo files
You now need to stitch each voice back together.
Look at line 4 in each pgo file. it will show the last frame that each file used.
5) Copy corofaces.awk, coroface.bat,merge_xml.awk from C:\Program Files (x86)\xLights and put them into your show directory
6) edit the batch file corofaces.bat
rem
rem edit these next lines and put in what your voice dat files were from Papagayo
rem First copy coroface.awk, coroface.bat and merge_xml.awk from the
rem
rem Syntax" gawk -f coroface.awk -v m1=xxx -v v1=nnn voice1.dat voice1b.dat > voice1.xml
rem edit these next lines and put in what your voice dat files were from Papagayo
rem arguments
rem "-v m1=face1". -v means you will define a variable. m1 is the variable. it contains
rem the model name for voice1
rem "-v v1=3649". This is the last frame in the first pgo file. Look on line 4 of
rem your pgo file. it will show the last frame.
rem
gawk -f coroface.awk -v m1=face1 -v v1=3649 -v v2=910 voice1.dat voice1b.dat > voice1.xml
gawk -f coroface.awk -v m1=face2 -v v1=3649 -v v2=910 voice2.dat voice2b.dat > voice2.xml
gawk -f coroface.awk -v m1=face3 -v v1=3649 -v v2=910 voice3.dat voice3b.dat > voice3.xml
after you have made an xml light sequence for each voice, we want to join them all together
em edit this command to contain all of the xml files you need to combine
rem Syntax: gawk -f merge_xml.awk {firstfile} {secondfile} {thirdfile} {.etc.} > {output xml file)
gawk -f merge_xml.awk voice1.xml voice2.xml voice3.xml > voice.xml
7) Now open the original wav file (before you chopped it itno 3 pieces)
Use the 3rd open option. give it the wav file name and then the xml file. (voice.xml in our example)
8) save the sequence
-
You might want to change the name of the file README.coroface to README.coroface.txt as it will usually show up as an error since Windows does not know what to do with a file extension of "coroface"
-
You might want to change the name of the file README.coroface to README.coroface.txt as it will usually show up as an error since Windows does not know what to do with a file extension of "coroface"
i did change it.
thanks
-
I can't seem to get it to display two channels for a single phoneme....
I've tried 3.4.14, 3.4.15, and 3.4.16. It looks like starting with 3.4.15 corofaces replaces faces, but the behavior is the same for me. It will display a single channel just fine, but as soon as it specifies 2 channels it seems to ignore the second one.
Here are some screenshots to help illustrate....This is in 3.4.14, but it seems to be the same issue in all 3 just the change from Faces to Corofaces.
-
that is not working Dan. Don Julien has a fix in mind.
The Papagayo tab is NOT released yet.
If you need faces in the next week use thw awk scripts that i talk about on the tutorials.
In about a week those scripts should be able to be thrown out.
Prompting for multiple channels is one of the big UI issues we a re working'thanks
sean
-
I am using the scripts....I thought I could preview the results on the nutcracker tab which is what I was doing, but I noticed the preview is not real good since it won't turn on multiple channels....perhaps I misunderstood the tutorial?
-
ah, i do turn on the two x,y locations but the second one does not light up.
So when i do an AI, i only get the upper lip.
When Don does his testing, he gets both to work. Since we are expecting don;s version sometime next week, maybe use my scripts. check your animations work., refine your pgo file .
by that time don should have the coroface portion working. He just solvedhow to make the drop down in the papagayo tab allow multiple channels to be checked per phoneme
-
So I exported the sequence to Vixen and it does not look like the two channels are being turned on there either. I know the script had the option for two channels, and it looks like it sets things correctly in xLights/Nutcracker, but can't quite seem to get the final product out.
-
I guess I misunderstood your post and thought you meant it was just a display issue in xLights, but I think you may be saying what I am seeing is correct until Don get's his release out?
-
I guess I misunderstood your post and thought you meant it was just a display issue in xLights, but I think you may be saying what I am seeing is correct until Don get's his release out?
That is correct, my awk script was not turning on the second channel. I tried Don Julien's latest code last night, it does turn on both channels.
i just put up version 3.4.17. This has the corofaces part of the Papagayo tab working. This release is PRE-RELEASE. There are some minor bugs. One we know about, the outline shows in the NUTCRACKER tab but not in the PREVIEW tab.
1) Select a pgo file
2) Select an output file to receive the xml sequence
3) Select the custom models that you are using. Be user "Part of my Display" is checked for those models
4) Setup channel assignment to Phonemes. You can add multiple channels to a phoneme by doing ctl-click.
5) Create an output sequence
Now go to NUTCRACKER tab
1) Do a file open, 3rd option
2) check the XML box
3) pass in wav file used for the Papagayo process
4) Select the xml file that you just created above.
5) Save this new sequence.
-
First time this has happened, so before I turn off my AV software thought I'd ask if anyone else is having this happen....
Kaspersky is identifying the file as malicious with certainty and immediately deleting it. It is suggesting it has HEUR:Trogan.Win32.Generic.
I realize these can be false positives, but wanted to check in.
-Dan
-
My computer was reporting that the file did not have a digital signature to validate it. I had to go around the system to get it installed.
Larry
-
Me too. Win8 64 didn't trust it.
-
I would be careful installing that software.
-
i think it was a false alarm.
I downloaded kaspersky and ran it. 9 hours, 3 1/2 million files later it did say that it found one file with a virus, the binary file i had for 3.4.17.
We have worked on 3.4.17 to fix the face outline in the PREVIEW window. I recompiled and created a new 3.4.17, it shows clean now.
I have not run any software to clean my machine, just change some of the source files around. I must have created a binary signature that looked like a virus.
The new 3.4.17 has been posted (This has the fix for the Outline of a face for Papagayo).
Download this and see if it is clean with your virus software.
BTW, 3.4.17 is not released officially until later today. I put it up for a few people to try it.
thanks
sean
-
Mouth channels look good in preview mode (after generating an xlights sequence)....they don't seem to show on the nutcracker tab though.
Random eye blinking does not seem to be working, though I may not have done it right. Actually I'm not sure how to make my eye channels come on at all. :)
Also on the Papgayo tab when you are defining the channels for each phoneme if you hit enter while the channel list is dropped down it will crash xlights.
Looks good though, and will be a huge time saver!
-
Mouth channels look good in preview mode (after generating an xlights sequence)....they don't seem to show on the nutcracker tab though.
Random eye blinking does not seem to be working, though I may not have done it right. Actually I'm not sure how to make my eye channels come on at all. :)
Also on the Papgayo tab when you are defining the channels for each phoneme if you hit enter while the channel list is dropped down it will crash xlights.
Looks good though, and will be a huge time saver!
we are not done. to make the eyes blink, you need to have one channel for the top and another channel for the bottom. in my case channel 8 is top of eye, channel 9 is bottom.
When I define eye open it is 8+9
eye closed 9
we are wroking on the fade out. if a voice is not singing for xx seconds, fade it out.
we should be doing another release by monday
-
i think it was a false alarm.
I downloaded kaspersky and ran it. 9 hours, 3 1/2 million files later it did say that it found one file with a virus, the binary file i had for 3.4.17.
We have worked on 3.4.17 to fix the face outline in the PREVIEW window. I recompiled and created a new 3.4.17, it shows clean now.
I have not run any software to clean my machine, just change some of the source files around. I must have created a binary signature that looked like a virus.
The new 3.4.17 has been posted (This has the fix for the Outline of a face for Papagayo).
Download this and see if it is clean with your virus software.
BTW, 3.4.17 is not released officially until later today. I put it up for a few people to try it.
thanks
sean
I have updated to 3.4.17, is there something I need to worry about? windows did not want to run it but I ran it anyway. that is not the first time I have had to do that.
-
Mouth channels look good in preview mode (after generating an xlights sequence)....they don't seem to show on the nutcracker tab though.
Random eye blinking does not seem to be working, though I may not have done it right. Actually I'm not sure how to make my eye channels come on at all. :)
Also on the Papgayo tab when you are defining the channels for each phoneme if you hit enter while the channel list is dropped down it will crash xlights.
Looks good though, and will be a huge time saver!
we are not done. to make the eyes blink, you need to have one channel for the top and another channel for the bottom. in my case channel 8 is top of eye, channel 9 is bottom.
When I define eye open it is 8+9
eye closed 9
we are wroking on the fade out. if a voice is not singing for xx seconds, fade it out.
we should be doing another release by monday
Yeah. I had the channels defined, but the eyes never showed up.
-
Stitched 8 PGO files together...Probably obvious, but wanted to add that you need to keep adding whatever is in line 4 of the PGO file as you continue to "stitch".
I did get some messages about duplicate phonemes even when I had the numbers right....have not had a chance to dig into that.
Any chance future versions will read line 4 automatically to know where to stitch and keep the running total for you? Not a big deal, but would be nice.
That said....still finalizing my workflow, but so far this was a huge step forward in time savings! Thanks! :)