Singing Faces No Timing track in Timing Track text box,

Blue Alloy

New elf
Joined
Mar 28, 2014
Messages
8
Singing Faces No Timing track in Timing Track text box,
I have spent three morning trying to get this to work. I have read the manual and I did try selecting another effect first but that did not help. The singing tree is in the model display window blinking at me I have the face effect on my time line. But there is no timing track in the drop down Timing Track box to select? Can you help me get this to work? Please help
 

Gilrock

Full time elf
Joined
Jan 4, 2013
Messages
438
Location
Tucson, AZ
Have you created any Papagayo timing tracks where you did a breakdown of phrases into words and phonemes?
 

Wolfie

Full time elf
Joined
Nov 22, 2015
Messages
121
Location
Milwaukee, WI
I feel your pain. I just went through this and there will be no timing listed in the face effect unless you have gone through the prerequisites first.

You need a timing track with markers on it.
Put the lyrics in the track by double clicking the segments and typing or copy/pasting the lyrics.

Its still not going to show in the list yet because its not a Papagayo track...yet.
Right click on one of those segments and select "Break down Phrase".

That will convert the track from a basic timing track to a Papagayo track and you will see the small green parrot icon appear on the track title block on the left of the sequence window.

Once you have the phrase broken down into words and the parrot icon, you can select the face effect and the timing track(s) that have parrots will appear listed in the dropdown.

It won't do anything yet though. But it will be listed.

You must then right click on each word and "Breakdown word". That will generate the phonemes. At that point the faces effect will drive the faces segments defined in the layout for that prop.



Thats the gist of it but there is a bit more detail (from my very limited experience) that I would like to add.

You can paste in whole blocks of lyrics for a whole song if you wish. My recommendation is, don't. It looks like it will save a lot of time. But it, at least for me causes a lot more work getting it all lined up. You can break down a whole track's worth of phrases. Don't. Again, more work later with stuff in the way. About the only bulk one I would vaguely attempt is breaking down all the words into phonemes but I don't do that either.

Here is my method and it works for me. Maybe for you but try it and see.

First, hit a lyric site and pull up the lyrics. Dual monitors helps here.

Depending on the song, it may be helpful to run the song through Audicaty and do a vocal filter on it. It will sound like crap but filter the hell out of it so you can hear the vocals only. Don't trim the length or anything. Now replace the track in the sequence with this butchered version (keep the original!). Now you can work with the vocals much better in xLights. Obviously not every song will need butchering like this. Some don't have heavy instrumentation behind or on top of vocals. If thats the case, just roll with the original.

Now, play the song at slower speeds (3/4 or 1/2 speed). With your new lyric track selected, listen carefully as you read along with the lyrics being sung. At the beginning and end of each line of lyrics hit "T" to set timing marks at the beginning and end. Don't worry about getting them dead nuts on right now. Roughly hew out the segments in the song. You are grossly marking out areas right now not looking for millisecond accuracy. Yet. That comes later.

Now with all the phrases marked, rewind. Now listen carefully for the vocals. Pause, rewind, play as a loop. Adjust the timing marks for the phrases so they are right on the money. Lots of the time you can look at the waveform (especially if you filtered the heck out of it for vocals) as indicators where things start and stop. Once you got the start/stop markers precise, double click it and copy/paste the line from the lyric site. Click the spacer sections between phrases and delete that section of the track leaving gaps. Move to the next phrase and repeat until you have the whole song's phrases spot on for timing.

Now play it at 3/4 or faster and double check your work. You will know when you got one wrong by watching the scrubber as it crosses the marks while listening to the song.

Rewind. Right click the first phrase and "Break down Phrase". This generates the words within it. Problem is, it will distribute the words evenly, not based on length, but based on the number of words. For example a phrase like "In my neighborhood" will be distributed with each word consuming 33% of the time of the marked phrase despite the fact that the first two words are extremely short compared to the last word.

Now, again monitor the song and scoot the timing between words so they are dead nuts on with the vocals. Some phrases will have vocal pauses or spaces between words. For example the phrase "Hey you! Get away from there!" of Pink Floyd's The Wall has a distinct pause between Hey and you. You can rip apart the start/end of words when the cursor is a pointer finger just like when dealing with effects on props. Click and drag and you can add spaces between words so they match with how its sung, not how its written. This will make the face go to a rest state between words which makes it look better IMHO.

Another reason to break words apart with spaces is for words that may use the same phonemes together. If one word ends in a O phoneme, and the next word starts with an O, split the words so you get a rest state between the O's. Say "Go Over". Feel your mouth with your finger when you speak the words. Feel that slight change in your mouth at the end of Go and the beginning of Over? Thats the rest state you need between the words in the timeline! Split the Go and Over on the timeline (or split the two O phonemes) and you will see a difference in realism. Again, IMHO.

At this point I right click and break down each word into phonemes for a single phrase. And again, more tweaking based on the song. When singing, many words have syllables that are carried longer. Now scoot the timings of the phonemes so they match how the words are sung. For example, "Its beginning to look a lot like Christmas". The first S in Christmas is carried longer sounding like Crissssmaas. So spread out the phenoms so they match. Compress some, expand others to maintain the word timing.

This is when I first put a face effect on a prop. Once I have the first phrase down all the way to phonemes. Now you can see what the face is doing and you can compensate your timing with a visual aid now. I have a controller in the computer room with a singing face so I can actually see blinkies. Playing the song at 100% speed and watching the display. If you "feel" something isn't quite right, its NOT! Fix it!

Now comes the advantage of not breaking stuff down in bulk....

Chorus sections of songs are frequently so darn close if not identical, copy/paste is your friend! Once you mark out a chorus section, break down the phrase, and get it all the way to phonemes, you can copy it. Problem is I don't have any clue how to drag and highlight a whole section of a timing track. If it can be done, I haven't figured it out. So using CTL and clicking on the phrase, then all the words, then all the phonemes you can select them all. I know, its tedious, but its worth it and can a few minutes of tedium can save you hours of repetitive work later. Then copy. Open notepad (or any plain text editor) and paste into it. You will get a bunch of text out of xLights. This is the gold nugget! Now you have a whole chorus section in your notepad!

Continue to work down the song breaking down the phrases, words, and phonemes. When you get to a repeat of the chorus again, go to your notepad and CTL-A to select all. Then copy. Return to xLights and delete the chorus phrase (don't break it down!). Make sure your scrubber is at the beginning of the chorus phrase(s). Now right click and Paste. Poof! You have the whole chorus in place and broken down all the way to phonemes. Do any minor tweaking as needed for variant timing.
 

Gilrock

Full time elf
Joined
Jan 4, 2013
Messages
438
Location
Tucson, AZ
Unselect the timing track when you need to select a block of timings to copy/paste. That will allow you to drag a yellow dashed box around them for your selection.
 

Wolfie

Full time elf
Joined
Nov 22, 2015
Messages
121
Location
Milwaukee, WI
Thats sick! I will try that when I get home tonight! Thank you Gil.

Any other tips on the process I am using?
 

Wolfie

Full time elf
Joined
Nov 22, 2015
Messages
121
Location
Milwaukee, WI
Gilrock said:
Unselect the timing track when you need to select a block of timings to copy/paste. That will allow you to drag a yellow dashed box around them for your selection.

Ok, tried that. Didn't work. I had to unselect ALL timing tracks. Even the non-lyric ones. Then I get a yellow rectangle. But if ANY timing track is enabled, even though I am not trying to select from it, the draw rectangle doesn't work.

But at least I got a much easier select method now! Thank you again Gil.
 

Blue Alloy

New elf
Joined
Mar 28, 2014
Messages
8
Thanks for you reply’s
I do have the Papagayo Icon on the timing track. I just started a new Sequence, I added a Timing track, added a label to it with, “If you look up in the sky” I right clicked on the words and selected break down phrase. I got the Papagayo Icon on the timing track. Then I added the Face Effect, to the Model Singing Tree 1. the tree is blinking at me. But when I select the Face effect on the time line there are no Timing Tracks in the selection box to select. The Timing track is not in the Timing track Selection box?
 

Attachments

  • Xlights Faces.JPG
    Xlights Faces.JPG
    362.5 KB · Views: 20

Gilrock

Full time elf
Joined
Jan 4, 2013
Messages
438
Location
Tucson, AZ
A timing track is not valid to be used in the Faces effect until it has 3 layers. Sounds like you did a breakdown of phrases to words but you didn't break down words to phonemes. You can right-click the row heading to breakdown the entire row at once. I usually work one layer at a time. I arrange all my phrases then right-click row and breakdown phrases. Then I arrange all my words and then right-click row breakdown words. If you have a repeated section just don't spend any time arranging the words...copy the words from the previous area first. It's still quicker to do only one breakdown step per row.
 

Gilrock

Full time elf
Joined
Jan 4, 2013
Messages
438
Location
Tucson, AZ
Wolfie said:
Gilrock said:
Unselect the timing track when you need to select a block of timings to copy/paste. That will allow you to drag a yellow dashed box around them for your selection.

Ok, tried that. Didn't work. I had to unselect ALL timing tracks. Even the non-lyric ones. Then I get a yellow rectangle. But if ANY timing track is enabled, even though I am not trying to select from it, the draw rectangle doesn't work.

But at least I got a much easier select method now! Thank you again Gil.

I'm sorry I forgot with you I gotta be specific. Yes you need to unselect "ALL" timing tracks to get into that selection mode. Otherwise you are still in grid selection mode. Technically you were never supposed to be able have more than one timing track selected at a time but it happened by accident and users threatened revolt if we "fixed" it.
 

Blue Alloy

New elf
Joined
Mar 28, 2014
Messages
8
Thanks Gil, that fixed the problem. I have just started using Xlights this summer, I have so much to learn, It is a great program Thanks for all your work. Are there any free example sequences, that can be download? Are the file at this link http://nutcracker123.com/nutcracker/sequences/ sequences or timing tracks? How do you use them is there a how to video somewhere?
 

Wolfie

Full time elf
Joined
Nov 22, 2015
Messages
121
Location
Milwaukee, WI
txt files are timings (or notes)
xml files are sequences
aup files are audacity project files (you need the correct mp3 to use it)

Basically many will trim songs down removing sections, trimming lead in or outtro stuff to shorten the song. You need the original MP3 they used, then you can load the AUP file in Audacity and you can then export the revices MP3 for use with the timing tracks and sequence. If there is no AUP file, they used the MP3 as is.

You can import the text files into timing tracks.

XML files being directly what xlights uses, you can open those directly.

I posted my timing tracks to Witch Doctor along with the lipsync tracks:
http://doityourselfchristmas.com/forums/showthread.php?43957-Witch-Doctor-%28Radio-Mix%29

I found these free ones a few months ago:
https://github.com/cp16net/xlights-sequences

That should get you started.

Sean has a ton of videos out on xLights here:
https://vimeo.com/nutcracker123
 

Blue Alloy

New elf
Joined
Mar 28, 2014
Messages
8
Hi Wolfie, Thanks for the links, and the Information. I will check the links.
 

Chad

New elf
Joined
Dec 27, 2018
Messages
31
Location
Brisbane, South
OMG, last night I spent 2 hours manually creating faces for the first part of my song. And today I found this thread.
Now I have it done in 2 minutes. What a time saver.
 
Top