Expansion of What Can Be Bound to Faces Effect / Phoneme Tracks?

[TastyHamSandwich]

Hi all,

If this can already be done, I'd love to know how to do it:

Is it possible to expand on what can be bound to a Faces definition, for use on Phoneme Timing Tracks? Such as, for a given model + faces definition, bind each phoneme to an effect with set parameters? Like, for every AI phoneme, do a Plasma effect with such and such parameters, or do a DMX effect with such and such values for every MBP phoneme, etc. etc.

For my purposes, this would expand greatly what I can do with the lyric tracks for our fixtures, and, I believe, simplify and enrich what I can do in terms of controlling DMX fixtures on lyrical timing. Presently, I'm working on integrating a DMX laser to draw singing faces, but changing multiple DMX values on multiple channels for each different phoneme is laborious.

Just curious on how this can be done, or the feasibility of implementing something along these lines.
Thanks!

[Gilrock]

Research how to use the State effect to control DMX models.

[TastyHamSandwich]

Aye, but the State effect does not follow Phonemes. That's what it comes down to, for me. If there is a way to make it do so, that'd be fantastic. I know I can use Force Custom Colors on a Faces effect to accomplish sending a DMX value based off phoneme timing, but it's just a single value. Being able to stack and send multiple values across various channels without messing with lots and lots of individual fixtures and one-off Faces definitions is my goal.

[Gilrock]

Of course it can follow Phoneme's if you think a little creatively.

[Gilrock]

Paste the phoneme layer effects into a new timing track and then name your state's after the phoneme's.

[keithsw1111]

There is no chance I would build it the way you describe. As a tool maybe but even then I am a long way from being convinced of the utility. Do you have an example of a completed partial sequence showing what you think it would look like?

[Gilrock]

So I actually tested my idea and it works.  When you enter the state names apparently capital letters are not allowed but don't be discouraged I just typed in lower case and it still worked using the upper case phonemes.  So how did I do this?  I created a normal lyric track and broke it down.  Then I unselected that timing track and did a dashed yellow box selection around the 3rd phoneme layer.  Then I clicked where I wanted the start location to be in an empty destination timing track and hit control-v to paste.  You won't see any user feedback when you click to select the target location for the paste but trust it cause it will work.  Then you create a State set for the model and type in the phoneme names and define what you want set for each phoneme.  Then drop a state effect and point it to that timing track.

[TastyHamSandwich]

Big ups, Gilrock. I'll have to mess with that way and see what I can come up with. I *was* discouraged by the fact I couldn't use upper case letters when I tried this in the past, because my experience using phonemes and faces was that if the phoneme didn't use upper-case letters it wouldn't work, so I assumed this would be the case here too.

I've had much experience with the lack of user feedback when pasting timing track sections as well, as I've just sequenced singing faces for 14 songs by manually inserting all the phonemes myself, one paste at a time. I did it this way because our singing faces are A/C and only have a small number of mouth shapes. The auto-generated stuff was too far-off from correct for it to be worthwhile to generate it then fix it each time, so I just brokedown a one-word lyric track for each one and pasted the phonemes where I wanted them, and it turned out to be a much faster and more accurate process for me that way.

The reason I was asking about a feature such as this was because we've begun implementing a DMX-controlled laser (a Skywriter HPX Tour, to be specific) into our programming layout, specifically doing laser-based singing faces, and there are several blocks of DMX channels that need to be addressed for it to run properly, and it would've been more efficient from my end to be able to be able to handle it all from one face/state effect based on the lyric tracks, rather than dropping in some dmx effects globally and using the faces effect/force custom colors to modify the ones I need to change on the fly. I managed to get it working this way, still, since I had to, and it's fine for now, but I just thought it would be a cleaner and more elegant way to handle this beast.

I hope that made sense, haha. I have a tendency to ramble incoherently, I know. It's been something of a journey getting this thing patched in and controlled by xLights, but i DO have it working now, and it's pretty cool, if I may say so myself

[TastyHamSandwich]

Here's a screenshot of the described method I have used, and a way-too-detailed explanation of how I set it all up:

There is a DMX effect handling the first 7 channels, the first 6 of which are the laser's "Master" channel block, which handles things like scaling, offset, and intensity, plus the input source. The 7th channel is actually part of the "Builder" fixture in the laser, but it's the page of effects to draw from, so it won't change, and thus I handled it here.

The "Builder" fixture is 2 channels of the laser's "Builder" (actually 28 channels, in total). The Faces effect is just handling these two channels, as it is the pattern to draw and the intensity. The Outline channel is setting the intensity (just a 255 value, as the Master has its own intensity I can set as well) and the phonemes themselves are setting the pattern to draw.

Following that is the rest of the laser's Builder fixture, which I labelled somewhat incorrectly as Master Fixture 2, and handles a bunch of other parameters, most of which should be static based on the song itself.

My ideal goal would have been to be able to handle almost all of this based on a single DMX effect, mapped to phonemes as I had described originally, so I can try it as Gil showed and see what happens. The downside there is that it adds more stuff to keep track of in the songs themselves, whereas in my mind it would've just all ran from my pre-existing work as is.

The laser manufacturers should be providing me with some of their tools to get down into the laser and set it up pretty much exactly as I'd like, so I should be able to streamline the setup within xLights even further, now that we've proven the theory as to how we can control it in this manner. I had it running off of an FPP last night, even. It was beatiful. *tears up*

Thanks guys