Setting up a 30 x 12 matrix in LSP for 12 30LED ridged pixel strips

[taybrynn]

Ok, I must say that setting up a large matrix in LSP seems extremely tedious, or (I hope its because I don't know what I'm doing.  

I set up (3) 120 node SSC (smart string controllers) using RGB, and pixelnet, to control (4)  ridged LED pixel strips with 30 LED(s) each.  I want (3) of these sets of 4x30LED strips, to create a combined 12 LED tall x 30 LED wide matrix (12 strips stacked vertically, each is 1 pixel tall, so 12 x 30 total).

So I have 3 SS Controllers named: top4, mid4 and bot4.  Each will control a set of (4) 30LED ridged strips.

What I'd like to do is define a 30 wide x 12 tall matrix and associate all the channels with it, so 360 matrix pixels, or 1080 RGB channels across the 3 SSC(s).

Here is how the 12 ridged 30LED strips would be layed out (top to bottom) in the matrix.
I used 001->120 channel numbering for this example, but it might actually be all unique
channel #'s in real life, like 0001->1080 RGB channels. (12 strips x 30 LEDS x 3(rgb) total)

top4 ch001->030
top4 ch031->060
top4 ch061->090
top4 ch091->120
top4 ch001->030
bot4 ch031->060
bot4 ch061->090
bot4 ch091->120
bot4 ch001->030
bot4 ch031->060
bot4 ch061->090
bot4 ch091->120 (really through channel 1080)

Observations thus far:

- Looks like I could do it by manually assigning each of the 360 pixels the correct controller+channel
  [ conclusion: seriously tedious and what a pain, but may be the only way? ]

- Looks like when I assign first create the matrix, it asks if I want to assign all the channels on a controller to the matrix., and if I say yes (i.e. select top4 controller)  ... and it get the first (top) row correct (like top4 1 -> 30) and then assigns every other row in the matrix as top4 1->30 as well ... so the only correctly assigned row is the first one.  It also spends a LONG time doing something (not sure what) ... but basically numbers all 12 rows with top4 01->30 ... which means the top row is right and the next three need channels re-selected and the last 8 are wrong or won't allow me to auto-assign them.  And if I never allow it to 'assign for all' ... it asks every single time I assign an element of the matrix [ annoying ].

- Then I can't seem to "auto add" other controllers to this same matrix ... I can only manually add one element of the matrix at a time.
  [ conclusion: the auto add feature basically doesn't work, does one row only and is limited to one controller only ... smells like slop code ]

So I'm kind of left thinking that in order to do a large matrix, I'm basically going to have to right-click and select the controller, say no when asked to add for all ...  and then right click and selection the channel ... and repeat for all 360 channels in the matrix.  I could auto generate the first row and get that done with a few less clicks.  And perhaps the best case is to auto-generate from the drawing for the first, then manually enter for the other controllers, since the auto-generate seems to work there, but not in the other place. See below for more on this **, but problem is the matrix is then defined as 30x4 instead of 30x12 ... so doesn't quite work as a starting point since I can't redefine the matrix size later (or don't know how to).

I know in the video of using LSP matrix, he just opens a pre-defined matrix ... and its a pretty small one at that.

Is there a better way to do this?  And does LSP have any smarter way to auto-assign channels to a matrix?  Is this a known bug?

** It seems to allow me to create a matrix for one SS Controller  (say top4) when drawing the controller in the visualizer ... and using MATRIX type, where I can define a 30x4 matrix and it then does auto-assign properly for just that controller ... but again, not sure how I'd do that for one combined matrix with 3 different controllers, when drawing seems limited to per-channel or per controller.

[dmoore]

It should only take about 5 minutes, including adding the controllers to setup the matrix (for someone with LSP experience).  Check this video on how to do so in LSP:

You are not allowed to view links. Register or Login

[taybrynn]

Thanks David.

Yes, I've watched that video a bunch of times now ... but it doesn't show you how to actually build the matrix, it just loads one thats already been built, then focuses on using the matrix.

There two methods I know of to build a new Matrix in LSP.

Method#1
[ initiated from right-click (on controller) -> Draw Controller Track Images -> Matrix ]

Method#2
select a range on timeline ---right click--> Create Matrix Animation in Selected -> [select|new|edit|delete]  

[FredS]

There is an easier way to set up a matrix. I can't remember if I found it in a post or figured it out by trial and error. I am not at the computer that I sequence with so this is from memory.

Start a new sequence and create or insert the smart string controllers. 3 controllers with 120 nodes in your case.

Right clik on one of the controller tabs. About half way down click on draw images.

On the draw image screen select the matrix drawing tool and draw a matrix on the screen at the location where you want it. The matrix tool will allow you to select the range of channels to assign to the matrix.

Select channel range then sleect multiple or single pixels. then click next.

The next screen should let you set the size of the matrix. (12 x 30, 30 x 12, etc.) It will also give you a choice of 6 paterns to use for the order of the channels in the matrix. Set size of matrix and select pattern then click next.


The channels will be assigned to the points on the matrix. It will save the matrix with the date and time of creation.


I am not an expert, but this method has worked for me and does not take much time. You may have to expermient to make it work correctly.

[taybrynn]

Thanks!!  I can live with this. 
 
Your right, I just needed to use Method#1 ... and then define it as 12x30 (type those #'s in) and then pick more channels which span across all 3 controllers ... since
its not limited to the controller you initiated the drawing for.  I thought it was, but was incorrect in assuming that.

Case closed.  I would say that there is still a bug in Method#2 ... and Method#2 is the only way to view if the grid was created ok or not ... or to manually tweak it later.