XL4 Networks Setup Help

[Phrog30]

This will be my first year with RGB.  I left LOR for DIY and have enjoyed learning Vixen and XL4... very impressed so far.  I'm sorry if this has been discussed before, but I would like some clarification on setting up XL4 networks.  From what I see, most of you setup each universe to 512, regardless of whether you use the entire 512 or not.  From my experience, if you didn't do this and then had to add in the middle this would mess things up in XL, am I on track with this thinking?  If setting up 512 is the norm, then my next question is I assume I would need to setup my controller (PixLite) for 512 (510 channels/170 pixels), regardless if I have that many pixels or not, correct?  I just want to make sure I am setting up XL4 correct up front so I don't have to redo things down the road.  Everything else seems to be straight forward.

I have seen some of the tutorials, but didn't see anything regarding this (could have missed it though).  So, if there are some instructional videos on this topic, please point me in the right direction.

Thanks,
James

[flyinverted]

Welcome to the dark side!

First off, adding models, increasing channel counts to models, removing models does not affect Nutcracker sequences. You merely need to re-save the sequence so the effects are rendered to the new channels. Thats all you have to do, that's it, nothing else, Really!

That is the beauty of NC, it is MODEL CENTRIC,  not Channel Centric like el-oh-are is.
Effects are generated toe MODELS.  Output is generated to channels.   This allows us to share sequences with each other and not have to worry about channel configurations.  If you sequence a 1200 channel megatree and share your sequence and I play it on a 3600 channel megatree, it will look the same, just bigger on my end.

Now going 510 vs 512 depends on your controller.  Some can handle the 512 split, some can't. Just set your NC config to 510 for all universes and you should be good to go. Others may have input on this one. I run all mine on 510 channels.

I sequence now with models I may or may not have. I sequence to channel counts that are unrealistic to my budget. When it's show time, I re-save to what I really have and go from there.  For 2014 I included models in my sequences that I did not yet acquire or build. This year those sequences are ready to go as I now have those lights.

Enjoy!

[dkulp]

It really does depend on your controller.   The SanDevices controllers really do well with 510.   The outputs can span universes just fine.   The j1sys p12x series, on the other hand, cannot have an output span two universes.   Thus, a "string" going out has to be fully in a universe.  Thus, you would only configure the channels you would be using on that output/universe.   

Don't know anything about the PixLites.   If they can span universes, then 510 is likely the right number.

I'd avoid 512 and use 510 everywhere.  Using numbers that are a multiple of 3 kind of helps as that way and RGB nodes would stay lined up on a multiple of 3 boundary which helps if you use the "Test" tab. 

[Gilrock]

My philosophy is you define the setup for the max the controller supports.  I use 512 because I run things that aren't divisible by 3 like 16 channel LOR A/C controllers and LOR CCR's use 157 channels since they have 7 macro channels.  I had one situation where I ended up using all 512 channels on one of my outputs.

[jnealand]

+1 to all the above.  I use 512 because all my controllers use 512. 

[Gilrock]

Yeah I didn't even use 512 for any particular reason.  Using 510 would work 98% of the time.  I had just blindly used 512 for everything and it saved my butt this year.  Mid-December I added a DMX snowflake to the end of one of my DMX chains.  I had 2 LOR 16 Channel controllers and 3 CCR's so I did the math and said ok it's 16*2 + 157*3 = 503 channels so I programmed it to 504 to use 504, 505, and 506.  I had pre-tested on the bench and it worked fine.  So I mounted it on the roof and it wasn't working and I was like crap how do I troubleshoot this thing.  I couldn't hook the programmer back up because the wiring for the programmer didn't match the wiring required by the LOR DMX wiring I used on the connector.  Then it hit me.  LOR CCR's can only be programmed to begin at DMX addresses that are divisible by 16.  I was panicing as I raced inside to do the math.  So my last CCR began at channel 353 so it ended on 509 leaving me exactly 510, 511, and 512 to run the snowflake.  Reprogramming was a PITA because since I didn't have a pigtail to convert the LOR backwards wiring I had to cutoff the RJ45, wire up a new one, program the snowflake, cutoff the RJ45 again, and re-wire it back to LOR.  Just another case where LOR has to be different in everything.

[Phrog30]

First off, adding models, increasing channel counts to models, removing models does not affect Nutcracker sequences. You merely need to re-save the sequence so the effects are rendered to the new channels. Thats all you have to do, that's it, nothing else, Really!

Does this include using layers from LOR and Vixen.  I ask this because in my limited experience, if I changed something in the setup, the models would remain the way I had them setup, then nothing would line up.

James

[Phrog30]

One more quick question.  I have noticed that most of the setups I see in the tutorials are multicast.  I was planning on using Unicast.  Any reason why this would be a bad idea?

James

[Gilrock]

I only use Unicast.  Multicast didn't work when I tried it and I liked the idea of Unicast anyways.

[Gilrock]

First off, adding models, increasing channel counts to models, removing models does not affect Nutcracker sequences. You merely need to re-save the sequence so the effects are rendered to the new channels. Thats all you have to do, that's it, nothing else, Really!

Does this include using layers from LOR and Vixen.  I ask this because in my limited experience, if I changed something in the setup, the models would remain the way I had them setup, then nothing would line up.

James

If you import data it's locked into the channels the data came from except that you can use the channel offset feature but that moves everything up or down.

[Phrog30]

First off, adding models, increasing channel counts to models, removing models does not affect Nutcracker sequences. You merely need to re-save the sequence so the effects are rendered to the new channels. Thats all you have to do, that's it, nothing else, Really!

Does this include using layers from LOR and Vixen.  I ask this because in my limited experience, if I changed something in the setup, the models would remain the way I had them setup, then nothing would line up.

James

If you import data it's locked into the channels the data came from except that you can use the channel offset feature but that moves everything up or down.

Ok, so my original assumption was correct, that if I alter my setup things will get messed up?

If I were to redo the channel config in LOR or Vixen, then import the layer, then save or force a render, will this correct things?

Sorry for the questions, just concerned that I will put a lot of work into sequences, only to realize I messed up something in my hardware configuration to force me to change, then mess my sequences up.  I plan on having at least one layer, which will be my pixel tree and globe (Brian from SuperStar).  I don't like LOR but just can't leave those sequences, they are too good.

Also, can somebody tell me how to add beat timing marks into XL4.  I know you don't need them, but I find it easier to program with them.


James

[Gilrock]

Yep we don't have any way to push imported data around if new channels are inserted.  What I did was made my LOR sequence take up all the lower channels so whenever I add new stuff I'm not impacting any of my existing stuff.

To add timing marks you can export them from Audacity and import them into xLights under the sequence settings dialog / timing tab.  Or you can manually click in the waveform and type 't' or you can start playing the sequence and type 't' as it plays.

[dkulp]

Yea.   I have my first three universes (my non-pixel, purely DMX universes) set to WAY more channels than I actually have props for.  Thus, if I add anything, I don't need to adjust anything.   There is space in those universes. 

The other nice thing about that is then there is space in those universes in the FSEQ file as well.    Last year, I couldn't get one of the controllers working right that I had for Universe 1.   I grabbed one that was setup for Universe 3 (my komby wireless universe) and then used the channel mapping feature in FPP to just copy the channels from 1 to 3.   Worked great. 

[ValentinoRx]

Welcome to the dark side!

First off, adding models, increasing channel counts to models, removing models does not affect Nutcracker sequences. You merely need to re-save the sequence so the effects are rendered to the new channels. Thats all you have to do, that's it, nothing else, Really!

That is the beauty of NC, it is MODEL CENTRIC,  not Channel Centric like el-oh-are is.
Effects are generated toe MODELS.  Output is generated to channels.   This allows us to share sequences with each other and not have to worry about channel configurations.  If you sequence a 1200 channel megatree and share your sequence and I play it on a 3600 channel megatree, it will look the same, just bigger on my end.

Now going 510 vs 512 depends on your controller.  Some can handle the 512 split, some can't. Just set your NC config to 510 for all universes and you should be good to go. Others may have input on this one. I run all mine on 510 channels.

I sequence now with models I may or may not have. I sequence to channel counts that are unrealistic to my budget. When it's show time, I re-save to what I really have and go from there.  For 2014 I included models in my sequences that I did not yet acquire or build. This year those sequences are ready to go as I now have those lights.

Enjoy!

Ok, so I just started on my XL mission this week. Since I'm wasteful and to make it easier for me I will use 50 pixels per output of my controller. so I set up my universes in the setup tab an I even drew out a few models. All good. How do I configure the channels so they output correctly?

[Phrog30]

Yep we don't have any way to push imported data around if new channels are inserted.  What I did was made my LOR sequence take up all the lower channels so whenever I add new stuff I'm not impacting any of my existing stuff.

To add timing marks you can export them from Audacity and import them into xLights under the sequence settings dialog / timing tab.  Or you can manually click in the waveform and type 't' or you can start playing the sequence and type 't' as it plays.

I did exactly as you, put the LOR stuff at the beginning. 

I created a bars label track with Audacity.  I imported into XL4 and I see it, but there is nothing there.  I was hoping it would stick a mark, in this case, 4 per bar, which is what I set in Audacity.  What else am I missing?  I really like having these marks as reference to the song, doesn't mean that's where I will put effects, but just reference.

James