Need help converting to xLights from vixen3. Issues with import

[dkulp]

This is kind of a side effect of only having Vixen 2 created .vix files to use as examples.   Vixen2 didn't really have any concept of RGB things so there wasn't a way to determine if 3 channels should be combined into one or not. 

Looks like we'll need to add some detection logic to look for  RGB things (names that and in " Red" with matching b/g) as well as then taking it one step further to look for the strands.   

That said, not really a priority for me right now.   The new layout effort is huge.

[stampedeboss]

This is kind of a side effect of only having Vixen 2 created .vix files to use as examples.   Vixen2 didn't really have any concept of RGB things so there wasn't a way to determine if 3 channels should be combined into one or not. 

Looks like we'll need to add some detection logic to look for  RGB things (names that and in " Red" with matching b/g) as well as then taking it one step further to look for the strands.   

That said, not really a priority for me right now.   The new layout effort is huge.

Is there a way to use fseq files on import, to be honest that would be my preferred method?

[Gilrock]

You can import an fseq file as a data layer but your channels will need to be a match.

[stampedeboss]

You can import an fseq file as a data layer but your channels will need to be a match.

Thanks everyone, looks like I can't make the jump at this time, Between Halloween and Christmas I have over 40 sequences I don't want to have to completely redo, plus with moving I really can't simply use a data layer.  With 12K+ channels, convert to effect won't work.

Nice program, impressed with where its headed.

[Phrog30]

I still think in your case the best choice is a data layer. To my knowledge, the only direct import is a LOR SuperStar sequence. This is because one of the developers added SS effects. There aren't direct Vixen 3 effects. So, even if the import would work, you would end up with a bunch of node level, on/off effects. This would be painful to manage.

It sounds like you don't want to redo your sequences, which is understandable. My advice is keep what you have, import as a data layer, and add to them in xlights. Then any new sequences can be done solely in xlights.

Whatever avenue you choose, good luck.

James

Sent from my SM-G900V using Tapatalk

[dkulp]

Data layer is definitely the way to go.  Also, once you have it as a data  layer, you can right click on a node in the row headers and have it convert that nodes data into xLights effects.    I wouldn't do that for something like a mega-tree, but if you have a few simple things (faces, AC things, etc...), you could get the data into xLights to then manipulate.

[Gilrock]

You can import an fseq file as a data layer but your channels will need to be a match.

Thanks everyone, looks like I can't make the jump at this time, Between Halloween and Christmas I have over 40 sequences I don't want to have to completely redo, plus with moving I really can't simply use a data layer.  With 12K+ channels, convert to effect won't work.

Nice program, impressed with where its headed.

What's funny is your reason you say you can't use a data layer is actually the reason I believe you need to use a data layer.  If you imported 12,000 channels as effects do you seriously think you will be manually editing all those effects?  Also when you try to use xLights in that way I think it makes our program look bad.  You will be complaining about the poor performance and slow render times because you have over 10,000 effects being rendered.  Imagine dropping a single spiral effect in xLights.  That generates a couple lines in the xml file to produce all that data.  If you then had to import that same data we can't figure out that its all one effect so you end up with 3,000 effects to execute the same thing.

I started out the same way at first and then finally decided it wasn't worth it to try to baby that old data along.  Plus having the old data locked into certain channels can actually be a big roadblock when you want to expand your show.  You are always having to deal with not wanting to impact where channels reside.  I'd say you have too many sequences.  I only ran 5 for Christmas.  Not because I didn't have more but because people will stay too long if you have too many songs.  Just trim it down to the ones that are actually good quality and not outdated.

[stampedeboss]

I actually run a different show each night of the week.  My plan was to rework the sequences, focusing on the big parts like matrix and megatrees at first.  Should be easy since most are nutcracker effects in Vixen.  Coming from Vixen3, same concept as higher level effects lower overhead.  Outside of those, most of the the others are easily moved to the model level after import, on/of and chases.

Will probably do all new development in xLights and retire the others as time permit.  That's why I started early.  Working on how to do similar effects between the two platforms, like chase is morph etc. 

To be honest, I am going to invest some time to see if I can come up with a conversion routine for Vixen3 xml to xLights.  Where can I find a decent description of the xLight xml layout/requirements?  It seems it really should be straight forward,  Vixen3 elements are equal to xLights Models.  Group are groups and so forth.

I think the biggest issue is going to be lack of "gradients" in xLights and what to do with those colors.

[Phrog30]

...like chase is morph etc. 

I think the biggest issue is going to be lack of "gradients" in xLights and what to do with those colors.

The morph is equivalent to a wipe.  SingleStrand most closely resembles chase, however, there are some things that Vixen allows in those effects that xlights can't reproduce easily, extend pulse to the end is an example.

When I want to mimic gradients in xlights I produce a gradient png "x" x "x" pixels.  I then use a picture effect with vector control.  See attached.  It's not as much work as you might think.

James

[Gilrock]

It looks to me like you could do what I see in that recording using Morphs.  You can turn the head duration to 0 on a morph and then the tail colors create your gradient.

[Phrog30]

It looks to me like you could do what I see in that recording using Morphs.  You can turn the head duration to 0 on a morph and then the tail colors create your gradient.

I tried. I couldn't get the duration the way I wanted, basically the tail was moving before I wanted. It was definitely close to what I was looking for. To my knowledge, I've only used this kind of effect a couple of times. 99% of the time the native effects/controls do exactly what I need.

James

Sent from my SM-G900V using Tapatalk