Changing a fixture type in a recorded show

Have you tryed to first erase the DMY adress with [At] [Enter], than change the Fixture Type and after that Patch them to the desired Addresses?

I only made a small test file and could not see any problems.

It should just work. You'll need to first [AT] [ENTER] the addresses for the existing units and then changing the fixture type. The problem you're encountering is likely that the Katanas use a different number of addresses than the Generic alternatives.

The colours should also work fine - though you may find if youre moving from a different colour mixing system (ie: the original units had a lime chip and the replacement units have a white chip) that the colours don't quite match up as well as you'd like.

If you're planning to go back to your other venue with the DTS units later, you might actually be better off to make a copy of the units in patch and then assign them as a new type. If you select the units, say they are channel 1 [THRU] 3  and then [COPY TO] 11 [THRU] 13 {PLUS SHOW} - you can change the type of 11-13 to be the touring venue units and they'll grab all the show data for channels 1-3. If you then end up adjusting any colour to make it match better you don't impact your original units.

No, I had not tried that, but it didn't work for me. I have a simple test file consisting of two cues:

1: katanas are Red 100, other colors 0.

2: katanas are Blue 100, other colors 0.

Even trying it like you said, after changing it to a Titan Solo Full Color 6ch (just to test it), the colors become:

1: unchanged

2: Red 12, Green 12, Blue 100.

I have no idea why that is happening but appreciate the help.

I believe that it should, but it isn't, unfortunately. And yes, I agree that the likely problem is that they very diferent fixtures, but that's what I want to know, what to do in this case.

Like I said, software like QLC+ links the new fixture channels with the old fixture channels based on their function, regardless of their position (first channel, second etc) or how many extra or fewer channels there are, so I'm assuming EOS would do the same.

Thanks for the tip on {PLUS SHOW}, I'll look it up and read more about it. Right now, I simply keep the original version saved elsewhere and return to it if needed.

I did a bit more testing. I added a few more color changes, as well as zoom, tilt and strobe, to see if it would jumble up further. In some cases it didn't, in others it did.

What I programmed in a Katana:

https://i.imgur.com/3W7ACaa.jpg

After erasing the address, changing the type to Titan Solo Color 6ch and re-addressing it (it jumbles up after White is used):

https://i.imgur.com/O0NOcrZ.jpg

To see if the problemas was a lack of white, I changed again. After reloading the file back to Katanas, erasing the address, changing the type to Generic IRGBWAS and re-addressing it (note that this time the Strobe was changed too):

https://i.imgur.com/z1JizZD.jpg

Still super confused as to what I can do to prevent this. Any help is much appreciated.

If anyone wants the actual test file: https://drive.google.com/file/d/1pxxrf2LD1dTPGvGiZCKqDgMeNO_5Xsns/view?usp=sharing

I dont think its going wrong - at least looking at the fiirst two screenshots.

You are mapping an RGBW fixture to a RGB fixture.

EOS is clever enough to try to work out what the RGB values need to be to get the same colour as the original fixture got (or at least a best approximation).  Its not like some of the basic systems out there that would just drop the white channel and copy the RGB values unchanged which would give you the wrong colour.

So for example when you had R=100 and W=100 it had to represent that by generating white by mixing RGB values of say 50,42,35. and then adding R 50 to that which gave you the 100,42,35 RGB values you saw.

It will also be using some calibration curves (for non generic fixtures) in the conversion so the % on each emitter are not the same value to get to white (you can see that it used 100,96,85 to get the same "white" as the original fixture gave at W=100.  So probably its white emitter was a bit towards a lower colour temperature than a straight 100,100,100 RGB would give you.

Hope that makes sense

Like I said, I had considered that the problem would be the lack of a white channel. But if you notice, in the Titan it starts jumbling up everything _after_ the white was used, even in cues 5 and 6, where it was only red+green and red+blue. 

Also, in the generic IRGBWAS (the third screenshot), there is a white channel, and it still gets all mixed up.

If your explanation also covers cues 5 and 6 for the second screenshot or the whole situation with the third, sorry, but I don't understand. Especially cue 6 in the third screenshot, 'cause changing R100 B100 to R100 B27 does not seem to be intentional.

Like Mike pointed out,
EOS is more than "just copy numbers".

It is hard to discuss the colors, only by looking at the numbers.
In the Background there often is a algorythm working to make the color to look as consistent as possible.
That is one of the Key-Features of EOS and ETC.

So most likely there was some magic in the Background that makes your first change happening like it was.

-> with the first change, you have a new starting point

After that, chaning the Fixture back, is likely to be NOT the same as before.
Because the fixture before was something different.
So the numbers will be a bit different. But the result, hopefully will stay close.

---

Matching all the different fixtures with each other is not a easy task.
So sometimes there is a little glitch inside.
But the most of it will fit. Like intensity, pan/tilt, ...

So don't be to fixed to the numbers.   

OK, I think I understand it better now, thanks. However, if this is a feature and not a bug, is there a way to turn it off? Right now I'd much rather see consistency in numbers, 'cause that's all I have while modifying the show at home, without the specific fixtures, which will only be available to me only on the day of the performance.

Ideally find out what the  will be in advance if possible but don't be over concerned as its highly likely that if they are a branded fixtures then ETC will have got the colour mapping correct. 

So you are much more likely to get the result you need ie with the show looking the same. letting the software do its stuff that just setting the colour numbers to be the same as they were for the old fixture.  

Even the one where the colour was original just and R and B or G and the values have changed is probably because either the original fixture has a stronger red than blue emitter or the new fixture has a stringer red than blue or green (it looks like its probably a stronger red as its pulled the red down by 10%) and the software is matching the resulting output so it looks the same as when you used the show with the original lights.

I don't know if you can tell it not to do this,  but as this is very thing that people using ETC do all the time, ie move shows from venue to venue with changes to the fixtures,  I think you should trust that its going to get it way more right than you trying to guess the RGB values at home for the fixture you cant try them with as the colours you get can be so completely different for the same RGB for different fixtures. 

Thank you for your report.

I tried to follow your instructions but can't see a problem. Yes, there are situations where Eos does color translations but this isn't one of those.

I started with a DTS Katana Standard, recorded in cue 1 with RGBW = 100/0/0/0 and changed the type to IRGBW.
I left it patched to an address (this does not affect the translation) and the result is still 100/0/0/0.

Can you please specify which software version you're using and which exact mode of Katana? If you have a showfile that demonstrates your problem feel free to send it to eos(dot)moderator(at)etcconnect(dot)com.