Color Mixing inverts along with image inversion

[samsc]

no.

nothing in catalyst is cmy.
its all rgb.
its all rgb because the output device is rgb.
it makes no sense at all in computer->projectors to talk about cmy.

but computer cmy is simply 1-r, or 1-g, or 1-b.
there is nothing complex about this.

what you saw with inverted colours was simply a little bug, i liked the result, so i left it in.

i added inverted colour modes that do this correctly yesterday.

---

what you need to be very very wary about is talking about lighting cmy or printing cmyk or computer cmy in the same way- they are quite different.

operating in different image colour spaces is totally different from whats going on here internally in catalyst.

[litemover]

no.

nothing in catalyst is cmy.
its all rgb.
its all rgb because the output device is rgb.
it makes no sense at all in computer->projectors to talk about cmy.

but computer cmy is simply 1-r, or 1-g, or 1-b.
there is nothing complex about this.

what you saw with inverted colours was simply a little bug, i liked the result, so i left it in.

i added inverted colour modes that do this correctly yesterday.

---

what you need to be very very wary about is talking about lighting cmy or printing cmyk or computer cmy in the same way- they are quite different.

operating in different image colour spaces is totally different from whats going on here internally in catalyst.

Thank you Richard for changing it back,

Firstly, I appreciate your candor and advice, Richard, but I very clearly understand The differences between lighting CMY and computer RGB to produce CMY secondaries etc, so I hardly think that I need to be "very very wary" as I've been mixing colors for a very very long time now and know very cleary what is and was going on in this situation.

Now, weather or not Catalyst, computers, or projectors ever output CMY in a computer sense without being mixed by RGB levels to produce this illusion hardly matters to lighting designers, lighting programmers, or any end user or Catalyst that only knows about what is going on visually on the stage - which translates that the color mixing coming out of the projector, video wall, plasma screen, or catalyst, visually - is Cyan on one channel, Magenta on another channel, and yellow, on the other.

Lastly, if it "was a simple little bug that you decided to just leave in because you liked the effect", that's your perrogative, you wrote the proggie. For all I care, leave it in as you like it but please understand that I reported it to here because *it was a bug* that I noticed and *this is the bug report forum*. I don't care to participate in a helpful effort if I'm to be repaid by a sour response.

My time is very valuable, as is all of ours, and I don't charge for my advice or perspective on this product. It's a genuine interest that I have and something that I care about the future of. I hope to keep it that way.

Sincerely,
Christian Choi

[samsc]

Thank you Richard for changing it back,
Christian.

I have to explain this to people other than yourself.

You know whats going on.

Others dont.

Dont feel patronised.

I have to try and take a more general explanatory tone for those who dont understand.

----

This is for you:

The difference between a bug and functionality in this case is non-existent.
There is no 'correct' way for this function to operate.
There is no reason why it has to work the way it does.
I didnt 'change it back' - i added another colour fx that works the way you describe.
All these things are conventions that become established or changed through usage.
For me -- v1 is not the right way to do anything. It is not a standard or benchmark.
Software establishes its own conventions.
when i can incorporate differing operating conventions - i try to do so, but i have to also try and limit the confusion that arises from adding features.
For me - none of this is at all ideal.
DMX is such a limitation. Trying to do things with a handful of channels, and avoiding too many modal states is very hard.

I am a mediator between the users and the underlying technology.
Sometimes i have to tell users how it really is, sometimes i can hide all that from them.
And in this case the underlying technology is an rgb system.

I had wished that i could set the defaults on the fader to 128 and allow users to add or subtract colour like they can in photoshop -
but the underlying technology did not work on the last generation of graphics cards- and i could not do this - and it was a very low priority to make this work.

This is how it should have worked - you should have been able to do something like the colour balance function:

[samsc]

This is how it should have worked - you should have been able to do something like the colour balance function:

And this colour balance function would need 10 dmx channels to simulate.

10 channels just to do colour balance?

What would we call them - and how would we display such functionality with 8 characters names on a hog 2 display?

RedHigh
GreenHigh
BlueHigh
RedMid
GreenMid
BlueMid
RedLow
GreenLow
BlueLow
PresLumn
???

How the f**k would the user figure out what these do?
and how they interact?
And as i cant reuse the names, i would be stuck with these function names that only worked for 'color balance'

---
Its hard trying to make something usable.
there are a lot of compromises.
sometimes they need explaining. sometimes not.

[litemover]

And this colour balance function would need 10 dmx channels to simulate.

10 channels just to do colour balance?

What would we call them - and how would we display such functionality with 8 characters names on a hog 2 display?

Complex color correction could open up a can of worms but could be cool to have. If you ever are confronted with the opportunity to color correct a bad video, be sure to charge by the hour as no one person percieves color the same way.

Many lighting programmers organize all their colors chromatically for all fixtures into the same palette, i.e. red for VL3000 is Red for studio colors, cyan is cyan for VL5s and Studio Beams, etc... I do the same with Catalyst colors in the cmy image color space mode. There are very few lighting fixtures that don't mix using CMY subtractive, the VL7 uses HLS and the old color pro actually used RGB, VL5bs are completely proprietary as are some areas of regular 5s so it makes sense to keep color values for this modal uniform across modes. Even Versatiles, though LEDS, mix in the CMY image color space mode.

Probably the easiest way to have a semi-advanced method of color correction would be to have a separate color correction channel that would give you RGB, HSV, HLS, CMY, and TMV image color space modes. Choose accordingly and adjust using your encoders to achieve the desired effect but this would really only be different ways to mix if you couldn't stack the final effect.

A more effective approach could be if you had a dedicated channel each for Low1 Mid1 High1 (which could be the current cyan mag yel encoder channels), L2, M2, H2,(3 channels for the second range) L3, M3, H3 (3 more channels) and had your color corrector channel(7) dictate which color mode you wanted to correct in, along with an extra gamma(8) channel and Alpha channel(9), then you could really color correct on the fly. An example, you could select RGB mode and adjust the RGB for the Low, mid and high ranges of the video or you could do the same with Hue, Level, Saturation. You could just do this in post (AE) rather then adding a 9 more channels for correction, though 9 channels isn't all that much these days and the effects you could create be pretty cool.

I would also be interested in more transfer modes in the standard color fx channel such as difference, stencil luma, Sillhouette luma, darken, multiply and the like, to be able to do more interesting interactions with multiple layers in wild, transparent ways. For example, you could do things like allow transparency of the layer beneath to show through based on its luminosity or intersecting color values. This would allow for more on the fly creation of new content based on old content. Recycling content becomes a neccesity after you've done a few jobs for the same client.

A new trend with some content providers is to provide the buyer several elements along with After effects files for you to output the default combinations and then create your own semi-original content. A good example of this is Digital Vision's Geolectrics, form, flux, nexxus, light forms, and many more. These elements on their own can be uninteresting and dull but with more transfer modes available, you could mix them to create new interesting content on the fly all the time.

Sorry for the book, just too much to think about.
Christian

[samsc]

I would also be interested in more transfer modes in the standard color fx channel such as difference, stencil luma, Sillhouette luma, darken, multiply and the like, to be able to do more interesting interactions with multiple layers in wild, transparent ways.
Christian

Christian - have you used shake yet?

the advantages of using dmx to do any of this very quickly evaporate.

i really dont want to end up with hundreds of channels to do things that are better done in another application.

i mean i really like the shake or maya shader networks. they remove the obstacles of linear networks of filter, but are unrepresentable and unusable in this application space.
http://www.apple.com/shake/

the learning curve on these programs is just mind boggling.

i did a number of these modes - like darken subtract and add in the colour fx, but some of these filters require additional parameters.

http://chaldee1.gotadsl.co.uk/~richa...read.php?t=123

[litemover]

Christian - have you used shake yet?
http://chaldee1.gotadsl.co.uk/~richa...read.php?t=123

Yes, I have used Shake and am currently in the process of learning it for the pro cert test. I love the way the Color correction node works or color x node. That's actually where I got the idea that I just described above, you must use Shake too.

You could do it with 9 more DMX channels and it would be easy to understand if explained right. 255 levels on the color channels are fine enough for me. Like I said, post or preprod in this case, is where i would usually do this but it would be pretty handy to be able to do this from within Catalyst. It would really help with drab images or overexposed DV. The alternative for me is to find the file drop it onto a firwire drive or usb key, edit it in AE or Shake, and then put it back into all my machines. Most of the gigs I do are at the very least 4 machines and I also program the lighting along with the video so this proccess of correcting files is a real time consumming one and means going to my room instead of the bar after the day is finally done. Danny has been a big help, though some shows we just cant afford another crew member.

If I were granted one color function, it would be one that would enable alpha channel actions with either travel mattes and separate fill files or premultiplied video with an alpha chnl(millions +) so that I could accept all of this logo and overlay stuff that is so frequently used in TV. I just got a call from the tech director of MTV for a Hamish Hamilton, Dunn and Dusted production inquiring about using catalyst for this (trackmatte and separate fill file in HD). I turned the gig down. Not because catalyst couldn't do the trackmatte stuff, I was just too tired.

I am right on the cusp of driving this technology to direct live/live broadcast (not via screens or projector but direct to air) that it would really help to have this function.

Any suggestions?

CC

[samsc]

the new colour modes are in f24

[litemover]

Excellent, thanks Richard.

I will make good use of them.

Best,
Christian Choi

[samsc]

Thanks.