Color Mixing inverts in color inv mode - opposite from v1.2. Color mixing preferably should stay in tact when inverting movies so you can use your color mixing palettes to further color the inversions
Christian Choi
Printable View
Color Mixing inverts in color inv mode - opposite from v1.2. Color mixing preferably should stay in tact when inverting movies so you can use your color mixing palettes to further color the inversions
Christian Choi
how about both?Quote:
Originally Posted by Christian Choi
It's up to you, when I invert an images color, I often color it further and found it awkward to try to backward mix the colors in my palettes. It also seems I can't achieve the same degree of coloring as I used to in 1.2 for some odd reason everything looks washed out. Prolly my imagination from trying to mix backward.Quote:
Originally Posted by samsc
CC
Did you set up your screens correctly?Quote:
Originally Posted by litemover
Calibrate them - change the dispay gamma ( the default is 1.6 on macs )
http://chaldee1.gotadsl.co.uk/~richa...hread.php?t=80
OS9 handles colour differently to OS 10.3
Apple also changed the way it handles colour in 10.3
Colour rendering depends on the output device - not catalyst.
i think the invert modes compress the hilites into a narrow range.Quote:
Originally Posted by litemover
I added invert modes with colour the right way round.
Richard,
Maybe you have done it the right way around but the fact of the matter is that you are employing 2 different color mixing systems RGB(additive) and CMY(subtractive) now, whereas the better and less confusing way across the lighting platform was to just use the one more employed by lighting (subtractive) as you did before.
Here is exactly what is happenning in a more detailed description:
The standard (uninverted) color mixing system in the current and V1.2 Catalyst is based on the subtractive color mixing system i.e. cyan, magenta, yellow where 2 FILTERS combined = secondaries(RGB) and all 3 combined = black; however, since it is a video system and not white light being subtracted via filters to render color, the card's default must be additive color mixing (RGB). You must have cleverly devised a system which reversed and employed the standard lighting color mixing standard so that the colors would match up with encoders and logically mix more like a lighting fixture for us lighting folk. This all works Dandy in the Standard Catalyst *RGB default* mode even though it is really CMY not RGB.
Now, when the image is inverted, The color mixing system reverts back to True additive color mixing AKA - RGB where all colors mixed together = white and the encoders turn to red, blue, and green. This is the problem/bug.
In V1.2, the inverted color mode wouldn't affect the subtractive color mixing system and would actually color the inverted media without it approaching whiteout, instead it would darken the inversion.
In conclusion, you did do it the right way around, which is the problem, it needs to be back to the wrong way around in order for it to work properly and maintain continuity with the default color mixing system.
This fix would aid greatly with the Catalyst user's current color mixing palettes AND more importantly, not to wash out to white so quick by mixing in secondary colors.
Hope this helps, sorry if my first description was so vague.
Christian Choi
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,Quote:
Originally Posted by samsc
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
I have to explain this to people other than yourself.Quote:
Originally Posted by litemover
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:
And this colour balance function would need 10 dmx channels to simulate.Quote:
Originally Posted by samsc
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.
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.Quote:
Originally Posted by samsc
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
Christian - have you used shake yet?Quote:
Originally Posted by litemover
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
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.Quote:
Originally Posted by samsc
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
the new colour modes are in f24
Excellent, thanks Richard.
I will make good use of them.
Best,
Christian Choi
Thanks.
I havent used shake. I went to a seminar, and played with the gui.Quote:
Originally Posted by litemover
Shake has 2 huge advantages over what i do. It doesnt have to do dmx, and it doesnt have to work in realtime.
alpha channels are total performance hog.
They have to be uncompressed to work properly.
You need to play around with the animation codec.
When we did the RMA's i rendered the RMA - on screen logo using the animation codec.
Live to air scares me....
This is a good point Richard, performance must be able to keep up with things like uncompressed alpha channels. This is why I am trying to find the fastest method of storage available for uncompressed.Quote:
Originally Posted by samsc
In the TV world, all of the video servers have this capability and it is very difficult to compete with them or even coexist with them if you don't have this ability with Catalyst. Catalyst is heaps more flexible than most of them but this one feature would give those who program Catalyst in the everyday TV world to be able to matte a logo or lower 3rd layer on another piece the proper way. Trans black or White has been a godsend but with pieces that have black or white in them...
What ends up happening is that the producers have to pay the editors or AE people to deal with this problem and the costs associated with it make it less appealing.
This is just one of those things that is a hinderance in the TV world because directors and producers would like to have the flexibility to matte things like sponsor logos and band names on whatever content one happens to be playing at the time. Does anyone else ever get this request or just me? I can see that this wouldn't be an issue in the touring or theater world as there is slightly more time to put together a premultiplied or already matted piece against a planned layer.
Christian
in a theatre they can pre-plan everything.
This isnt designed as an 'online' type compositing tool - something like Harry or flame or henry.
Broadcast people are used to those things.
And even these things didnt work fully in real-time last time i checked.
Im just trying to do simple compositing and dissolving. that has to work in real-time.
I cant do chroma keying either - properly - because those broadcast people have devised some kick ass keyers - that i cant possibly compete with.
There is no way i can make it idiot proof - keying is really problematic. and there are alsorts of tools available in the non-real- time world to make this acceptable.
This works fully in Profile and the EVS system in real time. I wouldn't need to do this or ask for this if it didn't. When editors deliver key packages, this is how they deliver them to the Profile crew.
Because I am unable to do this, it causes both confusion and bewilderment on their behalf. I'm trying to make it simple for them so that they don't go back to the producer and say "why are we using Catalyst when it can't even key how we usually deliver it, we should just use Profile, blah blah blah".
Profile operators aren't the only ones who are scared of Catalyst, the editors and content creators are too. It's a very difficult political situation to be in. Catalyst will need to be able to keep up with the most fundamental and rudimentary features such as this in order to succeed in the television world.
Many productions are prohibited from using video because of the cost of Profile. Catalsyt comes in at the right price but because of features like this not being implemented, it makes it neccesary to have a reduced Profile system and crew which defeats the cost to savings ratio.
I'm not asking for chroma key, the Globecaster does this real time and live rather well for 40 grand.
Christian
keying is a tough problem that needs to be done right.
maybe when i have a spare week or two, ill have another look at the issues,
Sorry for my outrageos post, Richard. You fixed the bug. Thank you. I understand the preassures you're up against. I must have been pretty tightly wound those days. I think I was just really surprised when I got on site and all my colors in my palette were off and I had to rebuild them.
Wow, I'm sure glad I'm not working my ass off as I was back then. I was constantly frazzled.
My appologies,
Christian