Chroma - a laser color blender

[White-Light]

Argghhh... and we go round again...!

Before this thread goes any further into the chasms of hilarity/stupidity/pejoratives/groundbreaking-colour-theory (delete as appropriate);
let's wheel out the horse again for light relief:

So the chroma hate continues despite the fact that it predicts a ball park white that can be fined tuned by small software adjustments rather than making expensive guesses when new wavelengths come along or when the novice is self building, and yet Briggs wonders why Tocket apparently left.

[stanwax]

http://www.youtube.com/watch?v=CqUZ04UWRk4

Rob

[White-Light]

http://www.youtube.com/watch?v=CqUZ04UWRk4

Rob

Again personal attacks just proves the point in #91 above once more.

You don't believe in Chroma yet you can't challenge the fact that for white balance it comes out with ball park ratios that are what you recommend anyway.

As RGB Laser Systems have a good reputation for white balance, lets put their projectors through chroma and see if we get any of the "wild colour concoctions" that Chroma recommends as stated by Mr. Briggs because if it comes up with the wild colour predictions that I have allegedly recommended to others, then with sensible power values the predicted white should be way out:

Compact RGB4

2W 638
1W 532
1W 450



Looks pretty much like a slightly warm (red sided) white to me.


Compact RGB6

2W 638
2W 532
2W 450



The golden 1:1:1 ratio it's a slightly cool sided white, exactly what most people see from this ratio.


Compact RGB8

4W 638
2W 532
2W 450




Surprise, surprise, a slightly warm (red sided) white again and exactly what you'd expect to see once more from an accurate prediction.


I could do this all day through every projector they make and the white balance prediction in chroma would be ball park accurate.

[dzodzo]

i don't have chroma installed (couldn't get it to work with Win7 64-bit ) but i'd like to know what would be the mix of

300mW 532
300mW 640
1W 445

in kvant. By the eye it seems blueish/purplish white, it's blue overkill after all, what would chroma say?

[White-Light]

I don't have the zip file to redistribute and I'm not sure where it installs plus you need Matlab which is like 1/2GB, so I can't give you the programme although I would be quite willing to do so if I did have it in a distributable form. (I'll have another scout around later).

However, I'm happy to run the figures and chroma says exactly what you're reporting seeing and its still within the white triangle:



BTW I'm on Win7 64 bit.

[aclasers]

http://www.youtube.com/watch?v=CqUZ04UWRk4

Rob

I like the bit at 1.18 in,

[White-Light]

Spoken to Erik today, you can download Chroma here:

http://blue.orgchem.kth.se/Chroma/distrib/

[borgqueenx]

We all need update so the app will at least start...

[White-Light]

It starts for me although I appreciate that's no comfort for you.

I've pm'ed you Tocket's contact details so that you can get in touch and hopefully find a solution.

Hope that helps.

[dsmudger]

So I've been using Chroma to evaluate various projectors I've seen, but can I just confirm something..

So I came up with Red 640nm/200mW, Green 532nm/100mW, Blue 637nm/200mW as an example of a good ratio (the circle's pretty near the middle) - I take it that means 100% power on each module will give (a predicted/YMMV etc) near-as-dammit white. (?)




However, due to the visibility function of each individual component colour being different, am I right in thinking that if you have only the 100mW green module on at 100% power (to display your 'greenest green'), it'll appear brighter than the red on its own at full power, which in turn will be appear brighter than the blue at full power? (from a visibility function chart I found: .862*100 > .175*200 > .11*200)

If so, isn't it reasonable (as an alternative strategy for choosing power ratios, and only in the case of analogue modulated projectors) to optimise for equal visibility of each colour in isolation, given that you can always achieve your perfect white in software?

So in this alternative scheme of choosing modules, 100% blue, 100% red, 100% green each in isolation will look equally bright, but producing white will require some unequal ratio of the three.

I say it's only an alternative strategy, no better or worse necessarily, as you can obviously compensate for this in the reverse way, i.e. (to achieve an image of uniform brightness) in software, don't drive your green to 100% to produce 'greenest green'; only ever drive it to the perceived (i.e. visibility-function-compensated-at-full-power) brightness of your perceived-dimmest module (the 640nm red in this case), same for the blue.

Of course this doesn't apply to TTL projectors, in that case if I understand correctly you get all-or-nothing of each colour in any combination. So to get a nice white, you'll have to trim one or more of your modules back to less than full power (in the case of analogue modulated modules driven by a TTL board). And of course the modules remain trimmed back when a given colour is used in isolation, so therefore you'll've have 'wasted' money on power you don't need in that particular module (assuming of course you care about making a nice white).

Or am I just getting it all wrong I'm new here