color balancing rgb and stuff...

[stephenhazel]

Ok this is probably the wrong forum, but i need eyes and this question ain't that hard (i think?)

I'm trying to get the right R,G,B values to my dang laser with my own dang software.

So i code up a lookup table.
ILDA files have one byte R,G,B values (0-255).

My EtherDream DAC wants values between 0 and 32767 (i am pretty sure??).
To send to my X-Laser Mobile Beat Mirage What A Long Name.

In my algebraic mind, that's mean i go

from 0 to 255 and put x*32767/255 into the table.

But that's a "linear response"...

My laser's docs say to use QuickShow and go to "color settings" and
pick "3 laser (RGB)" and "Analog with log response".

So that tells me i have to do actual math. Not just algebra.

Can someone who passed trig help me out here?
I mean, i passed it. I knowww that log is just the opposite of "to the power of".
But my brain can only grok algebraaa...

So what do i use instead of
from 0 to 255 and put x*32767/255 into the table.

I should know this one. But i'm lookin to cheat sigh...

Help ?

[araugh]

Going with a linear color response in your own software will work fine, the main reason we specify log for Quickshow is because that setting gets the user a bit closer to the colors used in the stock patterns in quickshow that rely on modulation towards the low end, like the gecko, without having to do anything with the min/max sliders. That guide was written for beginners to be able to get up and running with a decent result quickly, advanced users like yourself are sortof beyond that. If it's easier for you to code linear, try it and see how it looks. If you don't like it, try log, or your own custom algorithmic curve, or a lookup table.

I hope that helps

[stephenhazel]

Going with a linear color response in your own software will work fine, the main reason we specify log for Quickshow is because that setting gets the user a bit closer to the colors used in the stock patterns in quickshow that rely on modulation towards the low end, like the gecko, without having to do anything with the min/max sliders. That guide was written for beginners to be able to get up and running with a decent result quickly, advanced users like yourself are sortof beyond that. If it's easier for you to code linear, try it and see how it looks. If you don't like it, try log, or your own custom algorithmic curve, or a lookup table.

I hope that helps

i mean, if the laser r,g,b respond best on a logarithmic curve... I'd like to use that...

so linear is ... y=32767*x/255

what would logarithmic be? natural log? log base 10?
0 becomes 1 with log, right?

Well, i'll try linear again. But i coulda swore that with r=32767,g=32767,b=32767 that it wasn't white...

[stephenhazel]

well SHOISH.

i mean with the kajillions of .ILD files on the ftp here everything looks fine.
And i =mean= kajillions.

But they tend to show stuff at max brightness, just primary colors for older lasers.


But with my first cursory stab at it, things look pretty ok linear.

But when i start gettin out the smoke and tryin a beam show I'm gonna want intensity and fiiiiiine color variations...
I'll try that tomorrow eve. It's gf night.

Sooo if anybody (anybody anybody echo) can tell me how to log up a trig table... please