PDA

View Full Version : Photometric ligts mesures?


Hans-CC
05-15-2007, 05:56 PM
hi i've been reading masterZap article of photometric and radiometric !! ,even thou it is interesting i get a little confused about how to used this mesures in a practical way i dont now much about luemens ,cd , etc.. but i now watts !! so my question is this :
lets suppose that a director hires me to make an effect on a shot ,they film the shot an tell me that they use a 1000 watt ligth to lite the secene , then suppose i have a 3d scene that perfectly match the real shot and i know the distance in wich the light was in the real shot from the main object in wich i have to incorporate something , so i replicate the ligth in maya ,position it an oritented it to match the real light !!! what i have to face know is wich its the rigth intensity for that ligth on watts units ? o how do i get a cd ,lumen mesures or wathever that i have to use in maya ?

i ve seen that when you u use a physical light in mental ray conected to your ligth the mr shader has a value of 1000 in the color, does this values means watts ? o what does it mean exactly and how to convert this values to watts to match the real life 1000 watt light ?

this is an expample of what i am talking about !! i made this shot based on messures of the table and the elements in the shot ,and i knew the lights they used and their distance to the spot in wich i had to insert my cg elemnts but i what didint know is wich was the rigth intensity of my cg lights so i made a guess process to match the ligthing (that is what i normally do), but what i want its to know for the next time the exact messures i have to take care of and how to match them correctly in maya - mental ray ?


real shoot

http://www.martestudio.com/hans/Shot.jpg





CG Elements Rendered in Mental Ray
http://www.martestudio.com/hans/CG.jpg


Final Composition + Color Correction in After FX
http://www.martestudio.com/hans/Composite.jpg


Thanks For the Help !!

HANS

sixbysixx
05-16-2007, 01:18 PM
In my opinion don't bother: 1000Watts doesn't tell you how much light will reach your scene in any way: It also depends on kind of light (HDI, KinoFlo, Tungsten) and on the kind of reflector (fresnel, Parabolic, Softbox etc...)

So basically teh parameters to consider are too many in order to figure out a simple calibrated system.

If you still want to do it, get somebody on set to take light measurements, but I'm not sure how practicable that would really be.

Keep matching it by eye - seems to work rather well:)

Hans-CC
05-16-2007, 03:54 PM
Keep matching it by eye - seems to work rather well:)

well yes !! i actually I enjoy alot doing this but my goal was to be as exact as posible at leat as an initial stage of my ligthnig process.

and my doubt perhaps is what does the ligth intensity value mean in maya cd,lm,watts or it is just a maya stuff that is not related with anything ?

i think it would be nice to now and usefull for at lest initial setup!!

thanks

Hans

sixbysixx
05-16-2007, 04:53 PM
I don't have the slightest clue to be honest:D

But still, even if you should find it out, I don't think it will ever match, not even closely, because different light shapes produce completely different decay rates (a bare lightbulb might give you a quadratic decay, while a Parabolic Xenon spot has barely any decay at all) , so even if you knew how much light a lightsource emits, you still won't have a clue how much of that light actually reaches your object.

As I said: in my opinion the only halfwafway reliable and easy way would be to measure on set how much light reaches the spot, where your CG object is supposed to go.

Hans-CC
05-16-2007, 05:04 PM
I don't have the slightest clue to be honest:D

But still, even if you should find it out, I don't think it will ever match, not even closely, because different light shapes produce completely different decay rates (a bare lightbulb might give you a quadratic decay, while a Parabolic Xenon spot has barely any decay at all) , so even if you knew how much light a lightsource emits, you still won't have a clue how much of that light actually reaches your object.

As I said: in my opinion the only halfwafway reliable and easy way would be to measure on set how much light reaches the spot, where your CG object is supposed to go.

thanks for the help anyway !!! so it seems that the best way is match it by eye!!

thanks alot for the replies sixbysixx

Hans

MasterZap
05-16-2007, 06:10 PM
Glad you enjoyed the read of my article.

Now keep in mind it's written primarily with architectural users in mind, people who use lights with IES profiles (= real world measurements) and expect to click on a surface and "ask" it "so how many cd/m^2 ends up here"?

To your question:

1000W tells you almost less than nothing about the pixels in the image.

First of all, you do not know the luminous efficacy, or the spectra, so you don't know the number of lumens, or the spectral distribution of those watts.

Secondly, you neither know the subtended solid angle of the light, nor do you know the emission profile, so you have no idea how many of those watts fly off in the direction of your object.

If you had an IES profile for the light, you would be "better" off; you would know how many candela flies off in the direction of your object (because it had been measured), but then you need to know
- the exact reflectance of the surface
- the exact f-stop, shutter, film ISO (and gain, if digital) and complete film response curves in the camera or CCD
- the exact spectral response of the film or CCD cells

Until you have all that, you can't really get "accurately" from "1000W" to a pixel value on screen.

So... eyeball it ;)

/Z

Hans-CC
05-16-2007, 07:51 PM
Glad you enjoyed the read of my article.

Now keep in mind it's written primarily with architectural users in mind, people who use lights with IES profiles (= real world measurements) and expect to click on a surface and "ask" it "so how many cd/m^2 ends up here"?

To your question:

1000W tells you almost less than nothing about the pixels in the image.

First of all, you do not know the luminous efficacy, or the spectra, so you don't know the number of lumens, or the spectral distribution of those watts.

Secondly, you neither know the subtended solid angle of the light, nor do you know the emission profile, so you have no idea how many of those watts fly off in the direction of your object.

If you had an IES profile for the light, you would be "better" off; you would know how many candela flies off in the direction of your object (because it had been measured), but then you need to know
- the exact reflectance of the surface
- the exact f-stop, shutter, film ISO (and gain, if digital) and complete film response curves in the camera or CCD
- the exact spectral response of the film or CCD cells

Until you have all that, you can't really get "accurately" from "1000W" to a pixel value on screen.

So... eyeball it ;)

/Z

well it is actually more complex than i thought , after done the scene above the idea that it was to much guessing kept in my mind.

thanks MasterZap for the help !!
i really enjoy your articles !!
can't wait next !!:)

Hans

CGTalk Moderation
05-16-2007, 07:51 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.