ies real world lighting simulations & vray

[Carrotious]

hi yall,

 

am about to head down the path of attempting to work with real world lighting for a refurbishment project. working with lights from a few different manufacturers, recessed, hanging, on foot and exterior and with vray 1.5 sp1.

 

Has anyone with experience with this got any advice is it a valid tool or is it more pfaff than its worth.

 

I'm not new to ies but have always had to boost intensity by a substantial value to get them to work.

 

The point is that since I'm heading down this path I would love to get it right and be able to use it as a tool in the future...

 

Thanks for any input

 

Giles

[Carrotious]

ok so here goes, first attempt...

 

so kinda using a float method to expose vray camera. the material is a 255 pure white with a .5 vray color map.

 

exposure at 650 iso, 5.6f 30 shutter speed

 

color mapping linear with gamma 2.2

 

I have a 5m x 5m room 2.5m high and 9 ies lights.

 

the lights are 75 watt spots with what i'm guessing is a 24° radius(?)

 

All this to say this is not a very convincing representation of what i imagine 9 x 75w would do to this room.

 

Where next ?

 

Anyone got any input on this?

 

Cheers

 

Giles

[Carrotious]

Same scene with pure white walls.

 

exposure 246 f 8 shutter 30

 

maybe workable...

[Carrotious]

two versions with textures and previous settings. f8 for first f4 for second.

 

obviously massive colour problem with floor colouring the whole scene...

[Carrotious]

well here goes with a touch of tweaking. i guess that there is something there, but i obviously have no idea it these specific lights would actually light up the room like this. at least all the values make real world sense(ish). f4 shuter 30 iso 250... and the lights are behaving in terms of shape...

[rogue3d]

well here goes with a touch of tweaking. i guess that there is something there, but i obviously have no idea it these specific lights would actually light up the room like this. at least all the values make real world sense(ish). f4 shuter 30 iso 250... and the lights are behaving in terms of shape...

 

I've played with this myself and don't think it's accurate to the point of sacrificing the time it takes to render IES lights vs. final output.

 

I have a few questions (and it's because I don't buy into the VRay camera) not anything you did.

 

first (and correct me if I'm wrong) the whole point of IES data is to physically represent the light produced as close as possible to the manufactures exact calculations. illumination, decay, color, spread, etc... what is this data based on? I understand it from the architectural side, but from the cg side there is too many variables especially when using Vray Phys. cam.

 

For example, changing the ISO, F-stop, & exposure is all going to change the way the light is perceived in the rendering. You've proven this by having your multiplier constant and getting different results when changing camera settings. So you're negating all the info generated by the IES data as the light is not illuminating per its specifications.

 

To me it would seem that lighting manufactures would need to include extra info to allow us to set our cameras correctly. obvisously, this won't happen anytime soon, or for what we would need in Vray. It seems to me that you get a more accurate result (or reliable to the standard of the manufacturer) if you stick with standard cams.

 

But even using standard cams and IES still takes a bit longer to render then using Vray's own lights. Perhaps, a more accurate (or deeper) menu to the units in which vray lights work would be the best solution? Just asking, i've played with vray lights using both 'radiant power (W)" units and "luminious power (Lm)" and setting them per cut sheets from our lighiting designer only to have him say it's way wrong......in the end the lights were set to nearly 3 times their initial rating to get the proper lighting.....again thereby negating the cut sheet info.....this was without any artificial brighting (exposure control, gamma adjust etc..)

[BrianKitts]

To be honest with you I've never scene a decent looking lighting analysis rendering. At least not since when college a guy next to me did some stuff with Lightscape that knocked my socks off at the time, but knowing what my own skill level was back then, I would bet today I might not be as impressed.

 

With our new office that we just designed the lighting engineers did some luminance testing with Lightscape using a basic model version of the office that I provided to them. I wasn't impressed at all with the renderings, I just kept my vray interpretations to myself, but it gave them the answers they were after mathematically speaking.

 

I guess what I'm getting at is, If you want a lighting calculation I would be aiming for something that is a bit more directed at doing light studies. I'm not sure you can get both with one engine. Vray is a biased renderer, the majority of your adjustments are up for interpretation. I'm sure a technological genius could do it, but without my masters in lighting technology I would look elsewhere for a solution.

 

Dare I say perhaps an unbiased engine like maxwell could be a solution? I'm only guessing, someone with more experience than I would have to confirm that one.

[Crazy Homeless Guy]

I have a few questions (and it's because I don't buy into the VRay camera) not anything you did.

 

3 times their initial rating to get the proper lighting.....again thereby negating the cut sheet info.

 

to get near real world simulations using Vray, you would need to be working in linear color space, and i don't think linear color space in Vray works with standard cameras.

 

are you working in a non-linear color space, maybe a gamma of 2.2? it would make sense that you would have to increase the intensity of your lights, because the environment you have set up in Max is much brighter than its real world equivalent. you are going to need to add multipliers to your ies lights which i believe are set for real world environments in order for them to overcome the gamma curve your monitor has on it.

 

also, your results will not be accurate because working in a higher gamma effects the fall off of your lights, and is also made less accurate because to get over the curve in the first place, you increased your lights, which means you bounced more light into the space, so you have more secondary light in the room.

[Crazy Homeless Guy]

well here goes with a touch of tweaking. i guess that there is something there, but i obviously have no idea it these specific lights would actually light up the room like this. at least all the values make real world sense(ish). f4 shuter 30 iso 250... and the lights are behaving in terms of shape...

 

there is a mathematical equation to perfect exposure with the Vray camera. I will explain it later today, I need to get back to work now.

[rogue3d]

to get near real world simulations using Vray, you would need to be working in linear color space, and i don't think linear color space in Vray works with standard cameras.

 

are you working in a non-linear color space, maybe a gamma of 2.2? it would make sense that you would have to increase the intensity of your lights, because the environment you have set up in Max is much brighter than its real world equivalent. you are going to need to add multipliers to your ies lights which i believe are set for real world environments in order for them to overcome the gamma curve your monitor has on it.

 

also, your results will not be accurate because working in a higher gamma effects the fall off of your lights, and is also made less accurate because to get over the curve in the first place, you increased your lights, which means you bounced more light into the space, so you have more secondary light in the room.

 

 

When I was playing with it, I was using linear, gamma was set to 1.0. basically the default color mapping and no exposure control active.

 

but what I was wondering is if there are factors that need to be adjusted in order to make the results accurate (i.e. not adjusting until it "looks" right, but "knowing" it's right.........if that makes any sense.)

 

I guess in short; I wouldn't stand up and say if you use this light, this is exactly how it's going to look...(in the real world)......to a client. I say this becasue I was put on the spot about this the other day. I didn't use the IES lights, but I was asked how accurate (realistically) is this image.

 

Without being a lighting designer and knowing the small differences between light types. I don't know if a rendering is accurately representing the photometric data. For me, if I can't be 100% sure, then is there really a point to spend the extra time it takes to calculate and adjust the ies lights?

[Crazy Homeless Guy]

When I was playing with it, I was using linear, gamma was set to 1.0. basically the default color mapping and no exposure control active.

 

but what I was wondering is if there are factors that need to be adjusted in order to make the results accurate (i.e. not adjusting until it "looks" right, but "knowing" it's right.........if that makes any sense.)

 

I guess in short; I wouldn't stand up and say if you use this light, this is exactly how it's going to look...(in the real world)......to a client. I say this becasue I was put on the spot about this the other day. I didn't use the IES lights, but I was asked how accurate (realistically) is this image.

 

Without being a lighting designer and knowing the small differences between light types. I don't know if a rendering is accurately representing the photometric data. For me, if I can't be 100% sure, then is there really a point to spend the extra time it takes to calculate and adjust the ies lights?

 

Beyond just having Vray setup properly, Max needs to be set up properly in the Gamma and LUT settings under the preferences, and your monitor needs to be properly calibrated.

 

...and I agree, I would never state that my work is accurate to what it will look like. It is an interpretation of what it will look and feel like.

[Crazy Homeless Guy]

Also, why not post the Max file and IES's so we can all play?

[jackb602]

Because you're dealing with human perception, there is no "correct" setting to produce a rendering that predicts exactly how a scene will look in the real world. If you need to objectively measure light levels in a space, I think that can only be done with a false color rendering, which shows how much light falls on a given surface, measured in lux I think. Here's an example:

 

 

I don't believe Vray has this feature, but I bet it would be relatively easy to add.

 

Jack

[Crazy Homeless Guy]

To add to Jack's post... If the goal is a Lux measuremnet, or a other sophisticated analysis, then you need to be using something besides Max. It is my understanding that Max as a whole uses IES data poorly compared to proffesional software, some of which I beleive is free.

[rogue3d]

Beyond just having Vray setup properly, Max needs to be set up properly in the Gamma and LUT settings under the preferences, and your monitor needs to be properly calibrated.quote]

 

Ahh, that makes sense. Never thought of that.

[rogue3d]

To add to Jack's post... If the goal is a Lux measuremnet, or a other sophisticated analysis, then you need to be using something besides Max. It is my understanding that Max as a whole uses IES data poorly compared to proffesional software, some of which I beleive is free.

 

 

that's how this all got started, the OP was asking about IES settings. My whole view was that it's not handled well or easy enough to do in the end in order to justify longer render times and additional variables to have to play with / set in order to get a decent image.

 

My own personal experience, though limited, was from the postion that my client was making a decision on lighting based on my images. I was not comfortable with this, in the end I suggested he have the lighting design firm create some analysis images.

 

I'd be interested in that mathamatical approach using Vray you wrote about yesterday. If you have time someday to post it, I think a lot of people would be interested.

[Carrotious]

wow, was beginning to think no one was interested in the subject...

 

But thanks for the sudden rush of input.

 

I am working linear, and did have the 2.2 gamma, and I am calibrating my exposure using LELE's method of measuring the hotspot in float and bringing it down to 1 by adjusting the camera settings.

 

Of course once i start adjusting camera settings then who knows what's happening to real world light situations.

 

One good thing is that my ies multiplier is 1.

 

what can be seen now is when i change the ies file to a different light eg. a 75watt spot 8°, the light variation is very apparent. very usefull but of course this a change in an arbitrary environment...

 

I'm working with the lighting co on this project and the engineer is doing a 3d lighting analysis. i will be interested to see what he comes up with and who representative of a real world environment his version is.

 

The thing is these little vray renders give u a very easy to understand representation of the light within the room, its very helpful for the client, unfortunately is also fiction...

 

files below Modular lighting (http://www.supermodular.com) Bolster light with available bulbs. strangely the naming convention is wrong or i don't get it. i would have thought that a 75-8 was 75 watt 8° and a 60-8 a 60watt 8° but looking at the lm's of each apparently not 75-8 = 850 lm and 60-8 = 1400 lm

 

strange will take that up with the manufacturer...

 

cheers all

 

G