IES Lights Too Dark

[avvid]

Hello everyone.

 

I've modelled my kitchen as accurately as I can, and I've used IES lights, with the correct model for the bulb. This is the rendering result:

 

[ATTACH=CONFIG]45545[/ATTACH]

 

This is an actual photograph, with exactly the same settings (aperture, shutter speed, etc.) as I've used in my render:

 

[ATTACH=CONFIG]45544[/ATTACH]

 

As you can see, the lights in the render don't seem to be anywhere near as powerful. Since V-Ray attempts to be physically accurate, shouldn't I get at least a similar result (I know it could never be exactly the same)?

 

All materials are V-Ray materials, the lights are V-Ray IES, the camera is V-Ray physical camera, dimensions of all objects are pretty accurate, and all the intensity multipliers in the render dialogue are at defaults.

 

 

Thanks.

[Matt McDonald]

Could be a gamma issue...

 

I'd also be curious to see what would happen if you opened the 32 bit image in photoshop and exposed it (by eye) to match the exposure of the photo. Just for grins.

[Dave Buckley]

just crank the multiplier/intensity up or adjust exposure.

 

has the photo been post-processed?

[avvid]

Thanks Elevation, I don't think it's gamma. I'm working in 2.2 space and I've burned it into the image. Do I need to worry about gamma and the IES file? I'll do a raw render in a bit and adjust it in Photoshop, as you suggest.

 

Dave, thanks for your advice. That would indeed achieve the desired results, but the reason for this project, is so that I can learn things like IES lighting. If I can't get physically accurate lights to look almost correct in a physically accurate renderer, then surely I must be doing something wrong. Isn't that half the point of IES? That you don't have to guess how bright a light should be?

 

The photo hasn't been processed at all. It's a JPEG straight off a cheap compact camera. I suppose there could have been some processing within the camera. Might that explain it? I took the settings from the EXIF information and applied them to my V-Ray camera.

 

I really don't like to guess at things. I prefer to have things set-up correctly, and then I can rely on them in future. Otherwise every project would involve endless hours of experimentation until things looked more or less right for that particular shot. If I can rely on a 60 watt bulb to look like a 60 watt bulb, then I don't have to worry about adjusting it; I know it's right. Am I wrong about that? Surely if I do things accurately, they'll look correct, no?

Edited November 1, 2011 by avvid

[avvid]

Okay, well I don't have much experience in working with raw images, but I rendered this out with "Don't affect colors..." checked, then opened it in Photoshop, assigned Adobe RGB color space, set the color mode to 8 bit, to export to JPEG, used the "Exposure and Gamma" dialogue, and set exposure to 3.00, and gamma to 1.00, and got this:

 

[ATTACH=CONFIG]45552[/ATTACH]

 

As you can see, the lamps look about right, but the shadows are now too bright.

 

Am I getting this process right? If not, you can tell me to find some tutorials.

Edited November 1, 2011 by avvid

[John Dollus]

keep in mind that IES data takes into account all interaction with the fixture itself so you need to exclude that from your processing to get accurate levels. for example, if you are using an IES file from a lensed recessed light, you would want to exclude the 3d lens geometry from casting shadows or otherwise impeding the light.

[avvid]

Thank you John. I'd never actually understood whether the measurements were somehow taken or calculated directly from the filament, or from the whole lamp, so I finally have my answer.

 

However, I went as far as hiding the whole structure of the lamp itself (glass, reflector, etc.) and while it's a little brighter, it's still not quite right. I had tried forcing these things to not cast shadows, but it didn't seem to make any difference:

 

[ATTACH=CONFIG]45555[/ATTACH]

 

Do you get the results you expect from IES? Have you ever been able to compare to real-world accurately?

Edited November 2, 2011 by avvid

[NeWAcEiN]

i dnt know if u have modeld bulbs and have assigned and/or using glass on lamps on u r ceiling if u used any geometry in front of light it will block light frm emmiting full energy though go to object properties and uncheck cast shadows plus ull need to crank up intensity on lights since vray and mr works differently on intensities this tune's might help!!!!!!!!!!!!!!!!!

[avvid]

Thank you NeWAcEiN. I did indeed model the bulb, lens, etc., and I did assign a glass material to them. However, I have tried disabling any kind of shadow-casting for them (all methods I could think of), and it doesn't seem to make much difference.

 

As for V-Ray and MR's different takes on intensities: they're both measured in standard units, aren't they? Surely a V-Ray lumen is the same as a MR lumen, or a physical lumen.

 

Anyway, I've been doing some more research, and I think that the power output of the GU10s is quite varied. I've looked at many different IES and Eulumdat files for GU10s and they're all slightly different, so although the photograph that goes with the one I used looks exactly the same as the lamps I'm modelling, it may be slightly different. I'll take a guess at the power on this one, and assume that there isn't something inherently inaccurate about V-Ray's IES system.

 

Thanks for your help everyone.

Edited November 3, 2011 by avvid

[Ismael]

Worth considering:

 

http://download.autodesk.com/us/3dsmax/2012help/index.html?url=files/GUID-7FDACE20-5B5D-4DEF-ABDC-84B2732B0DF-2399.htm,topicNumber=d28e471875

[avvid]

Thank you, that's certainly an interesting tool, I'll take a look at that when I get a chance. I guess I could use it to check how much actual luminance I'm getting from the lamps.

[NeWAcEiN]

Thank you NeWAcEiN. I did indeed model the bulb, lens, etc., and I did assign a glass material to them. However, I have tried disabling any kind of shadow-casting for them (all methods I could think of), and it doesn't seem to make much difference.

 

As for V-Ray and MR's different takes on intensities: they're both measured in standard units, aren't they? Surely a V-Ray lumen is the same as a MR lumen, or a physical lumen.

 

Anyway, I've been doing some more research, and I think that the power output of the GU10s is quite varied. I've looked at many different IES and Eulumdat files for GU10s and they're all slightly different, so although the photograph that goes with the one I used looks exactly the same as the lamps I'm modelling, it may be slightly different. I'll take a guess at the power on this one, and assume that there isn't something inherently inaccurate about V-Ray's IES system.

 

Thanks for your help everyone.

i know they work on same units like lumen lx cat etc but the efficiency handler is different mr and vray engine takes different params to get same effect u cant test that use vray light meter in u r scene to calculate the energy efficiency in u r scene vray IES works gr8 in Vray 2.0 change u r colormaping to reinhard and test render again

[avvid]

Thank you, that's very interesting. I will try the light meter, then I can figure-out what sort of settings I need to get the correct result. And yes, I learnt about Reinhard recently, so I'm using it now; thanks.

[danb4026]

This may sound silly, but are you sure you have the lights pointing in the right direction. Sometimes ies lights can be a bit funky to figure out whether they are pointing up, down, right or left!

[Pat712]

I was/am having the same issue. After playing with the render settings for awhile I eventually decided the image was just underexposed. My solution was just to turn down the f-stop on the vray cam to about 2.0. Maybe try that.

 

Since no one else seems to have any trouble with IES lights I assume that I'm making some noob mistake..

[Dambiel]

According to this link:

http://www.pfbreton.com/2009/08/when-photometric-data-goes-wrong/

sometimes the data in an IES file is not multiplied properly by max.. You may try checking your ies file and seeing if you need to set a multiplier on the light to get the proper value.

[fadi3d]

"I really don't like to guess at things. I prefer to have things set-up correctly, and then I can rely on them in future. Otherwise every project would involve endless hours of experimentation until things looked more or less right for that particular shot. If I can rely on a 60 watt bulb to look like a 60 watt bulb, then I don't have to worry about adjusting it; I know it's right. Am I wrong about that? Surely if I do things accurately, they'll look correct, no? "

i have the same attitude (and i think it's correct) but dam do i struggle to get to this point.Although my renders are acceptable but i'm not convinced the are "Real" looking .

 

everyone does it differently. your point on Ies is true i sometimes crack the power up and estimate how much.i Haven't read the link by Yancy Small and it might be true but the thing is it all gets too complicated sometimes when it should be simple:create A Vray Ies load the Ies Profile and there u go.... heck it doesn't work especially if you work in LWF and u start wondering if there is something to do with this issue.

i really think autodesk should start simplifiying things a bit, i mean set up gamma to 2.2 and everything works accordingly.

[Ismael]

"Yes, I know. I've read over 2,000 pages on the subject of color management and over 10 years of hands on experience to get to the point I am."

Jeff Mottle, 11/8/2011

 

"In digital imaging systems, color management is the controlled conversion between the color representations of various devices, such as image scanners, digital cameras, monitors, TV screens, film printers, computer printers, offset presses, and corresponding media."

Color Management, Wikipedia

 

Using IES files in a render is no guarantee that such a generated image will be Photorealistic or Physically correct. Lights, Geometry, Materials, their proportions and the perceiving medium; they all play in the final image.

 

"Photographer's use a light meter to measure the amount of light needed to successfully capture an image on film/sensor."

About-com 'How Do I Start Taking Photographs'

 

Even knowing the illumination of a given frame, next comes the task of capturing it accurately with the viewing media;i.e., print, screen, plain viewing.

 

I never tire of recommending sorces as this: http://www.fredparker.com/ultexp1.htm

 

 

 

"Thanks everyone for your comments.

 

Abdullah, generally speaking, I put as much effort and detail into a production work as I possibly can. But there are two determining factors to level of detail: the negotiated fee and the project deadline. But you should keep in mind that experience in 3d work isn't a requirement for determining what looks realistic or what is an accurate interpretation of a designer's work. My husband has absolutely no computer experience, but he can immediately spot when something doesn't look correct.

 

Something to consider is how to efficiently use materials and minor geometry tweaks to add realism to your scene. Things that can ruin a scene are:

 

- poor camera work

- unrealistic lighting

- sharp edges on geometry

- bad uvw mapping (noticeable tiling, out of scale, or improperly oriented)

- poor materials (over-saturated bitmaps, low-res bitmaps, etc)

 

Regardless of the negotiated fee, my cameras, lights, objects and materials will always be the best I can possibly do, since it takes about the same amount of time to do things correctly as it does to do things poorly."

 

Frances Gainer Davey, "Bathroom With White Wainscot" Architectural Visualization Gallery