IES profiles on mesh lights - Iray renders them in wrong direction

mwasielewski1990

I have been researching this problem for months now with no avail.

Anyone who ever used IES Light profiles inside of DAZ knows, that if you use it on a point light, the default direction of the light is wrong - you have to rotate the point light +270deg on the X axis, meaning the default light direction points towards the negative X, even despite the IES profile just points downwards (so it should be negative Y from DAZ's perspective).

So, when using point lights that's not really a problem. Just rotate them and poof - problem gone.

However, emissive surfaces in IRAY also have the option to cast light via an IES profile, and such option exist in the surfaces pane when you edit the emissive surface.

But, let's say the emissive surface is just a material zone from a model, and not a standalone mesh light which you can rotate however you want.

What then? Did you guys find ANY solution to, for example rotate the IES profile itself? IES profiles are text files with some math proportions, maybe there's some obscure open-source python script to rotate the whole "formula" by X amount of degrees?

For my workflow this would be an incredible breakthrough if we could FINALLY use IES profiles correctly with IRAY. For any type of architectural renders etc, that would be HUGE.

NorthOf45

There is this: Emission Profile Master, with a user guide on the Readme page. Not sure it can take just any IES Profile, but Totte (Code66) is very helpful. Product thread here.

mwasielewski1990

NorthOf45 said:

There is this: Emission Profile Master, with a user guide on the Readme page. Not sure it can take just any IES Profile, but Totte (Code66) is very helpful. Product thread here.

WOW, this might be what I'm looking for! Thanks, gonna buy it later and test it out.

EDIT: Okay, I read the manual, and at least I "solved the mystery".

Iray does NOT support .IES profiles on just any shape. It only supports it on:

• Photometric liights (such as point light, either as a point, or geometrical like a disc or square)
• Mesh lights that are: rectangles, discs, and cylinders. So if you have any other geometry than this, you can only use the emissive shader "as-is", without an .ies profile.

So the whole idea is somewhat a bust.

Totte

(Sorry I missed this thread)

I can shed some more "lights" onto this problem...(pun intended).

This is what I've learned since I wrote EPM, which just confirms what I found out back then, and still seems valid.

Internally Iray uses Z axis for vertical axis, while Studio uses the OpenGL standard of Y as vertical axis.
Iray supports Type C (street light type) IES profiles, that light downwards, along the Z-axis. So, the complications are that you need the Mesh to be oriented using Z for vertical axis, and then the light will go "Downwards"., then you can realign that mesh so the light goes the way you want, like Esha did for her new Invisi Lights: Ghost Lights and Light Profiles. 

In the IES file you have a definition line like this

1 1000 1 19 1 1 2 -0.1 0.01 0.01

The line can be decoded like this:

    // 1 1000 1 19 1 1 2 -0.090 0.010 0.010
    // 1 = number of lamps
    // 1000 = lumen per lamp
    // 1 = Cancela Multiplier
    // 19 = number of vertical angles
    // 1 number of horizontal angles
    // 1 photometric type
    //     1   Type C photometry
    //     2   Type B photometry
    //     3   Type A photometry
    // 2 Unit type
    //    1   Feet
    //    2   Meters
    // -0.090 0.010 0.010 Luminous Opening Dimensions

 

mwasielewski1990

Totte said:

(Sorry I missed this thread)

I can shed some more "lights" onto this problem...(pun intended).

This is what I've learned since I wrote EPM, which just confirms what I found out back then, and still seems valid.

Internally Iray uses Z axis for vertical axis, while Studio uses the OpenGL standard of Y as vertical axis.
Iray supports Type C (street light type) IES profiles, that light downwards, along the Z-axis. So, the complications are that you need the Mesh to be oriented using Z for vertical axis, and then the light will go "Downwards"., then you can realign that mesh so the light goes the way you want, like Esha did for her new Invisi Lights: Ghost Lights and Light Profiles. 

In the IES file you have a definition line like this

1 1000 1 19 1 1 2 -0.1 0.01 0.01

The line can be decoded like this:

    // 1 1000 1 19 1 1 2 -0.090 0.010 0.010
    // 1 = number of lamps
    // 1000 = lumen per lamp
    // 1 = Cancela Multiplier
    // 19 = number of vertical angles
    // 1 number of horizontal angles
    // 1 photometric type
    //     1   Type C photometry
    //     2   Type B photometry
    //     3   Type A photometry
    // 2 Unit type
    //    1   Feet
    //    2   Meters
    // -0.090 0.010 0.010 Luminous Opening Dimensions

Yeah, I kind of already figured out the axis thing. The problem is, there are situations where an emmissive surface is a part of a larger mesh (i.e. a building, a room, etc.), so you can't just change its axis orientation. With a single object, it's easy.

Anyhow, perhaps we could change DAZ's vertical axis to Z, and this would solve the problem?

Totte

mwasielewski1990 said:

Totte said:

(Sorry I missed this thread)

I can shed some more "lights" onto this problem...(pun intended).

This is what I've learned since I wrote EPM, which just confirms what I found out back then, and still seems valid.

Internally Iray uses Z axis for vertical axis, while Studio uses the OpenGL standard of Y as vertical axis.
Iray supports Type C (street light type) IES profiles, that light downwards, along the Z-axis. So, the complications are that you need the Mesh to be oriented using Z for vertical axis, and then the light will go "Downwards"., then you can realign that mesh so the light goes the way you want, like Esha did for her new Invisi Lights: Ghost Lights and Light Profiles. 

In the IES file you have a definition line like this

1 1000 1 19 1 1 2 -0.1 0.01 0.01

The line can be decoded like this:

    // 1 1000 1 19 1 1 2 -0.090 0.010 0.010
    // 1 = number of lamps
    // 1000 = lumen per lamp
    // 1 = Cancela Multiplier
    // 19 = number of vertical angles
    // 1 number of horizontal angles
    // 1 photometric type
    //     1   Type C photometry
    //     2   Type B photometry
    //     3   Type A photometry
    // 2 Unit type
    //    1   Feet
    //    2   Meters
    // -0.090 0.010 0.010 Luminous Opening Dimensions

Yeah, I kind of already figured out the axis thing. The problem is, there are situations where an emmissive surface is a part of a larger mesh (i.e. a building, a room, etc.), so you can't just change its axis orientation. With a single object, it's easy.

Anyhow, perhaps we could change DAZ's vertical axis to Z, and this would solve the problem?

That would break like everything....