## LuxC4D 0.06c

Discussion related to the 'LuxC4D' exporter plugin for Maxon Cinema 4D.

Moderators: Ratow, coordinators

### Re: LuxC4D 0.06c

abstrax wrote:I mean the path of the .lxs scene file. The LuxRender path can have non-ASCII characters, the scene file mustn't. (I think it's a bug in Lux for Mac OS). So what is the path of the scene file?

Oooh. /Users/james/Desktop/CarPaint/carpaint2.lxs

No spaces, nothing crazy.
forgeflow

Posts: 137
Joined: Sun Sep 13, 2009 3:51 pm

### Re: LuxC4D 0.06c

forgeflow wrote:
abstrax wrote:I mean the path of the .lxs scene file. The LuxRender path can have non-ASCII characters, the scene file mustn't. (I think it's a bug in Lux for Mac OS). So what is the path of the scene file?

Oooh. /Users/james/Desktop/CarPaint/carpaint2.lxs

No spaces, nothing crazy.

Spaces would be fine too.

Hmm, but that's bad as I don't have a clue, what's going on here. I will build you a version that dumps the paths out - in the hope that I can spot the error. Thanks so far for your help.

Cheers,
Marcus
abstrax

Posts: 247
Joined: Sat Dec 08, 2007 3:52 am
Location: Auckland, New Zealand

### Re: LuxC4D 0.06c

Dumb question time...

How do I go about using an HDR image as a background/light source? I tried a simple scene with a Lux Light tag set to "Distant Light" and exported it, but of course LuxRender (0.06) barfed on the file (Unimplemented Light). I think I read somewhere else that these light types are broken? Not adverse to manually editing the .lxs file, but there aren't any examples of how specifically do to this. LuxBlend was of no use to me to "cheat" an example setup file, probably because I have no idea what I am doing in Blender. The scene file format wiki is curiously devoid of examples of scene file formats...

Mapping the .exr file to a giant sphere in Cinema 4D seems to work ok, but it's not really an ideal solution. What do I do?
forgeflow

Posts: 137
Joined: Sun Sep 13, 2009 3:51 pm

### Re: LuxC4D 0.06c

forgeflow wrote:Dumb question time...

How do I go about using an HDR image as a background/light source? I tried a simple scene with a Lux Light tag set to "Distant Light" and exported it, but of course LuxRender (0.06) barfed on the file (Unimplemented Light). I think I read somewhere else that these light types are broken? Not adverse to manually editing the .lxs file, but there aren't any examples of how specifically do to this. LuxBlend was of no use to me to "cheat" an example setup file, probably because I have no idea what I am doing in Blender. The scene file format wiki is curiously devoid of examples of scene file formats...

Mapping the .exr file to a giant sphere in Cinema 4D seems to work ok, but it's not really an ideal solution. What do I do?

You use an "infinite light": http://www.luxrender.net/wiki/index.php ... rmat#Light

The infinite light has also a parameter "gain" (float) with which you can control the strength.

EDIT: I can see from the code that you can specify the mapping type with the parameter "type" (string), too, which can be set to "latlong", "angular" or "vcross". Don't ask me what they mean - you probably have to try out.

Cheers,
Marcus
abstrax

Posts: 247
Joined: Sat Dec 08, 2007 3:52 am
Location: Auckland, New Zealand

### Re: LuxC4D 0.06c

abstrax wrote:You use an "infinite light": http://www.luxrender.net/wiki/index.php ... rmat#Light

The infinite light has also a parameter "gain" (float) with which you can control the strength.

EDIT: I can see from the code that you can specify the mapping type with the parameter "type" (string), too, which can be set to "latlong", "angular" or "vcross". Don't ask me what they mean - you probably have to try out.

This worked:
Code: Select all
LightSource "infinite" "float gain" [0.1] "string mapname" ["/Volumes/storage/Photos/hyperfocal/HDR_Sunset01_sample.exr"]AttributeEnd

The mapping type would depend on what your .exr image mapping was - "latlong" is the typical rectangular HDR image that wraps around 360 degrees like a panorama (exactly like the one lux produces when you use an "environment" camera. "angular" is a "chrome sphere" type of HDR image, and "vcross" are the "unfolded" box type of HDR, like you can see here: http://www.debevec.org/Probes/ (scroll down to see the different probe mappings).
forgeflow

Posts: 137
Joined: Sun Sep 13, 2009 3:51 pm

### Re: LuxC4D 0.06c

forgeflow wrote:
abstrax wrote:You use an "infinite light": http://www.luxrender.net/wiki/index.php ... rmat#Light

The infinite light has also a parameter "gain" (float) with which you can control the strength.

EDIT: I can see from the code that you can specify the mapping type with the parameter "type" (string), too, which can be set to "latlong", "angular" or "vcross". Don't ask me what they mean - you probably have to try out.

This worked:
Code: Select all
LightSource "infinite" "float gain" [0.1] "string mapname" ["/Volumes/storage/Photos/hyperfocal/HDR_Sunset01_sample.exr"]AttributeEnd

The mapping type would depend on what your .exr image mapping was - "latlong" is the typical rectangular HDR image that wraps around 360 degrees like a panorama (exactly like the one lux produces when you use an "environment" camera. "angular" is a "chrome sphere" type of HDR image, and "vcross" are the "unfolded" box type of HDR, like you can see here: http://www.debevec.org/Probes/ (scroll down to see the different probe mappings).

As I will release a version 0.06d with tweaked lighting behaviour (at the moment it's usually waaaay too bright), fixed launch problems on Mac OS and a few other changes, I plan to add the export of infinite lights, too. It was high up on my to-do list anyway. And no worries I'm also working on 0.07

Cheers,
Marcus
abstrax

Posts: 247
Joined: Sat Dec 08, 2007 3:52 am
Location: Auckland, New Zealand

### Re: LuxC4D 0.06c - Materials again

Hi,

I've been trying to play around with materials and stumbled upon a little strangeness...

Any color setting from C4D seems to be totally exaggerated, when taken into Lux. In my first material test, a 100% red looked totally out of this world... so I digged a little deeper into this.

I created a test file with 20 spheres one area light, which lights all spheres evenly and some (matte) materials. They only differ in the brightness of the color. the darkest sphere has 0% (equals 0.0.0 in RGB) and the brightest one is 100% (255.255.255).

Now this is, what the standard C4D renderer makes out of this:

And this is, what Lux makes out of it:

You recognize, that even at very low values, Lux renders the sphere below 10% (25.25.25) already very bright. While I know that this may have to do with the physical approach of Lux, I find this very confusing (or even annoying) from a user perspective, as I need a certain reliability from the previews in the editor, color and material settings...

Any idea, how this can be solved???
Dikkker

Posts: 7
Joined: Sun Nov 15, 2009 7:28 am

### Re: LuxC4D 0.06c - Materials again

Dikkker wrote:Hi,

I've been trying to play around with materials and stumbled upon a little strangeness...

Any color setting from C4D seems to be totally exaggerated, when taken into Lux. In my first material test, a 100% red looked totally out of this world... so I digged a little deeper into this.

I created a test file with 20 spheres one area light, which lights all spheres evenly and some (matte) materials. They only differ in the brightness of the color. the darkest sphere has 0% (equals 0.0.0 in RGB) and the brightest one is 100% (255.255.255).

Now this is, what the standard C4D renderer makes out of this:

And this is, what Lux makes out of it:

You recognize, that even at very low values, Lux renders the sphere below 10% (25.25.25) already very bright. While I know that this may have to do with the physical approach of Lux, I find this very confusing (or even annoying) from a user perspective, as I need a certain reliability from the previews in the editor, color and material settings...

Any idea, how this can be solved???

Which tone mapping do you use? In a nutshell: If you are using "Reinhard", then Lux will pull up the brightness, until the total average brightness is a middle grey. If you are using "Contrast", I think the average contrast gets maximised. If you use "Linear" you can set the image brightness directly in the form of real camera settings. If you use "Maxwhite" Lux stretches the brightness to the highest value, without blowing the highlights.

Please use "Maxwhite" for these kind of tests and tell me, if the problem persists.

Cheers,
Marcus
abstrax

Posts: 247
Joined: Sat Dec 08, 2007 3:52 am
Location: Auckland, New Zealand

### Re: LuxC4D 0.06c - Materials again

Dikkker wrote:You recognize, that even at very low values, Lux renders the sphere below 10% (25.25.25) already very bright. While I know that this may have to do with the physical approach of Lux, I find this very confusing (or even annoying) from a user perspective, as I need a certain reliability from the previews in the editor, color and material settings...

Any idea, how this can be solved???

I recommend using Linear and adjusting the exposure settings to get the image you want. When you get the settings the way you like them, go back to the Exporter settings and plug those values in. That way, each time you redo the rendering setup, Lux will start up with the same settings. Reinhard is sort of an "auto" exposure and really only meant to be a reasonable start. Linear is the way to go - it acts the way a camera and film would react.

Sensitivity is like your film ISO setting - higher numbers mean "more sensitive film" - image gets brighter
Exposure is like the shutter speed - higher numbers mean "more light" - image gets brighter
FStop is like the aperture setting on the camera - higher numbers let "less light" in - image gets darker
you can safely ignore Gamma unless you want to alter the contrast of the image.

Its a great way to approach it if you are familiar with cameras.

One thing to make note of though - 100% white doesn't exist in the real world. That would be a surface that reflects 100% of all light hitting it. The whitest white you'll encounter in the real world reflects *maybe* 80% of the light hitting it. Keep this in mind - C4D's conventions don't really simulate real-world conditions. I've done perfectly reasonable renders of "white walled" rooms in LuxRender where the wall colour was 50% grey in C4D.
forgeflow

Posts: 137
Joined: Sun Sep 13, 2009 3:51 pm

### Re: LuxC4D 0.06c - Materials again

forgeflow wrote:One thing to make note of though - 100% white doesn't exist in the real world. That would be a surface that reflects 100% of all light hitting it. The whitest white you'll encounter in the real world reflects *maybe* 80% of the light hitting it. Keep this in mind - C4D's conventions don't really simulate real-world conditions. I've done perfectly reasonable renders of "white walled" rooms in LuxRender where the wall colour was 50% grey in C4D.

Ok - I'll play with the different modes and post the results later today.

Anyhow - the question (from a usability standpoint) is, if there should be some sort of "adapted conversion" between a non-physical editor and a physical renderer. Otherwise you can never use any kind of (editor) preview for your final rendering, which would turn the renderer useless IMHO.

If Lux needs a white value of ~80% (or less) for the brightest natural white, then a 100% C4D-white should convert to that. I know this depends a lot on other factors such as lighting and camera settings, but if a renderer should be useful for any person from a creative rather than a technical background, than this is inevitable... IMHO.

How does this work in other renderers, such as Blender or Maya?

Cheers

Dirk
Dikkker

Posts: 7
Joined: Sun Nov 15, 2009 7:28 am

PreviousNext