All posts related to V2
#368252
Hello, I am trying to create a material that could reproduce this incandescent effect, in the tip of a metal tool.
The effect should decrease gradually blending softly with the metalic surface at the other end.
Image
I can create metalic materials, and I can create an incandescent effect, but cannot imagine how to recreate a gradual blend between one and the other.
Any help will be greatly appreciated!
Ernesto
Last edited by Ernesto on Sat Jun 01, 2013 1:42 am, edited 1 time in total.
By hatts
#368257
Pretty easy actually,

First method:
1. Single material that has two layers.
2. On the top layer, make an emitter. This can be a color (like bright orange) or maybe a gradient (red -> orange). This layer gets an opacity/clip mask, which singles out the "glowing" area on the object.
3. On the bottom layer, do the BSDF(s) for the regular metal.

Second method, for if you need custom UV mapping:
1. Duplicate the object (e.g. pliers)
2. Object 1 has only a glowing emitter material, with clip mask.
3. Object 2 must be *slightly* smaller or "behind" Object 1. It gets your regular metal material.
4. This way both materials can have different UV maps. Certain apps have multi-materials, which supports multiple UV maps, but I believe C4D does not support this, so you need the duplicate geometry.

Anyway here's my basic example (method 1)
Image
User avatar
By Ernesto
#368258
Thanks Hatts!

It looks simple, I will work on it, and will post the results soon.
Anyway I have a question about Clip Maps:

Perhaps I had a wrong idea of clipping maps, but I thought they had to be black and white only maps without gradients.
By what you explained it seems that I was wrong, and it is possible to use a black-gray-white clipping map.
Pleasse correct me in case i am wrong!

I think I will use the first method, since my metalic material will not need an UV.
Anyway i like the idea of two UVs, and asked the question in the Maya forum: http://www.maxwellrender.com/forum/view ... 53#p368253 just in case.

Thanks again!

Ernesto
User avatar
By Ernesto
#368260
Well, here I have the result:
Image
It is a golden ball, with certain incandecent parts.

I have played with the emiter intensity, and could see that the gradient works best for certain emiter intensity.
If the emiter is brighter the gradient gets lost, and if the emiter intensity is lower the contrast is lost.
I assume this is because the clip map is not HDRI and cannot adapt to the light variations.
I was planning to create a lighting animation using multilight adjustments, showing that the piece of metal gets more and more hot, and the gradient would be moving from one tip to the other, but not lost.

I wonder Is there any way to do this perhaps using HDRImage?
I have tried, and it works with a black background, but I cannot use a clipping map, and a Gloden material behind, because Maxwell cannot accept HDRI clipping maps... In case I use a clipping map, the limit between the incandescent and cold metal, gets visible without a gradient in between, for certain intensities.
Another way that i thought of is using an animated clip map...
Any other idea, is greatly welcomed!

Ernesto
#368264
My conclusion about this topic, is that the mask is limiting the emiter range.
As it could be observed in the material samples, the only way to get a gradient in the emiter, which is controled by the mask, is using a specific intensity, which may not be the desired or needed in the scene. In this specific case, the ideal intensity to get the best behavious of the mask, is totally insuficient to simulate a burning metal.
If a higher emiter intensity is choosen, the gradient is lost, and if a lower intensity is coosen, there will not be a good contrast.

It seems that the mask and the emiter have totally diferent ranges, being very narow in the mask, and extremely wide in the emiter.

The emiter range seems equivalent to an HDRI, but the mask is similar to a Low Range image.

1) The solution could be to enable HDRI or MXI format masks. In this way , the range of the mask will be wider because of the file format.

2) Or to ad a multiplier for the gray mask in order to cover the wide range of intensities of possible emiters. In this way, the user could choose the desired intensity, and after that, you could adjust the multiplier for the mask, in order to show the full range of gradients.

Perhaps Next Limit Team could find another and better solution.

Ernesto
#368266
I think what you're seeing is rather the effect of exposure clipping. You increase the strength of the emitter = you have to adapt your exposure or highlights will start to appear clipped (just a white blob), just as it happens in real photography. It wouldn't matter if you have "HDR clipmaps", because you don't have an HDR monitor - the monitor shows you values within a certain range, much less than what is actually in the HDR image. So you would still get white clipped areas if the emitter is too strong for that certain camera exposure.
#368268
Mihai wrote:I think what you're seeing is rather the effect of exposure clipping. You increase the strength of the emitter = you have to adapt your exposure or highlights will start to appear clipped (just a white blob), just as it happens in real photography. It wouldn't matter if you have "HDR clipmaps", because you don't have an HDR monitor - the monitor shows you values within a certain range, much less than what is actually in the HDR image. So you would still get white clipped areas if the emitter is too strong for that certain camera exposure.

No Mihai, it is not exposure related in my opinion.
In the lower case (materials previews) where the limit between the strong emiter, and the plain metal is a line, instead of a gradient, there is no way to fix that adjusting the exposure to get the gradient back.
In the first example, (the photograph of the metalic tool) at the tip, the light is overexposed, but you can still see the gradient.
That is impossible to recreate with curent Maxwell Materials, because if you want to overexpose the tip, the gradient dissapears.
And if you want to keep the gradient, you are forced to use a dim light as in the example by Hatts.

Image
This is an image edited in Photoshop, where you could see the burned area, and at the same time the gradient.
This is what cannot be done because the range of the mask is narower than the range of light intensities.

Ernesto
#368271
You know, part of this work we do is try to find solutions to problems, and most times there are. You would be a lot more productive if you stop these constant claims that something is impossible to do, especially in a case like this when it isn't and you don't even need to find a workaround. I don't understand why you can't use the example Hatts posted. The tip is almost white, you can't over expose higher than that. Then it's just a matter of using the right contrast in your weight map...

Here is a quick example, I just created a default gradient in PS, no contrast adjustments necessary even. I also exposed the rest of the image to bright sunlight so you won't say it only works for dark scenes. The emitter is strong enough so it looks white even when exposing for bright sunlight. So the tip is white, don't need any stronger emitter, it's white already. Do you see any sudden clipping in the layer weightmap? Ofcourse if you set your emitter strength to very high values, it's going to be so strong it "fills" out any gradient you apply to it. But as you can see you don't need high values to make a glowing white tip and it looks just fine. A glowing metal rod doesn't exactly put out 40 000 lumens of light.....

Image
#368272
Another example to better mimic your first example photo. Here I've made the gradient map more tight, hence the smaller zone of falloff. I just added a sceond emitter layer which has a more red emitter. Using correlated color setting for this is the fastest. I even think these emitter settings are too strong, I doubt a glowing metal would put out this much light that it's so visible in broad daylight exposure.

Image
#368278
That looks a better aproach... Mihai
As far as I understand you are using several clip maps, and one emiter for each of them, right?

Regarding my critic to the way current materials are limiting the light intensity in the emiters, it is not to bother you, at all.
I know you are creative enough, and I do not doubt that you will find ways to get the right look, but considering Maxwell claim, of being a phisically correct renderer, the best aproach would be exactly that: Being Phisically Correct.

The use of several clipping maps with a diferent gradient assigned to a diferent emiter, could make things better.
Let´s assume the range for each clip map, is 10, if properly set, using 10 clipping maps it would be possible to get a range of 100.
I really do not know if that would be enough or not, but one thing is sure: if we could get a phisically correct behabiour would be much better.

Just to be mathematically correct, here is the problem:

The ideal behaviour would be the following:
Using a gradient between white and black.
In the white areas, the emiter should show the full specified intensity
In the black areas, the emiter would not act
In the 50% gray areas, the emiter would show 50% of the specified intensity.

Actual behaviour:
Using a gradient between white and black.
In the white areas, the emiter show the full specified intensity
In the black areas, the emiter do not act
In the 50% gray areas, the emiter is not showing half of the emiter intensity. If the intensity of the emiter is high, the 50% area will work as the white area, showing the full intensity of the emiters. If using a very low emiter intensity, it seems to work as expected (showing half of the specified intensity)

Ernesto
#368279
Mihai wrote:Another example to better mimic your first example photo. Here I've made the gradient map more tight, hence the smaller zone of falloff. I just added a sceond emitter layer which has a more red emitter. Using correlated color setting for this is the fastest. I even think these emitter settings are too strong, I doubt a glowing metal would put out this much light that it's so visible in broad daylight exposure.
You're right, Mihai, building a red hot steel material this way works excellent and is easy to fine tune. One for the library :)
#368285
Ernesto wrote: Just to be mathematically correct, here is the problem:

The ideal behaviour would be the following:
Using a gradient between white and black.
In the white areas, the emiter should show the full specified intensity
In the black areas, the emiter would not act
In the 50% gray areas, the emiter would show 50% of the specified intensity.

Actual behaviour:
Using a gradient between white and black.
In the white areas, the emiter show the full specified intensity
In the black areas, the emiter do not act
In the 50% gray areas, the emiter is not showing half of the emiter intensity. If the intensity of the emiter is high, the 50% area will work as the white area, showing the full intensity of the emiters. If using a very low emiter intensity, it seems to work as expected (showing half of the specified intensity)

Ernesto
I think, Ernesto, it's already doing what you want. Everything has to do with the exposure of your scene. If you put a very high emitter intensity, for example 10000W that burns the image at that point, it can happen that 50% of that super high intensity also burns the image at that point and maybe 25% too, and maybe 2% too. But if you give less sensitivity to the camera (less ISO, more SS) you would eventually see the gradient.
#368288
I still see something wrong here.
Image
Look the soft gradient as previewed in maya, and compare with the same version in Fire.

In this test, a lineal gradient as default in photoshop was used, but the light gradient is acting in a non lineal way.
The light part is longer, and the dark side reduces its intensity abruptly

I do not know why it goes very abruptly to black.

Ernesto
Last edited by Ernesto on Mon Jun 03, 2013 5:43 am, edited 2 times in total.
#368289
In an attempt to correct the gradient, in a way that it would look lineal, I tried to use a non lineal clipping map.
Here, I had to force the texture (look at the curves in PS) so that it will compress the light part and will expand the dark part, in order to get a more lineal behaviour.
Image
If you use lineal gradient maps, the falloff behaviour gets slow in the light and abrupt in the dark border, as it can be seen in the previous post.

In the present example, you can see some problems in the dark side, where the gradient seems not enough to map so dim light.
I think someone expert at NL should look at this, she or he would know what is happening here.
I am not an expert.

Ernesto
#368292
The clipmap is not so much the issue, the graduation of the 32bit emitter is what you need to fix. Mihai and Fernando have already explained this..

Remember your screen is only 8bit (at best) so you cant see the intensity of the 32bit data, but you can adjust that intensity slider back and forth in Photoshop to see the full range of intensity in the hdr emitter.
Will there be a Maxwell Render 6 ?

Let's be realistic. What's left of NL is only milk[…]