Global illumination (2)

[moss]

Hi,

Another GI question and hopefully the last. When I render one frame just to create a photon map, do my settings (accuracy and radius) have to be as high as my final renders? Thanks.

Astro

[atyss]

Hi tonyastro,

Indeed, if you want to use this photon map when final rendering, you have to use the final output settings. Otherwise when you'll render your final images, the renderer will have insufficient data (taken from the rough photon map), resulting in flickering noise and artifacts.

I know it is long to calculate, but since the map is saved and reused everytime you render, you only have to do it once.

Hope this helps
Bernard Lebel

[moss]

Thanks Bernard. I have a question though: if I create my photon map with high settings, why do I still need to render my final images with these high settings when all of the calculation's already been done? Thanks again.

Astro

[moss]

Hi,

I made an experiment with two renders. One with high settings with "rebuild photon map" checked and the other using same settings but using the created map. They both rendered at the same speed and quality. Why is this? Also, I tried creating a photon map using REALLY low settings and then using that map to render a frame with high settings. The quality and render speed was the same as the first 2 renders. Can anybody help?.. Ed?

Astro

[atyss]

Hi Astro

---

I have a question though: if I create my photon map with high settings, why do I still need to render my final images with these high settings when all of the calculation's already been done? Thanks again.

---

When you created the photon map, you do not need to render with high settings anymore, as long as you do not check Rebuild Photon Map. When the option is checked, mental ray will perform a whole new calculation at each new frame. When this is not checked, and the path in the space leads to your photon map file, mental ray will take data from this file, thus avoiding any need for recalculation.

---

I made an experiment with two renders. One with high settings with "rebuild photon map" checked and the other using same settings but using the created map. They both rendered at the same speed and quality. Why is this?

---

I answered your question above!

---

Also, I tried creating a photon map using REALLY low settings and then using that map to render a frame with high settings. The quality and render speed was the same as the first 2 renders.

---

You mean that your new low-settings map gave excellent results with Rebuild Photon Map checked? I'm not sure to understand clearly. But a real good experience that you could do to test your FG settings you be to create a scene with a moving camera and a moving objects. If your frames do not have artifacts, then your settings are good.
I invite you to take a look at this XSI thread. That topic has been heavily discussed. The settings discussed are almost the same as in SI3D, so you should not have problems to use them:

http://www.highend3d.com/boards/showflat.p...=collapsed&sb=5

Hope this helps
Bernard Lebel

[moss]

My low-settings map gave me identical results (as the first 2) without "rebuild photon" checked. Can you tell me if I'm doing something wrong: here's how I do GI :
- I render 1 frame with "rebuild" checked (with filename/directory for map), accuracy at 500, min/max radius at .1 and .18.
- Render rest of frames with "rebuild" unchecked, same settings
Thanks.

Astro

[atyss]

Ok...

Finally, what you do is that you render frames using the same photon map you created the first frame.
BUT, correct me if i'm wrong, you are saying that this photon map - with low settings - gives you as good results as the very first test (with high settings) you did?

If this is the case, well, you're a lucky man! Achieving high quality with low settings is quite cool. What are your "low" settings?

Again, I encourage you to make a short animation to see if there are moving artefacts in the scene. Keep me updated if you go this way.

Bernard

[moss]

I think I see what my problem is. When I render the rest of the frames using photon maps, I don't uncheck the min/max radius so it recalculates everything (I think). That's why when I rendered using the low-setting map, it overode the map and used the high radius settings. If this is what I had to do, why do none of the tutorials tell you to do this? Correct me if I'm wrong. Thanks.

Astro

[moss]

I'm wrong. Radius has to be checked.

Astro

[atyss]

The Radius values define sizes. This is the minimum and maximum sizes of the "photon particle", so to speak.

When I work on my scene, I use 30 as maximum and 5.0 as minimum, with an accuracy of 100. This give big artifacts square, but at least the preview is fast and the lighting is quite accurate.

When I render (final) my scene I use Max Rad 100, Min Rad 0.01, and Accuracy 3000. This give a smooth render, with only a barely noticable artifact flickering.

My only concern right now is that you might be trying to use Global Illumination with Final Gathering... are you? These are two completely different things.
-The Global Illumination, in the Render Settings, only needs to check Global Illumination to work, not Rebuild Photon Map and other things in the upper part of the window. But the problem is that you have to assign GI properties to every object and light you want to receive/transmit GI.
-Final Gathering, on the other hand, is a subset of GI. It does not require to assign GI properties to lights and objects. When you check Final Gathering, it works, period. But, FG is totally dependant of your materials. If you have only desaturated colors, you wont see much of FG. FG also allows you to use objects as light emitters.

Finally, I am not sure what your question is. In your initial post your question was if you need to render with high settings when creating the photon map file. The answer is yes. But now you made tests and you seem even more confused. In your last post it seems (to me) that you do not understand what impact has the Accuray/Minimum Radius/Maximum Radius on the photon map....... I'm trying really hard to help you, but you have to clarify your question!

Thanks
Bernard Lebel

[moss]

Thanks for differentiating between GI and FG. It's all clear now.

Astro