X3 Ray Trace

[Dan Park]

12 core??? Oh dear!!! I do have an antique!

I'm going to have to sell a lot of renders to afford that.

I do have a quad core desktop but I use my laptop to do my Chief work so I can take it with me when I go out of town.

I would definitly use the quad core to do your ray traces.

[Pat1217]

Here's the test I ran. Standard Med quality no Photon map.

Notice the bump on the floor...it doesn't match. I re-did the bump map to make sure it was the same image as the floor and yet the bump is off.

The lighting needs work and I haven't figured out how to get it to look right.

Also I can't get a good stainless look (stove and dishwasher). I tried several different ways to get it to look shiny but haven't come up with anything yet.

I couldn't get the sunlight to come through the window either.

I tested the water glass (on the island) and didn't come up with anything that will make the glass look like glass. I used "flint" and also the normal material.

Allen..... how do I send you the plan with textures? I can't find a way to do it in X3. It's not the same as X2.

[Allen Brown]

File>Back Up Entire Plan.

[Susan Jeffery]

Here's the test I ran. Standard Med quality no Photon map.

Photon Mapping calculates how light bounces around the room. For a truly realistic ray trace, you will need photon mapping.

The lighting needs work and I haven't figured out how to get it to look right.

My magic recipe is to turn my interior ambient light way down (15% or less) and then start placing lights in my scene. This will give you a good idea of how the lights you are placing are lighting the scene. Also, a little post image correcting can work wonders.
Select the adjust image properties button then adjust the sliders (image attached). I almost always increase the intensity and contrast to brighten the scene.

Also I can't get a good stainless look (stove and dishwasher). I tried several different ways to get it to look shiny but haven't come up with anything yet.

Use your rainbow tool to adjust the material properties. Select predefined metal and stainless steel. Lighting in the room will bring out the "shininess."

I couldn't get the sunlight to come through the window either.

For sunlight through your windows, you will need a high level of caustic photons. Place a parallel light outside your window, angle it down 45 degrees and crank the caustic photons.

[Clark Johnson]

I just recieved my new computer the day before X-3 beta was released for ssa. I went right to the ray tracer with a plan I'm working on in X-2. This plan has no lighting or electrical in it so my shot was outside with only natural sunlight. I cranked the settings up for a full page render at high quality, wanting to test the "Wazoo new 6 core Phenom processor". It took only about 45 min. but the quality was far less than desired. I'm used to seeing povray results that are much more realistic. My next test will be a similar render with povray. In all honesty I never use ray-trace views for selling plan sets. My work is all pre-sold and I just don't need them. It is only for play that I even try.
Clark

[Allen Brown]

Susan,

Can you give me an idea on what number range of photons to experiment with?

For the three different photon boxes; caustic, indirect, & lookup, I don't know if I should start with 5, 50, or 500.

Without some sort of basic idea of what number range is "standard" or in the software designer's "Expected" range, I am really shooting in the dark here.

Thanks,

[Keith Jeffery]

Can you give me an idea on what number range of photons to experiment with?

For the three different photon boxes; caustic, indirect, & lookup, I don't know if I should start with 5, 50, or 500.

The wizard should give you a pretty good idea.

For caustics, I'd generally do something between 10,000 and 50,000.

For indirect I sometimes go up to 2,000,000. The limitation here is usually memory -- the photons can take up quite a bit.

For lookup, I'd generally stick to 50-100.

[Susan Jeffery]

Susan,

Can you give me an idea on what number range of photons to experiment with?

For the three different photon boxes; caustic, indirect, & lookup, I don't know if I should start with 5, 50, or 500.

Without some sort of basic idea of what number range is "standard" or in the software designer's "Expected" range, I am really shooting in the dark here.

Thanks,

Usually the defaults that the ray trace wizard sets will work fine. For sunlight through the windows, I would start with 25 caustic photons.

Indirect photons: calculate how light bounces in the room. In real life, light doesn’t just hit a surface and stop, it bounces off of that surface. So, if we increase the indirect photos we’re getting a more accurate calculation of how light would bounce in the scene. I usually stick with the default (8 for standard), but if I want a little extra "oommff" I will go between 10-15. This will add time to your ray trace. Essentially, you are telling the ray tracer to be more accurate in its calculation and that takes time.

Photon Lookup: we’re determining how many photons the ray tracer will look for. Between 50 and 100 will work for most ray traces. If you use too few, the resulting ray trace can look splotchy. Too many can make it blurry. The Ray Trace wizard does a good job of determining this for you. For most cases, you won’t have to change this number at all.

Maximum look up distance determines how FAR the ray tracer looks for photos. Again, if the number is too small the resulting ray trace can look splotchy. The ray trace wizard setting is probably best here.

For any ray trace, you’ll want to make sure Final Gathering is on. The more samples you set, the more realistic the scene will be. If you lower the number too much, the ray trace can get a “noisy” effect. With no Final Gathering, you may see a “disco ball” effect. Most times the ray trace wizard setting will work great, but if you would like even higher quality – this is a good number to increase.

Hope that helps, let me know what other questions you have.

On Edit, Keith's answer is much more simple that mine! I probably gave more than you were looking for.

[David J. Potter]

Susan and Keith,

There is no such thing as too much knowledge or useful data, keep-er coming please and thank you!

DJP

[Allen Brown]

Susan & Keith-DITTO!

Pat, I wanted to go ahead & run a raytrace without making any changes whatsoever, just to get a timed benchmark from your settings. I chose "Indoor-Standard Quality 4" since that was the one that came up.

When I started it, the estimated time was 2 hours, 40 minutes. Actual time was 1:11:09.

Obviously the lighting needs work, and I will try to use some of Susan & Keith's suggestions. The first pic is what I got, the second pic shows what it looked like after adjustment, and what my adjustment numbers were.

Obviously we want to get closer on the lighting before we make adjustments. I will continue to work on it.

[Susan Jeffery]

Allen,
I prefer spotlights with the drop off rate expanded to 120 degrees. They don't produce that glowing orb on the ceiling.

Susan

[Pat1217]

Allen,
It looks like some of my materials are missing. The color of the walls, the bamboo window shade, the flooring....

Wonder why that happened?????

[Pat1217]

Keith and Susan,
I originally was going to use photon mapping but the time showed about 13 hours and if the estimate was over by half, it still would have been too long.

I know I'm not suppose to compare render engines but here's one I just did in Thea....15 min 25 sec on my laptop and it's about double the image size.

I used the same bump map on the floor and it came out fine but the one I did in Chief didn't match up with the diffuse material.

I used the same lights as in Chief, I just upped the intensity.

[sutcac]

Allen,
I prefer spotlights with the drop off rate expanded to 120 degrees. They don't produce that glowing orb on the ceiling.

Susan

Then why do all the can lights in Chief come default with point lights!!!???

[Allen Brown]

Don't believe the initial time estimates given to you by the ray tracer. It almost always overestimates the time at the beginning, sometimes by an extraordinary amount.

I started a high end raytrace about 2 hours ago. The initial estimate was 34 hours to completion. Now that I'm 2 hours into it, it says I have 11 hours to completion. I bet it's done in less than 6. But no, Pat, Pheobe is not going to raytrace anywhere near the speeds of Thea, but I think She'll will beat POV.

I mostly got the materials fixed, but for the life of me, I can't get the stove to raytrace as anything but black.

Anyways, here's a preview: By the way, it looks a lot better than the jpeg attached here, it loses a lot from the compression.