I've always wondered how come changing resolution(dpi) increases rendering time if the size in pixels remains the same? If monitors and TVscreens use pixels for measuring sizes and resolution is mainly used for printing then why do they (monitors and screens) have a resolution of 72 dpi? 3000 x 2745 pixels at 300 dpi equals to 10" x 9.15" like you thought, Shonner. It doesn't matter what resolution you use while rendering. Printing at 150 dpi creates still quite sharp images if you want to get fairly good big prints with less rendering time. You are lucky to use inches in everyday life, we use the metric system here and I must always do more calculating when printing images because every program shows resolutions as dpi.

I've animated a talking puck with bones and it came out pretty well if I may say so. I put one bone to the back of the mouth and from it I created eigth bones around the lips to control lip movement. Then another two bones to control teeth. Multiple morph targets would be cool. Didn't Zmorphics have a plug-in for this? Power pack?

Dual 2 GHz G5 with 1,5Gb ram did the milk file in 58 secs with full raytracing and gamma correction set on. Without them the rendering time was 1:30. Oh, I wish this beauty was mine.

Turning of shadows speeds up the rendering. If the viewing angle is chosen correctly there is no need for cast shadows. Surfaces that aren't facing the light source aren't illuminated and they seem to be in shadow. For some reasons, bumpmaps appear to react better on non-shadowcasting lights. At least in this test scene.

It would certainly explain a lot but no. I do like to have my textures sharp but not that sharp. The Carrara file is 19 megs. I understood that cckens would have liked to have both the C-file and the tiff-file so that he could test if the problem occurred also on PC. So, I tried to make the C-file's size a bit smaller by exporting just the object that has the errors. I closed the original file and opened the file with the exported object and rendered. The flaws were gone, even with the original tiff-file. Of course I can send anyone the original tiff-file but it is probably no use. The problem derives either from the C-file's size or something else that is in the file for also the curtains and the walls have been filled with similar errors.

cckens, I'm afraid the filesize is about 19 megs. Which in fact might be one reason for the errors. As I exported just the one object that had the flaws, the flaws disappeared. I also started to think that the face of the clock is a part of a larger vertex object that has about 10 shading domains. It might be easier for the program if the shading domains were separate objects. I'm beginning to feel that this is something that has occurred when the computer has crashed while I've been working with the C-file. Computers tend to do it when I'm around.

If anyone still remembers the original topic... :)

This morning I reopened the C-file and the flaws were back even with the grayscale image. For those who thought it was the jpg-format that caused the errors, I must unfortunately tell that the original file was tif that was saved from psd and this morning I saved a grayscale tif file from the original psd again. The problem still occurs, not when I set the texture file but when I reopen Carrara and the file with the texture.

So, grayscaling the image may have not been the answer after all. What strikes me wierd is that if the jpg image has errors on its black areas, shouldn't the errors repeat when I just change the minimum value of colours from 0 to 1 and save the image. I did the value change by altering the output range in Photoshop's Levels-window.

It might be a mac related problem too, so those of you who use PCs know nothing of it. I don't know how much differences are the between mac and PC version. With bugs I mean, the programs should be the same.

• Hannu Kulju

Yep, it does the trick and the flaws disappear. I'll use grayscale images from now on for they are a bit smaller too. Thanks. I tried to recreate the problem with a simpler model with less facets and it didn't happen so maybe it has something to do with file complexity. The file the problem occurs in has over 250k facets and 500 objects. As you can see from the picture there are a few objects in the scene and there are more details inside the cuckoo clock on the back wall. This is a scene of an animation where camera shoots first at the clock then glances around the room and finally flies in through the door of the clock viewing the inside of the clock. -Hannu Kulju

My 10 minutes were more like an hour and and a half. Still I would like to add many more details but had to draw the line somewhere. The ground and the snowman are a single metaball object.

I'll try that out. Thanks. HK

I've used Rhino and I recall it has a material(shader) property called Clear finish. It is quite like the fresnel effect and is used for making the surface more reflective as it turns from the viewer and look like lacquered. So I think you quite definately need the Shader Ops and its fake fresnel to create similar materials. Once you-ve got it add the fake fresnel into the reflection slot Just as Ewinemiller has it. In Rhino the material blends into full colorless reflection as the material turns away. I have no idea what the value for the fake fresnel should be but maybe Rhino can answer that if you check the carpaint-shader. There might something else to Rhino's materials but I cannot remember them right now. And probably you need something more to reflect as Hoofdcommissaris said. I wish I could test this first myself but I haven't been able to get the shaderops demo working (what does that tell about my suggestions)...

Attached Link: http://www.renderosity.com/messages.ez?Form.ShowMessage=1102480

I had that problem too and I sought for help from the older messages. I found that it can't be done easily directly. Instead, you can save the goose into another file with the animation and no other objects in it. Then import it back to the original file. It is a bit complicated but this way the mesh and the skeleton get grouped(which you cannot do by hand either) and then you can change the position for the other geese. Maybe someone else has newer information and better ideas but I couldn't figure out any. At least it worked this way for me. I've attached a link to the old messages for you to check for yourself if you want to. -HK

GEE thanks butter! It did the trick! And many others too. I hadn't been able to do a sweep either in VM. I was sick for a while so I couldn't it until now Thanks again

I think the problem is that the program rotates cross-sections' drawing planes around the cross-section centers. As you can see in the picture the plane is perfectly diagonal in the corner but its grid is rotated. The grid should be perpendicular to the imaginary plane of the lines around the corner or one should be able to rotate it. Then one could calculate the change in the form by hand. For example in a corner of 90 degrees the cross-section is in a 45 degrees' angle to the paths' sections and its width or height(whichever changes) would be divided by the sine of 45 degrees (The value is bigger than the original). I managed fix some of the corners in this pipe I created with a cross-section in every corner by doing the corners the way I explained( with dividing by the sine of 45 degrees). In the other corners the grid was rotated and changing the width and height values would have only deformed the pipe in wrong directions. I also tried to set the values first and then rotating the shape but it is hard to get it right that way too. Anyway one shouldn' have to do this kind of shapes by hand. Hope this helps you even a bit.

I could imagine that it would burn me even more having something like that in my hands and not being able to use for there not being enough time or some other reason.