robcat2075

53 seconds! Steve @Shelton has gifted me with a fabulous new computer to further my A:M computing. Thank you, Steve! At the very top of this thread, in 2009, that benchmark scene was taking 19 minutes and 59 seconds to render. Now it's less than 1 minute!

I'm afraid the class had too many real-life things impeding their progress so we had to give it up. I might try again in the future if there is interest.

Welcome back to A:M @CaptBobo !

Any good news on this?

new render

Here are Yves' Photon Mapping tutorials for A:M users who want to get started with "radiosity"... https://forums.hash.com/topic/50594-yves-radiosityphoton-mapping-pages/

Radiosity is for mostly enclosed interiors but scale should not be the problem Here is an image from Yves' radiosity tutorial page... The lightness/darkness/contrast result is all adjustable. The strategy for success is not requiring that what works for exterior lighting is what has to be used for interiors and vice-versa. A universal light set up that satisfies that need probably isn't possible to do unless the huge difference between "outside" and "inside" is what is wanted. They don't do it in live-action Hollywood movies and they don't do it in animation studios. I can't think of a Hollywood live-action film that does such a shot without substantial trickery. When you see the outside from the inside in a movie it is rarely real. It is usually either rear-projected or green-screened or a model or a painting that has been placed outside the window of the in-studio "set". When they can't shoot in a studio and real daylight outdoors has to be seen through a real window they typically have giant sheets of grey film they put outside the window to darken the outside view to mate with interior lighting they do inside. In studio animation they design lighting for each shot individually and only after it has been finaled on the animation. Back in the 20 'oughts one of our forum members was working at DNA studios on "Ant Bully" and he gave me a tour of the studio. Everything the animators had to work with was in gray. No lighting or texturing was done until the shot was finished. If I had to do a shot such as you describe I might render one pass lit for the outside with AO, with the train interior all-black... Another pass lit for the interior lit with radiosity or bright AO, with the outside all black... And composite the two together... There would be other ways to do it and get a good result without needing the radiosity.

BTW, if you want to talk A:M live we have Live Answer Time every Saturday.

I'm afraid I never got used to using it. If someone else messaged me it was an interruption to whatever I was doing in A:M and i felt bad about always saying, "I can't chat right now". I like interacting on the forum.

Yes, i think not using partial darkness at all is best. I avoid it for any "finished" work because it creates painfully flat shadows. It's really just a cheat left over from the stone age. Ya know... you can animate that Global Ambience Intensity setting to be higher when the camera is in the train car. But for these interior shots you are doing, even more satisfying results could be had with radiosity. I doubt the render times would be much more painful than the 100% AO times.

I forgot to mention that AM can also do the compositing...

A one-pass solution is to use two lights, one Diffuse-only with the shadow darkness set as desired, and another that is Specular only with Darkness set to 100%. For some reason the Specular light must be ray-traced and have more than one ray. In this sample PRJ I constrained the Spec light to the Diffuse light so that only one needs to be manipulated in the Chor. I also gave the Spec light Expressions to take intensity, cone width and width softness from the Diffuse light so those parameters only need to be set in the Diffuse Light. SpekDarkness004.prj Diffuse-only light... Specular-only light... Both lights ON...

No, I haven't seen that happen before. Try this... On the first machine, save the chor out as a separate .cho file. Then save the PRJ as a new PRJ with a new name. Take both to the new computer. Open the PRJ and import the .cho file into it Do Embed All and save the PRJ with a new name.

My limited understanding of how Shadow Darkness works is that it renders the shadow as if object were less than 100% opaque. So with 80% Darkness, 20% of the light is still passing through the object and you get 20% of whatever the specular effect would be with no shadowing object at all. I'm doubtful there is a simple fix. The easiest work-around is to model and light and test as you normally do, then at render time ... Make one pass with your intended Darkness setting, but with the light's Specular Property OFF.... Then a second pass with Darkness set to 100% and the Specular ON... Then in Photoshop or other image editing app that does Layers and composite modes, composite the two together with the top Layer set to "Lighten" (sometimes known as "Maximum"). "Lighten" chooses the brightest version of each pixel in making the final image...

Steffen has marked this as "resolved" on the next update. Thanks for pointing this out.

Hurray! As long as the Action got baked on every frame the interpolation differences between Quat and Euler shouldn't be a big issue.

HI Charles, That Option setting seems to be broken and broken back to v17 at least. So, what to do? First, If you just want three channels and no more, you want Euler, because Vector has X, Y, Z and Roll. But either way, do this: Start a new Action with your character. Set the Key Skeletal Rotations filter and Key Model filter Do Shift-Force Keyframe, this brings up an option window. Choose "In all filtered channels" and press OK. This creates rotation keys on all the bone without needing to do it manually. Those rotation keys will be all Quat. Select the first bone in the Action, then in its Properties>Object's Properties, on Rotate and choose Convert Driver To > Euler (or Vector if you really want that).... 5. Repeat step 4 for all the other bones 6. Animate your bones. They will continue with the interpolation that has been set. If you plan to make many Actions with this character, save this Action out before you animate further. You can re-use it as a starter action that has your desired interpolation method already established.

That percentage for "lifetime of the particle" refers to a percentage of the life that is set in the "Life" parameter a few lines up , "00:01:00" in the image above. A keyframe that happens at 50% will happen at 0.5 seconds in the particle's life. If you changed the particle's Life to 3 seconds, the keyframe at 50% will now happen at 1.5 seconds.

The horizontal axis is the lifetime of the particle. The vertical axis is the property being controlled, in this case the opacity.

Try this... 1 make sure Curve View is selected 2 Use the magnify tool to squeeze the top scale so that 100% is visible 3 put values into "Over Life"

Is this something all the sprites will do... be born at 100% and fade over time?

HI Tom, Eliminate the Keyframes for Opacity in "Sprite Test 1" and try keyframing 100% --> 0% for Opacity in "Sprite System" The value in the Sprite Image ("Sprite Test 1") parameters changes the opacity of sprites currently being emmitted The value in the System changes the opacity for all the particles at once. Note that if you want each particle to have a fade or change over its own life regardless of when it was emitted, that needs to be key framed in the original material.

You're not missing anything. You can't keyframe volume level changes in an A:M Chor. It has been a requested feature but it hasn't happened yet. Typically you will prepare your sound track in some dedicated editing program like Audacity and export a single file to use in A:M When I've done a shot with several wavs that I had to move around to get them synced to my animation, I'll render that out to get the rough mix out of A:M, then plop that in Audacity and use it a model to make a version with the same wav parts but with the volume levels (and other effects) I really want. I mute the original A:M wave, export my good mix from Audacity to a new Stereo wav and swap that into my A:M chor.

Here is the same test, except that the EXR maps were rendered with 1 pass instead of 4. One pass actually got a smoother progression of gradient values, even though the number of pixels in the decal is the same either way. I can't imagine what math mishap would cause the difference. But there it is.

This image compares seven different narrow grayscale gradients rendered to EXR and used as displacement maps. These results roughly align with what I've read... the precision of floating point variables decreases as the value gets farther from zero. From RGB 0 to RGB 1 the increments are so small that the stepping is nearly invisible, but from RGB 254 to RGB 255 the precision only allows a few steps. I can't explain why these have uneven steps and don't match the previous examples. Possible difference... these were shot with an orthogonal camera instead of a perspective camera... ??