Looping question revisited
by Ron Wolfe
on
Feb 17, 2004 at 9:27:01 pm
Sorry if this pops up twice in the list. I thought that adding to the original thread about this would move it to the top, but my mistake. The following is about the thread from the end of December that discussed making a looping animation in pI. RW.
I realize this threads about 6 weeks old, but after working on something else for awhile I am still trying to accomplish a looping particle effect. The technique described does not seem to work real well for me because I want to loop the particles when they are at the height of their growth, not during the initial 50-60 frames. Maybe I'm misunderstanding something in the tutorial.
Here's my problem. I need a 30 frame, looping fire/smoke sequence output as individual frames, complete with alphas. The problem I'm running into is that I can figure out how to cross-fade in pI to get a nice looping animation, but - even though I am using a sequence of 32-bit tga frames for the background - pI does not output an alpha for those background frames.
I found a spot in the particle effect I liked between frames 130 & 160. So I rendered out 32-bit tga frames for frames 100-129 - I used the visibility feature to fade the sequence in. Then, I loaded these rendered frames as the background, but starting at frame 131. I then used visibility once again to fade out the particles from frame 130-160. Well, this looks good, and loops, but pI only outputs an alpha for the particles - which is killing me. I'm hoping I don't have to dish out several hundred dollars for a decent video editing software - I don't even know if it would preserve the alpha or not.
Then I thought of another idea. I placed an emitter at frame 0 and then I placed a copy of it, but starting at frame 30. My thinking was that I would fade out the first emitter from frame 130-160 using visibility again, and then fade in the second emitter over the same frame range. Since the second emitter should be 30 frames behind the first one, I thought frames 160 and 130 should match up, then I could just loop from 130-160 and render out. But, 160 and 130 don't completely match up.
Is my thinking off here? Shouldn't the particles get generated the same with the duplicate emitters, but just 30 frames apart?
I'm making these sequences for a 3D Game Engine project, hence the reason they need to be frames, and so forth.
If someone could point out my error or offer another idea, I would sure appreciate it.
Re: Looping question revisited by Alan Lorence on Feb 17, 2004 at 9:31:08 pm
No, the second emitter will not be a duplicate of the first -- there's no way to do that in pIllusion.
I suggest a different approach, which may work, or may cause problems:
Instead of using the first sequence as a bg image, create a "single" particle emitter from the frames (like the "mouse" emitter). This may eat huge amounts of RAM, and could really kill performance if the sequence is long, and the number of frames is high. It should work though.
Re: Looping question revisited by Ron Wolfe on Feb 17, 2004 at 11:04:50 pm
Thanks, Alan. I'll give that a try. It is only small, 30 frame - or smaller - sequences that I will be working with.
I hope it works, because the alternative is doing it manually, frame-by-frame in Photoshop. I wish the alpha wasn't such an issue, but with it, the effects in the 3D Game Engine look almost identical to what I see in pIllusion.
If you have any other ideas at all, feel free to suggest them, but I'll give what you did tell me a try.
Re: Looping question revisited by Ron Wolfe on Feb 18, 2004 at 4:24:02 pm
Alan,
I am trying the point emitter method but - for a reason I cannot yet figure out - it will not match up.
My stage size is 256x256 and I add a line emitter for fiery_smoke. I want to work with frames 130-160. I save out as .tga files frames 100-129. I then, using the mouse emitter as my example, make a point emitter with those 30 frames. I take this new point emitter and place it at frame 130. For some reason, after I place it, I have to increase the size considerably because it is about half the size of the rendered frames, but it still looks fine after doing this.
Then, using visibility, I fade out the line emitter from frame 130-160 and fade in the point emitter. Now, I would swear that when looping 130-160, that frame 159 or 160 should look like the frame right before frame 130 and loop (it should be identical to frame 129), but it just isn't.
Any problems seen with what I've described? Is this what you meant to try?
Re: Looping question revisited by Bas Wolters on Feb 18, 2004 at 6:25:50 pm
I haven't played enough with alpha channels in PI and I don't have PI installed on the machine I'm on now but does PI not preserve the alpha channel of a background sequence as long as the PI output is saved with an alpha channel as well?
If this is not the case maybe you want to try and first make a perfect finished loop without any alpha channel against a black background, which I understand you were able to create without problems. Then use Alan's technique and import the entire finished loop sequence as a new single particle emitter and make the black background color transparent. This way you may be able to avoid the problem you described in your last post. Export with alpha channel.
However, a preserved background alpha channel would be most ideal and much quicker.
BTW - the new (PI 3.0) background image options (start & end frames, etc.) are great! They really help to create perfect PI loops FAST!
Re: Looping question revisited by Ron Wolfe on Feb 18, 2004 at 7:37:14 pm
Thanks. I will try your suggestion, also. These background image options you mention - this is in 3.0? I have 3.0 and do not see this. Perhaps I need to update, or maybe it is in beta? It sounds good, regardless.
