Wet Map innacuracy

Discussion about fluid simulation in Realflow 4
shaun_michael
Posts: 4766
Joined: Sun Sep 10, 2006 8:04 am

Wet Map innacuracy

Postby shaun_michael » Fri Mar 05, 2010 11:28 am

The last possibility might be that the UV map isn't being exported correctly with the plugin - which I'm about to check.

All fine on the UV exporting in the sd file.

Shaun


User avatar
deckard
Posts: 51
Joined: Wed Sep 26, 2007 9:46 am

Wet Map innacuracy

Postby deckard » Thu Mar 11, 2010 8:59 am

I'll try a new UV when I get unbusy again.

But I see this as RF being picky, when I can paint a texture in Photoshop for this 'unclean' UV and it will map and render just fine. Technically, there is nothing wrong with it, but RF just wants it done a particular way, which is fine.

Am finding other wet map glitchiness. The manual does not adequately explain how to set them up and neither did the tut vid I first saw. One really has to scour this forum to find out the gotchas.

Like the when I first saved TGAs they all rendered black for some reason. I read in another thread where you advised to always use BMPs. So am wondering if TGAs dont work then why are they even an option, and why does RF default to them? Then for some reason, Fusion wont load the RF saved BMPs for processing. I have to first convert them to TGAs thru LW.

If I simulate to say frame 100, stop and scrub back, the wet maps really ought to scrub but they dont. If I scrub back to say frame 50 and start simulating again, the wet map for frame 100 will render out on frame 50. I have to first jiggle settings to get it to update by closing and reopening RF and disabling and re-enabling the wet map.

Also, if I start out in 'wireframe' shading and switch to 'textured' it never works and instead continues to show only a wireframe(?). I have to first switch to 'smooth' then to 'textured' before the UV appears in GL, but it usually shows the original random image you have to apply to get it to work in the first place, instead of the current wet map. If I start simulating it does not show the wet maps, although they are saving out, but I usually see a jump in the way the wet map appears, on that frame I restarted on (this may be about my wacky LW generated UV). If I click on 'textured' again, it instead shows 'smooth' (?). I have to disable and re-enable the wetmap before the wet maps UV shows up. Sometimes I have to do this a few times before it updates. Glitchy.

shaun_michael
Posts: 4766
Joined: Sun Sep 10, 2006 8:04 am

Wet Map innacuracy

Postby shaun_michael » Thu Mar 11, 2010 9:50 am

Yes, very glitchy.

I did a test using the model above and whilst in the scene everything was working as expected, and on XP64 I exported my wetmaps as tga files, no problem. Initially I thought this may have been related to a minor bug in the 32bit build, but having tested that, on native RF geometry it works fine, exporting as tga files.

So what could be the problem. Could it be the UV map on bespoke geometry that RF is, as you say, being really finicky about - possibly. The map used on the above model worked just fine, first time, exporting tga files.

Whilst in the same session of RF I had no problems scrubbing the timeline and the wetmap images updating in GL, but having shut RF down and started a new session with that scene, the images are no longer updating.

So yes, glitchy. Not sure what more I can do to help I'm afraid. Ensure your UV has as few seams in as possible and ensure that substeps are sufficient enough for there to be a decent level of accuracy in the sim (I'm assuming here that the accuracy of calculating each particle on the wetmap is also related to the accuracy of the particle position).

Shaun

User avatar
deckard
Posts: 51
Joined: Wed Sep 26, 2007 9:46 am

Wet Map innacuracy

Postby deckard » Fri Oct 07, 2011 5:40 pm

Am back trying wet maps again with RF 4. They absolutely will not work right, even in a simple scene with a card made of two polys. RF always insists on rendering wet spots where there are no particle collisions.

RF is rendering wet spots for each particle that collides with the collision object, exactly as it should. So there is nothing wrong with the UV or the geometry. But the problem continues to be that RF adds many extraneous wet spots where there are no particles at all in the sim. I've tried adjusting every setting. I dont see how wet maps are possible.

QT example- http://www.box.net/shared/static/xam8bkxtg05v9vrtv0md.mov
project- http://www.box.net/shared/static/xc32u8cderx1l886ivz3.rar

Another wacked QT example- http://www.box.net/shared/static/a72nmm3bih268es4skl7.mov

User avatar
deckard
Posts: 51
Joined: Wed Sep 26, 2007 9:46 am

Wet Map innacuracy

Postby deckard » Tue Oct 11, 2011 4:40 pm

Well I figured out that if I reboot my computer (not just restarting RF), then wet maps start working right (?!) After awhile they get screwy again and I have to reboot. Someone more tech minded may have an idea WTH causes this.


Return to “RF4: Fluid simulation”

Who is online

Users browsing this forum: No registered users and 1 guest