Planetside Software Forums

General => Image Sharing => Topic started by: nbk2 on April 07, 2012, 12:00:02 PM

Title: A new Tornado approach
Post by: nbk2 on April 07, 2012, 12:00:02 PM
I had a look at Old Blaggards Tornado and thought I start a new approach. Just so my mind doesn't fall asleep.

I simulated the Tornado vortex via an expanded very dense helix with a slight sine wobble. (See image tbasics) which is again the helixed to a slight degree

Biggest problem is that the density of the helix fizzles out at lower cloud densities see 1 and 2 and lighting the thing is a nightmare.

30 mins on making the nodes. 3 days lighting..

But a nice rope tornado and not one of the big ones maybe got a step closer.

In the moment the basic shape comes from a circle image and the ground thing is a doughnut.

I haven't tried yet to make the warper eat instead of the image Mr Lampposts reverse cone.

I tried that but couldn't displace it and went back to my approach.
Title: Re: A new Tornado approach
Post by: dandelO on April 07, 2012, 12:19:37 PM
Nice job, Nbk2! The first image looks great. 8)
Title: Re: A new Tornado approach
Post by: nbk2 on April 07, 2012, 12:38:45 PM
Quote from: dandelO on April 07, 2012, 12:19:37 PM
Nice job, Nbk2! The first image looks great. 8)

Thanks a lot.  ;D
Title: Re: A new Tornado approach
Post by: Kadri on April 07, 2012, 01:06:20 PM

Looks good  :)
Title: Re: A new Tornado approach
Post by: nbk2 on April 07, 2012, 01:37:19 PM
Quote from: Kadri on April 07, 2012, 01:06:20 PM

Looks good  :)

Thanks a lot Kadri!
Title: Re: A new Tornado approach
Post by: Oshyan on April 07, 2012, 08:38:13 PM
Very promising!

- Oshyan
Title: Re: A new Tornado approach
Post by: mhaze on April 08, 2012, 05:56:42 AM
Took the words out of my mouth Oshyan! Keep at it this could be good!
Title: Re: A new Tornado approach
Post by: nbk2 on April 09, 2012, 01:15:28 AM
update

Still to clean .. but well  ..

There is now an inner funnel and and outer one and a feeder cloud .. 
Title: Re: A new Tornado approach
Post by: Dune on April 09, 2012, 02:55:26 AM
This is looking dramatically good, much better than the warped simple shape I used once.
Title: Re: A new Tornado approach
Post by: elipsis1 on April 09, 2012, 09:38:35 AM
Stunning!  ;D
Title: Re: A new Tornado approach
Post by: nbk2 on April 09, 2012, 03:53:32 PM
Thanks a lot :)
Title: Re: A new Tornado approach
Post by: Kadri on April 09, 2012, 03:57:18 PM

Looks very good ! I like the layered look  :)
Title: Re: A new Tornado approach
Post by: dandelO on April 09, 2012, 06:58:17 PM
I like the dust kick-up and the inner/outer funnel gives a nice effect, too. Great work on a tricky subject. :)
Title: Re: A new Tornado approach
Post by: nbk2 on April 10, 2012, 02:20:53 PM
I think it's ok if you look at what I was aiming at.

http://www.markdmckinney.com/images/Tornado2.jpg
http://www.uvm.edu/~inquiryb/webquest/sp08/pmontgom/Tornado.jpg
http://www.weatherstock.com/slides/Tornado%203-RF-CD.JPG

I have roughed it up since then a bit  but couldn't help myself introducing the next level ...

as an idea from this image
http://noticeablyfat.com/gallery2/d/2289-2/Tornado+Lightning.jpg

I used one of cyphyrs lightnings which become to broad. In the moment I use a lightning that isn't visible to the camera at 100 luminosity and one visible at 1. I think I need to exchange the model with a plane and a real lightning to get thinner lines.

At least a lecture in light propagation and clouds.

The next two weeks I have a lot of other work todo so can't donate too much time to this but I hope I can pick up afterwards.

Title: Re: A new Tornado approach
Post by: cyphyr on April 10, 2012, 02:32:17 PM
Cool nice image :)
I've not seen that lightening model in a while :) Surprises me that its come up so fat. I remember it being thinner. What AA are you using? Also I wouldn't try to use the model to light the scene with its luminosity. Use standard lights for that. One at the cloud base and one near the ground where the lightening strikes.  Leave the model with a luminosity of 1, just so it self illuminates. Of course you'll get a much better bolt from a image but in that case why not just comp it in afterwards?
Cheers
Richard
Title: Re: A new Tornado approach
Post by: mesocyclone on April 10, 2012, 02:43:27 PM
Wow, thats a nice set of renders!!

The tornado with inner vortex is amazingly done.
I suggest you to google for "wall cloud" and "mesocyclone" (my nickname  ::)) to give you some idea on the "feeder cloud".
Title: Re: A new Tornado approach
Post by: nbk2 on April 10, 2012, 07:17:14 PM
Quote from: cyphyr on April 10, 2012, 02:32:17 PM
Cool nice image :)
I've not seen that lightening model in a while :) Surprises me that its come up so fat. I remember it being thinner. What AA are you using? Also I wouldn't try to use the model to light the scene with its luminosity. Use standard lights for that. One at the cloud base and one near the ground where the lightening strikes.  Leave the model with a luminosity of 1, just so it self illuminates. Of course you'll get a much better bolt from a image but in that case why not just comp it in afterwards?
Cheers
Richard

This image was AA 5.


If you download your models they come with a very high luminosity. So I assumed thats how it was done. But the brightness values seem to jump wildly around. I will try what you said  should I reattempt this again.
Title: Re: A new Tornado approach
Post by: nbk2 on April 10, 2012, 07:24:15 PM
Quote from: mesocyclone on April 10, 2012, 02:43:27 PM
Wow, thats a nice set of renders!!

The tornado with inner vortex is amazingly done.
I suggest you to google for "wall cloud" and "mesocyclone" (my nickname  ::)) to give you some idea on the "feeder cloud".

There are so many forms around, just googling that doesn't really help. Which one do you mean? You are right I didn't use yet enough time on it :) Too obsessed with funnel.

But one of the very definite ones would be
http://pattersonpages.wikispaces.com/file/view/Supercell_Thunderstorm.jpg/33761313/Supercell_Thunderstorm.jpg

would be cool. :D
Title: Re: A new Tornado approach
Post by: mesocyclone on April 10, 2012, 09:30:46 PM
Quote from: nbk2 on April 10, 2012, 07:24:15 PM
Quote from: mesocyclone on April 10, 2012, 02:43:27 PM
Wow, thats a nice set of renders!!

The tornado with inner vortex is amazingly done.
I suggest you to google for "wall cloud" and "mesocyclone" (my nickname  ::)) to give you some idea on the "feeder cloud".

There are so many forms around, just googling that doesn't really help. Which one do you mean? You are right I didn't use yet enough time on it :) Too obsessed with funnel.

But one of the very definite ones would be
http://pattersonpages.wikispaces.com/file/view/Supercell_Thunderstorm.jpg/33761313/Supercell_Thunderstorm.jpg

would be cool. :D

Yeah, you can search for "supercell" too!  ;D

Well, i'ts just to you to get an general idea of the thing... Not necessarily be exactly like one of them! Storm structure is something so variable...
But they have some common traits.
In fact, it is somewhat common to see a tornado under a round-shaped cloud like that one of the image you've linked...
The mesocyclone is an ascending and rotating column of air, it is the "heart" of the storm... And the area where most tornadoes will form.

The image you've linked is a mesocyclone base...Sometimes it will spawn a rotating lowering attached to it called "wall cloud"

(http://www.extremeinstability.com/stormpics/svr3794.jpg)
(http://i15.photobucket.com/albums/a351/CALEBE_BATTOU/w-416-25.jpg)

Wall clouds are known for being a sign of possibly forthcoming tornado...

So, what I'm trying to say is not that you necessarily have to insert these things in your image... It's just an idea to make your work even more awesome and realistic !  ;)
Some more images:

(http://static.ddmcdn.com/gif/tornado-5.jpg)
(http://skip.cc/chase/100617/100617label02.jpg)
(http://i.ytimg.com/vi/xl6yZoKMVe8/0.jpg)

;D
Title: Re: A new Tornado approach
Post by: mesocyclone on April 10, 2012, 09:33:18 PM
PS.: I know it can be difficult to do something like that in TG2, but I trust the skills of the people who read this forum  ;D
Title: Re: A new Tornado approach
Post by: nbk2 on April 11, 2012, 01:25:22 AM
a starter on that wallcloud business .. some on the web nearly looked like that..  :P


(http://farm2.static.flickr.com/1046/578213749_09521561fd.jpg)
Sadly cancelled

Quote
Not Acceptable

An appropriate representation of the requested resource /forums/index.php could not be found on this server.

Additionally, a 404 Not Found error was encountered while trying to use an ErrorDocument to handle the request.
I am busy the next two weeks..  so will update when I can .. 

well used tinypic as said it's a starter ...  :)

(http://i42.tinypic.com/x5duvm.jpg)
Title: Re: A new Tornado approach
Post by: TheBadger on April 11, 2012, 02:26:44 AM
Quote from: cyphyr on April 10, 2012, 02:32:17 PM
Use standard lights for that. One at the cloud base and one near the ground where the lightening strikes.  Leave the model with a luminosity of 1, just so it self illuminates. Of course you'll get a much better bolt from a image but in that case why not just comp it in afterwards?
Cheers
Richard

I would like to see examples of both of these suggestions! Have wanted to do something with lightning strikes for a while, but had no idea how to start.

By the way, where is this lightning model coming from? And, how would you use a photo of lightning?
Title: Re: A new Tornado approach
Post by: nbk2 on April 14, 2012, 11:27:39 AM
hmm, dunno why the lower bit turned out so dark it's the same cloud as above.  It seems to be a real shadow. Still I like the progress it has more drama now.  :)
Title: Re: A new Tornado approach
Post by: nbk2 on April 27, 2012, 03:52:34 PM
Well the newest mesocyclone attempts..

I made about 30 inbetween those ..  I think I will take a step back .. I feel like I am aiming for something quite inprecise.
Title: Re: A new Tornado approach
Post by: Oshyan on April 27, 2012, 03:54:01 PM
Woah, you've got some crazy iridescence going on at the edges of the top (sunset) one. How did you do it?

- Oshyan
Title: Re: A new Tornado approach
Post by: nbk2 on April 27, 2012, 04:06:44 PM
Quote from: Oshyan on April 27, 2012, 03:54:01 PM
Woah, you've got some crazy iridescence going on at the edges of the top (sunset) one. How did you do it?

- Oshyan

I hope I saved the right one.

The mesocyclone cloud is a cloud thingy called mesocyclone in the upper left corner.. (I do surprise myself sometimes by naming things logically)


I tried a gathered project but the upload limit prohibits it.
Title: Re: A new Tornado approach
Post by: dandelO on April 27, 2012, 04:08:23 PM
Quote from: Oshyan on April 27, 2012, 03:54:01 PM
Woah, you've got some crazy iridescence going on at the edges of the top (sunset) one. How did you do it?

- Oshyan

Come on, Oshyan! You know how to do that. ;)

NBK, great, now animate that wiggler! :)
Title: Re: A new Tornado approach
Post by: dandelO on April 27, 2012, 04:15:29 PM
Quote from: Oshyan on April 27, 2012, 03:54:01 PM
Woah, you've got some crazy iridescence going on at the edges of the top (sunset) one. How did you do it?

- Oshyan

It appears NBK has another way that I can't work out from looking at the .tgd! I'd just output an extremely slight amount of colour from the cloud layer's fractal/surface shader. But, then again, I am simple. :D

Maybe one of your distribution image maps has a slight hue to it instead of just b/w, NBK?
Title: Re: A new Tornado approach
Post by: Oshyan on April 27, 2012, 04:16:37 PM
I know of a way to do it, yes, but this effect looks subtler and better integrated than I would have expected. So, quite curious. I'll take a look at the TGD. ;D

- Oshyan
Title: Re: A new Tornado approach
Post by: dandelO on April 27, 2012, 04:22:00 PM
I've got some quite promising spectral cloud .tgd approaches in progress, I'm finding quite a difficult task localising it to where it needs to be, though. My best attempt has been to limit the spectral effects by distance from camera but it's clunky and awkward still. I see NBK has it quite limited to specific areas but I assume that's because the effect is only applied on one cloud layer here. Looks cool, whatever the method. :)
Title: Re: A new Tornado approach
Post by: nbk2 on April 27, 2012, 06:09:51 PM
Hmm, I wasn't really aiming for irradiance. But I played A lot around with densities and sharpness and try to set under lightning ;ight propagation at 4 and higher and environment light at 2.

I saw in the file it's close to the picture. I was mostly form chasing so might have not saved at the right point.

When I had a form I liked I tried to give it a better tornado colour. Mostly by playing around with above mentioned settings.

I'll attach the converted bmps, I am quite sure they should be B&W.
Title: Re: A new Tornado approach
Post by: dandelO on April 27, 2012, 06:22:39 PM
Then I can't see where you've got that funky green cloud-lining from at all!
I'd assumed you had some slight colour input to one of your cloud layers but, yes, the maps are all b/w indeed.
Cool construction, anyway, I tried to do a big heavy twister storm like this before but gave up. You're definitely on the right track. :)
Title: Re: A new Tornado approach
Post by: nbk2 on April 27, 2012, 06:41:21 PM
All I have done is basically taken below cloud (a helix) and twisted it again (another helix) made it a bit wider and used the depth modulator to bring more form to it.

I would try with density 333 and sharpness 0.0006 and then light propagation 4 and enviro light at 2.

IMO the irradiance is probably the propagated light which pics up enviro light. I occasionally also did the same with the cloud layer above.

What you see is a very tight helix that is helixed again. It depends how light propagation etc is programmed.

The whole reason is that a cyclone seems to be compacted internally but the edges fizzle out.

Is there a way to introduce univariate random noise? I tried to pick out one axis from a perlin but it put 3 axis through.

Title: Re: A new Tornado approach
Post by: mogn on April 28, 2012, 04:44:14 AM
Put your axis into  x, y or z input of a build vector set the other inputs to non zero constants.
The output of the build vector to the input of the Perlin noise.
Title: Re: A new Tornado approach
Post by: nbk2 on April 28, 2012, 05:05:00 AM
Quote from: mogn on April 28, 2012, 04:44:14 AM
Put your axis into  x, y or z input of a build vector set the other inputs to non zero constants.
The output of the build vector to the input of the Perlin noise.

Cool thanks..  ;D
Title: Re: A new Tornado approach
Post by: Kadri on April 28, 2012, 05:26:31 PM
Quote from: dandelO on April 27, 2012, 06:22:39 PM
Then I can't see where you've got that funky green cloud-lining from at all!
I'd assumed you had some slight colour input to one of your cloud layers but, yes, the maps are all b/w indeed.
...

I don't know the inner working of TG2 but the masks are not pure gray scale images...They are RGB files but i do not know if this is the real reason .
Title: Re: A new Tornado approach
Post by: nbk2 on May 02, 2012, 09:41:18 AM
Light propagation 4 enviro light 2 standard 0.006 density and 5 sharpness

To the left of the tree there is hopefully the searched effect going on.