WASasquatch's Terragen Flames / Campfire

Started by WAS, June 23, 2018, 06:48:21 PM

Previous topic - Next topic

WAS

Here is the Terragen campfire example. Scales are at 50cm (and even slightly under 50cm for placement and masking some boundary flames), which is actually a bit under scale, but for example purposes it works nicely, especially pushing Terragen's Ray Tracer to catch as many voxels (?) as possible.

The scene is built at 0,0,0 for ease of example, though the flames can be translated via the cloud layer position and Fire Position shader. The Flame Density Fractal seed is an easy way to get a new flames. You can also play with the Flame Warper seeds, though I'd be careful, it is easily bugged to create noise if not perfectly smoothed. I look forward to seeing your fires. Especially those with paid Terragen and optimizing it for a "usual" Terragen render rather than trying to obtain better quality above limitations.

Render time on the example scene image is about 1h 20m. You may be able to get away with less internal cloud quality of GI with higher [than 6] AA.

As a note: I do not intend any of my work for commercial use or gain without explicit permission from myself. These are for personal, and educational purposes.

luvsmuzik

Thank you!
This is a great share. Not that it would ever happen for me personally, but commercial gain means "profit/pay for goods or service". not to include an award as a prize for an entry in competition? As if one would include this or any other w/o permission or citation?

BTW, have you applied for the educational version of TG. (A  reliable unit might be first though, if you are still struggling) Life is such fun. :(

WAS

#2
Quote from: luvsmuzik on June 23, 2018, 07:41:46 PM
Thank you!
This is a great share. Not that it would ever happen for me personally, but commercial gain means "profit/pay for goods or service". not to include an award as a prize for an entry in competition? As if one would include this or any other w/o permission or citation?

BTW, have you applied for the educational version of TG. (A  reliable unit might be first though, if you are still struggling) Life is such fun. :(

Thank you, it's much appreciated. And yes, you are correct, an award would not be commercial gain. Or even donations from a website/portfolio where it is featured. The reason I'm adding this on my more "quality" items is because I have lost "credit" for things I have done in the past, including a PHP framework that imo, and it's current use, was revolutionary.

I have not looked into educational version. It was my understanding the educational use licenses were granted to educational establishments themselves, though I may be wrong. Though it's not like I wouldn't love to be able to use all TG has to offer and discover new techniques for everyone. And as a note my current computer, while not what you'd want is probably well within TG minimum spec with 8GB RAM, and soon, a new CPU, though the current does well enough. 1h 20m on this I think would have been 3+ hours on that intel stick. I also don't really use objects, as I don't have money for ones I think actually look good, so don't really suffer to objects. I also allow TG to cache to a USB stick which so far apparently saves from crashes other people would have during RAM peaks. The Oasis peaked at 10gb RAM, and would have most certainly crashed without ReadyBoost.


WAS

Thanks Kadri.

Was doing some testing and it seems you can omit the depth modulator for little softer edges.

Dune

It's a good setup, and gives nice flames. But you really need a huge amount of voxels/quality/AA to get grainless results on such a small scale, and thus is slow (breakpoint for me). Maybe doing a crop and a total without fire combined would be faster.

WAS

#6
Quote from: Dune on June 24, 2018, 12:50:04 AM
It's a good setup, and gives nice flames. But you really need a huge amount of voxels/quality/AA to get grainless results on such a small scale, and thus is slow (breakpoint for me). Maybe doing a crop and a total without fire combined would be faster.

Considering people talking about render times upwards 8+ hours, I honestly don't see how 1h 20m on a minimum spec system is really that long. Most my usual renders themselves take 8+ hours, so that added hour for a small flame area doesn't seem like much of a impact (especially when you add post work, or making flames elsewhere, and apparently the issues in setting up cards). Also I assume in a regular scene a small campfire won't be a focal point unless a scene like your neanderthals, which is a really cool angle.

And as I said though, my settings are mainly to get by the limitation in AA, meaning with AA alone, considering it's exponential denoising, can probably due the trick with lower settings.

luvsmuzik

#7
Here is a larger render of original 50cm campfire scene. File share by WASasquatch in this topic.
1920x1080,
MPD 0.8 AA 10
Render time 7hr45min. (16GHP, TG Creative)

I love the results and render time, considering, is on par with similar files with similar setups. imho

EDIT No postwork

WAS

#8
Quote from: luvsmuzik on June 24, 2018, 07:50:10 AM
Here is a larger render of original 50cm campfire scene. File share by WASasquatch in this topic.
1920x1080,
MPD 0.8 AA 10
Render time 7hr45min. (16GHP, TG Creative)

I love the results and render time, considering, is on par with similar files with similar setups. imho

EDIT No postwork

And considering this scene specifically MPD probably doesn't need to even be above 0.5 since it's all pretty small stones in the darkness. Flames look nice bigger, thanks for sharing. Am pretty pleased with the colour. Was having issues before with it looking to rich like cartoon flames. Also lots of reflections on surfaces and rocks which could simply be disabled since specular highlights don't seem to be brought out with these low light-sources.

luvsmuzik

I have not messed with lighting yet as I wanted to do it straight first.....but I will probably try something. It is just my nature. ;D

bobbystahr

Thanks for the share, will mess around on the workstation later, post email time.
something borrowed,
something Blue.
Ring out the Old.
Bring in the New
Bobby Stahr, Paracosmologist

WAS

#11
Quote from: luvsmuzik on June 24, 2018, 08:32:57 AM
I have not messed with lighting yet as I wanted to do it straight first.....but I will probably try something. It is just my nature. ;D

Still though thanks for showing me in higher resolution. I added the Terragen Action I shared in discussion to yours. Really love the contrast and bloom on the flames.

Quote from: bobbystahr on June 24, 2018, 09:06:28 AM
Thanks for the share, will mess around on the workstation later, post email time.

Thanks Bobby, can't wait to see.

I wonder if these could work at all with 2D or v2 clouds. May try some things later. If it's possible, in theory, it'd cut render time in half.

Update: for fun I have made a mask to use for the coal and flame locations to make a "fire field" to see how it looks. I am using a separate Depth Modulator to achieve different heights in the flames though I wonder if a surface layer could somehow do this better on the density.

Dune

It works with v2 (and is a bit faster, as far as I tried), but 2D is a nogo. Then you might as well put in a bunch of cards and project the shaders onto them, which might work nicely, btw, and be a bit faster perhaps.

bobbystahr

Quote from: Dune on June 25, 2018, 01:14:01 AM
It works with v2 (and is a bit faster, as far as I tried), but 2D is a nogo. Then you might as well put in a bunch of cards and project the shaders onto them, which might work nicely, btw, and be a bit faster perhaps.

there's a thought!
something borrowed,
something Blue.
Ring out the Old.
Bring in the New
Bobby Stahr, Paracosmologist

WAS

Quote from: Dune on June 25, 2018, 01:14:01 AM
... Then you might as well put in a bunch of cards and project the shaders onto them, which might work nicely, btw, and be a bit faster perhaps.

Not a bad idea really though in general, if it looks nice. I've never played with card pops and 2D projections yet, but I'd imagine it'd be able to "Slice" noise by position of the cards giving unique flames per card.