A natural situation

[yonosoy]

.

[yonosoy]

.

[yonosoy]

Thanks to CTZn for his incredible ISL folder (snow material).
Agpapan-apx-400CD developed .
Tip: don't use posOS() over a pre-displaced terrain ! Thanks for the patient...

And this two from exr raw (atmosphere situation is optimal for this). Pink and orange has more than 3200 s/p. River only 450 s/p.
Now 6-7 millions of polygons are good, no artifacts at the moment (twisted code maybe), etc...
Test for fbm3Valued and mie cumulus phase function.

zenith 91.1º

zenith 89º

still not proper scale...

Thanks in advanced.

P.D. a test (something strange exporting pictures sometimes, there are rules too...).

Well. things change from programs to internet but this is closer, to correct more over.)

P.D.2: Pink is mie_cumulus (is slow to converge), river don't know; not the same behaviour in the two computers actually; Linux > wine (a "not emulator" for windows) > indigo, etc = some problems, frequently ...

P.D.3: this one from the previous day...

[yonosoy]

We "announced" this post because there is a new....
Finally, a positive test for PP AD with the presence of water (with cauchy and SSS). Perfect sphere equation instead of polygons and a low turbidity (1.2) with no reactions. Phase function definition is missed when the scene is resaved/relaunch (a local problem (?)). Mie is a little bit noise; beautiful...
Whit this is hope to move forward with ISL shading. Canyons at proper scale at in order...
Near indigo's output (in need to start with tone-mapping/developt/post-production from scratch now !!!).
So many thanks !
(Great work OnoSendai).

[StompinTom]

Beautiful! Will have to have a go at it one of these days

[kklors]

Awesome!

[yonosoy]

So many thanks. Expecting your beautiful hands (in general).
Three still inconclusive ISL tests (all fbm at the moment). This is the way!
Canyons + tectonic +detail. Subtle variations of code... 80x80 km.
Tonemapped EXR (buffff!!!).
Thanks in advance. .

Maybe with help could be a rig!
Please; is possible handle shared parameters in blender?

[OnoSendai]

awesome stuff!

[yonosoy]

Is all your succes, thanks again.
Best try this week (under investigation...); in a basis, CTZn work's. ALbedo is a crazy not working equation seems, and since the parameters are not connected maybe bump works also bad. All the data in the pigs (only to taste how works; perfect in a middle range).20x20 km patch (a polygon). You can try to move the camera or sun positions or turbidities... look at the time!!!! Feel free to change every word, but expect disasters... be patient.
Purely indigo's output.... really forgive all the mess around this little garage.
All bases working .

P.D: this "fallback" is really near to the original code; thanks too to Fakeshamus (normalWS, etc,...).

alive_canyons.pigs

(4 KiB) Downloaded 379 times

A rectification. This is the original, very far for this above (begining to found the scale for patches). Try it in a no seriously big patch... The difference is ....

And the wireframe to test how much excess of polygons are there. For big big scale the code has low definition.

P.D: The wireframe!!

[yonosoy]

.

[fenerolina]

Nice 3 de 9 amb folre i manilles :)
A bit late too but wau! those isl canyons and zenit 91.1 clouds leaved me bouche bée.
congrats.

[yonosoy]

.

[fenerolina]

Pues aqui hay uno que se ha quedado con las ganas!
I want to know more!

[yonosoy]

.

[yonosoy]

And a surprise. Linux from 4 threads of 2100 that runs not over 1800. Windows from 4 threads of 2400. Both native's systems whitout problem...
Another big surprise around the exportation, but we want to give as many info as possible (it means pigs exported from blendigo (linux and windows) and cindigo (windows and linux "emulation" (this is an incomprensible local problem)); views from all opengl and the pictures.
A month over this problem and in an hour (native windows 64), solved from cindigo.
No more complains since both monitors are OK. Tonemapping is clear again.
A milimeter camera is perfectly optic. There are tips for this kind of macros.

Note: Linux could be excesively complex to manage... is not an invitation but the confirmation of an old suspect.

The confirmation from the AMD machine, windows above (in this one linux 3.8 don.t run because a segmentaion fault...), Is the 3.6.26 version but works like 3.6.12 since the same version.

EDIT: A new installation gives correct results (windows is faster). Don´t ask...
A lot of problems...