Very cool thread, looking forward to seeing more of your renders!
Silverwings Render Corner
-
Polinalkrimizei
- Posts: 648
- Joined: Sat May 02, 2009 6:59 am
Re: Silverwings Render Corner
Watching this closely, Raphael
Very cool thread, looking forward to seeing more of your renders!
Very cool thread, looking forward to seeing more of your renders!
Re: Silverwings Render Corner
Hi silverwings.
So many thanks for share the scene. Wonderful!
Trying to push the test further... I play with the volume, ISL from scratch. Not homogeneus and not uniform medium.
The picture is taken from v3.8. This means that the material of the mirrors is not the same (I aproximate your definition with an exponent value of a million, that is the maximum value posible). 15 h of rendering for 4000 spp over 24 threads of 2,6 Ghz. Still not perfect; I will let it render other 10 hours. The speed decrease until about a quarter of the original scene.
The development is from the exr output (photoshop and after effects).
I feel that this is the correct place for the picture. Hope you like it.
Tomorrow I will update the result over the same picture.
So many thanks for share the scene. Wonderful!
Trying to push the test further... I play with the volume, ISL from scratch. Not homogeneus and not uniform medium.
The picture is taken from v3.8. This means that the material of the mirrors is not the same (I aproximate your definition with an exponent value of a million, that is the maximum value posible). 15 h of rendering for 4000 spp over 24 threads of 2,6 Ghz. Still not perfect; I will let it render other 10 hours. The speed decrease until about a quarter of the original scene.
The development is from the exr output (photoshop and after effects).
I feel that this is the correct place for the picture. Hope you like it.
Tomorrow I will update the result over the same picture.
- Attachments
-
Last edited by yonosoy on Mon Jun 26, 2017 2:28 pm, edited 3 times in total.
"... nor 0.2 galaxys, nor 0.8 little chikens ..."
Re: Silverwings Render Corner
Very interesting advance yonosoy
Re: Silverwings Render Corner
Wow, this is really getting challenging to render, thanks Silverwing for the gift that keeps on giving Not many engines out there that can do this, if I may say... excellent tweak, yonosoy!
Can we maybe get the tweaked scene file please?
Not many engines out there that can do this, if I may say... excellent tweak, yonosoy!Can we maybe get the tweaked scene file please?
-
Silverwing
- Posts: 71
- Joined: Sun Aug 23, 2015 1:16 am
- Location: Ludwigsburg Germany
- Contact:
Re: Silverwings Render Corner
This is really cool. Thanks for posting it here. I really like the structure of the noise. Keep up the good work!yonosoy wrote:Hi silverwings.
So many thanks for share the scene. Wonderful!
Trying to push the test further... I play with the volume, ISL from scratch. Not homogeneus and not uniform medium.
The picture is taken from v3.8. This means that the material of the mirrors is not the same (I aproximate your definition with an exponent value of a million, that is the maximum value posible). 15 h of rendering for 4000 spp over 24 threads of 2,6 Ghz. Still not perfect; I will let it render other 10 hours. The speed decrease until about a quarter of the original scene.
The development is from the exr output (photoshop and after effects).
I feel that this is the correct place for the picture. Hope you like it.
Tomorrow I will update the result over the same picture.
Cheers,
Raphael
You don´t dream in cryo.
Re: Silverwings Render Corner
So many thanks. 
The final picture runs 26h for 6900 spp.
Changes:
1 Radial/Sharp as options for splat and downsize filters (more consistent results, maybe another user could say us why...).
2 Lens material. This is not terminant, but helps in some circunstances. I connect the SSS with a value of 0 and down the maximum_extinction_coeff to 0. Zero absorption plus zero SSS equals to 0 (this value is supposed to be the summatory of the values for the absorption and the scattering). Subtle faster process and a cleaner result is acquire sometimes in this way (maybe,...).
3 ISL for the volume of fog in absorption and scattering with Henyey-Greenstein phase function.
----------------------
4 Connected in the pigs the physicall aperture difraction, is in time (sadly I do not in the render).
5 Large mutation probability set to 0.65.
6 Increase the SS factor from 2 to 8.
7 Tonemapped EXR true.
The last three steps are for a shorter run ,I hope.
The most difficult thing to calculate in this scene is the reflection of the lighted volume in the mirrors.
The version for V4.0 has no value in this moment. The drasticall change in the fbm section makes the structure of the noise completily down. Please, take a look at this soon! Also slower than in V3.8.
Edit: The last one without filtering the noise (I don´t advice that I loose some caustic details).
Because of this I reached in the pigs the SS to 8. I belive that in V4.0 the scene is completely free of noise in the lenses. Only the mirrors seems to be hard to obtain...
Edit 2: To connect the SSS is a wrong clue.
Definitively this.
1 Radial/Sharp as options for splat and downsize filters.
2 Lens material. Down the maximum_extinction_coeff to 0.
3 ISL for the volume of fog in absorption and scattering with Henyey-Greenstein phase function.
4 Connected the physicall aperture difraction.
The final picture runs 26h for 6900 spp.
Changes:
1 Radial/Sharp as options for splat and downsize filters (more consistent results, maybe another user could say us why...).
2 Lens material. This is not terminant, but helps in some circunstances. I connect the SSS with a value of 0 and down the maximum_extinction_coeff to 0. Zero absorption plus zero SSS equals to 0 (this value is supposed to be the summatory of the values for the absorption and the scattering). Subtle faster process and a cleaner result is acquire sometimes in this way (maybe,...).
3 ISL for the volume of fog in absorption and scattering with Henyey-Greenstein phase function.
----------------------
4 Connected in the pigs the physicall aperture difraction, is in time (sadly I do not in the render).
5 Large mutation probability set to 0.65.
6 Increase the SS factor from 2 to 8.
7 Tonemapped EXR true.
The last three steps are for a shorter run ,I hope.
The most difficult thing to calculate in this scene is the reflection of the lighted volume in the mirrors.
The version for V4.0 has no value in this moment. The drasticall change in the fbm section makes the structure of the noise completily down. Please, take a look at this soon! Also slower than in V3.8.
Edit: The last one without filtering the noise (I don´t advice that I loose some caustic details).
Because of this I reached in the pigs the SS to 8. I belive that in V4.0 the scene is completely free of noise in the lenses. Only the mirrors seems to be hard to obtain...
Edit 2: To connect the SSS is a wrong clue.
Definitively this.
1 Radial/Sharp as options for splat and downsize filters.
2 Lens material. Down the maximum_extinction_coeff to 0.
3 ISL for the volume of fog in absorption and scattering with Henyey-Greenstein phase function.
4 Connected the physicall aperture difraction.
- Attachments
-
- silverwings_Indigo_Lenses_scene_(tweaked_by_yonosoy).pigs
- (174.96 KiB) Downloaded 492 times
-
- not filtered.
-
-
"... nor 0.2 galaxys, nor 0.8 little chikens ..."
Who is online
Users browsing this forum: No registered users and 62 guests