Planetside Software Forums

General => Image Sharing => Topic started by: GxMew on August 23, 2009, 09:28:01 PM

Title: Images At Quality: .1, .2, .4, .6, .8, 1.0
Post by: GxMew on August 23, 2009, 09:28:01 PM
I did an image test and this is what I got...if you want HQ image go to: http://gxmew.deviantart.com/art/Triangles-Can-Be-Art-134513560
Title: Re: Images At Quality: .1, .2, .4, .6, .8, 1.0
Post by: old_blaggard on August 23, 2009, 09:54:46 PM
Good test. It's interesting how the detail of the clouds seems to top off at about .6, but the lighting is very noticeably improved all the way up to a detail of 1.
Title: Re: Images At Quality: .1, .2, .4, .6, .8, 1.0
Post by: Oshyan on August 24, 2009, 05:12:22 AM
Render times for each would be helpful too, if you have them. :)

- Oshyan
Title: Re: Images At Quality: .1, .2, .4, .6, .8, 1.0
Post by: GxMew on August 24, 2009, 02:39:10 PM
.1 = 28 Seconds
.2 = 57 Seconds
.4 = 2:42 Seconds
.6 = 4:54 Seconds
.8 = 7:37 Seconds
1.0 = 11:12 Seconds
Title: Re: Images At Quality: .1, .2, .4, .6, .8, 1.0
Post by: mt_sabao on August 25, 2009, 05:44:55 AM
(http://www.eso.org/~lcalcada/terragen2/tg2-render.jpg)
Title: Re: Images At Quality: .1, .2, .4, .6, .8, 1.0
Post by: GxMew on August 25, 2009, 03:47:25 PM
Very nice graph but the outlier would be 1 why is that?

(sorry, I edited your post accidentally! I've put it back now! - Matt)
Title: Re: Images At Quality: .1, .2, .4, .6, .8, 1.0
Post by: Matt on August 25, 2009, 10:13:37 PM
I don't think it should be treated as an outlier.
Title: Re: Images At Quality: .1, .2, .4, .6, .8, 1.0
Post by: rcallicotte on August 27, 2009, 10:31:37 AM
Thanks for showing this.  Good illustration.
Title: Re: Images At Quality: .1, .2, .4, .6, .8, 1.0
Post by: PG on August 27, 2009, 11:01:22 AM
There is quite a difference between 0.8 and 1.0 though. The latter gives the sun much more power which affects the rest of the scene, the atmosphere is much lighter and the godrays are more visible than in the former.

Edit: Also, maybe the whole Render detail maximum of 1.0 should be removed from the free version restrictions on the website as anything higher than that seems to make no visible difference to the quality, being that the removal of this was one of the main reasons I paid £200 for it, it doesn't exactly live up to expectations
Title: Re: Images At Quality: .1, .2, .4, .6, .8, 1.0
Post by: FrankB on August 27, 2009, 11:42:29 AM
no PG, that really depends on the scene. This render has a focus on sky, which does not need / benefit from render detail 1, but other scenes with small scale detail and vegetation certainly do.

Title: Re: Images At Quality: .1, .2, .4, .6, .8, 1.0
Post by: PG on August 27, 2009, 11:53:49 AM
So what does it actually effect then? The number of microtriangles used to render the image?
Title: Re: Images At Quality: .1, .2, .4, .6, .8, 1.0
Post by: FrankB on August 27, 2009, 12:35:01 PM
I cannot specify that in technical terms, but my own render results tell me that.
If that test were to be repeated with a small scale displaced terrain, also with shadow areas, you would see what I mean.
Also, remember that for example the GI setting is "relative detail", and it's relative to the master detail setting.
This is also the reason why in this sky render, you see that the shadow side of clouds is brighter at detail 1, compared to detail 0.8.

However, these days I rarely use detail 1, because I can get almost equal, subjective render quality with detail 0.8-0.9
Another reason is that I'm not exactly patient :) so these numbers seem just the right compromise for me personally.

Lastly, when you want decent underwater detail, you even need to increase to beyond detail 1. So you're not going to see many scenes with visibleunderwater terrain from me, that's for sure :)

Cheers,
Frank