Thu, Nov 28, 10:51 PM CST

Renderosity Forums / Poser - OFFICIAL



Welcome to the Poser - OFFICIAL Forum

Forum Coordinators: RedPhantom

Poser - OFFICIAL F.A.Q (Last Updated: 2024 Nov 28 11:20 am)



Subject: Better eye glints in P6


spedler ( ) posted Thu, 05 May 2005 at 3:01 PM · edited Thu, 28 November 2024 at 10:37 PM

I've been looking at rendering better eye glints without using painted glints onto eye textures. I don't think I have any definite answers, but here's what I have learned so far.

The test bed for all these images was V3 with the DAZ medium-res textures, a single infinite light casting equal light on both eyes, and the Front camera used for renders. In each case the cornea material was altered to make changes. Using the specular channel only, I got these results:

specular_only.jpg

The default settings are useless - the highlight is barely visible and the eyes look dead. It's possible to increase the highlight size, but that doesn't really help and the specular value needs to be higher. By turning down the size, and upping the specular value, you can get small bright highlights, but increasing the size makes the edges softer, which might not be desirable. In the end I settled for the final render as a comparison, keeping the left eye at these settings for comparing with the various specular nodes in the right eye.

The next image shows the effect of the specular, blinn, and glossy nodes plugged into the alt-specular channel of the right cornea.

spec_blinn_glossy.jpg

Note that the specular node with a specular value of 3 and a roughness of 0.1 gives identical results to the specular channel alone with the final settings in the first image, so I didn't look any more at this node. Blinn in the default settings gives no highlight at all, and I had to turn the specular rolloff right up and increase reflectivity, too. The eccentricity needs to be very low to avoid completely blurred highlights, but with a low eccentricity you get a very small highlight. In the end it's a trade off, and the two blinn images show the difference in eccentricity with other settings the same. The glossy node is interesting because of the sharpness attribute but if this is turned up too high, the edge of the highlight is jagged. The default size is OK, but the brightness needs to be turned up with the Ks setting.

Finally, there are the phong and anisotropic nodes:

phong_aniso.jpg

Phong doesn't seem very good here. To see it at all the size and specular value need to be turned right up and then the highlight appears to be too 'high' on the cornea. Anisotropic produces non-regular shaped highlights and if you increase the size the specular value needs to be turned up as well. But it does give nice, hard-edged highlights.

So which gives the best results? It's not clear-cut from these short tests, but phong doesn't look helpful and the specular node gives the same results as the specular channel alone. I prefer the blinn node with a small bright highlight, but glossy could also be useful in many cases.

Sorry about the length of this post - hope it's of interest.

Steve


Rance01 ( ) posted Thu, 05 May 2005 at 4:04 PM

Interest? BOOKMARKED. Good work spedler! I love these comparison threads. Best Wishes, Rance


byAnton ( ) posted Thu, 05 May 2005 at 4:26 PM

Cool comparisons! I like the smaller highlights personally. I've found adding noise into the Anisotropic node can be sweet on the eyewhites. Anton AKmaterialroombookmark

-Anton, creator of Apollo Maximus
"Conviction without truth is denial; Denial in the face of truth is concealment."


Over 100,000 Downloads....


quinlor ( ) posted Thu, 05 May 2005 at 4:44 PM

Very interesting! Myself, I use the anisotropic node with turbulence pluged into the bump channel. But your comparison shows me that I should experiment with blinn for situations with softer lightning.


Dave8 ( ) posted Thu, 05 May 2005 at 5:40 PM

Attached Link: http://www.laroo2.com/tut/ibl/

look here for another way


face_off ( ) posted Thu, 05 May 2005 at 6:59 PM

I personally like anisotopic, with a node generated bump on the EyeWhite and no bump on the Cornea.

Creator of PoserPhysics
Creator of OctaneRender for Poser
Blog
Facebook


DistantWorlds ( ) posted Thu, 05 May 2005 at 8:51 PM · edited Thu, 05 May 2005 at 8:54 PM

Attached Link: Eye Glints in P5

file_233076.jpg

The best I've seen 'til now is this (see attachment) unfortunately I was never able to reproduce this, although there's a Screenshot of the entire Node Setup! :( face_off, could you post your Node Setup and maybe an example render? ;)

Message edited on: 05/05/2005 20:54


face_off ( ) posted Thu, 05 May 2005 at 9:34 PM

file_233078.jpg

DistantWorlds....

The attached render is an example (taken from Costello in my gallery). Unfortantly the node setup is pretty experimental, so I can't really post it at the moment - and it wouldn't make sense if I did. But in summary....Eyeball and cornea handle all spec (turn off spec on iris and pupil). Use Anisotopic for spec at about 0.05 highlight size (smaller if you use bigger bumps). Eyeball has a spots bump - find a good eye reference shot and experiment with the spots node until you get the right bumpiness. The experimental stuff is getting the subsurface scattering into the eyeball, which is working well in this shot. And also getting the nice eyelid shadows which is done using some ugly maths that you don't what to see.
One of the challenges with all this is that specular reflects a point light. But in reality the light reflecting of our eyes comes from larger objects (windows, walls, tv's, big lights etc). So you don't want pin-point light reflections in the eyes - they have to be sharp (since the eye is wet), but also big. That's why I go anisoptopic with a large highlight and add a bump. A better option would be to have lit object in the scene to provide reflections for the eyeball. I've been experimenting with having an ambient image pasted to a skydome - to provide things to reflect in the eyeballs. Haven't got anything to show yet.

Creator of PoserPhysics
Creator of OctaneRender for Poser
Blog
Facebook


Virtua36 ( ) posted Fri, 06 May 2005 at 3:34 AM

.


Spanki ( ) posted Fri, 06 May 2005 at 4:13 AM

.

Cinema4D Plugins (Home of Riptide, Riptide Pro, Undertow, Morph Mill, KyamaSlide and I/Ogre plugins) Poser products Freelance Modelling, Poser Rigging, UV-mapping work for hire.


DistantWorlds ( ) posted Fri, 06 May 2005 at 6:32 AM

Ok, I just reconsidered....the best I've seen 'til now was definitely yours, face_off! ;) (Not that I'd be suprised to get high quality solutions from you! Keep up the great work!) ;)


Casette ( ) posted Fri, 06 May 2005 at 8:19 AM


CASETTE
=======
"Poser isn't a SOFTWARE... it's a RELIGION!"


Dave8 ( ) posted Fri, 06 May 2005 at 9:04 AM

gave the eye thing a try eye16sk.jpg also while your here face of any idea what went wrong with this render using real skin shader and P6


face_off ( ) posted Fri, 06 May 2005 at 10:47 AM

Dave - common problem you have here. Where bump + displacement > certain threashold (depending on amount of camera zoom), displacement artifacts occur - as has happened here. Solution is to tweak the render setting - increase the Min Displacement Bounds (try 0.01 or 0.1). In rare cases Firefly does not respond, in which case turn off displacement mapping, since at this level of zoom the granite displacement does not show anyway. Oh, and I love the shape of her upper lashes - very nice.

Creator of PoserPhysics
Creator of OctaneRender for Poser
Blog
Facebook


Dave8 ( ) posted Fri, 06 May 2005 at 11:50 AM

Attached Link: http://www.renderosity.com/viewed.ez?galleryid=949666&Start=1&Sectionid=1&filter_genre_id=47

ty its working now here is link to my original render just incase ya want a look


unzipped ( ) posted Fri, 06 May 2005 at 1:13 PM

bookmark


spedler ( ) posted Fri, 06 May 2005 at 2:17 PM

face_off, those are excellent glints in post #8 - I'll have to try them this weekend, see if I can get it to work. Many thanks!

Steve


Privacy Notice

This site uses cookies to deliver the best experience. Our own cookies make user accounts and other features possible. Third-party cookies are used to display relevant ads and to analyze how Renderosity is used. By using our site, you acknowledge that you have read and understood our Terms of Service, including our Cookie Policy and our Privacy Policy.