Hair Rendering Difference Viewport Picture Viewer

[Jim Scott]

Hi Paul,

I don’t use hair very often, and can’t think of any reason why you are getting the different results. Could you post your project file so that we could try to confirm and troubleshoot the issue?

[Paul Nord]

Thanks for your response. I’ll attach the file below.

12406_hair2.c4d.zip

[Jim Scott]

My initial thought is that the difference in images is due to the high resolution of the rendered image in the Picture Viewer versus the one in the View window which is always rendered to fit that window. Below is an image from the Picture Viewer at 100% from the original camera view compared to one from the View window which has been zoomed in to approximate the same magnification. They hairs in each image look basically identical, meaning that the hairs aren’t really softer in the View window but the image has been softened by squeezing a high res image into a lower resolution window. Hopefully that makes sense. I am probably just demonstrating my own ignorance, but this is the only explanation I can come up with so far.

Picture Viewer at 100%

View zoomed in to approximate above

As I said before, I don’t work with hair much and am still playing around with how to get the soft hair texture you’re looking for. I will report back if I am successful. Hopefully some hair experts out there will have some useful info.

[Steve Bentley]

What about the openGL being on or off? I get very different looks from FXy stuff depending on whether I’m letting the hardware give me my preview or the software.

[Paul Nord]

Thanks for your help! OpenGL differences might be a good hint. I don’t have a lot of experience with openGL but used it for this one because the resolution is so big. I will look into that now.
Do have any other hints on how to get a “softer” look for hair in die HiRes rendering?

[Steve Bentley]

We’ll often run a transparency channel with a fresnel shader in it along the edges of the hair. This way you can have each strand be slightly transparent either at the edges or in the core and it takes the “edge off” and reduces sharpness. You just have to fiddle with the fresnel gradient to get it right, often using the twirl down to get actual positions of the nodes in the gradient. Sometimes the black and white nodes are only 3 or 4 positions apart which is hard to to by eye.
I recommend turning off refraction though to keep render times down or using a refraction index of 1. You can also try this in the alpha channel and reverse the fresnel gradient, but we’ve had better luck in the trani channel.

You could also run a multi pass render so the hair has its own object buffer channel and then post process the hair only with a slight blur, or run a motion blur pass for just the hair so that you can amp up its motion blur and keep the blur of the character “correct”. This way the blur isn’t just an overall fuzzyness, it has some connection to how the hair is moving. This idea can produce (in the extreme) what I can only describe as “hair that Monet would have painted”.