Biased V Un-biased Renderers.
-
Hi guys,
I confess, I don't when the difference... can someone humor me?
-
One makes a guess based on a sampled area. (Biased)
The other doesn't. (Unbiased)
So, if light is bouncing around a room the biased will render a spot then assume from this calculation the surrounding is somewhat similar and then move to another spot. Thus it's faster.
Whereas, the unbiased is looking in every area and rendering based on the light source/s.
That's my agricultural understanding of it.
-
Thanks Rich, good to see the beards' healthy....
-
If you can't tell the difference, then both have fulfilled their purpose.
-
@ecuadorian said:
If you can't tell the difference, then both have fulfilled their purpose.
hmmm... not sure I agree.
If an unbiased render looks like a Vray render then you over Photoshopped it, if a vray render looks unbiased it must have rendered for days.
-
@utiler said:
Hi guys,
I confess, I don't when the difference... can someone humor me?
From a user's point of view (and grossly generalizing):
- unbiased renderers have relatively few options and are quite easy to use. Downside: they're slow.
- biased renderers have many options and are relatively hard to use. Upside: they're fast.
Both are quite capable of producing goodlooking results.
-
@unknownuser said:
Both are quite capable of producing goodlooking results.
Yup, this is what I meant. If the result looks good, it doesn't matter how it was produced... it doesn't even matter if it's physically accurate at all.
All real photographs you see in glossy magazines are photo-shopped. This means real-life accuracy is not necessarily a good thing... We love to enhance real life, to bring it closer to art. So what really matters is your artistic skills, no matter if you work with biased, unbiased or real life photography. Just find something that works well with you and your workflow. -
@ecuadorian said:
@unknownuser said:
Both are quite capable of producing goodlooking results.
If the result looks good, it doesn't matter how it was produced... it doesn't even matter if it's physically accurate at all.
Agreed! Doesn't matter in the slightest! (In a lot of cases, that is. I can think of a few exceptions.)
-
Some people seem to be "render puritans" who believe all post process is cheating and that a unbiased render shows the truth.
IMHO, 3D is cheating. We only model to a certain degree of detail. We use textures and bumpmaps to give the illusion of infinite detail. We use simplified mathematical algorithms to simulate how light bounces. Yes, even unbiased render engines.
To quote Giannis (writer of Thea renderer: "Even if a technique is theoretically unbiased, this does not mean that it will resolve to 100% correct results. I am going actually to insist here that there is no unbiased technique that will deliver you 100% accuracy in practice." "At the end of the day, a good looking render will be the one that is pleasing to our eyes; it is usually the skills and effort that make that, the tool can be a great engine or even windows paint!"I'd say, cheat in every possible way to get your image as good as you can and ready in time.
A collegue once told me that you can improve a render by 50% just by mixing it with a photo. I think he was right. -
Pixero,
-
@unknownuser said:
@utiler said:
Hi guys,
I confess, I don't when the difference... can someone humor me?
From a user's point of view (and grossly generalizing):
- unbiased renderers have relatively few options and are quite easy to use. Downside: they're slow.
- biased renderers have many options and are relatively hard to use. Upside: they're fast.
Both are quite capable of producing goodlooking results.
Cheers, Stinkie, love the hat!!
And thanks to you other guys for the input....
Advertisement