Page 1 of 3
If Octane is Physically Based ...
Posted: Sat Mar 22, 2014 12:05 pm
by pumeco
.
... then why can't I do Anamorphic bokeh in it?
It's like there's a whole element of features missing from the camera of the program (lens physics/geometry). We have such things as Aperture and ISO, all stuff that relates to the
capture of an image, be we have nothing really to control the
glass we're actualy capturing it through!
Lens physics/geometry is missing, I need Anamorphic bokeh and lots of it

Re: If Octane is Physically Based ...
Posted: Sat Mar 22, 2014 1:31 pm
by kavorka
ATM, we do have bokeh, just can't change the shape of it.
Re: If Octane is Physically Based ...
Posted: Sat Mar 22, 2014 2:15 pm
by pumeco
.
That's what I mean; there's no control over the physical geometry of the virtual glass we're shooting through. For Anamorphic bokeh we need to be able to stretch the lens vertically. The renderer really ought to simulate the effect of different amounts of shutter blades as well. Nine blades, for example, will give more rounded bokeh than five.
I don't know whether they're taking suggestions, but that's more than a suggestion, it's a must if the renderer is ever to be matched to a real-world lens because some lenses, especially in the movie business, are Anamorphic. Not sure what the proper term is for it in a 3D sense, probably 'Lens Modelling' or something, but I really hope they will add it.
Seems really odd it isn't there already!
Re: If Octane is Physically Based ...
Posted: Sat Mar 22, 2014 2:44 pm
by dionysiusmarquis
You are aware that the rendering would look like this:

Re: If Octane is Physically Based ...
Posted: Sat Mar 22, 2014 4:28 pm
by riggles
Part of what you would need in Octane is support for film back size and pixel aspect ratio — which I am surprised they don't have. Also surprised to see a lack of lens blade support, and the way they implemented DoF (as the aperture radius in centimeters, oddly). But, honestly, the need for anamorphic lenses has been reduced to specific artistic effect, as opposed to it's former and primary role of filling up a film negative. Anamorphic bokeh and lens flare is often done in post, and is likely not a high priority for the Octane Team. Certainly not higher than displacement support ... I would hope.
Re: If Octane is Physically Based ...
Posted: Sat Mar 22, 2014 4:55 pm
by ROUBAL
That's what I mean; there's no control over the physical geometry of the virtual glass we're shooting through. For Anamorphic bokeh we need to be able to stretch the lens vertically.
If you are meaning anamorphic flares, they are not directly due to the geometry of the anamorphic lens which is partially hemy cylindrical. The anamorphic flare is due (for what I know) to the horizontal diffraction of light on dust and imperfections of the surface of a non spherical lens (vertical hemi cylinder). The non sphericity leads to directionnal (horizontal) light rays instead of isotropic behaviour.
As a proof : with a real camera the image is deformed by the anamorphic lens at capture (when filming) and projected
without deformation by using a second complementary anamorphic lens. The light passing through the lenses has a normal behaviour. The flare effect is a surface effect and only its image is captured by the lens.
So, a perfect anamorphic lens into a physically based render engine would not give anamorphic flares. This is an imperfection which has to be simulated.
Re: If Octane is Physically Based ...
Posted: Sat Mar 22, 2014 5:22 pm
by dionysiusmarquis
I just tried to simulate a anamorphic lens with an glass object. Its possible to squeeze the image and the bokeh deforming looks right. The problem is that you'll lose the camera focus. I think its not done by changing the lens geometry. A Anamorphic lens/adapter is more an interaction between different lens geometries.
Re: If Octane is Physically Based ...
Posted: Sat Mar 22, 2014 5:36 pm
by ROUBAL
Imho, you can't get it because as I described previously, when you use anamorphic lens (the purpose of anamorphic lenses is to capture images on a narrow film and display on a wide screen) you need one lens at capture and one lens at projection with opposite effect . In case of video images, se second lens is a software simulated lens for display on monitors (stretching the image but not simulating a real lens).
In a render engine, I think that you can't simulate both lenses at the same time.
Re: If Octane is Physically Based ...
Posted: Sat Mar 22, 2014 7:30 pm
by pumeco
.
@dionysiusmarquis
Can you please post a render of that test, I would really, really,
really love to see that!!!
@All
Thanks for the discussion. I see some truths and some myths in what I read. It's true that initially, Anamorphic lenses were used to squeeze the image into the aspect of the film, which then had to be de-squeezed to be played back at the right ratio. Nowadays it's used to pull more detail into the frame and give you a wide angle of view. Most importantly though, people are on to it because of the magical bokeh it gives - there's nothing else like it.
Anamorphic lenses have that magical cinematic look when you shoot through them, and the
reason they look
sooooooooo cinematic is because generally, the only time we ever see Anamorphic bokeh is in the movies. The effect is so in your face that even if you were to attach an anamorphic filter to a DSLR such as the one demoed here, it
totally transforms it from looking like video into something much more impressive, something you associate with a real movie.
Here's a test video of what Anamorphic bokeh looks like (please watch it through if you have time):
http://www.youtube.com/watch?v=ftMmEzbKwRA
Having a real Anamorphic shaped lens isn't the only way to get Anamorphic bokeh. The video above actually uses a trick, it's a filter which screws on to the camera like any other filter would. The only difference is that it has a black oval on it. You shoot through that oval and you get Anamorphic bokeh even without a real Anamorphic lens. Anamorphic lens flare, now that's a different thing (but equally important). Anamorphic flare, as with any flare, is indeed an artefact, but the Anamophic
bokeh is not an imperfection, it's intentional.
The filter above has a horizontal scratch engraved into it to create the anamorphic flare, but the bokeh itself is very real even though the lens isn't anamorphic, so there are two ways to get the bokeh.
One final thing to note about anamorphic lenses is the way they distort the bokeh
depending on how far things are out of focus, it makes things look more stretched the further out of focus they are. The further, the more you see the effect in the bokeh. A movie that springs to mind that had
very noticable anamorphic bokeh, was the first Jaws film.
It's an awesome look, miles away from that boring round bokeh

Re: If Octane is Physically Based ...
Posted: Sat Mar 22, 2014 8:13 pm
by dionysiusmarquis
pumeco wrote:.
@dionysiusmarquis
Can you please post a render of that, I would really, really, really love to see that!!!
As I said I wasn't able to focus anything in the Scene. You will need another lens to get it focusable again like in this pdf on page 6:
http://www.schneideroptics.com/projecti ... istory.pdf
Some real world diameter would help I think. To setup these lenses with trial and error wouldn't be easy.
Here's my very first try. The orientation of the lens wasn't right. But you can see that the circle in the back has this uneven dof (getting stronger sideways). I think that this is what will cause the bokeh you meant. As I've seen your Movie i'm not 100% sure anymore. But If it's the right way, i don't think it'll be a easy setup.