The “Standard” works as before. Edit: not in all flavors of R16/edit.
Typically if nothing shows up in the rendering if PR is used—this is a question of scale if nothing happens in the Physical-Render: a typical test if it is a scale problem is to set the aperture to 0.01 or even smaller. The Depth Pass from that, while in Physical will not include the DOF blur, as a depth pass is a information or data pass, not an image pass (I edit the message here, as Physical was not your concern). However, you can have the depth-pass in both, but the DOF (Effect) is not available in Physical as it might double the effect.
If you provide a scene file, you can send a copy and delete all objects, but place a cube where you like to have the “focus” and a cube that should be out of focus.
I assume that the Render Settings>DOF is set up. You find it under “Effects”, it should be there if you had it in previous releases.
Let me know if that works for you, otherwise please provide more information. :o)
I am ok with the physical renderer although I don’t think the scale and aperture
settings required to get a decent effect bear much resemblance to the real
world.
My question was about using the old method with the standard renderer where you
set the focus distance then the distance in front and behind that where full
blur is achieved - does this method still work?
Just to fill it in, the DOF of the Physical render was fine tuned based on many lenses and the result of those. Scale matters a lot. Yes, being a lens collector since many decades (Zeiss, Petzval, etc) I know that between practical glass and the digital representations thereof, that there is a difference: No questions. But the measurements are in real world measurements.
The “old” depth-map blur version (vs “Physical”) works here (R16-Studio) and compared to older versions, e.g., R12, I can’t see a difference in the GUI.
The old version has no scale dependencies as it is a pure depth map based blur effect and based on the settings in the camera (distance only). In the Render Settings>Effects>Depth Of Field, you can set even gradients to have multiple focus points, or a very unique depth change of it.
What is not working on your side, what do you miss?
In my version of Cinema 4D - I cannot see render settings- effects - depth of field option.
See attached image.
As far as the phisical renderer goes if I open a new scene and place a standard 100cm cube in it then add a camera - which comes in roughly 500 cm away from the cube. Then make that camera active and go to render settings turn to physical and activate DOF.
Look at the camera settings it comes in as a standard 35mm sensor size with a 36mm lens at f8
If I change this to f1.0 and move the focus distance either a long way behind the cube or just up in front of the camera - I get no blur worthy of mention.
In the real world with either of these settings the cube would be severely blurred.
In order to get any sensible blur with this physical setup I have to increase the sensor size to about 80mm with a corresponding adjustment to the focal length to bring the angle of view back to something similar.
And remember f1.0 is something very few of us ever use in the real world.
What I am trying to achieve is the camera moving towards and past a speedtree tree model. I want the tree to be in focus until the camera gets close and only then for those bits close to the camera to become blurred. I will then output this with alpha channel and composite inside after effects with a previously shot background.
Now I have figured out how to get the effect I want in terms of depth of field using the physical renderer (by manipulating sensor size focal length and aperture) - I am faced with a new problem which is that the DOF function seems to be destroying the bark texture on the tree and making the leaves become almost invisible well before the camera has got close enough to the tree to cause any blur of the tree model itself.
This makes it unuseable.
Any help you could give me on this problem would be greatly appreciated.
To get to the Render Settings Depth Of Field, you have to add this via the Effects button. (I can’t see an image, perhaps it was too large and was rejected? (800w x600h is the max.)
In the way you describe it, 100cm cube/500cm-distance, F/8, full-frame sensor: Your cube should be in the focus. The Hyperfocal length is even at 5.3m (The point after that everything is in focus.
A wide angle lens shows not the effect you are after with that set up, not even if you use a F/1.0 (See both images.)
I talk here about focus, not about sharpness, which would be a longer theme. Most books are “blurry” (pun intended) in that point. I have only seven different 35mm lenses in my collection, but they do not differ in the illustrated focus distance (Again, sharpness is different).
If you can set up an image that shows me a real world blurry ness with that setting, I will be surprised (I photograph professionally since 1978). A F/1 or even smaller as used in Stanley Kubick’s Barry Lyndon (F/0.7) showed even useable results) This was a 50mm lens, which should have a shallower DOF than a 35mm or 36mm. So, this is not doable in the practical world. To increase the sensor size will change that of course.
As F stops are relative and mathematical set ups, some believe even that the ratio is measurable, take an image from a cut lens, the area where the aperture sits is typically surrounded by the smallest elements/groups of glass. It is a relative value. Hence why we use in film/movie T-stops (the real light Transmission, not an simple F value), but I digress.
Hawk BTW produces a complete series of T/1.0 but yes for the S35 format. I certainly agree, an F/1.0 on a large sensor size and 35 or 36 mm is absolutely uncommon.
To set the lens to 80mm might help here, but on the end, you have to match the background plane that you mentioned to composite with it. If you don’t match this exactly, you end up perhaps with a cool effect, but not with a good and seamless result.
Geoff, my suggestion would be here, based on the text above, to animate the focus point and the F-stop so the effect has perhaps even more impact.
I follow here your target, not that I think it might work nicely with the practical footage, but I haven’t seen anything, so I can only suggest.
The difference to the practical world is certainly that we can play with the aperture without loosing or changing the incoming amount of light. Typically with cinema cameras the aperture ring has no “clicks”, so a seamless animation with it has some reality. Some. If I had to do it with my RED Dragon, I would set the camera to HDRX and with 22,5 tops I can certainly play with the aperture and animate the “gray-point” later back on target in NUKE or After Effects. Sounds like the Vertigo effect from Hitchcock, but instead of Aperture and Zoom, it is Aperture and exposure. (Auto Exposure in Film is a big NO. So I do not even suggest that here.
Please have a look at my screen-shot. I miss even the Cylindrical Lens in yours, which is certainly already (!) part of the Prime, according to the Help Content, see image. With that certainly part of everything above it. (Edit, I have to install the Prime version these days and explore it in detail again, it seems not in sync with the Help-Manual. I’m really sorry about the confusion in the material. /edit)
It seems like the installation was compromised, or the hard drive is buggy, guessing here, I know, not the best way nor helpful.
The “disproportionate effect” that you describe, without seeing anything, this is not simple to say. I really don’t like guess work on my part, as it feels always not pro like to me to provide you with that. However, here we go: Samples too low?
I assumed that you had them before, as your question pointed in that direction. From there I concluded—that your update should have it as well. See, why I think less of guess work, and there it hit me. Assumptions are no pro-support. My bad.
It is really difficult to tell from that list if those should be included. So, I checked the Help Content—hence the Cylindrical lens is there in the Prime Part. To my understanding anything Prime is inside of Broadcast. Obviously not helping either. Next step:
To get really clear about it, I installed a Broadcast version and I get the same list as you are. I’m surprised that there is so much available, and that part is not.
=====
Which leaves us to explore your problem with the Physical. Perhaps copy a small part of the bark the light and textures, as well the camera, a simplified scene, I guess these trees are otherwise copyright protected. I will have a look into it. Please mention in which frame the tree is in focus and in which frame(s) the tree is out of focus.
I’m trying to create some realistic narrow depth of field blur on a tree model.
The tree is regular polygon object and the leaves are polygon objects with a texture and transparencey map for cutting out edge of leaves.
When I set up the physical renderer to cause blur on this tree the leaves become almost transparent but the original sharp edges are still visible - whilst the branches are not really very blurred at all.
The same happens to the bark texture on the tree - it disappears well before the tree itself gets very blurred.
You can see the effect on the leaves on the attached image
There are two options, you contact the support or set up a little scene with one leave and the a twig plus trunk with (!) your textures and the camera. Simplified and small, but representing your set up.
P.S: please test the scene file. Then set up the camera as well as the render settings in the way you have it in your scene. Then attach it here in a new post. In this way we have something to discuss here. If you have a different idea, let me know.
BTW, please check if the alpha channel in your image (post #1) if it has in the mask part (dark) really no value at all - just zero! Anything else, as in my example (spots) will more or less fail!
