I am a photographer and I am starting to integrate CG assets into my still photos. To do this, however, requires that my rendered objects/images match the quality of my images. Therefore, I have to have hyperrealistic 3D renders.
Please keep in mind that I am mostly rendering cars, planes, buildings, and other non-living objects so we are not talking humans or organic materials minus the occasional plant life.
I am wondering if blender combined with other software would do the trick or if I should consider moving to Maya or cinema 4D once I have blender skills down.
P.S. I am not sure if this thread belongs here so my apologies if this is misplaced.
I used to use Cinema 4D, and in fact was a long time advocate, having used it since V4 on the Amiga. Although the workflow is different, it’s possible to get equally good results with Blender, and in fact, because Blender is open source with an easily accessible python interface, probably easier to get the results.
On the downside, the UI is not the most intuitive (though once you get used to it you wonder why everything else doesn’t work the same), meaning the learning curve at first is much steeper than most.
My advice, give it a go for a while, watch some tutorials from the likes of Andrew Price, BornCG etc. to get you going.
At the end of the day, if it doesn’t suit you, apart from your time, what has it cost you?
I have actually been learning/using Blender for going on three months now. I have successfully modeled a car, several small objects, and other small things. The modeling is coming along quite nicely. In fact, I find myself following tutorials but branching off once I have the hang of it, which to me is a huge step.
I was just wondering if after learning blender, I should transition to a more industry standard program to produce more realistic results.
Maybe with keyshot or other outside material/rendering programs I can achieve the same results?
Thought on using keyshot or other external renderers?
The quality of a render depends on two things: the skill of the operator, and the capabilities of the render engine. Hyper-realistic 3d renders can be done in Blender and Cycles (the built-in render engine). As a matter of fact, Cycles has been integrated in other applications such as Rhino3d because of its quality output.
Cinema4d’s render engine is aging, and not quite up to the standards of currently available render engines anymore. Someone even ported Cycles to C4d, and many other C4d users rely on external render engines to create photo-realistic output. Not saying it cannot be done with C4d, just that the quality of the output of render engines like Cycles, Vray, Octane, Maxwell, Luxrender, Arnold, Renderman, etc. is arguably better compared.
The 3d app itself is less important in these cases (as long as they can communicate with one of these). This means you would needlessly spend thousands of dollars on C4d or Maya while you, as a photographer, are predominantly interested in achieving hyper-realistic photographic render output. Focus on the render technology, and less on the 3d app environment.
I would suggest to stick with Blender, and test the various render engines. Cycles is more than up to the job. However, Vray is considered ‘top dog’ by many (particularly for architectural rendering), and it is also available for Blender. So you could invest a couple of hundreds of dollars in Vray and use Blender to feed it the scene information it requires (the output from Blender, Maya, Cinema4d, and Max will be identical with identical material, light and render settings).
Having said all this, without good knowledge of how to set up a scene with lights and materials you could have access to the best render technology in the world, and still come up short with lackluster results. Luckily you are a photographer, so you would be familiar with how light behaves, and good light setups - which will speed up your learning.
And a lot can be achieved in post - so good knowledge of how to composite the render into your existing photos is important (Photoshop, etc.). Also, if you are taking photos in which you will be placing CG assets, you should be taking a light probe and generate a HDR that is used in the render engine in order to emulate the exact lighting conditions at the time you took a photo. This will make life much easier during compositing - but will take more time when doing the actual shoot, of course. But the results speak for themselves.
Also take note of “filmic” to achieve much better dynamic range in your renders and a much more convincing “photo realistic” result. And be aware that most 3d applications make similar mistakes with the default render settings.
So, in a nutshell:
You do not need to invest in either Maya or C4d to achieve hyper-realistic rendering. It would actually be a waste of good money.
You should focus on the render engine instead.
create HDR probes during your shoots: this will infinitely simplify the compositing.
be aware of filmic to get great results ready for compositing.
use a good compositor (Photoshop, for example) to combine your render with your image at 16 bit per channel.
I was wondering about creating HDRI’s on my own since I have all of the equipment. It is great to know that I can accomplish this and get more realistic renders using the photos of my backgrounds.
I will sift through these videos and add them to my viewing list.
Honestly, the modeling has come pretty easily but the materials have been giving me trouble. Additionally, I find that I don’s composite in Blender. I just take everything into photoshop and add my own effects that make for a more realistic blend.