I was wondering if I always need to use the maximum value of the options from the elements of "Add Modifier". For example, do I need to use the view 6 in Subdivision surface? I know it depends of what we are trying to achieve, but we all want the same. Make something look the most believable. So do you always use the maximum value Blender offers?
$\begingroup$
$\endgroup$
4
-
1$\begingroup$ I think you were right in saying that "it depends on what we are tying to achieve", that could be something don't much subsurfed (e.g game assets, low poly style, for rendering performance...). Anyway what you are referring is not the maximum, but just the first of the stop-steps of the slider: you can go beyond that. See for example: How can I set sliders to values outside the limits allowed by their slider? $\endgroup$– CarloSep 26, 2016 at 17:44
-
$\begingroup$ Thanks Carlo. So what is your advice for a realistic cinematic rener? Push all the options to their maximum ? $\endgroup$– GrobbySep 26, 2016 at 18:07
-
$\begingroup$ If render time (and therefore quality) is no object, raise settings like this as long as it makes a noticeable impact. No need to overdo it just to overdo it :-) $\endgroup$– JakeDSep 27, 2016 at 1:00
-
$\begingroup$ More broadly, Pycoder, this is a question from a newbie seeking advice from someone with experience (like you). My question is what is the best approach to model an object for a realistic cinematic render for exemple? Because if I can compare with photoshop that uses layers, here with blender all we do is not reversible. Even for the "Add Modifier", I need to apply it before moving to the next step. So for exemple, if I uses the Subsurf, what is the best if I want a realistic look.? What are your advice based on your experience?. $\endgroup$– GrobbySep 27, 2016 at 6:03
Add a comment
|
2 Answers
$\begingroup$
$\endgroup$
My advice would be the exact opposite:
Use the minimum amount of subdivision that will give you acceptable results.
How much that is depends on the complexity of the shape and the resolution of the images you want to generate.
Needlessly increasing the amount of subdivisions comes at a price: RAM (memory) usage and computer processing. A higher number of vertices will impact your computer's performance. Keep in mind that for every level of subdivision you are increasing the number of vertices exponentially, so it's quite easy to end up with geometry that exceeds your computing capacity.
For more details read Render crash using subsurf modifiers.
3d realism requires taking clever shortcuts and fooling the viewer. Use high level of detail only on those areas that are featured prominently on the final image(s), objects that are going to be far away need less detail than those up close.
$\begingroup$
$\endgroup$
With subsurf, you don't want to apply it before you move on. This is meant to stay throughout the project (and even in final renders) to keep your base mesh simple enough to edit. Here is a great example of using the subsurf modifier all over the place:
https://cloud.blender.org/p/gallery/57e5084f0fcf294119c5055c
You can look at nearly every production-quality file and see just about everything with subsurf modifiers still on a simple mesh. Here is another production file (a benchmark for cosmos laundromat) for you to look at:
https://gooseberry.blender.org/gooseberry-production-benchmark-file/
