Cineversity

I have discovered over the last year or so that many of the default settings in C4D, while they may look good, end up causing unnecessarily long render times.  Often, slight modifications to these defaults result in huge speedups in both render speed and editor redraw while having little to no effect on the render quality.

Obviously, most folks (above beginner level) know that things like primitive segments can be decreased depending on factors such as size and distance from camera, but there are other settings like the default interpolation for spline intermediate points being set to Angle 5-deg that can seriously bog things down if one has several of them.  Also, things like Ambient Occlusion, which produces a nice effect but is slow to render at default can be sped up with a few simple tweaks.  These are only the ones I know about, I am certain that there are many more.

So, I propose a tutorial, or perhaps a series of tutorials that focus on maximizing the speed of both the editor windows and the render engine simply by deviating from the defaults. 

I know that there are many situations, and there is no such thing as a “one size fits all approach’.  However, I doubt that the default values that exist were just randomly chosen.  They were likely picked to service a particular scenario or group of scenarios.  So, this tutorial could cover different situations (which is what has me thinking it might be best as a series).  There could be one on render optimizations for photorealistic images, another on modeling optimizations for medium to wide shots, or on project optimization for sketch and toon rendering.  These are just a few that pop into my head, I am sure there could literally be dozens.

Such a series would certainly appeal to users at all skill levels (I’ve even seen fairly seasoned users unaware of C4D’s ability to save custom default states, for example), giving us tools and ideas for making stunning works with this amazing program in faster, more efficient ways.

Thank you for your consideration,
-Will

Hey Will,

Certainly a larger theme and not simple to cover it completely.

The problem starts already that some parameters are mostly ignored. Some are critical for the results, as you wrote. Just to give you a number, in the e.g., past five years I have supported way over thousand projects. As usual I get files and have a look. I have some areas where I check as well, and I see in 99% of the cases defaults. Especially in light and scene illumination (one of my favorite parts in movie-making/Cinematography) some parts were established to overcome limitations, but over the years the initial idea is somehow lost and so the default is there and ignored. I think my observation horizon is statistically stable (being a scientist as well, I like to see the world as well in numbers)

However, the button line is, part of that knowledge is in the training already, admittedly not prominent all the time, as it results out of the explanation of the tutorials. I certainly see 3D artist as people with good skills to adapt to tools, which is certainly a characteristic of these field of art.

In a nut shell, the idea to just have a run down with all “tricks” to optimize, can be given in a short time, if done in a responsible way, is not really possible. To share that bag of tricks with no background leaves the recipient with “empty rules”. I leave that for the people who do not care too much about educational value. (I hope that sounds not arrogant) But to do tutorials, I do my research since nearly a decade for 3D, based as well on 19 years working in this field. That is why I see it as a major challenge to do such thing. On the other side it is not really possible to move through (let me guess) 3,000+ pages of manual, combine all the options given in it, and provide for any field in that 3D can be used the right tip. As I normally say in my hands on classes, if you take on 1,000 of the options and functions given C4D, and combine them with just one you are roughly in the millions, take 30 different fields where it can be used and combine 4 tools each time, you get a number of theoretical options of around 3*10pw13. Quite something. There I feel a little bit overwhelmed to get this for anything done.

However, since a half year I explore some new ideas and I certainly will add this requirement to the list. Thanks for sharing your thoughts. I take education seriously and I’m amazed that many people talk me out of fast food only to consume fast “food like information”, because it is free sometimes. (I just read Poynton’s 2nd edition, “Digital Video and HD” to cover that part for an upcoming series.) I like quality.

All the best

Sassi

Thanks for your reply, Dr Sassi,

  My intent in making this suggestion was not to create a fast-food type of experience.  Rather, to start with a few common scenarios, break down where one has found that artists are using default settings that don’t really benefit that scenario, and discuss the hows and why’s of changing some of those settings to produce excellent results in a more efficient manner. 

  A great example of this was our recent interaction on spline intermediate point interpolation, wherein I learned that the default Angle - 5 deg is serious overkill for the majority of situations, and leads not only to slower render times, but also to slower editor redraw response.  I took that information into a grand piano model I made some years ago.  The model is so intricate that I can barely move in the editor window at all, even on my 12-core with 32GB of RAM, and a top shelf GPU.  Simply changing to the ‘Natural’ setting made it possible to easily move around the editor view.  I really had no reason prior to that to not trust that the default wasn’t set up with efficiency in mind.

  And that’s what inspired this idea.  I think that first, there are defaults like that one, which are just not suited for most needs which could be addressed.  But to go even further, one could take specific situations that are common in this type of work, and make a focus not so much on how to build this, and how to animate that, but rather on,  “Ok, so now that you have this, here are some scene optimizations you can perform to improve upon it.”. 

  Some of these things may indeed be covered as a part of other tutorials, but the issue there is that it can be difficult to sit through an hour of instruction, 99% of which you already are more than familiar with, trying to sift through it all in the hopes of finding some little nugget of gold.  This is not to disparage in any way those tutorials, as there are always people who haven’t yet learned ‘X’, but for more advanced users looking to understand ‘Y’ it can be time consuming and difficult to find ‘Y’ embedded at the 45 min mark of ‘X’.

  I realize it is a big task, and I’m certainly not suggesting it’s something that could just be banged out in a few weeks or months.  Rather, I see it as something that starts small, and grows with tips shared from a variety of users and instructors.

  Or, maybe a better approach would be a structure applied to all the tutorials that are henceforth made in which the author has a chapter dedicated specifically to optimizing the particular scene he has created for the purpose it was created for.  In this way, one has context from which to form the instruction, and the end user has a means to focus their search for a technique in a more specific way.  Think of it as an index appended to the end of the lesson, in it’s own self-conatained movie that goes through the deviations from defaults that helped make the scene work.

Anyway, just thinking out loud… I just feel there is a need here for users that no one has yet figured out how to best address, but that I feel certain there is a solution for.

Thanks again,
-Will

Hey Wil,

Yes, no fast food for the brain! There is no real “diet” for wrong data, hehe.

You mentioned a very specific point, where I do not have found an answer to: The more one knows, the less attractive any material becomes that is not specifically tailored to the high skilled one. Well, the way currently is “hands on training” or “1o1”.

As we all learn “our” stuff based on interest and client demand, we create a very individual mix. Having taken way over 80 courses at FXPHD, and many thousand other tutorials. I know that finding a “gem” become less often. To find a “Gold Nugget” finally is great, but the price to pay (time) becomes expensive after a while. I haven’t seen any solution for that so far. A complete structure education (from A-Z) takes even more time and is less compatible with the specific needs. (Note: I do not talk about introduction training here, which is rather simple to design, compared to provide training towards top level skills)

If you find the key to such, I’m open!

Interactivity is certainly the key for this them, and Cineversity Live might be an option for some rough and intense pointers inside of that theme. Which would require a prepared audience, and a more professional use of the chat-options.  You might suggest that theme there. It could be even a series, modeling, light, rendering, texturing, GI etc.

All the best and thanks for your input.

Sassi