Cineversity

One more quick thought about the mirroring issue and then I promise I’ll let it go.  The thing that was so confusing to me about that part of the video (and I know it was in the section about first modeling the fork) is that I followed Patrick’s instructions about sizing the reference image to 1000 units.  I also plugged in the same values as his for the amount of rotation and the (x) offset.  Then when he applies the mirror command it just lines up perfectly with his photo of the fork where as mine is way off.  I know the application produces a mirror image but It just seems odd that my copy of Cinema would produce such different results.  It just seems like my mirror command results in a mirrored image larger on the (x) axis than his.

Anyway, I enough said.  Sorry to keep belaboring the point.

Take care,
Paul W

Hey Paul,

Let’s do a quick elaboration on the model you have. As long as the mirror axis is spot on the world axis (it can be on the object axis as well, but for the exploration now, lets go that route), which means that the values of one side of the axis will be negative, the other side has positive values. This is since the 17th century based on René Descartes a standard (Cartesian coordinate system), so I think we agree on that point. (World uses the default axis, Object the axis of teh model, in both cases teh axis origin is set where the “mouse” finds a reference point)

Back to CINEMA 4D, if you have mirrored the half fork around the world axis, the resulting points on the other side must have the same value, except if the were positive, they are now negative, if they had to cross the mirror axis. (Or vice versa) This is not a given in Object mode, if the object is far away from the world axis, hence the exploration in world, to prove the point. If in model mode, you can later move the result to line up with the world axis and get the same precision. Just to clear that “point”.

Go in Point-Mode, and with the fork-spline selected, switch from the Object Manager to the Structure Manager. If you click on any point of the fork, you will see that the values of the points have one highlighted line. Memorize the point number (the first in the line. Now click in the editor view on the point that was mirrored. A new line in the Structure Manger will be highlighted. Compare now the numbers for X,Y and Z. (If there are no lines, inside of the Structure Manager, go to the menu “Mode”, and set the entry of that menu to Points.) You can of course select two symmetrical points at once, if the screen is large enough, and it should be for the fork, the lines will be shown—as in my image. Points following the axis “up and down” will stay negative or positive, so typically one point changes the polarity only—but the values (absolute) will be the same.

CINEMA 4D works in floating point values, with (to my knowledge) double precision, since some years as standard. When I got my first version of C4D in 1996, I could chose which precision to install, as computer time was valuable. If I’m not mislead, this option was only given around that time. Today, we have for Computer Graphics certainly a precision that is not really a problem—at least not visually.

As I mentioned earlier, the image was the problem, as practical photography has not that precision as C4D has. I speak from 35 years photography with a strong emphasis on reproduction photography. But I think no-one would argue that aligning an object with a camera or even putting it on a scanner might have some tolerances in the alignment. One could argue that the photographer could have rotated the image to have a great results. But if details are the key, then every starter in Photoshop will know, that moving pixels by pixel-increments will lower the image quality by at least a quarter. Same is BTW true for lens correction methods, moving a pixel by a none full amount, renders the new value as an average with the neighbor pixels. Simple math, nothing to argue here I hope. Working in Ps since 1993 I think I have seen all tools and cheap tricks, some beginner might have the idea to fix it with sharpening or similar horrible tools (speaking of reproduction photography), I hope we can take the point home, that we don’t do this in reproduction work.

So, even if the image was not perfect as a C4D render, why would use Patrick such imperfection in the first place? We talk about tutorials here. To take a “perfect” image and model from there, would not have caused Patrick to talk about alignment problems. Tutorials are there to get the procedure clear, and while doing so, introduce the tools necessary. A good tutorial shows these kind of problems that one can have. Bad tutorials run just as fast as possible to the result and doesn’t case what kind of problems could have possibly shown up by doing the next similar—but different—project. It shows Patrick’s experience having modeled tons of objects, to point that out. Again, a tutorial done by a beginner would not even have bothered to go that path. Patrick’s work is clear, structured and well thought out. I don’t see a problem with the tutorial in question.

I like to work on road-blocks with people here inside their use of 3D. I can’t see a problem related to the use of C4D or any problem that has to do with 3D. With that, I can’t go further, having been a mentor since many years from time to time, I know my borders and my place.

For the Polygon/N-Gon problem, please use a new thread.

My best wishes.

Sassi

Hi Sassi,

Thank you for the response.  I will let the mirror thing go.  I hate to bug you more but did you see the post before that last one?  Re: n-gons totally different than Patrick’s result.

I truly apologize for posting so many questions about this tutorial series.

Thank you again,
Paul W

You’re welcome, Paul W.

The main idea in this series is the procedural of how things work and the way to manipulate the “objects” in the scene. If your only focus is a precision that you have personally defined and which goes beyond the series here, you might have a complete different target. Patrick’s tutorials can be reproduced, to the point. If there is a problem, we can check it out.

Modeling is like playing Chess, each “new” move results in a different outcome of the result. Call it “Chaos-Theory”. Stay 100% precise to the instruction and the result will be the same. You need to follow to the point if precision is you main idea about. It is a computer, the same application, with the same input will result in the outcome. If the CPU has floating point imprecisions, we might talk about some differences, but so far I haven’t seen a lot of such complains in the past ten years.

If there is something else: We can hunt down the problem. If there is a problem, that disables your progress of the tutorial, please share a scene file and perhaps a screenshot, it has helped to see where the last problem was located. If you think the application has a precision problem, please check this with the MAXON Support, they are there to help and make certain that everything works as expected.

For a new problem, please open a new thread. Please share your results with the N-Gon. What can I say without knowing what you have so far?

Best

Sassi

Hi Sassi,

I didn’t mean anything with the comment about letting go of the mirroring differences.  I just meant that I need to continue with the process of learning Cinema.  Regarding the floating point issue with my computer I hope that isn’t why my n-gon and resulting polygons look so different.  As I mentioned in the earlier posts I’ve tried this tutorial several times trying to replicate Patrick’s results and so far have been unsuccessful.  I will, however, keep trying because as you stated if you either don’t understand why something is working or not working it can come back to haunt you later. 

Again, please excuse my other comment regarding the mirror issue as I more than appreciate all the time you have invested already trying to help me with this.

I have to run off now as my wife and I have an engagement but I will keep you posted on my progress (unless your sick of me that is).
Take care,
Paul

No problems here, Paul. Thanks for the reply. I like to discuss a problem until it is solved (if possible at all) or it might be not needed anymore, e.g., by finding a new workflow.

I’m very interested in solving problems, as I have the philosophy that any problem will come back as long as it is not solved or replaced with a newer problem/solution.

I don’t think that a floating point issue is the answer here. I mentioned it only as it is common knowledge that this leads to different values in render-farms. Just to name it.

What I don’t like, and I hope I make sense here, is to mention something and then offer to “let go of it”. Why do I have that idea about it in that way? Because we talk in a forum here. Anything that is mentioned as a problem and is not handled then, might leave the impression that there is something wrong (in the application, in the tutorial, etc.). This leads for other readers (and learners of C4D) to the impression that there is something known as a problem, but not acknowledge or solved. After a while this might pile up and as a result, a “blurry” impression is created that we have the wrong tool.

If a problem is mentioned I want to solve it or direct it to the people who can (hopefully). In that way I have answered in fora (here and elsewhere) in the past ten years now 25,000 times. I’m not aware that I ever said, no idea-let go of it. This is not my philosophy nor my idea of how to handle things.

Yes, it takes time to get to the point of a problem. But I hope that any single problem solved once, clears the space for many other artists. Example, any question I got so far about C4D, has helped me to check if I know it in the first place, then has made it clear to me how people understand the material to learn C4D, and last but not least to get to the point how “others” can interpreted instruction in a complete different way.

This has lead me to look beyond the problem and try to understand the underlaying structure how people explore 3D. Some techniques that I have developed over the years are certainly a product of these interactions, and not just pure knowledge of C4D. I read often in the manual and check each scene/problem (if possible) with the current version. I don’t know everything, but a lot. What I miss, I explore. I have six degrees and never stopped learning. I know—there is always something new. That is why I ask to understand the target here, and inside of it to mention EVERY problem that occurs. We need to solve it, or it will put a gray tint after a while over everything. Sloppy learning leads to sloppy techniques and skills. Not my taste nor my idea about.

Having said that, If you feel uncomfortable with a problem, please let’s go to the core of it, let’s try to solve it. A good base is only achievable if we start this way. There are no shortcuts, nor any little things that we can ignore. In fact, 90% of all problems that I have received from users in the past ten years are solvable with accurately applying the basics. There is the core of nearly everything in C4D. Its magic is in the combination of basic things. Missing out a little bit, takes this magic more and more out.

As this thread is already long, please open a new one for the Polygon/N-Gon problem, as mentioned please with file. Thank you.

Edit: I have the feeling I went along all points with the Spline and the fork, but whatever I find and demonstrate in a hopefully re-producable way, e.g., checking the numbers in the Structure-Manager I have the impression it is not what solves it for you. /edit.

My best wishes to a nice and long weekend.

Sassi