Hello,

The knurling tutorial is online:

http://vimeo.com/31532890

Enjoy.

Very good tutorial OSTexo!

I learned that you could select on the description text for the size of the selected object and change it. Even with numerical components like Pi.
I don't know why I didn't know that, but hey!

Very good and well paced description.

Thanks!

Hello Magic,

Thanks. I actually didn't know I could do that either until I began thinking how users like bemfarmer make these really neat programs that might use those sort of functions that MoI could understand. I was pleasantly surprised that it worked, it will save me some time. It also indicates that these things that take half an hour sometimes could be scripted out and really take a couple of minutes, screw threads and knurling alike.

That could be a nice script:

The ability to cover any single surface with one repeated pattern.
For instance, you could select just the one knurl element, tell the script dialog things like size, count and the configuration, such as staggered or grid and the script would smartly "grow" the repeated pattern on a surface.
Of course, discretely using the Flow command.

This would lead to other fantastic abilities, such as a random Greeble option...

wow.

Hi Mike, could you describe a bit more about what the "smartly grow" part of the script would do?

Often times requests that say stuff like "just do it in a smart way" can basically equate to some kind of artificial intelligence being present in the script.

But unfortunately scripts can't think or use their own judgment, they are only programs that execute a strict sequence of code where what they do has very exactly specified...

Program code is really dumb - it's just a list of instructions to execute and "being smart" is not one of the instructions that is available.

- Michael

Oh my... my intentions of a script like that would be of one made by a third party person. It was just an inspired notion, and not really a wanted request - please continue ahead with the super-blend request first. ;-) lol!
This idea is most likely even too esoteric of a function to warrant including in the Wish List. Especially since it's not hard to accomplish with the existing Flow tool.

By "Smartly", it would simply have the ability to judge placement and spacing, making duplications of the chosen object in a matter that would create the pattern.
As In the case of OSTexo performing the array to duplicate his object.
But in this case, the script would know to place them in a staggered "checker-board" pattern, allowing for proper spacing.

The script could, I imagine perform the Flowing of this array to the target surface. Matching the surfaces point to point as usual.
But one more thing I thought of - If the target surface has trim areas, the final flowed object group would also be trimmed. Cut outwards perpendicular to the normals of course.

Hi Mike, that's sounds like probably a really quite difficult script to create though, so chances of someone else cooking it up are probably pretty slim.

Actually something that needs to analyze shapes might really be beyond what can really be done in a script and instead would need some kind of more full fledged C++ plugin that was able to work at a lower level.


> By "Smartly", it would simply have the ability to
> judge placement and spacing, making duplications of
> the chosen object in a matter that would create the pattern.

Well that "judge" part is again maybe implying some kind of artifical intelligence.

I don't think that there is any simple sequence of steps that will produce equal spacing of any kind of shape even just on a plane, let alone trying to do something that goes across a potentially squeezing or expanding bendy surface.

Stuff that tries to produce results like that often times goes through a kind of seeking or exploratory process in its code, trying to make small adjustments and optimizing things gradually in steps. That's a quite sophisticated level of stuff to try and accomplish within a script.

- Michael

I agree also,
And MoI does not need to get bloated and complicated.

It's just that OSTexo makes it look so darned easy! ;-)

I found myself looking at a pen tonight analyzing the diamond knurling on it.

This is an interesting modelling exercise, but in real world CAD, knurling is just indicated with shading. As you say it is an inexact process, so there is no real point in drawing it. If you want to render knurling, it is best done as a render material, controlling sub-poly displacement.

Hello SteveMacc,

On some level I have to agree with you about that, although there are applications I need this process for that cannot rely on rendering for the knurling result. For example, I rely on MoIs AI export for image insertion for product guides, a process that provides much better and more flexible output options than a poly render and export ever could.

While trying to come up with an effective process it was interesting to read information from machinists blogs about the challenges of getting knurling to behave on different diameter cylinders. While I didn't understand all the details of it since it's typically easier to see it live, the read was great to show how these problems are overcome in a real world sense and aided me in trying to figure out how to do it in MoI.

I also think the process itself is useful because other modelers can pick bits and pieces from the tutorial to incorporate into their process for completely different modeling tasks.

All that said I think I learn the most when doing tutorials, it get's me to think of different ways about the same problem and then whittle down the process to what happens to be the most balanced. In this process Michael Gibsons experience provides invaluable assistance, not only from an application perspective but from a modeling and process efficiency perspective as well, that alone is worth more than the price of MoI.

On the other hand you do bring up an interesting point about the knurling with polys, could you share and screenshots or video about your process?

Very well said, OSTexo.

I agree that a map to drive displacement or bump could be used if the knurling is not of significant portion of your render or further away from the Camera where LOD is not needed. I do think though that the majority of knurling applications lends itself to put it on small items which could drive the camera lens closer to those objects (unless it is an open toolbox render). In those cases to me it would be better to actually make the geometry and then you could get as close as you need with the camera. I'm also biased by not being very great with Photoshop so the creation of a 32-bit greyscale map to drive the knurling displacement correctly would take me almost as long to just make it in MOI.

Hello,

Just a quick doodle, tried using a rounded knurl shape.

Those are the ones I use on the backs of PC's. Pretty nice.

So... why go through the trouble of working at a material offset when MoI handles it so well. At least for close-ups of sharp detail like that.
Every angle and light-reflective glint properly represented, and detail even down to where the edge is cut off at a 45.

Now yes, a scene where you never get close enough to those surfaces to tell warrants a simple material effect, as well as not needing to replicate that detail if the end-purpose is to mill that part.
But OSTexo, you're doing a good job showing what can be done with MoI's new feature.
The same as with the speaker mesh made by Flow. As for education and fun's sake - If MoI can do it, why not exploit it.

Hello,

For me it's a matter of the purpose of the object. I wouldn't even need that level of detail for a guide, it would be a nightmare to manage that without the benefit of hidden line removal. I do think the little bit of added detail is helpful in cases where parts could be similar.

On a side note I had a meeting with a CM recently that was highly impressed with MoI, in fact the viewer they were recommending was more expensive for just a viewer than MoI. I think it could be a great presentation tool in that sense, and in that case the customer might be spoiled and want to see details earlier in the process before renders and RP.

Hello,

For example, much simpler geometry, more curves substitution. For a guide or simple procedure it is more than enough detail to have the user get the idea.

Looks very nice.. :-)

Peace,
Rich_Art. ;-)

ROFL.....you're all doing a great job of showing me just how badly I'd like a new motherboard, processor and memory ;) Oh and a good workstation class video card.....