Hi MrBraun,

> P.S:
> How i change the UI colors?

You can change the viewport background under Options / Background color.

But there isn't any easy way to change the color for the other UI elements - some of those are drawn using some little bitmap images, so you would have to edit the bitmap to change the colors.

- Michael

Tnx for the answer Michael! ;)

@Michael
Angle to Fill / Step to angle = Item count
So in this case Item count is automatic :)
No need to input it :) So in this case Item count must be disable at the start :)
I am agree I can take my poket calculator :)

Michael,

I like the new Align tool... it's like "Set Points" in Rhino,
only it's easier to use and it works with other objects as well.

Cool, now I don't need "Project to C Plane" anymore... :-)

Nice job!

Jesse

Pilou wrote:
> Angle to Fill / Step to angle = Item count
> So in this case Item count is automatic :)

I see.... But it seems like this creates an inter-dependence between those values that would be difficult to show in the UI.

For example if you entered both an angle to fill and also a step angle which then updated the item count, what would you expect to happen if you then went and edited the item count?

- Michael

Just don't permit to users to input a number in the Count item !
Only Step and Angle to fill :D
I agree it's not academic :)
maybe it's an too progressive idea ! ;)

It will be the same result than your but with an headhache so Forget it !

Pilou wrote:
> It will be the same result than your but with an headhache so Forget it !

So don't throw away your calculator yet!! ;)

- Michael

You're right. it was in R3, but only accessible through the command line. So i didn't know about it until R4. I've already experienced the "shadow" effect when using picture frames, so i do agree with you there. It's not a perfect solution for everything. But still useful for many situations... Also, i didn't realize you included alpha channel support for background images. Very nice touch, Michael :)

jonah

> Also, i didn't realize you included alpha channel support for background
> images. Very nice touch, Michael :)

Thanks! :) I thought it could come in handy for just this type of thing.

I was going to say that the alpha method would probably be a better way to go for people-type PictureFrames in Rhino as well... Except I just tried it and it doesn't seem to work, unless I missed some setting somewhere?

- Michael

Jesse wrote:

> I like the new Align tool... it's like "Set Points" in Rhino,
> only it's easier to use and it works with other objects as well.
>
> Cool, now I don't need "Project to C Plane" anymore... :-)
>
> Nice job!

Thanks Jesse, I'm glad you like it. Yes, it is kind of a combination of Rhino's SetPoint along with a 2D Illustrator type bounding box alignment tool.

It can work for Project to cplane, but you will have to turn on all the points and select them (unless your object is already planar). That will work for now, but I think eventually there should be an actual project command as well. But I haven't quite figured out where to put a project to cplane command, there is a little bit of a conflict with a different type of project (for projecting a curve to a surface).

I've been meaning to post some pictures of what you can do using Align, since a lot of people are not familiar with SetPoint in Rhino. So here we go.

Let's say you have drawn this curve and moved some points around:




Now you want to do a revolve of this curve, but you want to tune up the points so that they are all lined up - you want the end points to be lined up vertically along the axis, and you also want the first two and last two points to be in a horizontal line so that the revolved shape will be smooth instead of come to a sharp point.

So first, select the endpoints, then you can run Transform/Align, and use the origin snap to line them up to the vertical axis (switch from the default "Horizontal" alignment in the upper-right options area to "Vertical center"). That will line up the endpoints to the axis:



Now select the upper 2 points and do Transform/Align, with the Horizontal option, and pick on the end of the curve:



Repeat for the bottom 2 points:



And there you have all those points squared up with the axis and with each other. So you can use this anytime you want points to be lined up in a straight line - sometimes this might be the last 2 points of a curve so that the curve's ending direction is set to be exactly horizontal or vertical, and you can also do this to several points in the middle of a curve to flatten out a region and make sure that there aren't any little wiggles in it.

- Michael

Nope, you didn't miss it. Rhino still doesn't offer direct support for alphas on picture frames or background images. The workaround is pretty simple though. first make a picture frame of that image you need (to get correct proportion). Then make a new surface by snapping to it. Then assign a texture, and also put a second image into the transparency channel. And then set that object to always be displayed in render mode (this must be done individually for every viewport for that object). It's a lot of steps, but the result is perfect. I also just realized that if the environment mapping and the gloss finish are set to zero, then the image will not be affected by shadows & lighting. And the transparency of the object can be adjusted in the properties. WHEW!!!.... That was quite the workout!!! My fingers are ready for a massage and a shower... :)

jonah

> Then make a new surface by snapping to it.

I don't think you need to create a new surface, because a PictureFrame is already just a plane with a texture map set on it - you can go in and edit the texture map settings, etc.. it's just a normal object where the texture map property gets filled in. The command is just kind of an aid for drawing a plane with the aspect ratio of that image.


> Then assign a texture, and also put a second image into the transparency channel.

I tried doing this by just setting the transparency map to the same image. But this didn't seem to work - I guess the transparency map does not read alpha from an image but instead probably is intended to take a grayscale image and use the intensity as the transparency?

So I guess there is another step in here where you have to strip the alpha channel out of an image and put it as its own separate grayscale image and then use that as the transparency map?


> I also just realized that if the environment mapping and the gloss finish are
> set to zero, then the image will not be affected by shadows & lighting.

I couldn't seem to get these to do this - I think the gloss is specular highlights, which are added in addition to regular light shading, so that doesn't turn lighting off. If you put an environment map on an object, that does turn lighting off, but it also turns the texture map off as well. You need to turn lighting off but leave the texture map on.

What would work is an ambient light setting, you would crank the ambient light all the way to white and that would provide full lighting at all spots. But there doesn't appear to be a per-object setting for ambient lighting...


It's a bit easier to just support the alpha! :)

- Michael

>>It's a bit easier to just support the alpha! :)

>>- Michael


No arguments here :)

@Mr braun :)
and do you remember this :D
http://www.3dcommune.com/3d/galview.mv?ZBrush+hnisSElqI+1079923736+&v_index=6&nd=1169250342

Yes Pilou !!! hehehehe

Thosa are the pictures that u have realized for all the user of ZBrush! heheheh ;)

Really cool end Tnx again Pilou! ;)

Image import??!?!

Thank you thank you thank you!!! I just saw this update. BRAVO!!

-Dooki

Hello Michael,

Thanks for this new beta-version. The background images are really very useful. However, I needed a lot of adjustments before I got them right. Attached to this post you find my first setup with the background images. As the setup is difficult I probably use a ridiculous scale setting. I tried to enter the corner points of the images using the input fields, but MoI kept jumping back to the mouse position.

During modelling I encountered some problems. Perhaps you can explain these:

- The fuselage of the aircraft is a "loft" of a number of cross-sections. Some of these cross-sections are pure circles, other cross-sections are modelled using mirrored "splines". As you can notice when you select the fuselage, it seems to consist of four quarters instead of a complete surface or solid. Is it caused by an inaccuracy in my mirrored spline cross-sections? How can I detect such an error?

- I started noticing the problem of the non-solid fuselage as I wanted to add a fillet between the fillet and the stabilizer (the little wing at the rear). As I had problems with these fillets (during which MoI started to lock-up my computer!) I did some experiments. I noticed that fillets only work properly on solids. If I create a wing by extruding a wing profile, without using caps, it seems impossible to create a fillet with e.g. a solid cube. Therefore I recreated the stabilizer as a solid surface. However, as I still was not able to create a fillet, I suspect that the error is located in the fuselage.

What do you advise? Recreate the cross sections? Is it correct that a fillet between two surfaces does not work?

Yours sincerely,

Ray.

About the background images : is it possible to have a mixed files save "3dm + image" for an automatic reloading?

> Attached to this post you find my first setup with the background images.

Hi Ray, we won't be able to see the background images unless you attach those here too.

It may be useful to draw some lines or a rectangle before doing the background image, as a snapping guide.


> As you can notice when you select the fuselage, it seems to consist of four quarters
> instead of a complete surface or solid. Is it caused by an inaccuracy in my mirrored
> spline cross-sections? How can I detect such an error?

This is a side effect of blending some circles along with non-circles.

A NURBS circle is made up of 4 internal arc segments, although since they are fused together with a shared tangent, they will not automatically separate out with the regular "Separate" command.

For a lot of operations, this circle-segmentation does not cause a problem, but one area where it can cause a problem is when you loft between a circle and a regular curve - one part of the process of lofting is combining and blending together the different sections into a single surface and any kind of segmentation can become magnified at this point. What is happening is the segmentation is causing a small crease in the resulting surface, and MoI splits creased surfaces up into separate sheets.

This is something that I should be able to fix inside of loft - I should be able to detect the segmentation and re-approximate that curve with a single smooth non-segmented piece within tolerance which will then behave better.

Until then here is a workaround - turn on points for the circle and select one or more points and drag it. This is one area where MoI does the type of automatic re-approximation. If you drag any point on a circle and then drag it back (not undo it, that will restore the original circle which is not what you want) the "perfect but segmented" circle will be replaced with a fitted one that is one smooth non-segmented piece which will loft better.

So do that little "distrub point on circle" thing on each circle and you should then get a better result.



Filleting does work between two surfaces, but only if they are single individual surfaces, your fuselage in this case is made up of joined sections (due to the crease splitting), so that's why the surface/surface fillet is not happening there, in this case you probably should intersect the objects by trim/join or booleans and then do an edge fillet instead. But the little crease in the fuselage that is running right through that area probably is confusing things as well.

I'll see if I can improve the lofting to automatically handle this type of thing better.

- Michael

> About the background images : is it possible to have a mixed files save
> "3dm + image" for an automatic reloading?

Hi Pilou, do you mean to save the bitmap data directly into the .3dm file?

This would be possible to do, but it will certainly increase file size a bit... A couple of things might be a little complex, like I might imagine that someone who got a file with an embedded bitmap might want to get the bitmap out of it so they could edit it or something...

- Michael