Blending contoured shapes on a curve...

[Egg Bramhill]

>

[Ross Macintosh]

I don't know where I'll ever use these techniques but they sure are fun.

I created another one but posted it in the gallery because it seemed so lonely over there and the thread on CD's is bumming me out.

Regards, Ross

<a href=http://www.designstop.com/>DesignStop.Com</a>

[Egg Bramhill]

It was your intial posting started me along this track. I don't suppose it will help you on your architectural (?) work but it may be good for a graphics designer in logo's etc.
Egg

[mike engles]

Hello Ross

Try as I might I cannot reproduce your blend on a path. I wonder if you could post some more detail as a tutorial.
I do get a blend, but it always has the original object on top, which destroys the effect.

Thanks


Mike Engles

[Ross Macintosh]

Mike - I have to report I don't know how to solve the problem you report. I experience the same limitation. If you look carefully at my first post above you'll see where the blend doesn't come together perfectly.

In the second example I posted with the linked ribbons - I cheated if that's possible in this kind of thing. The parts covered over or trimmed out hid the ends of the blends. In the example posted in the gallery forum, I didn't have the problem. That circular path by coincidence worked out with the number of steps to produce a more perfect blend.

I've now tried hours of experimentation trying to find a successful work-around but it hasn't worked out for me. Egg mentions that he had to "mess about with the overlaps on the ends to get a smooth transition". I tried that unsuccessfully. Perhaps he can explain what he did.

I'm starting to think that Xara's developers might relatively easily be able to include a feature that can change the effective back/front relationship of the end points of lines. They now have given us the ability to "reverse paths" as an option in the shape editor tool. Perhaps next to that option could be a way to alter the back/front relationship of line end points. It seems to me that ability could have value not only when attaching blends to lines, but also when using brushes. It could make both functions more flexible and not clutter up the interface.

Regards, Ross

<a href=http://www.designstop.com/>DesignStop.Com</a>

[Mark (theKeeper)]

Hey cool tip here people!

This works good for making a "one-eyed willy" too! [img]/infopop/emoticons/icon_biggrin.gif[/img]

Please forgive the Jpg artifacts... [img]/infopop/emoticons/icon_rolleyes.gif[/img]

Mark...

[ivan louette]

This one was made with beveled blended circles on a curve. Each circle receives a repeating excentric elliptical filling and circular not repeating transparency. Sky was made of superposition of the same fractal clouds with color modified, a nudge and hue transparency.
Regards
ivan

[mike engles]

Hello Ross

I think that I have replicated your blend.
The secret is the size of the elipse and the angle which gives a blend that is not too overlapped. I used 50 steps.

Mike Engles

[Mickie]

I don't know what it is, but you guys have inspired me to start experimenting more. I've got a long way to go before I get close to what you're doing ... Eric's "paperweight" is wonderful!

I think i can, i think i can,... [img]/infopop/emoticons/icon_rolleyes.gif[/img]

Mickie

[Ross Macintosh]

In an attempt to work around the common problem of the order of the start/stop objects messing up an otherwise seamless effect I noticed the following:

I'll attempt to explain what I did in enough detail that what I did can be reproduced by others if they so desire.

I had the idea that by creating a flipped version of the same blend attached to a closed curve I would be able to create a "patch" from one that I could use to mask over the glitch in the other. It seemed like a good idea. I created one simple linear blend. I then drew an ellipse and rotated it 15 degrees. The ellipse was cloned and then flipped using both the horizontal and vertical axis flips. What that did was result in two ellipses that looked similar except that the single nodes that are displayed when you select an ellipse were in different positions on each ellipse.

I attached copies of the blend to each ellipse. The result looked very promising. They appeared identical except the "glitch" - where the overlap of the blend start/stop objects is jarring - occured in different locations. (Those locations corresponded with the positions of those single nodes that display when an ellipse is selected).

To make a "patch" I could apply over one of the glitches, I had to sacrifice the other blend. I converted one to editable shapes which made it a group. I then put a filled rectangle over the part of that group that corresponded to the location of glitch in the blend I planned on "repairing". Selecting both the group (former blend) and the rectangle I used arrange>combine shapes>intersect shapes menue item to crop the group, effectively creating the patch. Upon pulling this patch over the blend I realized I had a problem. The two did not align perfectly as expected.

On a conceptual level I couldn't understand why the approach didn't work. The problem appeared to relate to the angles of the blend objects and not their spacing. This was troubling because I felt it just didn't make sense.

To try to figure it out I attempted to simplify the blend to its most basic. I drew a short vertical straight line and then duplicated it. I applied a blend from one line to the other. That blend was applied to a circle with just the default 5-steps. Within the blend tools, I selected that the blend should be rotated along the curve. I then zoomed in on the blend's start/stop position and surveyed the situation.

The problem was evident. The image attached below shows the kind of thing I saw when I set the "view quality" slider to its lowest (hairline) setting. The xara blend doesn't put the start/stop objects in the same position. One is rotated relative to the other. This seems weird to me. I hope Xara Ltd. investigates this.

I know to many it must seem trivial to be concerned about such things. I think the stuff people are starting to do here really starts to demonsrate that the blend is a very powerful tool. Fixing this angular problem and providing a means to switch the relative order of a blend's start/stop image would stop constraining the potential of this tool.

Given the similarity, I suspect the brush tool is based on similar programming to the blend. Although I haven't investigated brushes much, I would expect they have similar problems and might also benefit by Xara Ltd. looking into this.

Regards, Ross

<a href=http://www.designstop.com/>DesignStop.Com</a>

[This message was edited by Ross Macintosh on January 11, 2001 at 08:12 PM.]