Stepped conicles - any cure?

[ajreed]

I used a conicle fill in the process of getting some shading on a flag (your fault Gary with your October tut [img]/infopop/emoticons/icon_smile.gif[/img] ) and was reminded of a minor imperfection I have noticed before.

As shown below you can see steps in the shading created as a result of using a conicle fill. The fill centre was some way from the object, but this still seems like a failing in a program that otherwise manages fills and graduations so well in other areas.

Is there a way around this or has anyone a plausible explanation as to why it cannot be better? As I said it really only manifests itself when dealing with the larger radii.

Alan

[ajreed]

I used a conicle fill in the process of getting some shading on a flag (your fault Gary with your October tut [img]/infopop/emoticons/icon_smile.gif[/img] ) and was reminded of a minor imperfection I have noticed before.

As shown below you can see steps in the shading created as a result of using a conicle fill. The fill centre was some way from the object, but this still seems like a failing in a program that otherwise manages fills and graduations so well in other areas.

Is there a way around this or has anyone a plausible explanation as to why it cannot be better? As I said it really only manifests itself when dealing with the larger radii.

Alan

[Egg Bramhill]

Do you notice the steps in XaraX itself or is it just in the exported image? If it's the exported image then you could try a jpg or a png.
Egg

[ajreed]

Its in the original xar file unfortunately, and for your edification (cor thats good at this time of the morning!) here is the xar file.

Alan

[Egg Bramhill]

Yes I see what you mean by the banding. An alternative to Gary's method is to create "intersects" of the basic flag shape and fill with black and apply an elliptical transparency. See attachment.
PS It's 5.30 and time for bed!!!
Egg

[Egg Bramhill]

The orange shapes are intersects of the original flag shape. They are filled black in the above image and an ellipitical transparent fill applied.

[ajreed]

Thanks for that. There are as you explain alternative ways of getting the shading and I don't think I really explored them to any great degree.

Coming back to my original query, why does conicle filling graduate so badly?

The linear fill below is up to the standards we have come to expect from Xara but the conicle fill seems to let the side down a bit.

Any comment from Xara? Or do we just have to try and work around it?

Alan

[BobInce]

hmm. Yes, conical fills always seems to have (by my count) 520 gradations, regardless of how far from the centre you measure.

A more continuous algorithm for conical fill would be much appreciated.

[Ross Macintosh]

Andrew - did you really count or do you have a clever way of determining the number? I imagine one way would be by applying a blend of little lines to a curve (with the curve being a circle and rotation along curve applied to the blend). Adjusting the number of steps in the blend would effectively adjust the number of equally sized subdivisions in the circle - which could be adjusted to determin the number of steps in a conical fill.

Regards, Ross

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

[Christine Farrelly]

Try making the fill 50% transparent, duplicate it and move the centre rotation handle a fraction. You can repeat this again if necessary.

Christine

[This message was edited by Christine Farrelly on October 04, 2001 at 22:31.]