Iso Metric Grid layout -- Possible Problem?

[nwilkinson]

Hey guys,

JohnC is absolutely right: what XaraX does is _NOT_ isometric! It's axonometric, more specifically dimetric.

As to Bryan suggesting that Xara were right to compromise on getting in line with pixels ... RUBBISH!!! For two reasons:

1. Most people when working in isometric simply aren't interested in pixel precision of the finished image, they are after precision ... and who said it was going to be output to screen anyway?

2. Xara is _infamous_ for its lightning-quick, superb antialiasing. If you don't think it's that good, take a look at http://www.isocalc.com/tutorials/antialias-examples.htm -- I gave XaraX first place, and that usurped its little brother CorelXara2. If Xara's so good at sub-pixel presentation (i.e. antialiasing), why should Xara worry about aligning the isometric [ahem] grid to pixels? Beats me.

That's just my two pennies. As John mentioned, IsoCalc's help files include a complete explanation of all of this -- even with the trial version.

Isometric is isometric. There is only one. Don't be sold short.

Nick Wilkinson.

[Peter Clifton]

Hi Nick

How's it hanging, I wondered if you had found this creative powerhouse section of the web. [img]/infopop/emoticons/icon_biggrin.gif[/img]

I'm back doing Isometric illustrations at work at the moment. If I can get permission I will try and forward an example or two.

Peter

[pduggan]

Having previously sworn by Xara in virtually every respect, but never tried its 'isometric' grid, I've been following this thread with interest.

So here's my thought for the day:

Try drawing a simple 'cube' to Xara's 'isometric' grid (attachment 1). Looks like someone's sat on it!

Now try a similar cube based on equilateral triangles (attachment 2). Looks like a cube, even if I cheated by scaling my original vertically by 112% (hope John's angles were accurate!)...

I still love [most of] Xara, but have to say I'm disappointed about this. A straight visual comparison of the two cubes shows how much difference it makes, but surely it would have been no harder to get it right?

(Part of original message removed here because I made a mistake!)

There might be a workround for the angles if you could group your whole drawing and stretch it vertically by the correct amount for printing or exporting, but 112%'s nearly 0.22% too much (assuming both John's angles and my maths!), it's a poor substitute for an accurate grid and probably useless for serious purposes...

If anyone can tell me how to draw the second cube by lining up equilateral triangles accurately with Xara's rectangular grid and/or 'snap to objects', I'd like to know! (Not as a substitute for a proper isometric grid, but because I tried to draw my second cube like this before just stretching the first, and found myself falling back on positioning the shapes by eye.)

Peter</p>

Peat Stack or Pete's Tack?</p>

[This message was edited by Peter Duggan on March 10, 2001 at 06:24 PM.]

[This message was edited by Peter Duggan on March 10, 2001 at 06:38 PM.]

[Soquili]

Hi Everyone,

Gary has a tutorial that shows how to make a guide for drawing cubes. The link to the tutorial doesn't seem to work, but you can download the zipped version of the August 99 tutorial at Gary's Tutorials


Soquili [img]/infopop/emoticons/icon_smile.gif[/img]

[Clements]

Thanks to all who took the time and effort to look at this. The grid can certainly be used as designed ... I JUST WANTED OTHER USERS TO BE AWARE THAT, technically, it may not produce what they might expect.

Yes, there are ways to create true isometrics in Xara (or any other graphic program that can skew and scale) by applying a couple of numbers. They were posted somewhere in John Morris isometric site, but I don't have them with me nor the link to his site.

If someone from Xara overlooks this thread and does not reply in the near future, then I will email a "bug" to them. Don't let that stop you from doing so yourself, however, if you so desire [img]/infopop/emoticons/icon_smile.gif[/img]

Regards, Joh

[pduggan]

Thanks, Soquili...

I'd already tried something almost identical with hexagons after giving up on my triangles, but must just have been careless somewhere along the line!

BTW, the point about my squashed cube is that, despite measuring equal numbers of grid divisions along each axis, it doesn't look like it. Of course the grid can be used as designed, but taking its divisions at face value (surely the whole point of a grid?) will result in the squashed effect.

Peter</p>

Peat Stack or Pete's Tack?</p>

[This message was edited by Peter Duggan on March 11, 2001 at 04:20 AM.]

[Peter Clifton]

Hello People

Since this thread is about my number one passion in illustration.

If you check the CorelXARA forum and look for a thread title called

Help creating a shaded, transparent cube

I posted an answer with a small acrobat pdf file attached, that contains some numbers I worked out a couple of years back.

I originally created it for CorelDRAW users but pages 1 & 2 are relevant to any illustration program.

Peter

[bputnam]

Thanks Peter, I'll take a look...

Like I mentioned earlier, I suppose Xara made this compromise in order that all the "isometric" grid elements, major spacing and subdivision marks, would lie on pixels. I think I'll take back my suggestion that this was the "correct" compromise, and as Nicholas mentioned, the isometric grid could be done correctly if antialising is used.

Anyway, as it is, the height of the triangles (in the X direction) is equal to the side length (in the Y direction). If you play around with pythagorean's theorem, you can see that an image using this grid can be made isometric by contracting its width by a factor of sqrt(.75), or scaling the width by about 86.6%.

I don't have Xara with me at work, but I'll give that a try when I get home.

Bryan