Chspace (change space) command and z values

[RyanGC]

Has anyone else noticed that chspace often (not always) shifts objects in the z plane?

 

For example, Object A, in model-space, has a z value of a, and, post chspace modification, Object A, in paper-space, NOW has a z value of b.

 

If this has been noticed, as I'm sure it would have been, does anyone know why this happens?

[irneb]

The problem is that paper space "has no z values". Or at least that's how it's supposed to be. I'd have guessed that ChSpace from MS to PS would set all Z's to 0.0, then a ChSpace from PS to Ms would set all Z's to the current elevation ... but as usual ACad's not always working logically.

[RyanGC]

An object can be placed at any z value in either paper-space or model-space. There are no restrictions in regards to this, and it is entirely up to the user.

 

...However I think that you've missed the issue. What is happening is this. Let's say, for example, I have a single-line text in model-space with a z value of 0 (zero). I then run chspace on that single-line text, changing the object to paper-space, and the modified single line text now has a z value of -10.841 (or whatever random number AutoCAD happened to serve me up that time). Like I already said, this doesn't happen every time, but it certainly does happen.

[irneb]

An object can be placed at any z value in either paper-space or model-space. There are no restrictions in regards to this, and it is entirely up to the user.

 

...However I think that you've missed the issue.

Note the quotation marks around the "has no z values" in my previous post.

Let's say, for example, I have a single-line text in model-space with a z value of 0 (zero). I then run chspace on that single-line text, changing the object to paper-space, and the modified single line text now has a z value of -10.841 (or whatever random number AutoCAD happened to serve me up that time).

Which is why I said ACad's acting illogically. It should either keep the same Z value (if they don't implement 2D paper) or change to the current elevation (if a PS acts as a piece of paper only) ... not some arb randomized surprise when it feels like it! But then maybe some ADesk programmer thought life's a bit boring and introduced this "feature".

[YZ]

The x & y values almost always shift when you use the CHSPACE command. (Unless by a freak coincidence the coordinates align in both paper and viewport.)

 

In the case of the z value, it will stay the same if the paperspace and viewport are at the same scale. (In my case paperspace is equivalent of a 1:1000 model, not sure how other industries set theirs).

 

If the viewport scale is different, then it will always change the z value by the viewport scale. It comes down to the comment in the command line: "Objects were scaled by a factor of 0.8 to maintain visual appearance." That includes block scales, rotation, x, y, & z values etc.

 

Sorry, I tried to reply earlier but the site was down when I clicked "Post".

[irneb]

That's how it "should" work. But every so often you can do a ChSpace from MS to PS, and do it again from PS to MS using the same lines and the same VP (no zoom factor change) and you end up with a strange looking Z value. This doesn't happen all the time, but it sometimes does (even through VP's at 1:1). It's as if there's some bug regarding these values.

 

I use several scales on my drawings, ranging from 1:1, 1:2, 1:5, 1:10, 1:20, 1:50, 1:100 ... 1:5000. It depends on the type of drawing (detail, section, plan, site plan, etc.) and the size of the building / site. But in none of those can I calculate a Z value changing from 0.0 to -10.841 as the OP has indicated in post #3. It might be an issue regarding rounding errors if the VP's scale is very small, e.g. -0.000010841 may show as 0.0 in PS, but when you CS it to MS through a VP of 1:1,000,000 you could end up with a value of -10.841. But that's a highly unlikely scenario. Not to mention the OP's example is the other way round, i.e. 0.0 in MS, but -10.841 in PS. With this possible error it means the VP should be something like 1,000,000:1 - basically modelling microscopic structures.