Sketchup is Inacurrate???

jeff hammond · 26 Mar 2012, 17:26

What happen with this ?
A circle is simulated or not ?

I don't really think it matters how accurate sketchup is for that type work (archviz).. as long as the picture looks good or conveys an idea properly, nothing else really matters.

accuracy mostly comes into play when you have to actually build what's being drawn.

dukejazz · 26 Mar 2012, 21:31

Displaying objects:

Accuracy and precision: Accuracy 999,999 to 0.000001 by factor of the units and the zooming changes the slue range of precision.

Sketchup can view the very small object with no limits: ?
Setting length to mm you can measurement to 1 nano-meter 10^-9 or you may change to another units scale.

You can square and resize sketchup objects by the scale tool
I use the meshtab to read measure unit only

Now if I wanted to show smaller units of measurement I would turn off the display unit format and then write-in the new units of measurement

Examples
.3 to 3 Angstrom

1,75 to 15 femto meters

and keep going to dx
Sketchup can view the very large object with no limits too: to infinity: Also

Now the bad new:
Inaccuratcy

Sketchup with more then one objects in a model must follow this rule: Smallest and biggest objects has to be with in a trillion of each other or they won't be format to GUI.
trillion = 1,000,000 x 1,000,000
Plugin and tools will margin the range of precision in their process so wactch out.
Have a nice day

Trogluddite · 27 Mar 2012, 08:48

Surely, the question really should be "Is sketchup accurate enough?"

For example, I create a lot of files for sending to a CNC laser cutting machine. The cutting head movements are driven by stepper motors. Thus, there is a maximum precision that can be reached by the machine (one step of the motor transformed by whatever "transmission" is used.)
So, just like SU, the machine does not draw true circles - there are a fixed number of points on the bed to which the head can be sent, with a touch of intertia smoothing the steps a little on the way.

So, if I want to obtain e.g. a sufficiently smooth curve, I must judge how many edges are needed based on the precision of my machine and behaviour of the material. 1000 edges would be silly - it would generate a massive CNC file when uploaded to the machine for absolutely no improvement in the end product. If I'm making a hole to clear a 3mm bolt, I may use as few as 10 edges - the bolt head and/or washer will easily cover any slight "raggedness" of the circle.

So it all comes down to tolerances - if you saw a plank of wood by hand, there is effectively no difference between 15mm and 15.00034532mm - the "precision" of the numbers is illusory, as you can't possibly cut that accurately.

thomthom · 27 Mar 2012, 08:51

@trogluddite said:

Surely, the question really should be "Is sketchup accurate enough?"

That is still just half the question. Enough for what? For modelling the universe (very large units) or atoms (very small units) then it's not accurate enough. Because it's not designed to handle such extremes.

Gaieus · 27 Mar 2012, 11:15

Don't be so theoretical, TT. "Enough" meaning the normal use of it starting from mechanical design to landscape/urban planning.

I have to say that for what I use SU - and that is somewhere between the scope of individual building architecture and landscape/urban size projects, it is accurate enough.

I noticed that when surveyors survey land and sites, their "accuracy" is somewhere within the 10 centimetre (~ 4") "tolerance". When I measure an existing building, even if it is new and (supposedly) was built as accurately as the bricklayers/builder can go, there's at least about 1-2 cm (~ 1 ") "tolerance", too.

Now why should my models be more accurate? (Of course, I can go for even more accuracy but other than getting things co-planar and arcs meeting in an arch, there's no reason for me).

thomthom · 27 Mar 2012, 11:28

Point being is that in these types of discussions people tend to argue based on their own usage. Usage which often is never mentioned, so people argue on generic terms based on their specific experience. So two participants arguing their case can easily be both simultaneously right and wrong.

pilou · 27 Mar 2012, 12:20

Sorry for the question but as I never use with it
On a plotter (image above) circles from SU are transformed (postscript or other) or it's just like on the sreen so tiny segments ?
From Jeff's answer seems yes (tiny segts)...

andybot · 27 Mar 2012, 13:07

@unknownuser said:

Sorry for the question but as I never use with it
On a plotter (image above) circles from SU are transformed (postscript or other) or it's just like on the sreen so tiny segments ?
From Jeff's answer seems yes (tiny segts)...

It's always segments as far as I know. Those circles in the mind of SU never translate into CAD circles. I find I have to redraw parts of an autocad export from sketchup because those line jagginess can look really awful on an otherwise smooth plot. Also, all the overlapping and partial line segments are just awful to work with. Don't know if Layout does anything different though, I don't imagine so.

Trogluddite · 27 Mar 2012, 16:37

@thomthom said:

Enough for what

He he, yes I was being lazy with my words! "Enough" meaning - enough for the model to meet its ultimate purpose, as decided by the end user (or his/her client/customer).
As you also hinted - the original question is unanswerable because we do not know the purpose of the model - without this information no definitive answer can be given, nor can the OP gauge the usefullness of any advice offered.
The "universe" to "atoms" comparison is interesting - the units within SU are only notional, we are free to interpret the text labels as we wish. I could arbitrarily decide that the letter "m" really means miles or microns (or Mega light years!) if it made my model making more simple - and this would require no adjustment to the internal number representation of the software. In fact, we can see that this is even advisable sometime - use metres as a millimetre substitute to avoid the "small faces" problem.

genma saotome · 27 Mar 2012, 19:48

I don't have a problem with SU circles and arcs being a collection of tangents. I do have a problem with the Followme tool not being able extrude a face that is actually congruent to that collection. That, IMO, is an instance of what I would call inaccurate... as in wrong enough that I must do manual corrections to make it right.

jeff hammond · 27 Mar 2012, 20:21

@genma saotome said:

I don't have a problem with SU circles and arcs being a collection of tangents. I do have a problem with the Followme tool not being able extrude a face that is actually congruent to that collection. That, IMO, is an instance of what I would call inaccurate... as in wrong enough that I must do manual corrections to make it right.

can you attach an example .skp to illustrate your issue?

(and yeah, I agree that there's all sorts of problems with follow me... I'm just curious as to what you're trying to do with it)

Wo3Dan · 27 Mar 2012, 22:00

Sketchup is inacurrate???

Hard to answer. As has been said by others here, it depends on what you model and on what you need as a result.

I read the question as: can you rely on SU if you know what you are doing?

Just saying that SketchUp’s limitation is 0.001 inch (0.0254mm) may be difficult to understand. Can’t it do better than that? Is it realy that inaccurate?
See John Bacus’s answer in this thread: http://groups.google.com/a/googleproductforums.com/forum/#!searchin/sketchup/precision$2Binch$2Bjohn/sketchup/MqyEMnjVXjs/rVPdazWSUX4J

At first (at that time) I didn’t fully understand the meaning of this threshold mentioned by John.
Precision goes way beyond that limit. (Although its setting only serves how to display dimensions, etc. It doesn’t have any effect on the modeling itself.)

Thomthom “proved” to me that SketchUp conciders length differences equal to or less than 0.001” as not existing.

To shed some light on what this threshold realy means when modeling I had to put some simple examples together in a model. I’ll share them (the model) here, together with some of my conclusions.

SU can’t have any two endpoints in the same context with a distance less than (or equal to) 0.001”.
So the shortest edge is 0.001”+…
to model with smaller distances between certain objects you can go way below that 0.001”. But the geometry has to be split over different contexts (groups) to allow endpoints to come as close as 0.000001mm. This can be demonstrated by setting the precision to 0.000001mm.
If somehow SU doesn’t accept the small input that you wish to feed, move away first and then back by [same distance+small] input.

FAIK SU itself uses that high precision (and even higher) for correct intersections etc. This is (to me at least) very good to know. It gives me the feeling that at this point I can trust SU’s inferencing engine and what is under the hood.

Even if you model in mm or round inches or more, the program has to do its job at way higher precision to still give you an accurate model.

A (maybe not the best) example:
Think of a stairway, floor to floor 2.75m, 15 steps. => rise 183.333333mm
You may not want to know the rise in three decimals. Millimeters is enough to check its correctness. But it’s good to know that SU divides correctly. Now you don’t have to care about chain dimensions. You can just measure from ground to any of these steps.

p.s. I haven’t played with angles in a similar way (yet).

1/1000" threshold

genma saotome · 30 Mar 2012, 07:25

@unknownuser said:

@genma saotome said:

I don't have a problem with SU circles and arcs being a collection of tangents. I do have a problem with the Followme tool not being able extrude a face that is actually congruent to that collection. That, IMO, is an instance of what I would call inaccurate... as in wrong enough that I must do manual corrections to make it right.

can you attach an example .skp to illustrate your issue?

(and yeah, I agree that there's all sorts of problems with follow me... I'm just curious as to what you're trying to do with it)

Sure, I've attached an example. My work flow is to import a template of what I want to extrude along an arc, then create a circle of the required radius, use the protractor to measure off a set distance (n the example it is 20d), clean away what's not needed, select the line of the arc and hit Followme.

All you need to do is download the model, highlight the arc and use Followme on the vertical face standing at 0,0,0.

What you will find is none of the four corners of the extruded shapre are where they belong. The two corners closest to the center of the circle are farther apart than they should be and the two farthest form the center are too close together. The example has a set of faces that should be where the corners fall. You'll see they fall short or have pushed thru.

Example.skp

Alan Fraser · 30 Mar 2012, 08:33

Dave, I wouldn't extrude in that manner for complete accuracy. If I'm understanding the problem correctly, this is an example of the limitation of SU when it comes to aligning faces with the beginning and end of an extrude path. The face to be extruded is not fully perpendicular to the beginning of the path...as witnessed by the slight gap between the original template and the extrusion.
In these kind of situations you need something of a lead-in and follow-through to get the shape to extrude properly. I'd be thinking in terms of extruding along a greater section of arc, then trimming either end.

Another solution is to edit either end of the extrude path so that it starts and finishes perpendicular to the existing faces. I have done this in the attached file. I have then welded the entire path back together with Recurve. You'll find that the profile now extrudes perfectly...even to cutting a hole in the diameter wall of the semicircle.

Example-2.skp

genma saotome · 31 Mar 2012, 01:02

@alan fraser said:

The face to be extruded is not fully perpendicular to the beginning of the path...as witnessed by the slight gap between the original template and the extrusion.

I think it is perpendicular Alan. The Y value of every point of the original face is zero... and so is the the centerpoint of the circle and the start of the arc. I'll recap my procedure:

Place the face to be extruded at 0,0,0.
Draw a vertical line up 2-5m from 0,0,0
Draw aline on the dotted red axis of the necessary length -- in this example I used 100m
Draw a circle at the end of the previous line, using a radius of 100m
Change the number of segments of the circle from 24 to 360.
Use the protractor at the center of the circle to map out 20d of arc.
Draw a line from the center of the circle along the dotted line, intersecting the circle.
Delete what is not needed.
Highlight the arc.
Use followme on the original face.

Considering the above, I see a mistake could be made at step 2, failing to draw a line on the axis, and at step 4, failing to center the circle on the endpoint of the line. Either would create an arc that is not perpendicular to the face. Wouldn't the proper quality check be to verify that both steps 2 and 4 have a Y axis value of 0?

I havn't had a chance to review the model you uploaded... I'll check later tonite or first thing tomorrow.

WRT your suggestion to use a lead in and follow thru, that is what I do and is what I had in mind when I wrote needs to be manually corrected. I have created several thousand of these extrusions... each one needing to be manually corrected. Either I'm using the tool improperly or it doesn't do what it is supposed to do.

Wo3Dan · 31 Mar 2012, 13:44

@genma saotome said:

...I think it is perpendicular Alan. ...

Both sides of the path are at an angle of 89.5 degrees (not perpendicular) to the connected faces. So the follow me starts and ends incorrectly.

jgb · 1 Apr 2012, 16:38

Let's keep in mind that it is called SKETCHup Not GoogleCAD.

SU was originally designed for architectural concepts, not as an engineering CAD.
Accuracy and precision in those fields take on totally different definitions.

Yes, I use SU as a Pseudo-CAD because I cannot afford nor justify SOLIDWORKS for what I am doing.
It does the job real fine, but I wished it could do more, especially in the area of solids animation.

I have to keep reminding myself constantly when defining my airplane parts that I an NOT making shop drawings; I DON'T have to concern myself with parts micro-fitting and dimensional precision beyond making it look right.

If it came to really making that airplane, I would be using SOLIDWORKS or more likely; CATIA, using my SU model as a CONCEPTUAL MODEL.

jgb · 1 Apr 2012, 16:47

@genma saotome said:

I do have a problem with the Followme tool not being able extrude a face that is actually congruent to that collection.

In the "DUH" thread this problem with Follow-me came up. Here is my (copied) reply.......

Re: The "Duh!" thread (aka the Doh! thread)

Postby jgb on Sun, 25 Mar 2012 12:03 pm

Another thing with Follow-Me I discovered a while back (and reported) is when you are defining a path that has curves, place the template face somewhere on a straight portion of any line segment, NOT at an endpoint.

If follow-me starts on an endpoint, the face template follows one of the subtended angles and royally screws up at the end, with an end face out of alignment. This is especially a major factor with closed loops.

However, if started on a straight section, it will end perpendicular to the last line of the path, and if the path is a closed loop, it will join properly.

mitcorb · 1 Apr 2012, 17:35

This is a good piece of advice. But would placing the profile on the segment rather than the vertex on a specified radius change the radius of the extruded figure somewhat, and thereby possibly creating a misalignment somewhere else?

jeff hammond · 1 Apr 2012, 17:53

@jgb said:

Let's keep in mind that it is called SKETCHup Not GoogleCAD.

so going off that train of thought, I should just buy AutoCad then all my drawings will be done automatically.