@pilou said:
Why not increase number of segments of arc circles for have a better precision?
@unknownuser said:
Sketchup gives 2118,27 6442 with circle 1000 segments!
Check my very first post you can make the figure with 75 segs very nicely. I do not subscribe to philosophy more is better. For the OP use one would probably want>=10% etc margin on veneer one buys.
160217.6 ellipse test.skp
@caul said:
@pilou said:
And what is its result in this case ? (Ellipse 400*900)
2118.2799288074484
The result above is for major = 900, minor = 400, if your input refers to the semi axes multiply the result by 2.
This is the correct value irrespective what Op thinks. On-line calculator spits out same number;
In addidition I used two different methods to model ellipse ( 10 x 20 ") to verify the model approach SU uses is ok ( Two circle and parallelogram, plus the one noted above but it is slight mod of parallelogram ) methods;
Attached the result I get for the 400x900 mm ellipse. The actual value is as above and I get 2112.236 => Good enough for me. I get same result if curve is exploded or left as curve. What I did seem to find is welding before scale helps and you have to be very careful to lock axis direction of scale.
bAZ;
Complete integral of second kind required for ellipse;
For drawing Wikipedia best for you. The have an animated approach but call it parallelogram approach, mine almost the same but equal for all practical purposes. Will find link for you https://en.wikipedia.org/wiki/Ellipse
I used online calculator & my length is with .02+% of what they calculate.
I only use 4 points for may drawing then did mirroring to get 1/2. Most of curve was made with Bezier cubic spline Order 3 with seven segments. The scale and my drawing over lay very close.
I think know what problem is but need more test. Think may be adding segments ad infinitinum(sp?)my not be the way one wants to go. As you know SU has two sided faces so deleted duplicate vertices so could see what scale does and raised some questions I want to investigate.
conjuate ellipse.skpBaz: as noted the model was just proof of concept so not worried about segments yet. the 10X5 (semi) " better model reports 24.227n inches for 1/2 , has only 72 edges and looks very nice.
Some way the number I posted above is not correct so forget about running you calculation. BTW that brings up a question. There are several eliptic intergals and the corresponding polynominal curve fits to those. Do you know which type your equation is for.
Attched is pic of ellipse drawn by the means of supplemental chords with no scaling. I included different seg counts just to show what can be done. If there is intersest I'll redo with more care.
Su reports length of 48.364" for size shown.
If you can run the polynomial curve fit equation shown I would appreciate. If this approach is useful for you I'll redo and show how to draw. It is a little more complicated than scaling but not much since you can draw 1/4 and then mirrow.
Oops see one dimension got cropped it is 10"
There is no closed form equation for perimeter of ellipse
it is approximated by p=2PIsqrt[[[ a^2+b^2]/2]^1/2]
The correct value p=4aE where E is the elliptic integral where k=1/asqrt[a^2-b^2]
My limited experience SU primitive geometry contributes as second order impact on processing speed:
Usually its is the graphics processor and computational load of the image textures;
I some times save model and use that, delete all materials and see what the speed a nd program size is;
That at least gives you some idea of the problem facing you. The actions can be done:
You are correct, terrible job on my part trying to explain;
For 32 float, you have 1 sign bit, 8 mantissa bits and 23 significant bits but because way the sign used significant bits is 24. The decimal equivalent digits for 24 bits is 7.2+ digits. Thus you see the basis of the display I was discussing earlier. Using an online converter 0.35 converts to .3499999994039..... and 0.4 to .40000000596046.... and rounding is done for those to read .35...... and .40......... and has nothing to change the precision use internal to the machine. Why the OP machine changes .35 to .40 I can only guess, but if a generic SU issue that would have many more folks flagging the problem so => probably issue with his machine or its operating system. There is zero info provided for it so no further action can be done except by the op.
why ask me when you can go to net do serach on 32 float, look at converter and see what it tells you. The actual precision of 32 bit float is equivalent to about 24 bit when you consider 1 bit sign , 7 bits mantissa. Depending on the mother board you have the actual bus may operate at more than 24 bits and conversion actually happens when the data is stored in memory.
Su 8 runs as 32 bit & does not make any difference what version of windows you use. Think su change to 64 bits at 2014. My 2015 with 64 bits reports as expected. The precision you are changing only effects display precision not any thing to do with internal calculation.
There are on line IEE floating point conversion programs you can run . Float cannot convert all the real numbers exactly so guess is you are seeing a rounding error, that is pain you pay for dynamic range vs accuracy.
BTW the task manager will report what processes are running 32 bits.
@aureus said:
@mac1 said:
Making assumptions is a basic voliation of general approaches to trouble shooting.
Cheap philosophy is of no help here.
It won't make any serious contribution to this thread.
Trouble shooting to help NSA and DOD that was a theme always in our minds and you see that same from folks from the FAA during accident investigations. I do not think they think it is cheap approach and we all have a tendency to fall in that trap that can lead one astray. So it is best to not make statements like made here in.
Making assumptions is a basic voliation of general approaches to trouble shooting.
Will try one more time, take the model you posted above and orbit until things go crazy, select one of those, look at the out liner that will show you which instance it is, select edit component and you should see all the instances of that definition selected in the out liner, select reset skew, that will change all the instances since you are in the edit mode. Now try the orbit again and you should see the problem is gone for that definition. Just to be on the safe since I think I exploded the lower level entity in your nested model so you have one so program deals with just the one item and possible different axis location.
The question is how did the skew happen??
Face Me Trees Behaving Weird Again_mac1A.skp
Very condescending and presumptuous comment especially when you have not tried the solution presented before I posted, it is not just replacement=> read the post! Here is the skp model I made before posting the picture. 
This was done probably in time it takes for you to do 1 tree almost. Suggest you are stove piped. There are two cases where one type tree was located at verties of rec with random rotation and scaling of 1 to 0.5, second case was trees form you model BTW copied along a Bezier curve and then a few selected ones replaced with different type tree. This was done using a plugin called comp spray plugin available at plugin store.
I have tried and done
Face Me Trees Behaving Weird Again_mac1.skpIt appears resetting of scale and skew on the ones with problems make them perform as expected. More test needs to be done , was wrong once already 
BTW it sounds like you are going to trouble not needed. The small circles can be used as proxies, you can then replace them with a target tree ( you select the proxies and then target tree in component browser, then select replace selected. Hope our friends have not changed this been several years ? ) and they will be copied in to the proxies scale and its axis aligned to the proxies.
This file simplifies you op model, I used all same trees, randomly select some trees, open component browser, select a replacement tree, it could be a motorcycle if you want. Click it an select replace selected see what you get.
revelever: FYI, You may not have found this but Shapeway has a free cloud based service wherein they run checks on your model, more extensive then NetFab free, check their web site. BTW they have a limit on model size and number of polys ( 64 MB and 1,000,000).
Here is link http://www.shapeways.com/tutorials/how_to_use_meshlab_and_netfabb. I think most printer have to use slicer to print the item since they lay down a small layer at a time. I was initially concerned about your model and possible issues running into small sections that are not supported.
Derie: FYI The attached pic shows both the orginal on left and your problem model on right. Each is locked in my skp. You posted model was flattened, then used PP to extrude up to height shown in original and dimension shown at same display resolution for each. As can be seen both correlate. This approach basically replicates your model and one would expect both to be same. However, if the contention SU randomly changed model it seems reasonable errors in x,y which is not case. Detail look at steps in model creation can hopefully find where error was introduced.
derie: Think you need to take a look at this step in the model:
<..... attached the original file - with the original axis orientation. Is part of a house that has a specific orientation, based on some topographic measurements and terrain.>
The model is " racked " some way + checking the evelation position shows confirmation of what Dave reported. I even confirmed there is no extraneous geometry causing the " problem". How and why that happened I have no idea but is high probability it is caused by OP. ( aka you). Before you proceed very far with model correcting now is a good idea??
FYI ref my post above: The 64 bits results in a almost 16 decimal digit precision ( 53 log (2) = 15.955 vs 7 one gets with 32 bits.
derei: Nothing in the digital world is precise. SU uses floating point arithmetic and it can not represent all the real numbers. For example, the legacy 32 bit systems 8 bits are allocated for exponent and one for sign and this results in the display precision of 7 digits.( That is 24 bit since the sign bit can used). When you select display precision in SU it has nothing to do with the internal accuracy. Why display was not changed when they went to 64 bits I guess was based on lack of need; The "snap to" is over ridden by input to the measurement box, but using is not helpful IMHO=> deselect and watch for the end point or intersection or other inference engine indications. Allowing a program to make unilateral correction without user control I do not do unless initiated by myself. Beore anyone comments I am aware error detection/correction is used every where 
@kamati said:
Nothing 'exotic'. Standard Windows 7 Home Premium, IE 11.0.9600.17801.
I reset all IE settings to default and opened SU as you recommend.
Same results; white screen when I "select region".
Windows 7 does not show the elevated admin login at the start screen. Enable that and see if it helps.
Better yet tell us what location you are using because in some cases terrain is not populated.