Posts made by pdonner

I wonder whether anybody has experience of work along these lines:

We intend to buy around 20 aerial photographs from the 60's to produce a base material for historically oriented rough modeling of buildings in our area. To do so, we should be able to transform the photographs and connect them to form an image of the area where one photo connects well to the other.

Assuming that the photographs were taken perpendicularly (straight downwards from the aeroplane), we would still need to take into account the fact that the outer edges of the photo are further away than the ones in focus. So apparently the outer edges should be stretched in order to connect with the neighboring picture.

We know that the professionals have software to make this kind of a correction, but I guess that some member of this community could have experience of this kind of a problem. Do you know of a method to stretch the material in a way which could be used as a replacement for that expensive orto correction procedure, which isn't within our reach. - Any ideas?

Excellent, Chris: Yes, that was almost useless.

There are other weird things that I've seen happening while using the Freehand tool, but unfortunately I forgot how to reproduce them. Sometimes while drawing and editing, I've got a dotted line at one end of the curve. It's not a guide since it has beginning and an end. Is there anybody who has seen such a thing? I wasn't able to do anything with the line, but maybe somebody has been more lucky.

Installed the BezierSpline plugin and woops... that was quite a meal for one day. The plugin is more or less a program within the program. It was amazing to see how interpreted ruby scripts can produce so smooth GDI behaviour.

To be honest I only managed to take the first two, three lessons on the BS class. Being full of features this wasn't the newbie tool we were looking for, but in due time it will hopefully be a good asset capable of handling many of the complex challenges that we will meet.

Thanks numbthumb for your hint to the interesting Ruby extension for SU. While searching for tutorials on the topic I also came across some material (such as the waterfall example). But I'm maybe a bit too cautious to install such extensions.

What I was looking for, was a tutorial on basic use of the drawing tool... and perhaps about paste or importation of freehand drawings from external vector drawing programs. What are the available options?

The curve menuitem on the Edit menu also raise a number of questions. What are their purpose? Apparently the curve can be exploded and after that the parts can be moved to improve the quality of the uneven line. What else can I do to improve the quality of the 'curve'?

[While on the topic of rubies, I could envisage a ruby which would convert the line segments of the (so to say) curve into arcs. I've been asking myself: Why doesn't the freehand tool create a series of arcs instead of these line segments?]

SketchUp is a wonderful tool for 3D modeling, but sometimes it would be useful to combine those marvelous models with some freehand drawing. Unfortunately very little has been written about the Freehand drawing tool. Are there ways of editing a Freehand shape? The tutorial video goes as far as encouraging us "to use other tools" "for more accurate work". I would like to know more about the Freehand tool and maybe about ways of interacting with other tools to be more accurate? Is there anything like a proper tutorial with useful advice how to benefit from the Freehand tool?

Thanks TIG for nice SU brain gymnastics.

@tig said:

The Geometry could be just a tiny line 0.1mm long and 'hidden' AND on a layer that is 'off' - the text is still visible.

This works, but at that point you must have already done some drawing (which makes the text visible).

@tig said:

Alternatively make the Text into a Group [prevents use of screen-text]

Interesting observation. The grouped text always faces the observer, but it moves according to the observation point. Essentially the grouped text behaves as if it would be attached to some entity... itself.

@tig said:

Or make the Text as 'flat' 3D-Text in a Face-me Component-Instance.

Yes, this one behaves like on-screen text, but it will still move in space while orbiting and panning.

@tig said:

Or add the Text as a 'Watermark' Image in a particular Style

This is an excellent tip for the one who wants to document a model, as it gives much more visual flexibility compared to what the text tool has to offer. The big limitation is that we are still limited to the Position pegholes offered by SU.

@tig said:

It's not a 'feature' it's a known 'bug'!

That's what they often call the bottom line.

This could happen to a newbie trying out the various SU tools or to an instructor, who might want to start his/her work by writing an explanatory note before working on the model. After finalizing the text, it seems that the result of our tedious writing effort disappears.

Maybe the designers of SketchUp don't want us to follow this kind of a procedure; or should this be described as a bug? Fortunately, the text hides itself only temporarily. By moving the cursor on the text, the content again becomes visible... only to disappear again when the cursor moves away. The remedy is near: Start work on the model, and the explanatory note will become permanent.

Can anybody explain the reason for this feature to me?

Hi. I downloaded the trial version of Simlab's plugin and activated it. Successfully exported a model which can be viewed in Acrobat Reader 9. Tried to open the file in the Acrobat 8.1 editor, but it repeatedly crashed. The trial version also seems to limit the number of exports to 15.

1. Are there any solutions to these shortcomings?
2. How does Simlab's software compare to the PDF export function of the SU 8 Pro version?

Hi again. Today when I visited Google 3D Warehouse I noticed a new item called 'Share' which triggered a dropdown menu with the missing Link and Embed items plus related code. For the time being they are available on the English page but our localized version of the site has the old style buttons.

---

The iframe was very easy to embed.

There are a number of iframe parameters which are more or less self explanatory. Then I should find out if the MID-parameter is generated in 3D Warehouse or if it is the one designated in my SU application?

@gaieus said:

Well, it utilizes the GE Plugin (with a little experience in html and javascript you could build your custom, much "nicer" applications this way) ...

I do agree with your prediction that the GE Plugin shouldn't pass away very quickly. We should apparently also look for the GE Plugin documentation to be able to do the javascript tweaking you are referring to.

Thanks anyway for all the details.

Thanks. I got the code snippet. What do you think, will this feature work in the future? I mean: Is this something I should share with the other team members or will the iframe functionality soon disappear altogether?

Hi. I don't see any 'link' or 'include' buttons on the details page of my model. Is this scheme still supposed to work?

@jean lemire said:

However, I made a test and created a group containing a rectangle with a dimensioned side. When I select the group from the outliner, only the dimension shows. The rectangle stays invisible. This is strange and meybe is a bug.

That's the behaviour I tried to describe. We run into the problem all the time, because the group containing a dimension is part of the procedure that we developed for our designs.

Have a look at the thread Scaling of distorted ground floor layout and you'll understand how we arrived at seeing these weird dimensions.

Thanks Jean for a good answer. That's indeed an elegant workaround for the strange dimension behaviour. I do agree that layers are certainly the most straightforward way of controlling visibility.

But the answer avoids the question. Why doesn't the hidden group hide its dimensions in all contexts?

The Dimension tool is very useful as a way of visibly stating critical measurements in our model. However, at times we would like to get rid of that data without deleting the measurements for good.

Let's say that we need a rectangle with a dynamically updated dimension at one of the edges. As the dimension is linked to the rectangle, it could be useful to group the two together. It is a good practice to name groups, among other things as a way of facilitating control of the grouped entities. Let's call our rectangle "DimensionedRectangle". Now, the rectangle and related measurement might turn out to be redundant, so we would hide away the group.

Later on, we make use of the Outliner Window as a handy way of controlling the entities of our model. While working in the Outliner the selection may pass over the greyed DimensionedRectangle, and now - quite unexpectedly - the measurement figure and associated graphics become visible.

It is hard to know if this is a designed feature or a wellknown bug. Did anybody else come across this strange behaviour of dimensions?

Thanks TIG. You produced a very nice enhancement of our scheme by offering the TIG two dimension scaling formula that was really missing from our scheme:

layout_dimension / model_dimension = scaling_factor

To be honest: There are still a couple of fine details that we pay attention to in order to get it all right. We play with the snapping and resolution settings of the Model Info > Units property sheet while setting the guides and moving back to scaling.

And yes, it is indeed a good idea to lock the group after successful scaling.

TIG's tutorial sheet is excellent. I summarized our findings in Finnish on the web site of the group: http://sites.google.com/site/iivantiira3d/rakennuspiirustuksen-mittakaavan-asettaminen.

Thanks TIG for your kind advice.

@tig said:

The tapemeasure tool is a quick way to 'scale' the whole thing in every axis...

I think that the tape measure is even more handy than that. You can actually scale any part of the image to an exact size by typing in the required dimension. The problem is that this doesn't give access to the other dimension.

@tig said:

Use Scale for one axis, and type in the required dimension for that axis [size], with a units suffix - e.g. 20'6" or 6m - it will scale in that axis to that size - no 'tweaking this and that' is necessary... / You can also scale in the other axis by any required dimension too...

Quite correct. The need for tweaking arises from the fact that the other dimension is in the image. This puts us in a situation where we don't actually know the other outer dimension (before it has been scaled).

I'll add an image to make the scenario I described a bit clearer:

But is there a handier way of fixing the other dimension?

Hi everybody. I'm working with a group of fellow newbie SketchUp planners who intend to model the houses of our village. We rely on architecture plans showing the layout of the ground floor as well as images of the facades. The design process starts by importing the scanned plan of the ground floor. The plan is scaled to approximately the real size. Then we place guides along the outer walls and insert dimension shapes to indicate the main measurements. This is good for the design process and we intend to use a 2D rendering of this layout to prove that the design has correct size when the model is submitted for publication in Google Earth. - At this point we group all these entities, to keep them connected while scaling.

It is very easy to scale the image to exactly correct dimension in one direction. Every newbie should know that this is done by using the tape measure tool. The distance between the two outer walls is being measured by toggling the tool with the Ctrl key into the measurement mode (no plus sign). Then the measurement visible in the architecture plan is entered. By pushing Enter and accepting the resize, the image will scale in both directions to accommodate with this value.

This should do the job. We actually do get the exactly correct dimension... in one direction. Unfortunately the architecture plans are never very precise and the scanning process may enter distortions as well. But we would like to have both main dimensions exactly correct (which will also please the GE evaluators).

It is possible to gradually iterate towards the correct measurement in the second direction by using the scale tool, but this is a bit cumbersome and tedious method. - Does anybody have a speedier procedure to arrive at the exactly correct second dimension?

Sorry for dwelling on a topic which doesn't fall under the heading.

Yes, the text versions are really neat and 'readable', but the second one seems to suffer from some CSS problems.

Many thanks Chris for the pointer to your highly useful video tutorials. It is quite obvious that these text oriented videos should be looked upon in HD format - as you indicate. Unfortunately I don't get there. When I select 720p HD and choose the Full Screen button this results in an audio only black screen presentation.

Found a thread on one of your videos with an embedded YouTube player at http://forums.sketchucation.com/viewtopic.php?f=18&t=25100&p=215249&hilit=+writing+sketchup+ruby+#p215249 That worked excellently without any problem.

Checked four of the videos with quite some pain. Assuming that I'm not the only one experiencing this problem, I'm courageous enough to ask if there is anyone out there who could guide me how to play the fine tutorial videos in HD mode?

@tig said:

The batch importer v1.1 here http://forums.sketchucation.com/viewtopic.php?p=229939#p229939
now works with images too...

Thanks TIG. I'm very grateful for being able to read your scripts as I haven't found a proper SU Ruby tutorial either. And I'm new in Ruby scripting as well, but I do have a useful Ruby tutorial at hand. - Nice language.