After a bit research on photo match I, to my surprise,I found a rather scarce interest in this feature of SketcUp that I use almost daily. It doesn’t seem to have been updated in a long while and, in my opinion, badly needs it. Does anyone know if Trimble plans to add a feature like FormZ does? Basically all it is, is the option to select points in the model and points in the photograph, and then let the software solve the equation for you (three points being the minimum, but more points can be added for increased precision). In my opinion this would be a much hassle free way of matching perspectives. SkectUp one works decently well, but only if you can find two sets of perpendicular lines, which is not always the case.
That video is showing their equivalent to SketchUp’s Project Texture feature. In SketchUp you could use the rotate, scale, and camera field of view features to line up a model with a photo, and then use the project texture feature to put the building textures on to the model. Their way would be useful in saving time lining up the model.
Match Photo is for a completely different need. It’s for when you have a photo but don’t have a model. Once you have set up the various perspective lines you can then model geometry that follows the photo, and it will get the texture automatically.
Thanks for your answer. I read your post twice, but I am not sure I get exactly what you are saying.
As far as I know, there are two ways of combining a 3d model with a photo in SketchUp: one is adding the picture as a watermark and then use the normal tools, as I think you meant, to line it up. The drawback of this would be that it would not be possible to project textures into the model, so for me mainly the issue would be that it would be difficult for my model to project shadows into the surroundings. The other way is match photo, that allows it, but is sometimes difficult to us if you don’t have parallel lines.
What perhaps we agree on is that it would be nice to have FormZ way of aligning things.
I agree that it looks like a nice way to match a photo - What Colin was getting at is that the video assumes you have a model to start with and want to import a new matched photo to line up with your model… SU allows you to do this, but the documented way of using photo match is to get the perspective before you start drawing and use the image as a guide for drawing.
I think perhaps a bounding box on the existing model (or parts of it) where you could move each corner (like you can when editing a 2d texture) might offer a similar amount of precision. Especially if you could have a “parallex” slider to add some fish-eye distortion like you get on real photos.
Now I understand what Colin was saying. Most of the time, what I do is to insert what I have previously drawn into a context, as you can see from my portfolio:
…which is very useful to show my clients how their structure would look once finished. However my problem is that is all fine if you have two sets of parallel lines that are perpendicular to each other, but if you don’t (the floor doesn’t count as it could be sloping!) you are left trying to eyeball it, and that can be time-consuming.
I appreciate what you and Colin are saying but, at least in my case, I don’t think it makes a big difference whether you have already modelled something or you are starting from scratch, or am I wrong?
