Hello! I am new here, so sorry if this is a dumb question.
I need to be able to visualize a model of a proposed new development in its context of surrounding buildings, the street, etc. Someone suggested cutting around a snapshot of a SU model and pasting it into a photo. Obviously, it would be impossible to ensure proper scale and angle.
The other two options I’ve seen are:
1- Importing the model into Google Earth.
2- Using the PlaceMaker plugin to bring OSM data into Sketchup.
It seems like the success of either of these options would depend on whether someone has already modeled and imported the surrounding buildings - and you’d have to trust that they modeled fairly accurately.
Am I missing anything? Are these the only ways to do this? I feel like this would be a more common application of SU, but am having trouble finding any examples online.
I do this with Match Photo all the time and just posted examples here. This works for still photos, but if you want a fly through or real time visualization in SU while you model, then yes, the buildings need to be modeled. There are various options for that.
Why not take a photo?
To find the same perspective from which the photo was taken, you can use the “Match Photo” feature and create a scene. Using this scene, you can use any renderer you like (or a SketchUp 2d export) and later integrate the render into the photo using a 2d image editor.
Those photos of the duplex are amazing(!), and exactly what I’m looking for. Did you do it the way Aerilius suggested? If you have a minute, do you mind breaking down the general steps? Or pointing me to where a tutorial to do this might exist? Thanks so much!
I don’t have a lot of time at the moment to give a detailed explanation, but yes:
• I had a survey from a licensed surveyor to model the site.
• Match Photo was used to match the photos to the model
• Within SU the result is OK looking, but the model sits on top of the photo
• The final result has to be assembled in Photoshop with:
a. The original photo as a background layer
b. Exported image of the SU model without the photo backround on a middle layer
c. A copy of the photo as a top layer with a lot of masking work to cut out the background and leave only forground stuff.
Thank you both for your help. I’ve been playing around with this for the last week, and I think I’ve got the basics down. However, I have one question - and maybe it has more to do with Photoshop than Sketchup?
Once I import the 2D graphic of my model (that was properly scaled using Photo Match) into Photoshop…it is not properly scaled to the same photo the way it was in Sketchup. So I resize it, just like I see this guy doing in this 10-year-old video from about 6:00-7:30ish here (BTW this is really the only video tutorial for this that I’ve been able to find online). But once I re-scale the 2D graphic of my model in Photoshop, haven’t I lost all guarantee that it is properly scaled in relation to its context?
The original photo was shot with a Nikon D70 (that was a while ago!) which has a native resolution of 3008 pixels by 2000 pixels high. Your SketchUp window matters for the output.
I size my SU window so that match photo fills the height of the window and there’s at least a little extra space on the sides. (You’ll never get a perfect match for both height and width adjusting the window size)
File > Export > 2D Graphic, and I use TIFF. I enter the vertical resolution of the camera, 2000 pixels in the Height field, and SU calculates the Width. Notice it’s more that 3008. That’s because of the extra space. Also be sure to use Transparent Background.
Open the file in Photoshop. You get a profile mismatch dialog. What the color space is for SketchUp output is a good question, but Generic, your-monitor-profile, or even sRGB probably all do well enough. (I would like to know the definitive answer, but this could open another topic of discussion.)
Now in Photoshop, go Image > Canvas Size and enter the horizontal resolution for the camera’s image (in this case 3008 pixels) and have it crop the excess from both sides.
Now you have a your SU output with pixel resolution matching the original photo. Select all, and with the move tool, drag this image to the original photo already open in another tab, PS will switch to that tab, then hold down Shift key and release. (The shift key centers the dragged image on the target image.) The SU image should drop as a new layer aligned with the SU model in the proper location.
For the foreground material, I duplicate the background layer, put it on top, create a layer mask and then start masking out what I need to.
Thanks so much for your help so far. I have another question. The steps you’ve outlined above have worked perfectly. However, if I try to render first (Shaderlight), the tiff output definitely does not retain the scale and location when imported into Photoshop (as it does when exporting 2D graphics directly from SKP), and therefore does not align with the same background photo. Have you/anyone run into this, and have a solution?
I haven’t use Shaderlight, but I have dabbled with SU Podium a little. Isn’t there somewhere when you set up the rendering to specify what the pixel resolution of the output will be? Set at least one or both dimensions of the rendering to match the pixel resolution of the original photo. That’s going to be camera dependent. For example, back then, my Nikon D70 was 3008x2000 pixels, but now I shoot with a Nikon D7100 which is 6000x4000 pixels. (It turns out that that high a resolution can make SU choke with Match Photo and projecting textures - discussed in another thread somewhere) You can always resample an image in Photoshop so they both match, but it’s better to nail it from the start.
