Here's the text that describes the procedure in PhotoMatch:

Matching an Existing 3D Model to a Photo's Context
Use Photo Match to match an existing 3D model to a photo's context. To match an existing 3D model to a photo's context:

1. Take a digital picture of the location where you will place your building. This photo might or might not have an existing building where the model will be located.
2. Select the Open menu item from the File menu. The Open dialog box appears.
3. Navigate to your model.
4. Select the model.
5. Click Open. The model appears in the drawing area.
6. Select the Match New Photo menu item from the Camera menu.
7. Navigate to the photo where you want to place your building or structure.
8. Click on the photo. The photo is selected.
9. Click Open. The photo will appear in the drawing area on its own scene in SketchUp. You are also placed in a Photo Match session where you will calibrate SketchUp's camera to duplicate the position and focal length of the camera used to take the photo . The words "Photo Match" appear in the upper-left of the drawing area. Finally, the Match Photo dialog box appears. Refer to Photo Match Session Controls and Context-Menu Items for further information.
10. Click and hold the cursor on the origin. The cursor changes to a hand.
11. Move the cursor to a distinct origin-like point on photo (where three axes might intersect, such as bottom corner of building).
12. Release the mouse button. The origin is established.
13. Uncheck the Model checkbox in the Match Photo dialog box. The model will be hidden.
14. Click on an red axis bar grip (). The cursor changes to a hand.
15. Move the cursor to the starting point of a position on the photo representing a line parallel to the red axis.
16. Release the mouse button.
17. Click on the other red axis bar grip (). The cursor changes to a hand.
18. Move the cursor to the ending point of a position on the photo representing a line parallel to the red axis.
19. Release the mouse button. The first axis bar is aligned to the red axis.
20. Repeat steps 9 through 14 for the remaining three (one red and two green) axis bars.
21. Check the Model checkbox in the Match Photo dialog box. The model will reappear, oriented correctly to the photo (but likely not-to-scale with the photo).
22. Click and hold on any of the axes. The cursor will change to two opposing arrows ( ).
23. Move the cursor up or down the axis to scale your model. Your model will scale to fit the photo.
24. (optional) If your photo contains an existing building that is now represented by your model, click the Project textures from photo button in the Match Photo dialog box to project.
25. Context-click to invoke the Photo Match session context-menu.
26. Select Done. You are placed in a sketch-over-image session. This mode, unlike normal SketchUp drawing mode, is a 2D drawing mode. The words "Sketch Over" appear in the upper-left of the drawing area. See Sketch-Over-Image Session Controls and Context-Menu Items for further information.
    Caution - Tools that manipulate your point of view force you out of a Sketch-Over-Image session into normal SketchUp drawing mode. These POV tools are the Orbit Tool, Position Camera Tool, Walk Tool, and Look Around Tool.

Note: Make sure that the model axes correspond to those of the photomatch before you import the model (rotate the model before you import, if necessary).

@mike lucey said:

Hi John,

In the case of Pavorotti lip-syncing, I am not familiar with this
story, but I imagine he was doing it with his own recorded voice.

This is a somewhat different matter. The China Olympic people
were misleading the World and selling the true singer short (excuse
the pun).

Mike

What about the rich tradition of voice dubbing, the most famous of which has to be Marni Nixon?
http://en.wikipedia.org/wiki/Marni_Nixon

If your city is shown in detail on google earth you could try the following:
Bring a snapshot into SU, then model a building whose height you know. Next adjust the shadow settings in SU (i.e. month, day, time of day -- a lot of trial and error but you can get close) to match the shadows cast on the photo. Finally, model the other buidings in simple form and pull up until their shadows match. Save these other buildings out as components and add detail as required, then re-load into SU.

This thread is making me nostalgic for the 9-11 conspiracy theory posts....

Profiles are essential for new users -- allowing newbies to tell easily when they are not making faces and helping to distinguish groups and components.

Check out http://www.realworldimagery.com
The best foliage I know - perfect for photoshop, and with alpha masks already made.

Drop shows up on the right click context menu - select a group or component, then right click.

Now that smokers are considered pariahs*, here's relief for all us die-hards....

*despite the junk science of second hand smoke.

This is ridiculous, the prostate exam is nothing. For that matter neither is a biopsy if required -- like being snapped with a rubber band. Of course it requires that you visit a doctor more than once in 10 years....

@remus said:

It varies i think, most TVs are 16:9 but i think 16:10 is fairly common in monitors.

Television pixels are not square, thus computer screen of 1920 x 1200 is 16:10, but an HD TV screen is typically 1920 x 1080 or 1368 x 780 or 16:9.

http://en.wikipedia.org/wiki/Aspect_ratio_(image

EDIT - for some reason the link doesn't copy correctly - the last parenthesis screws up. Try a google search for TV aspect ratio and pick the wikipedia link.

Here's some stuff from a few years back -- forklift I downloaded (not a formfonts item) and some drums and racks I modeld from an actual storage installation.

drum_racks.skp

To get to the old gallery section substitute f=9 at the end of the URL. For a good time go all the way back (about page 157) and look at the early examples by Grant Marshall, Ollie Bergman, Rein de Hart, Greg Laverdera, Allister Godfrey and others just to see what could be done with v2.

"If you want to de-select everything -- all your geometry, and dashed lines, and such, is there a command that is all about that? Like Apple "D" in funky Photoshop?"

CTL+T on windows - shortcut built into SU. Must be a similar one on the mac.

BTY, why would you want to learn rendering when your illustrations are far superior already?

Yes, this is a bug in SU6 that usually happens when you are working with an older file (i.e. V5). It's been discussed and acknowledged on the Beta forum, but never fixed. A work-around is to "save as" your component (give it a new name and you can delete it later). Then try reload and you will get the browser. If you have multiple components to reload you have to repeat the process. Can be very time consuming (not to mention annoying) if you have lots of components. Let's hope they fix it in V7.

Susan,
You might want to look at the Epson "all-in-ones" -- specifically the RX595 ($150) or the RX640 ($200). Limited to 8.5 x 11 scans and they do a decent job, but they also do color copies and are very good photo printers (6 ink cartridges).

Paul

Way too many to list them all, but:
Five Easy Pieces
My Dinner with Andre (all talk, but very moving at the end)
Defending your Life (Albert Brooks, Meryl Streep, Rip Torn)
Meet Me in St. Louis
Princess Mononoke (not quite sure to this day that I understood it, but stunning)
Shadow of a Doubt (favorite Hitchcock)
Pretty Poison (Tuesday Weld - incomparable bad girl)

Must be "The Eagle has Landed". Michael Caine did not make a very convincing Nazi. Although Donald Sutherland was great as amoral Irishman. Even better was he in "Eye of the Needle".

By any chance do you have integrated Intel graphics on your computer (as opposed to a separate video card)? This is a common problem with the Intel on-board chipset and SU. No real solution except a proper video card that supports OpenGL.

I must say I think the best stuff being built is in the Netherlands....

http://www.nytimes.com/2008/06/08/magazine/08mvrdv-t.html?pagewanted=1&_r=2&ref=magazine

It's your style setting that's causing the problem. Try using the default style 02 hidden line instead.