Here I rescaled an image you can see what happened to the resolution. As Vinoth points out.

\

When I click on an embedded image in a client's Illustrator file, the effective resolution is 269 x 34 ppi.

When you say "a client's file," does that mean it is a file you have prepared for the client, or it is a file you received from the client?

Where are you reading the PPI values?

Illustrator's screwed-up PPI readings will indicate different PPI on an embedded raster image after it has merely been rotated. So the readings you see do not necessarily indicate that the image has been disproportionally scaled. You can see this by performing a simple exercise:

1. Rectangle Tool: Click (don't drag). In the resulting dialog, enter 2 inches for width, 1 inch for height.

2. Object>Rasterize. 300 PPI.

3. Document Info palette: Turn on the options Selection Only and Embedded Images. Note that it indicates "Resolution: 300 x 300 pixels per inch."

4. DoubleClick the Rotate tool. Enter 90°. Click OK.

5. Look at the Document Info palette. It now indicates "Resolution: 600 x 150 pixels per inch."

Illustrator can, of course, resample raster images. That's what happens whenever you select Object>Rasterize or Object>Flatten Transparency, and the current selection involves raster images. People unwittingly resample raster images all the time in this program.

All raster images on a page do not necessarily need to be the same resolution. Examples: You don't need as high a PPI for a rasterized drop shadow as you do for a detailed photograph. You typically need a higher PPI for line art or text rasterized as 1-bit images. 1-bit raster images of quite low resolution are often used for texture effects.

The belief that "everything in print has to be 300 PPI" is a myth. It's a popularized corruption of the age-old rule-of-thumb to scale raster images so as to achieve a PPI that is between 1-times and 2-times the halftone screen ruling (i.e.; 150-to-300 PPI, assuming a halftone ruling of 150 LPI.) A PPI of under 1 x LPI risks visible pixelation. A PPI of over 2 x LPI is just wasted data that will be interpolated away. The expression of the rule-of-thumb as a range, rather than a fixed single value, is to accommodate a little bit of on-page scaling in either direction (and there's nothing really wrong with that, either).

JET

Megan-6543 wrote:

 

... Which number should I pay attention to for print quality? The 269 ppi or the 34 ppi? ...

To both. If you use Photoshop and have ever stretched a selection with pixels with the transform tool (Ctrl + T) or stretched the entire image with the Image Size, then you can think that in Illustrator this is the same thing. The image is rasterized again with the new dimensions by interpolating the original number of pixels and it will have less original detail along the side that interpolates more.

In Illustrator, to find out how bad or good it is, check the original pixel dimensions next to Size in the Document Info panel and compare to the number of pixels per side after scaling (the scaling that was already done by your client). Pixels per side after scaling is not the numbers next to Dimensions in Document Info unless you plan to rasterize or print at 72  ppi. Instead find the needed pixels per side by multiplying the current size (inches per side) by the intended ppi.

You can rely on the Resolution numbers in the Document Info panel if this is the ppi you intend to use and if you never rotated or deformed the image in a way that makes its original sides no longer coincide with the original sides of its reset bounding box. This is because Illustrator reports the effective ppi by mapping the original number of pixels per side to the current reset bounding box of the image. To understand what this is, create a box, rotate it say 35°, and choose Object > Transform > Reset Bounding Box (this is the transform box that lets you scale with the black pointer). Illustrator maps the original number of pixels to the sides of this bounding box. If the box was rotated 90°, resetting the bounding box would not show any visual difference but internally the sides are swapped which makes difference when mapping the original number of pixels and reporting the effective ppi. Because of this primitive way of reporting effective ppi, for rotated objects this information is no longer reliable.

It will also be a problem if the image was rotated and stretched and it is not obvious from the image what were the original sides, for example it will be obvious that a portrait was rotated but not abstract art. In such case if in doubt, zoom in and examine carefully for image stretching appearance and try to figure out how the image was originally oriented based on the original pixel dimensions reported for Size in the Document Info.