What determines size of a JPG ?

I purchased some supposedly high resolution JPG's of Monet artwork. In
a graphics program if I view 'image information' it says they are 2600
X 3600 pixels at 300 dpi. However if I expand them in my graphics
program to 100% they apppear to be pixelated. How can I tell if they
were truly shot at this resolution or if they have been manipulated
from a smaller image?

Thanks,

Allen

"allenc" <twice.redeemed DeleteThis @gmail.com> wrote in message
news:63aa15c8-3dfe-4fd5-a10e-8e3e3c27daf7@i7g2000prf.googlegroups.com...
>I purchased some supposedly high resolution JPG's of Monet artwork. In
> a graphics program if I view 'image information' it says they are 2600
> X 3600 pixels at 300 dpi. However if I expand them in my graphics
> program to 100% they apppear to be pixelated. How can I tell if they
> were truly shot at this resolution or if they have been manipulated
> from a smaller image?

This is an interesting question. If you are seeing pixellated data, that
may be an indication that the images were resized, or it could just be jpeg
artifacting. Jpeg not only has a finite cell size, but it also quantizes
color and brightness values.

I would do this by making several duplicates of the image in Photoshop,
resizing each one down by an increasing amount, and then resizing back
again. Then go compare the images and see at what point you can tell the
difference. This is a subjective judgment. Sharpening or other
manipulations may have been applied to the image, and you are really looking
for meaningful data, such as canvas and paint texture, brush marks, etc.

For example, create 3 copies, resize the them to 66, 50, and 25 percent of
the original. Then resize them again to match the original size. To make
comparing easier, create a new image and drag the layers from the original,
and the three copies to that image. Then click on the "eye-cons" in the
layer palette to hide and show individual layers for comparison with the
original. Add noise and/or sharpen with unsharp mask to see if you can
match one of the resized images to the original.

None of this constitutes proof that there was any deception involved, but it
will prove that the image does not really contain a full 9 megapixels. For
example the images may have been scanned from slides with less than 9mp of
effective resolution.
--
Mike Russell - www.curvemeister.com