50D vs. XTi = IQ tie

Started Oct 8, 2008 | Discussions thread
OP Johns a geek Regular Member • Posts: 188
Re: 50D vs. XTi = IQ tie

I saved as jpg to keep download sizes down for people who want to get the files. To test your point, I exported one of the images twice. Once as quality 7 jpg (312k), and once as a 16-bit TIFF, which creates a 2.9MB uncompressed file. I opened both in Photoshop, layered them on top of each other, and chose "Difference" as a blending mode, and then flattened the image. This subtracts the top (jpg) from the bottom (tif) data. The result is a black page, indicating visually that there's no visible difference. To put some numbers behind that, I opened the Histogram window, and chose "Show Statistics". Out of a possible range of 256 levels, the mean value is 1.52, and the std deviation is 0.80. The most different pixels in the entire comparison were 15 levels apart, and there were only three of them.

The JPEG compression isn't lossless, of course, but it is insignificant from the point of view of this testing.

Why Quality 7 JPG? Why not Quality 10 JPG? If you wish to see the
noise differences and the sharpness differences and even exposure
abilities, you do NOT want to lose anything in the JPG save. The
difference in quality between 7 and 10 can and will nullify the
differences between the two cameras' shots.

In the interpolation and compression phase of the save it will be
tossing out vital image information in order to conserve space. That
information includes much if not most of the noise. Its doing it for
both the XTi and the 50D. You need to save both as quality 10 (DPP
3.41 goes up to 10, or 12 if you're using LR/PS). A valid test does
not introduce a lossy compression variable.

This invalidates your test IMO.

Anyone agree?

-Alan

Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow