50D vs. XTi = IQ tie

Started Oct 8, 2008 | Discussions thread
ForumParentFirstPreviousNextNext unread
Flat view
Johns a geek
Regular MemberPosts: 188
Like?
50D vs. XTi = IQ tie
Oct 8, 2008

These kinds of tests have been somewhat beaten to death. I read them all nervously as I thought about my choice for a new body. But I'm now the owner of a shiny new 50D and I thought I'd see how it would hold up in the image quality dept. against my XTi.

Basically, I don't see that much of an advantage, if any, given the same light falling on the sensor.

As far as usability, the 50D is far superior, given the control options, frame rate, beautiful screen, custom settings, much better viewfinder, etc. etc. etc. The only real downside is the weight.

For all you pixel-peeping nerds out there, I've made available some test images. I did two series of tests, mainly aimed at higher-ISO performance. All the shots were handheld, so beware of focus and shake effects which degrade sharpness. This is not a resolution test!

http://rapidshare.com/files/152167325/comparison_shots.zip.html

First I shot a series of a small cactus under relatively dim indoor lighting. I let the 50D meter, and then manually set the same exposures on the XTi. The goal was to see how noise compared across the range of available ISO from 400 up. I also underexposed some ISO3200 shots and pushed them 1 or 2 stops in DPP to compare with the 6400 and 12800 in-camera expansion modes.

The second series was of the sun-backlit leaves of a coleus plant. I shot at ISO1600, to see how high-ISO would look under brighter lighting. I also wanted to see the effects of post-processing nr. I also tried downsampling the 50D data to match the XTi data. ie, I took a 960x720 crop from the 50D and exported it as a 800x600 jpg.

All pp was done with DPP 3.5.0. I checked that all settings were set to 0 or off, to leave the data as raw off the camera as possible. I also copied the 'recipe' and pasted it to all the sample images, so that even if something was being done, it's been applied evenly across all the images. All the test images were exported from DPP as quality 7 jpg.

To my eye, the output from the two cameras seems quite similar, both in terms of per-pixel noise, and in terms of resolution-independent noise. I welcome any comments.

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