XTi image quality

[CalgaryPix]

Picked up an XTi a couple of weeks ago to play with, I'm considering switching over from the Fuji S2 Pro to Canon (there's a system I want to play with). So I got the camera and three extra lenses to play with. What I got was the kit lens (which sucks - I know) a Canon 70-210 f4 and a 50 f1.8 along with the Tamron 28-70 f2.8 Di. I picked up the Tamron after reading several very favourable reviews and also after playing with it and several L and non-L Canon lenses. All the Canon's I tried in the same range seemed to be no better than the Tamron (I got sample images from each, although no definitive tests) and for the money was happiest with the Tamron.

However, the camera seems to be soft with each of the lenses - I did a shoot the other day for a flower magazine and tried all four lenses and got the same results from each.

I'm really impressed with the colour and tonality of the images - but the fine details were really soft and had to be over sharpened.

Something I haven't done is tried different software than the basic programs that came with the camera. I know that my Fuji doesn't put out nearly as good quality with Adobe Raw Converter as it does with the Fuji software, and even the beta of Lightroom did a pretty decent job.

Any thoughts? I exposed as close to the high end of the histogram as I could get away with, and took out the background in PS with curves and the eraser.

http://i170.photobucket.com/albums/u266/SPD_CGY/Bloom_11_M_LR.jpg

http://i170.photobucket.com/albums/u266/SPD_CGY/Bloom_11_M_LRcrop.jpg

These were shot with studio flash, f9 at 1/125.

[Carl Harsch]

Are you shooting RAW or JPG? If the latter, have you tweaked the preferences to suit your taste?

The XTi is certainly capable of producing some mighty fine images. The glass used will also have a major impact on the quality of the photo, as will shooting techniques and settings chosen.

[CalgaryPix]

Yes, I almost always shoot RAW.

[Strayz]

well, I know that the XTi will put out images that rock. I have a few shot from a kit lens that would shock people. The answer is like Carl mentioned above.

Glass maters, I did not think it was that big of a deal, untill I shot with some "L" glass.. Knocked my socks off.. down the trail and somewhere in to the Willamette river..

I noticed you shot the first image at F25? might want to bring it back to f16 or so. most lenses are sharpest from f8-16, this might help a little bit. also take a carefull look at waht apps you are useing for post processing. It can take a while to find what you liek to do and what gives best results. I myself go DPP>16 bit tiff > PS CS2 > output to what have ya, then save for what ever I am useing. The second image looks a little like a 100% crop zero post processing. Try useing DPP, adjust the white balance. Use the auto curves, and set the saturation to around 120-130. It will add a bit of pop and color to your images. Next save as a 16 bit Tiff, dont resize it yet. now open in photoshop ( I recomend getting a newer version how ever that is up to you how you aquire it). First thing, All color adjustments first, then size it down around 10-15% and then run a unsharp mask settings around 100, 1.7, 4 then resize it down to the final size. Run a final unsharp mask again if needed. Each time you size it down it will look to get sharpened but it is in the eye of the beer holder. Runing the final unsharp mask should be the last thing you do, if you need to blur, crop or somethign to it to the full size image then down size it, dont do that last as it will not look as good in the final product.


Also rember DSLR's have a huge learing curve, so keep shooting and if you can suceed the first time shoot again, and again, and a well you get the idea.

[CalgaryPix]

Daron, thanks for the input, should note that the aperture was actually listed as f9 and the speed was 1/125th.

As well the second image IS post processed (hense the little bit of jaggies), original was even blurier. I wouldn't think the only Canon glass you can use would be L series, their standard glass (outside of the kit lens) should be at least usable - but all four of the lenses I tried (including the Tamron as above) gave pretty much the same result.

For this image I used DPP, into PS7 (one day I must upgrade - and I only use authorized software since my living depends on my computer.) I used the images at full size from the camera since it was for pubishing. I'd done some testing with the camera, including using strobes, and had gotten some pretty good shots. Guess I'm a bit frustrated with the glass I got for this camera since my S2 Pro can get way more detail using even a 50mm f1.8 (a lens I also have for the Canon.)

Re: unsharp mask, I generally find better results for print use running around 180 to 200 per cent, radius .7 and raise the threshold as needed to reduce effect on grain (for the web I use around .3 radius and fade it back.) This gives less stepping on edges than running larger radius and lower percentage and was how I was trained years ago on big Linotype-Hell scanners. I'm one of those guys using PS since version 2.

[Strayz]

Hmm.. ok the exif information that I am pulling form th image itself is showing 1/160th of a second and F25.. how odd... I included a screen shot of it.

Yes, I come from the multi step down unsharp masking camp.. Both are valid ways of hadeling images. I use the way I was taught back in photoshop 5 so it is all good.

[CalgaryPix]

Well, I stand corrected (actually sitting and typing corrected, and no - there's no white out on the screen) - just checked the exif on the raw file and sure enough F25, although I was sure it was much lower - but I was probably looking at the Aperture Value.

So now, I have to go back and do some testing to find the sweet spot in my lenses.

That being said, the image shouldn't be that soft from diffraction should it?

This is a totally raw version of the photo - only lightened a touch in photoshop, no sharpening.
http://i170.photobucket.com/albums/u266/SPD_CGY/Test0000.jpg

Is this the level of detail you're getting? More, less?

Thanks for your help.

[Pure Image]

I would definately say I get more detail. I have the XT. I shoot a lot of maternity and newborn. I get nice detail right down to tiny hair/fuzz on those little ones.

Not sure what is going on with your camera/lens. Did you use a tripod. Could it be a little hand shake?

[CalgaryPix]

Very curious, I just did a test with the same camera/lens (although different lighting - from f2.8 right up to f32.) Sharpest seemed to be between f11 to f18. All the shots were way sharper than the original shoot pics. I was hand holding the camera then, which shouldn't have been a problem with studio flash and a shutter speed of 1/160th and the pictures don't have the usual camera shake ghosting - but maybe. Might be getting shaky in my old age.

http://i170.photobucket.com/albums/u266/SPD_CGY/NewTest.jpg

These artificial flower were shot at f14 at 1/6th of a second on a tripod with continuous lighting, no post processing.

[CalgaryPix]

BTW - this is the whole image scaled down and showing where the above was cropped from:

http://i170.photobucket.com/albums/u266/SPD_CGY/NewTestShowCrop.jpg

[Pure Image]

Much clearer! I have been told that a tripod can really make a difference even at higher shutter speeds. Also if the image is underexposed sometimes that can result in an unsharp image and some extra noise.

[CalgaryPix]

I've been finding the Canon almost always underexposes by about 2/3s to a full stop - not all the time, but often enough. The first shots were as close to "perfect" exposure as I could get without blowing highlights. The second shots actually were a little blown out, but not severly.

[Strayz]

Oh >>SNAP<< I forgot

One other thing, Makes sure you are not useing any filters. I know for a fact one of my lenses dont like them. Period. TO the point it borks all images shot with them on. SO I now dont use them. There is also a few lenses out there that are known as not likeing them the 100-400mm "L" is almost famous for it. BUt my 18-55 kit lens hates them as well.

[CalgaryPix]

Well, my testing setup is still set up so I just tried the same shot with and without the B&W filter - yes, there is a difference in sharpness, not huge but a definite difference.

Anyone want to buy a just barely used UV filter? Actually I'll keep it on for when I'm doing snap shots so I don't put fingerprints on the lens - that would probably be worse for sharpness than the filter.