Help with personal test of Tamron and Sigma lenses

[jwhee0615]

Seasons greatings all. I was wondering if I could obtain some help evaluating a couple of lenses. I have picked up a Tamron SP AF 28-75mmF2.8 XR Di Macro and the new Sigma 24-70 macro to test out on my 20D. I am comparing them directly with my kit 18-55 and an older Tamron 28-200 lens which I believe are both lacking in the sharpness department. I have taken a series of test shots using a news paper taped up to a cabinet and my camera on a tripod using mirror lock up and a remote release. Using the Tamron, Sigma and canon kit lens I took RAW shots at wide open, F4.0, F8.0, F16, and fully stoped down. I took each of these at 18mm, 24mm and 28mm respectivly as well as at 50mm each and then 55, 70 and 75.

Now I'm fairly new to digital and I am not completely confident of what my shots tell me. I am not at all confident that my camera is taking as sharp a photo as it should either. Kind of hard to tell since I have no confirmed tack sharp lens at my disposal and I know the Tamron and Sigma do not have a history of consistancy. I would like to post some examples to my Pbase account for some of you helpful and knowledgable people to evaluate but I am unsure of the best way to proceed. Do I convert to Tiff or Jpg? Sharpen? Adjust white ballance?

I have initially converted all of the shots to 16bit Tiff's and applied the full amount of sharpening that DPP has available. No other adjustments made from RAW. Initial results look to me to show that the Kit lens is at least as sharp as the Tamron which is still not that great, and the sigma is really not very sharp at all.
I sure would like some other opinions though. I would greatly appreciate a bit of assistance. Thanks in advance...Jeff

[pcasciola]

First off, a newspaper is very low resolution. I'd get something with higher resolution to start with, like maybe magazine cover, or maybe even the Tamron box like another member and I just posted examples of in another recent thread:

http://photography-on-the.net/forum/showthread.php?t=51644

Second, in my opinion, you don't want to apply any sharpening or other post processing.

Also, You need to look at the center of the image and then each corner seperately, because while some lenses appear nice and sharp in the center, as you approach the edges some become a lot softer than others. And some "bad copies" are bad in just one corner. Take 100% crops from both the center and then each of the corners, then start making your comparisons.

And, if you get a chance, make up some comparison collages and post some of your results here.

Good luck with your tests.

(http://photography-on-the.net/forum/showthread.php?t=51644)

[jwhee0615]

Well I read some other posts on lense testing that a news paper should be used as you get a wide area of text coverage. I only applied sharpening to see if it would make any difference. If I post pics to be evaluated I will not sharpen. Also, how do you make the collage?

[pcasciola]

For pure sharpness tests I guess the newspaper test will work, as long as the paper is 100% flat, but I prefer to use something with some color in it too. For a pure resolution test, you could print the USAF 1951 chart on a photo printer, which is available from many different web sites in PDF format, which will allow it to be enlarged without too much pixelization.

Not sure what software you are using, but in Photoshop to create a collage you just create a new black image, say 800x800 to meet the suggested image size for this board, and then paste in four 400x400 crops from the 4 different lenses, labeling each one.

[colblaha]

after reading about the USAF 1951, i found this site.

I've never tried it, but it's a start:
http://www.jimdoty.com/Tips/Equipment/USAF_Test/usaf_test.html

[ron chappel]

Some hints that may help...
Test at reasonable distances,maybe 10-15 meters for general tests.
If landscape,macro, portrait ,etc are important to you then test at those distances because all lenses have certain focal distances where they are good or not so good

Take several pics with each lens and keep the best! That will allow for the inevitable small focus errors that can easily ruin an acurate test

One very usefull thing i can suggest is to experiment by sharpening pics from both the bad lenses and the good ones. It's common these days for someone to choose a cheaper than they'd like lens then improve the images in photoshop type programs.
This can only work so far though !! Try deliberately oversharpening an image and you quickly see what i mean (or do it several times if your editing program doesn't alow high settings)
Interestingly one quick clue in finding out just how good a lens is,is to see how far it can be sharpened.A great lens will allow much more sharpening and still look good while a cheap lens only accepts a modest amount of sharpening before it starts to look WORSE
Hopefully this suggestion doesn't confuse the whole issue more.One quick way of cutting through all the variables is to take the optimomly sharpened files from both the good lens and the cheap lens and print them out.............then ask yourself-would i be happy with the cheap lens photos?


It's sometimes interesting to compare a zoom lens at the same magnification (vary the distance to the test target) so that you can see where a lens is best .
eg. most tele lenses much softer at the long end

[ron chappel]

Oh-just in case you were considering it-
Never be tempted to resize a picture to get a direct comparison with another.
Even the smallest,slightest,tiniest downsizing/upsizing completely invalidates a resolution test!!
If it must be done for some reason then downsize both/all but do them different amounts

[Nightcrawler]

I would be interested in the outcome of this experiment. I did stumble upon someone elses test - http://www.pbase.com/fstopjojo/sigma_2470.

[phili1]

I am curious, why do people have to do lens tests. What exactly is the purpose except to make yourself paranoid.

In 30 years I have read numerous reports from popular photography & others,where they critercise the lens for barrel distortion or something (which is noirmal for this type of lens) else but at the end they say but in field tests it performed fantastic. Pictures were sharp with lots of contrast.

Now read what I said ]field test[/U] is the no one way of determining that a lens is good.

Take your camera to a field with trees, make sure the light is right. Take a picture of the bark with both lenses and compare the detail in each. One lens will give more detail then the other.

I own the 28-75 Tamron and in my opinion it is a great lens. I also own 2 Canon L lenses and have heard bad ab out both of them and I think they are super.

Sharpness is what you see and we all see diferent.