How do I tell is a lens is as sharp as it should be?

singletrackmind · Jul 26, 2008

TamiyaGuy · Jul 26, 2008

Hmm... from the links I tried, I could only get the "web-sized" versions of each image. Try uploading the full, unedited versions to a website like www.tinypic.com or www.imageshack.us. Also, there are a lot of photos there. Try focusing from 2 feet at the lenses' maximum aperture, f/4, f/9 and perhaps the minimum aperture.

Sorry if that was a bit of a crappy answer, but it's nigh-on impossible to check a lenses' sharpness from tiny photos like that.

Overread · Jul 26, 2008

Agreed - websized photos go through compression and often get sharpened again after resizing so its very hard to judge how sharp a lens really is from any websized photo. Fullsized is the only way - if you can't upload them to a site like Tamiya suggested try taking 100% crops from the photos (corner and centre from a few). A 100% crop (incase you don't know) is when you take the fullsized photo and then crop (cut) a section out of it and post that as a photo - thus the section you post is at full size

singletrackmind · Jul 26, 2008

I will try to post them on tinypic. On flickr I can view them in different sizes. I may be able to change a setting to allow others to do the same. I will try that too.
Thanks John

Bifurcator · Jul 26, 2008

This might help you as well: http://www.graphics.cornell.edu/~westin/misc/ISO_12233-reschart.pdf

singletrackmind · Jul 27, 2008

Ok so I think I figured it out but you guys will have to sign up for a free account to view the different sizes. I think you will be able to look at the original size photos now. I have been using this site for more than a year and it is a clean site. No worries.
Thanks John

chris · Jul 27, 2008

Photographing resolution charts is fine but it will only give an indication of the performance of the whole lens, camera and viewing system (including the effects of any in or out of camera file manipulation). The lens performance may be apparently degraded by factors that have nothing to do with the lens.

If you really want to know how well your lenses perform then send them off to an optical laboratory and get the lens tested properly. Alternatively, get a life and go out and take some real photograhs and see if they are acceptable for your needs. After all, how many photographs of resolution charts do you see hanging on peoples walls.

Bifurcator · Jul 27, 2008

/raises hand.

singletrackmind · Jul 28, 2008

chris take a look at my flickr photostream and you will see that there are a couple of hundred real pics of real people and real places but most of them have already been manipulated in some way or the other with software. basically i was hoping that someone could decode the greek instructions on the chart. so that I could figure out how to use the numbers and bars to determine if my lens is acceptably sharp. the chart gives some formulas and such to determine some other number that is supposed to represent the resolution of my lens. What are the groups and what are the elements and what is an acceptable number? 1.41, 1.59, 2.52?
Thanks
John

chris · Jul 28, 2008

The chart attempts to show the lens resolution in lines per mm. I cannot decipher the text on the chart images so I cannot comment on that. The contrast of the images appears to vary quite a bit and this can have a profound effect on the perceived results. From memory, resolution figures within the range of 30 to 40 lines per mm used to be considered quite reasonable.

My point was that when you photograph the chart the image you end up with may be more degraded due to effects from the camera body, file manipulation, viewing method etc than from the lens. When you try to assess resolution from a chart the result is highly subjective and depends upon the viewing conditions and visual acuity etc of the viewer - there can be variations of 30% or more in the perceived resolution. This is why Modulation Transfer Function (MTF) is usually preferred instead of resolution charts when assessing lens performance. MTF is measured on an optical bench with no camera body to interfere with the results. From the MTF curves you can get an indication of the amount of contrast that a lens can handle and its resolving power.

Also, the image quality obtained from a lens is not necessarily dependent upon it having a high resolution. The best test of a lens is to use it to take the type of images that you prefer. If you want to compare lenses then take photos of real subjects using the lenses under identical conditions same lighting subject distance, same aperture, same camera body, same ISO and eliminate as many sources of vibration as possible by choosing a solid, static subject, mounting the camera on a good tripod or solid object and using a shutter release cable etc. Critical assessment of real images will tell you much more about a lens performance than any resolution chart.