Canon 24mm EFS

[joebranko]

I just recently purchased a Canon 24mm prime lens and decided to test the sharpness of the lens to see if it measured up to the manufacturers/retailers hype of being 'sharp as a tack'.

I have never tested a lens before so it has been trial and error.

First I downloaded a test chart which I found on the internet and printed it. I mounted this print on a 45deg angle (approx.) to the plane of the sensor. With the camera on a tripod, I took a shot at each aperture and at various distances from focal plane to center of target. Three of these shots are shown here #1& 2 When I viewed these shots on my computer at 100% I noticed that the edges of the lines were saw toothed ( is that even a word?) Since my print was on normal office stationery on a laser jet printer I wondered if what I was seeing was the ink bleeding into the paper? I did not think I could draw any conclusions from this test without a proper test chart printed perhaps on photographic paper.

I then decided to take some shots printed on book covers which should present a better target. I have included these also.

Today I decided to take a few shots of a landscape with the 24mm prime and same shot with the lens I most frequently use, my 17-70mm. These are also shown. The landscape shots have only minimal PP; white point/black point, highlight reduction and contrast and levelling.

It may be premature to make a judgement with just this data but I would still welcome any comments/advise.
#1 chart


#2 chart


#3 books


#4 books


#5 landscape


#6 landscape

[Black Pearl]

Your test chart is designed to see if the camera/lens combination is front/back focusing and won't give you any real indication of centre/edge/extreme corner performance. For that you need an actual lens test chart that you shout straight on.

The corners of your bookshelf shots don't have anything in them so won't display extreme border performance - which is where a prime should outshine a zoom.

Your landscapes have had various and we don't know if consistent processing, plus they're not available as full sized files for close analysis.



Download a lens test chart, set your camera perfectly inline with it and test both lenses at various apertures then compare the results.

Supply the results as RAW files so others can download them and give you feedback from their own experiences.

[dje]

Hi Joe

As Robin says, the test chart you've used is for testing AF, not sharpness. One chart commonly used for qualitative sharpness testing is the "Siemen's Star" such as the downsized version shown below. You can download a full-sized version here (suggest "Star Chart Full").




The idea is that the sharper the lens, the more detail you will see in the inner area of the chart where the lines get finer and closer together. You should do the test shots in RAW with no sharpening applied and view at 100% in ACR or LR if you want to get a fair comparison between lenses.

Dave

[DanK]

To test for AF accuracy, open the aperture all the way. Shooting at f/16 increases DOF and will make it harder to see a problem of front or back focusing, if there is a problem.

Re the sawtooth pattern: I didn't see it, but that would not be a lens issue. it sounds like pixelation. That can happen in printing, if the printer is not high resolution, or in editing a jpeg. I woudn't worry about it.

[joebranko]

Your test chart is designed to see if the camera/lens combination is front/back focusing and won't give you any real indication of centre/edge/extreme corner performance. For that you need an actual lens test chart that you shout straight on.

The corners of your bookshelf shots don't have anything in them so won't display extreme border performance - which is where a prime should outshine a zoom.

Your landscapes have had various and we don't know if consistent processing, plus they're not available as full sized files for close analysis.



Download a lens test chart, set your camera perfectly inline with it and test both lenses at various apertures then compare the results.

Supply the results as RAW files so others can download them and give you feedback from their own experiences.

Thanks Robin. I have downloaded a chart and will get it printed. One site suggested I print it on photographic paper20 x 30". I will try to get that done and post the results soon.

[joebranko]

Hi Joe

As Robin says, the test chart you've used is for testing AF, not sharpness. One chart commonly used for qualitative sharpness testing is the "Siemen's Star" such as the downsized version shown below. You can download a full-sized version here (suggest "Star Chart Full").

The idea is that the sharper the lens, the more detail you will see in the inner area of the chart where the lines get finer and closer together. You should do the test shots in RAW with no sharpening applied and view at 100% in ACR or LR if you want to get a fair comparison between lenses.

Dave

Thanks Dave. I downloaded this chart: http://static.bhphotovideo.com/explo...?itok=lze5dj9q and will get it printed. Not sure if I should use matt paper or gloss though. Perhaps one of each if not cost prohibitive.

[joebranko]

To test for AF accuracy, open the aperture all the way. Shooting at f/16 increases DOF and will make it harder to see a problem of front or back focusing, if there is a problem.

Re the sawtooth pattern: I didn't see it, but that would not be a lens issue. it sounds like pixelation. That can happen in printing, if the printer is not high resolution, or in editing a jpeg. I woudn't worry about it.

Could it also be a monitor problem?

[dje]

Joe the image you linked to in post 6 above has a resolution of only 960x540 which is too low for a 20x30" print. You need a good quality print for testing sharpness. The one I linked to in post 3 was 6299x4725 and has a more detailed star pattern!

Dave

[Dave Humphries]

Hi Joe,

I just recently purchased a Canon 24mm prime lens and decided to test the sharpness of the lens to see if it measured up to the manufacturers/retailers hype of being 'sharp as a tack'.

I have never tested a lens before so it has been trial and error.

If I were you, I'd save myself a lot of time and effort and drop this idea, unless you want to read up on lens theory instead of practising photography.

I don't feel you're going to be able to replicate the necessary technical standards at home to draw any meaningful conclusion, what's more likely to happen is that (in part because of the test limitations you are under), the results may appear sub optimal - and then you are going to be dissatisfied with your new lens (even if that's only a niggling feeling).

I don't blame you for trying, I think I did 'way back' when I was starting out (I think it is a natural phase we go through!), but, because I had an engineering background, came to realise the futility sooner rather than later - and gave up the idea.

Instead, just enjoy the lens; take it out, shoot great pictures - I wish I had a (Nikon) 24mm prime for my camera.

Just trying to save you hassle and potential heartache (if that's not being melodramatic), Dave

[Shadowman]

Could it also be a monitor problem?

Hi Joe,

How is your post processing software setup? Does it apply sharpening upon importing?

[joebranko]

Joe the image you linked to in post 6 above has a resolution of only 960x540 which is too low for a 20x30" print. You need a good quality print for testing sharpness. The one I linked to in post 3 was 6299x4725 and has a more detailed star pattern!

Dave

Thanks Dave. I will use your suggestion.

[DanK]

Could it also be a monitor problem?

conceivable, I think, if the monitor is relatively low-res. However, it's a distraction for your purposes.

I'd print the target on matte paper, to avoid reflections, even though the edges will be slightly less sharp.

I'm not quite as pessimistic as Dave. I think the odds that you will find a problem worth dealing with are pretty low, and it's a nuisance to test, but some of the tests aren't all that difficult. A reasonable test for corner sharpness is to set your camera on a tripod facing a brick wall from a normal shooting distance, parallel to the wall, and then take a series of shots, starting wide open and then closing down a stop at a time. You'll probably find that the corners are a bit less sharp, particularly wide open. Unless there is something dramatic, however--like one side very different from the other--it's probably just normal and not anything that will have much practical effect anyway.

If you want to test for front and back focus, there are several ways to do it, and the ones that are done close-up with a target won't necessarily give you precisely the same results as those done at normal shooting distances. For the former, you need a target with much more detail than the one you used, such as this one: http://regex.info/blog/photo-tech/focus-chart. And you need to shoot wide open. For shooting at a distance (I forget the usual guideline--I think it is at least 50 times focal length, but you will have to google it--there are a number of methods. One is called dot tune. You can also do it with moire patterns, e.g., http://www.komar.org/faq/camera/auto-focus-test/.

Some people find lenses that are enough off in one way or another to make it worth a fix or a trip back to the vendor, but I never have. After a while, I stopped testing lenses unless I noticed a problem.

[joebranko]

Hi Joe,

How is your post processing software setup? Does it apply sharpening upon importing?

My PP software is Lightroom. It does apply sharpening but I am not sure if it does so at importing.

[DanK]

My PP software is Lightroom. It does apply sharpening but I am not sure if it does so at importing.

just import the photos and then check the develop panel. By default, LR will apply a little sharpening, but you can zero it out.

[Dave Humphries]

Hi Joe,

Some people find lenses that are enough off in one way or another to make it worth a fix or a trip back to the vendor, but I never have. After a while, I stopped testing lenses unless I noticed a problem.

Indeed, shoot first, ask questions later


You said it was a Canon, I have more faith in their (and say Nikon's) Quality Control than certain other (third party) manufacturers, so as said, I can't see the point in worrying yourself unnecessarily.

It may be worth doing that AF test and using micro-focus adjustments in camera to fix any front/back focus, but do take the time to research how first, if what's been said above is not enough.

I prefer to wait until I see a problem in MY real world use before I start spending my time testing.