About a lens' sweet spot

[Dazz1]

Yesterday I took quite a lot of photos with my 18-55mm kit lens, mostly I left the aperture fairly wide, and with zooming, that means around f/4 - f/5 and I think I have noticed a trend. The sharpest photos seem to be when the focal length is around 30mm. Photos at 55mm are not as clear as at 30mm, but better than at 18mm. I was looking at dpreview's review of the lens, and they say sharpness is good 18 to 35mm "even wide open". A couple of questions come from this...

1. Since they say "even wide open", is it common for a lens to have sharpness problems wide open, and to benefit from stopping down?
2. Can different lenses of the same model, have individual characteristics. I mean, could it be that mine does perform better at 30mm than at 18mm, even though the review doesn't mention it.

Also, all photos were taken with auto focus. Could it be that the auto focus got it a little wrong at 18mm more often than at 30mm, or is more likely a lens property as I have been thinking above? Maybe I need to do my own controlled testing. What would be a good method, and test subject. I imagine a tripod and manual focus, then trying sets of focal lengths, each set at a different aperture, all of the same subject?

[Kym]

Great questions! There are a few excellent items that will help understanding the answers.

Any lens, esp. zooms will have a sweet spot.
A prime will be sharpest at a specific aperture, typically f/4 - f/8.

Stopping down to f/16 or more will generally invoke sharpness issues due to physics ( http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm )

A zoom lens is a compromise, so it will be sharper at a given focal length, also affected by aperture. ( http://www.cambridgeincolour.com/tutorials/lens-quality-mtf-resolution.htm )
Eg: My 50-500 is best at ~460mm and f/8

Lens MTF explained: http://www.luminous-landscape.com/tutorials/understanding-series/understanding-mtf.shtml

Take some time to read these and it will all become sharp :)

[Dazz1]

Thanks! I am off to read.

btw. a quick experiment on my tripod, manually focussed, confirms 30mm is best on my lens. More testing needed... Maybe I need a good test card.

[Kym]

Here are some charts .... http://pentaxdslrs.blogspot.com.au/2008/06/part-1-autofocus-adjustment-for-pentax.html

[Dug]

Assuming this is your lens -Canon EF-S 18-55mm f/3.5-5.6 IS II, here is DxOMark's findings on it.
http://www.dxomark.com/index.php/Lenses/Camera-Lens-Database/Canon/EF-S-18-55mm-f-3.5-5.6-IS-II/%28camera%29/619
Go from the link to the "measurements" tab, then "sharpness" below that, then the "Field map" tab. Click on the various apertures and focal lengths to see the changes in the representation of the frame size.
Green is good, red is not so good.

Specific to this lens:
f5.6 seems to be the best across most of the zoom range, with 35mm giving the best sharpness across the frame, but as you observed 55mm does not give the best results but going to f8 is where it works best.
Don't go past f11 if you can avoid it at any FL.

Don't let spec's get in the way of your creative aperture decisions, but if you can get the same result with a more favored focal length by zooming with your lens then you will get the best of both worlds.

[Dazz1]

Assuming this is your lens -Canon EF-S 18-55mm f/3.5-5.6 IS II, here is DxOMark's findings on it.
http://www.dxomark.com/index.php/Lenses/Camera-Lens-Database/Canon/EF-S-18-55mm-f-3.5-5.6-IS-II/%28camera%29/619
Go from the link to the "measurements" tab, then "sharpness" below that, then the "Field map" tab. Click on the various apertures and focal lengths to see the changes in the representation of the frame size.
Green is good, red is not so good.

Specific to this lens:
f5.6 seems to be the best across most of the zoom range, with 35mm giving the best sharpness across the frame, but as you observed 55mm does not give the best results but going to f8 is where it works best.
Don't go past f11 if you can avoid it at any FL.

Don't let spec's get in the way of your creative aperture decisions, but if you can get the same result with a more favored focal length by zooming with your lens then you will get the best of both worlds.

What a great tool. Thanks for showing me that. My eyes have trouble with the yellow/green but it seems to agree with what I was observing.

[jjphoto]

...
1. Since they say "even wide open", is it common for a lens to have sharpness problems wide open, and to benefit from stopping down?
...

In a word, yes, but it depends on the lens and what it is designed for. Many macro lenses are designed for optimal performance close to wide open. Most lenses, like yours, primes or any common photographic lens typically works better when stopped down at least one stop and generally at least 2 stops. Performance tends to continue to improve until diffraction kicks in and limits the lenses optical performance. For most lenses this is between F8 and maybe even F16. Diffraction occurs much earlier with macro lenses so they tend to be designed to perform best close to wide open, but are always sharper stopped down a little. Some lenses perform almost as well wide open as they do stopped down but these are never cheap and are certainly exceptions, such as Leica R 180/2.0 and similar high end lenses.

http://photocornucopia.com/images/Lenses/Leica/R180Cron/IMG_7782_1_pc_400.jpg


...
2. Can different lenses of the same model, have individual characteristics. I mean, could it be that mine does perform better at 30mm than at 18mm, even though the review doesn't mention it.
...

Yes, it's called sample variation but it's normally not a big issue. Some brands have very little sample variation due to stringent quality control, others have lots (anecdotal evidence only).

Roger Cicala from Lensrentals (in the US) has written some brilliant stuff about it and if you are going to read anything from anyone you should probably read some of Rogers stuff. He has access to vast amounts of high end lenses and optical test gear, writes well and knows what he's on about. Start with these:

http://www.lensrentals.com/blog/2011/10/notes-on-lens-and-camera-variation
http://www.lensrentals.com/blog/2011/10/the-limits-of-variation
http://www.lensrentals.com/blog/2012/11/canon-24-70-mk-ii-variation


...
Also, all photos were taken with auto focus. Could it be that the auto focus got it a little wrong at 18mm more often than at 30mm, or is more likely a lens property as I have been thinking above? Maybe I need to do my own controlled testing. What would be a good method, and test subject. I imagine a tripod and manual focus, then trying sets of focal lengths, each set at a different aperture, all of the same subject?

You should always test using Live View to focus manually. AF is not consistent or reliable, IMHO.

You are doing the right thing to test your equipment and know it's limits. I think everyone should do that to some degree. I've written a short guide on the subject as it's something I often do with my own lenses:

http://photocornucopia.com/1044.html

There's more to lens performance than sharpness alone so it's worth testing your gear and learning how it performs with respect to it's bokeh in various circumstances, vignetting, flare etc. depending on what matters most to you.

[Dazz1]

A couple of test shots with those charts, shot on a tripod with remote control for shutter release. While the charts are designed to measure focus (and I used auto focus), I think you can see the different in sharpness as well. Looking at the 0mm line, the first picture, the one shot at 32mm) is sharper (to my eyes anyway :) ).

http://ubuntuone.com/30NKZ4BY1Wr2Hs49irJE3e

http://ubuntuone.com/2KWbAuvhmSL65Xk4G14JYX

- - - Updated - - -


In a word, yes, but it depends on the lens and what it is designed for. Many macro lenses are designed for optimal performance close to wide open. Most lenses, like yours, primes or any common photographic lens typically works better when stopped down at least one stop and generally at least 2 stops. Performance tends to continue to improve until diffraction kicks in and limits the lenses optical performance. For most lenses this is between F8 and maybe even F16. Diffraction occurs much earlier with macro lenses so they tend to be designed to perform best close to wide open, but are always sharper stopped down a little. Some lenses perform almost as well wide open as they do stopped down but these are never cheap and are certainly exceptions, such as Leica R 180/2.0 and similar high end lenses.





Yes, it's called sample variation but it's normally not a big issue. Some brands have very little sample variation due to stringent quality control, others have lots (anecdotal evidence only).

Roger Cicala from Lensrentals (in the US) has written some brilliant stuff about it and if you are going to read anything from anyone you should probably read some of Rogers stuff. He has access to vast amounts of high end lenses and optical test gear, writes well and knows what he's on about. Start with these:

http://www.lensrentals.com/blog/2011/10/notes-on-lens-and-camera-variation
http://www.lensrentals.com/blog/2011/10/the-limits-of-variation
http://www.lensrentals.com/blog/2012/11/canon-24-70-mk-ii-variation



You should always test using Live View to focus manually. AF is not consistent or reliable, IMHO.

You are doing the right thing to test your equipment and know it's limits. I think everyone should do that to some degree. I've written a short guide on the subject as it's something I often do with my own lenses:

http://photocornucopia.com/1044.html

There's more to lens performance than sharpness alone so it's worth testing your gear and learning how it performs with respect to it's bokeh in various circumstances, vignetting, flare etc. depending on what matters most to you.


I will. Thanks for all that info.

Interesting what you say about auto-focus. Manual is still the ultimate :). I also read the other day that auto-focus in live view, while slower because it is contrast based, may also be more accurate than the faster focus used when using the viewfinder. Of course, this would depend on the camera design too.

[jjphoto]

A couple of test shots with those charts, shot on a tripod with remote control for shutter release. While the charts are designed to measure focus (and I used auto focus), I think you can see the different in sharpness as well. Looking at the 0mm line, the first picture, the one shot at 32mm) is sharper (to my eyes anyway :) ).

http://ubuntuone.com/30NKZ4BY1Wr2Hs49irJE3e

http://ubuntuone.com/2KWbAuvhmSL65Xk4G14JYX

- - - Updated - - -




I will. Thanks for all that info.

Interesting what you say about auto-focus. Manual is still the ultimate :). I also read the other day that auto-focus in live view, while slower because it is contrast based, may also be more accurate than the faster focus used when using the viewfinder. Of course, this would depend on the camera design too.

By the way, sharpness often varies with distance so a lens may perform poorly at close range but extremely well at infinity, or vice versa depending on the lens. Most times there's not a big difference but there certainly can be. Once you've tested your lens at close range do the same at infinity to see if there is any change.