Comparison of wide angle lenses - Tamron 7-35mm, Canon 16-35mm MKII and Zuiko 7-14mm
Posted 11-19-2007 at 10:58 AM by Paul Shields
I present for you here a fairly rough and ready test - more to try out the new blog features on Photocamel as much as anything else 
.
It's often been said that one of the things that the Olympus 4:3rds system would struggle at is ultra wide angle due to the smaller sensor requiring very wide angle (in 35mm terms) lenses. The ZD f/4 7-14mm was an attempt by Olympus to dispel this myth, and largely it does do that. Conversely, a common criticism of DSLR's using the full frame (35mm) based sensors is that they are poor at wide angles due to light fall off and soft corners when shooting wide open and at a wide angle.
I'm lucky enough to have access to a Canon 5D + their new 16-35mm mkII f/2.8 zoom as well as my usual Olympus gear and thought I'd try a quick test to see what differences there are between the two systems when shooting at such wide angles. Just for good measure, I'll also throw in the Tamron f/2.8-f/4 17-35mm lens - a cheap but fairly well-regarded lens but which as shown here is really more suitable for APS-C sized sensors than their full frame cousins.
As each lens has a different minimum focal length I used 17mm as the common denominator, though the EXIF for the 7-14mm shows 9mm (i.e. 18mm), but I think that's close enough.
Full scene overview - all lenses shot wide open:
Olympus E-400 ZD 7-14mm f/4
Canon 5D + Tamron 17-35mm f/2.8 17mm
Canon 5D + EOS 16-35mm MKII f/2.8 17mm
Center 100% crops at f/2.8 (this obviously excludes the Olympus 7-14mm as it's maximum aperture is f/4)
Center 100% crops at f/4
Center 100% crops at f/8
Center 100% crops at f/16
Conclusions on center shots:
The Zuiko is very even and pretty sharp even wide open. By f/8 it seems to have hit it's peak, and performance has tailed off by f/16.
I'm impressed with how the cheap Tamron is really quite sharp in the center. It seems to have a bit more contrast compared to the Canon. Of course these center shots hide a multitude of sins, so unless you use the lens on a crop camera you're going to have to deal with the results in the next round of images.
Corner 100% crops at f/2.8 (again excluding the Olympus 7-14mm as it's maximum aperture is f/4)
Corner 100% crops at f/4
Corner 100% crops at f/8
Corner 100% crops at f/16
Conclusions on corner shots:
As can be seen the 17-35mm is really rather poor on a FF body wide open, though unsurprisingly improves significantly when stopped down. The Canon 16-35mm mk2 is better but again does need stepping down (though not so much) before the detail starts appearing. The Olympus ZD 7-14mm has an advantage (if you can look at it that way) of starting at f/4 and puts on a good show straight away. By f/8 it's as good as it gets (I did take f/22 shots with all of these lenses but they do not show any improvement, and in the Olympus case it's a marked deterioration due to diffraction issues).
All of the lenses suffered with chromatic aberrations (CA) to a greater or lesser extent. The Olympus has it least and is pretty much unaffected by f/8. The Tamron is the worst offender (and I can back this up with real world shots that show this a lot). The Canon improves as the aperture decreases, but never quite removes it.
Of course it's not really fair to compare 3 lenses that fall into different price brackets not to mention different mounts. The Zuiko was also not at the widest angle (here at 9mm rather than 7mm) which makes quite a difference in field of view. I have tested this lens at 7mm though and it's pretty even across the board. I also don't have a 14mm FF equivalent zoom lens at hand so can't compare (do any actually exist?).
Even so, if I were going out with the aim of shooting some WA landscapes I'd plump for the ZD 7-14mm 75% of the time. It lets me shoot at wider apertures, so I can get possibly faster shutter speeds hand held (and the new E-3 with IS will add to this greatly). It also has the least distortion and propensity to show chromatic aberrations. It is really incredibly accurate in showing horizontal or vertical lines, regardless of focal length. A great architectural lens. The one let down is that it can not accept filters due to the protruding front element. It also flares very easily.
I'd use the Canon 16-35mm the remaining 25% - either high ISO requirements or if I had a tripod with me as it really does need stepping down quite a bit to hit peak performance. What it does show though (and the 5D sensor is probably helping here) is a greater amount of detail, but you pay for that with smaller apertures and slower shutter speeds.
The Tamron costs around 1/5th the price of either of the other two lenses, and it does show in all round performance. Center resolution is really quite excellent, but there is a lot of distortion (check the horizontal and vertical lines in the intitial shots - they're very bowed). It looks like it would make a great WA lens on a cropped sensor though, so long as it's not used for precision shots such as architecture.
Equipment used:
Olympus E-400 - 10mp CCD sensor
Canon 5D - 12mp CMOS sensor
Olympus ZD 7-14mm f/4
Canon 16-35mm Mark II f/2.8
Tamron 17-35mm f/2.8-f/4
Tripod
Mirror lockup and wired remote on Canon, infrared remote and shutter delay on Olympus
Lightroom
Link to original RAW files (note 170MB zip file)
http://photocamel.com/ldm/RAW_FILES.zip
Addendum:
I recently picked up the Olympus E-3 and took a few reference shots with the 7-14mm attached. First impressions - it resolves quite a bit more detail than the E-400 that I used in this test. I can't explain why for sure as they're both 10mp sensors, but I guess there's some more finely tuned processing going on within the E-3. It would be very interesting to test the E-3 + 7-14mm against the 5D + 16-35mm mkII. Maybe I'll get round to that at a later date.
.It's often been said that one of the things that the Olympus 4:3rds system would struggle at is ultra wide angle due to the smaller sensor requiring very wide angle (in 35mm terms) lenses. The ZD f/4 7-14mm was an attempt by Olympus to dispel this myth, and largely it does do that. Conversely, a common criticism of DSLR's using the full frame (35mm) based sensors is that they are poor at wide angles due to light fall off and soft corners when shooting wide open and at a wide angle.
I'm lucky enough to have access to a Canon 5D + their new 16-35mm mkII f/2.8 zoom as well as my usual Olympus gear and thought I'd try a quick test to see what differences there are between the two systems when shooting at such wide angles. Just for good measure, I'll also throw in the Tamron f/2.8-f/4 17-35mm lens - a cheap but fairly well-regarded lens but which as shown here is really more suitable for APS-C sized sensors than their full frame cousins.
As each lens has a different minimum focal length I used 17mm as the common denominator, though the EXIF for the 7-14mm shows 9mm (i.e. 18mm), but I think that's close enough.
Full scene overview - all lenses shot wide open:
Olympus E-400 ZD 7-14mm f/4
Canon 5D + Tamron 17-35mm f/2.8 17mm
Canon 5D + EOS 16-35mm MKII f/2.8 17mm
Center 100% crops at f/2.8 (this obviously excludes the Olympus 7-14mm as it's maximum aperture is f/4)
Center 100% crops at f/4
Center 100% crops at f/8
Center 100% crops at f/16
Conclusions on center shots:
The Zuiko is very even and pretty sharp even wide open. By f/8 it seems to have hit it's peak, and performance has tailed off by f/16.
I'm impressed with how the cheap Tamron is really quite sharp in the center. It seems to have a bit more contrast compared to the Canon. Of course these center shots hide a multitude of sins, so unless you use the lens on a crop camera you're going to have to deal with the results in the next round of images.
Corner 100% crops at f/2.8 (again excluding the Olympus 7-14mm as it's maximum aperture is f/4)
Corner 100% crops at f/4
Corner 100% crops at f/8
Corner 100% crops at f/16
Conclusions on corner shots:
As can be seen the 17-35mm is really rather poor on a FF body wide open, though unsurprisingly improves significantly when stopped down. The Canon 16-35mm mk2 is better but again does need stepping down (though not so much) before the detail starts appearing. The Olympus ZD 7-14mm has an advantage (if you can look at it that way) of starting at f/4 and puts on a good show straight away. By f/8 it's as good as it gets (I did take f/22 shots with all of these lenses but they do not show any improvement, and in the Olympus case it's a marked deterioration due to diffraction issues).
All of the lenses suffered with chromatic aberrations (CA) to a greater or lesser extent. The Olympus has it least and is pretty much unaffected by f/8. The Tamron is the worst offender (and I can back this up with real world shots that show this a lot). The Canon improves as the aperture decreases, but never quite removes it.
Of course it's not really fair to compare 3 lenses that fall into different price brackets not to mention different mounts. The Zuiko was also not at the widest angle (here at 9mm rather than 7mm) which makes quite a difference in field of view. I have tested this lens at 7mm though and it's pretty even across the board. I also don't have a 14mm FF equivalent zoom lens at hand so can't compare (do any actually exist?).
Even so, if I were going out with the aim of shooting some WA landscapes I'd plump for the ZD 7-14mm 75% of the time. It lets me shoot at wider apertures, so I can get possibly faster shutter speeds hand held (and the new E-3 with IS will add to this greatly). It also has the least distortion and propensity to show chromatic aberrations. It is really incredibly accurate in showing horizontal or vertical lines, regardless of focal length. A great architectural lens. The one let down is that it can not accept filters due to the protruding front element. It also flares very easily.
I'd use the Canon 16-35mm the remaining 25% - either high ISO requirements or if I had a tripod with me as it really does need stepping down quite a bit to hit peak performance. What it does show though (and the 5D sensor is probably helping here) is a greater amount of detail, but you pay for that with smaller apertures and slower shutter speeds.
The Tamron costs around 1/5th the price of either of the other two lenses, and it does show in all round performance. Center resolution is really quite excellent, but there is a lot of distortion (check the horizontal and vertical lines in the intitial shots - they're very bowed). It looks like it would make a great WA lens on a cropped sensor though, so long as it's not used for precision shots such as architecture.
Equipment used:
Olympus E-400 - 10mp CCD sensor
Canon 5D - 12mp CMOS sensor
Olympus ZD 7-14mm f/4
Canon 16-35mm Mark II f/2.8
Tamron 17-35mm f/2.8-f/4
Tripod
Mirror lockup and wired remote on Canon, infrared remote and shutter delay on Olympus
Lightroom
Link to original RAW files (note 170MB zip file)
http://photocamel.com/ldm/RAW_FILES.zip
Addendum:
I recently picked up the Olympus E-3 and took a few reference shots with the 7-14mm attached. First impressions - it resolves quite a bit more detail than the E-400 that I used in this test. I can't explain why for sure as they're both 10mp sensors, but I guess there's some more finely tuned processing going on within the E-3. It would be very interesting to test the E-3 + 7-14mm against the 5D + 16-35mm mkII. Maybe I'll get round to that at a later date.
Total Comments 10
Comments
-
I've just been very impressed with the new Tamron lenses. Great articles. Thanks for posting it.Posted 11-20-2007 at 11:56 AM by JDArt
-
Yeah, Tamron do make some excellent value lenses. I had one of their classic 90mm macro lenses when shooting film and it was excellent. It's still going strong now in the digital world.Posted 11-20-2007 at 02:15 PM by Paul Shields 
-
My Tamron 28-75 XR Di is a staple on the 1DS Mark II. It easily outshines many of the Canon lenses I own.Posted 11-20-2007 at 02:28 PM by JDArt
-
interesting findings
thanks for sharing.Posted 11-25-2007 at 11:42 AM by Unregistered -
thanks
This was a very interesting test for me. The 5D and the E3 are not far apart in price and I like wide angle stuff.
F16 is too much for 4/3rds cameras. I don't know why Olympus include it and F22 on their lenses. The difference in depth of field is so small compared to the loss of image quality.
Thanks for your work.
JonPosted 11-25-2007 at 05:21 PM by Unregistered -
Jon
I agree - f16 does not show maximum resolution of the plane of focus. However, I still think that such aperture settings are useful. It's all about showing the whole image in focus, and stopping down to f/16 or more can still have a perceptual effect of making the image appear sharper - especially for very close macro photography (e.g. 50mm + EX-25).Posted 11-30-2007 at 04:42 PM by Paul Shields
-
Nice article Paul. I didn't know you were so into Thomas the Tank...
Nice comparison and it looks like on a crop sensor, the Tamron is a pretty good investment, although I think this particular lens is a little weaker than some others in their line-up as mentioned in the comments above.
Regards,
PeterPosted 12-02-2007 at 02:38 PM by peterst6906
-
I gotta say...
This test if anything just proves how good the Zuiko 7-14 really is. Not impressed with the Canon lens, more so considering how much more you have to pay for that Body + lens. Those who get a 1dsMKIII better be ready to get the best lens money can buy for that mount or they will be throwing away megapixels out of the window.
Just imho, of course.
- RaistPosted 12-06-2007 at 01:05 AM by Raist3d
-
Great to see how well the Tamron performed.. not surprised at all since I own the Tamron 28-75 2.8 and use it for all my portrait work.. and it's tack sharp..Posted 12-06-2007 at 11:43 AM by JoRoFoto
-
nice test. Is it me or is the Tamron really having a heavy vignette in the corners?Posted 12-29-2007 at 07:13 AM by Rense
Total Trackbacks 0
