The Holding Tank

Home Page Photography

Photography RSS 2.0 Feed

| Photos & Themes RSS | Hardware | Technical | Links | Books and Magazines  | Downloads |

Hardware

| Lens Data and Reviews | Lens Selection | Long Lens Selection | Wide Range Zooms | Consumer Zooms |

Macro Equipment |  Tilt and Shift Lenses | My Bags | My Old Hardware |

| My New Hardware | Lens MTF Comparison | Lens Tests | Weather Protection | Canon AF System | Camera Tests |

 

Lens Tests

| EF 28mm f1.8 USM | EF 50mm f1.4 USM | EF 100mm f2.8 Macro USM |

| EF 180mm f3.5L Macro | EF 180mm f3.5L Macro + EF 1.4X II  | EF 180mm f3.5L Macro + EF 2X II |

| EF 200mm f2.8L II | EF 200mm f2.8L II + EF 1.4X II | EF 200mm f2.8L II + EF 2X II | EF 200mm f2.8L II + EF 1.4X II+ EF 2X II |

| EF 300mm f4L IS | EF 300mm f4L IS + EF 1.4X II | EF 300mm f4L IS + EF 2X II | EF 300mm f4L IS + EF 1.4X II + EF 2X II |

| EF-S 18-55mm f3.5-5.6 | EF-S 10-22mm f3.5-4.5 USM | EF 17-40 f4L USM | EF 24-105mm f4L IS USM |

| Tamron SP 150-600mm f/5-6.3 Di VC |

| Primes | 17mm Region | 24mm Region | 28mm Region | 35mm Region | 50mm Region | 100mm Region | 300mm Region | 400mm Region | 600mm Region |

| EF 180mm f3.5L Macro + TCs | EF 200mm f2.8L II + TCs | EF 300mm f4L IS USM + TCs | Tests With the Kenko TC | Macro Tests |

| Filter Effect on Sharpness and AF | Flare and Filters | Image Stabilizer Tests | TC Sharpness | Best of the Best | Method | Understanding the Results | Real World 100% Crops | Tripod Tests | L-Plate Test |

Lens Tests

Table of Content

Overview

Results

Interpreting the Results

Primes

Zooms

Groups

Specials

Summary Table

 

[Updated with the EF 180mm f3.5L Macro USM and Tamron SP 150-600mm f/5-6.3 Di VC]

Overview

The motivation for these tests is mostly acceptance testing of new lenses with secondary objectives of personal information and the potential of future health checks of lenses.

The Method, Target and Analysis Software

For those that want to do simple subjective tests, a method is outlined here.

 

Results

The results are organized in a top down structure with less detailed information and a greater degree of abstraction at the top level and more detail at a lower level.

So a zoom lens will have a summary of all focal lengths against f-stop. A menu at the top of that page links to a more detailed set of data for each  focal length. 

Then each focal length page has a menu at the top which links to each f-stop at that focal length. The f-stop page has fully detailed data and a link to a 100% crop page.

It should be noted that although the results are labeled with MTF the results are not referenced to the source target contrast as it should be. Like IMATEST results the ISO derived data is referenced to the average contrast of each region of interest. The USAF derived results are referenced to the contrast of the USAF reference region with the largest contrast over all the measurement regions.

In the case of the MTF 50% value often reported in internet test, these results display the value in line pair per millimeter and the often reported lines per picture width (height).

In all cases the for, maximum apertures of f5.6 or above, centre point auto focus point was used to focus the lens. Because the Canon AF system can stop anywhere in the maximum aperture depth of field some variation in test results may result. This issue could be reduced by multiple tests and selecting the best or using manual focus and a split-image focusing screen, I have not gone to that amount of trouble.

For most lenses some asymmetry in lens sharpness can be experienced, particularly at maximum apertures although this should largely disappear as the lens is stopped down. These effects are probably only noticeable in real world use for wide angled lenses where the soft areas may become very soft.

In these tests I have used my standard capture sharpening, similar to Canon's recommendation of USM R0.3, T0, 300%. The actual sharpening is a layer edge masked USM R0.7, T0, 300% blended at 65%, this provides almost exactly the same amount of sharpening pixel-for pixel as Canon's recommendation but with the advantages of edge masking and layer blend protection. Spectral analysis indicates this level of sharpening is about right for correcting the typical anti-alias filter.

If sharpening is not used all the tests will be doing is measuring the camera's anti-alias filter. A comparison of sharpened and unsharpened is given here. As can be seen this only affects sharpness (MTF and MTF50%) and not chromatic aberration. It seems my results come up slightly sharper than the PhotoZone results probably due to differences in sharpening, also my corner results are probably more "in the corner" that PhotoZone assuming they use the ISO 12233 target, so direct comparison is probably difficult. Additionally my tests are done at much closer distances and so are not directly comparable for that reason.

Judging from these results the maximum MTF50% seems to be about 60 lp/mm limited by the camera anti-alias filter. Considering the 20Ds Niquist limit is about 78 lp/mm this is quite impressive and surprisingly close to theoretical estimates

The results have been compared to the supplied Canon MTF data with some success here.

I have now adopted a best of 3 sets of results to try to reduce the effects of non-ideal focus.

See this page for tests with the Kenko 1.4X Pro,

See this page for tests at macro distances.

An example of test repeatability is shown here.

 

Interpreting the Results

If you have difficulty interpreting the abstract numbers you can always refer to the 100% crops that will be viewable from each f-stop result page. Look for the "View Crops" link as the lowest menu item.

The page linked below contains subjective judgments for various sharpness measurements from real lens tests from 100% crops and printed output.

Understanding the Results

 

Primes

EF 28mm f1.8 USM

This lens has good centre sharpness but poor edge and corner sharpness, additionally there are high levels of chromatic aberration. However the CA is easily correctable to similar levels as the 17-40 f4L in a digital workflow having a resulting improvement in APS-C corner sharpness of about 10%. The measured MTF data is a good match to Canon's published MTF so it does what it says on the tin.

Having said this in an A4 print the difference in corner sharpness between the EF 28 f1.8 and the 17-40 f4L @ 28mm is not that obvious at f8, it can be seen but you have to look for it. Comparing the stopped down zoom to to the prime wide open makes the corner sharpness easier to observe but the prime is still producing a reasonable image for this size print. This just shows how demanding pixel peeping is compared to moderate print display requirements.

It is worth noting that the performance of the EF 28mm f1.8 USM on APS-C suggests a more disappointing performance on full frame than is likely. Examination of the Canon and Photodo MTF curves for f8 30 lp/mm shows a dip around the corner of the APS-C frame that recovers for the remainder of the coverage of full frame.  

 

EF 50mm f1.4 USM

A very sharp lens, slightly soft wide open but soon sharpens up. (Results updated Jan 07; The originally posted result were prior to the practice of multiple AF instances. When this was applied one of the new results was better and so replaced the old results.)

 

EF 100mm f2.8 Macro USM

A very sharp lens. Retested with best of 3 focus locations. 

 

EF 180mm f3.5L Macro USM

Centre sharpness is very good by any standards but not as blisteringly sharp at the 100mm macro USM or 200mm f2.8L II, the 300 f4L IS is also sharper except above f8. The same comments apply to edge and corner. Sharp A3 prints should be possible wide open. The general characteristic flattens off early at about f5.6 indicating the AF accuracy of the lens may be the limiting factor. The lower left corner of this lens is slightly soft wide open indicating some mild decentering,  improving as stopped down finally recovering at f8.

 

EF 180mm f3.5L Macro USM + EF 1.4X II

Obviously not as sharp as the 200mm + 1.4X or the native 300mm, the performance is reasonable. Centre sharpness is close to the 200mm +1.4X from wide open to f8 but beyond this the 200mm +1.4X  pulls away and the extended 180mm falls off slowly. Well stopped down f11-f16 performance is comparable to the 300mm in the centre as the 300mm is clearly optimized for close to max aperture. Edge and corner sharpness of the 180mm is similar but slightly more at a disadvantage once again on a par with the 300mm around f11.

 

EF 180mm f3.5L Macro USM + EF 2X II

The performance with the 2X is very poor worse than the 300mm with 2X, but this may be due to imperfect manual focus.

 

EF 200mm f2.8L II

A very sharp lens. Retested with best of 3 focus locations. Sharpest at f11.

 

EF 200mm f2.8L II + EF 1.4X II

Very sharp performance from one stop closed. Wide open is slightly soft. Chromatic aberration is significantly increased at slightly less than twice the value of the prime on its own.

 

EF 200mm f2.8L II + EF 2X II

The higher TC factor is taking its toll here. Optimum performance is about one stop closed but defiantly soft wide open. Resolution is reasonable in the centre and not to bad at the edges, the corners remain a little soft. On the plus side chromatic aberration is well controlled and better than with the 1.4X.

 

EF 200mm f2.8L II + EF 1.4X II+ EF 2X II

With the version II teleconverters it is possible to stack them, 1.4X first on the body. The body ignores the 1.4X and a max stop of f5.6 was reported although it was really f8. So with the 200mm AF was still available and used in this test. Using AI Servo illustrated the under-damped response of the AF system under these conditions; with the AF activated the distance scale could be see to overshoot the correct distance and over a period of about a second chatter either side eventually homing in on the correct focus. In the field this would result in a number of images being poorly focused and can be expected of any situation where the actual effective aperture is not reported to the body.

Performance wise as might be expected is disappointing with a very soft centre performance and extremely soft edge and corner. Overall the contrast was poor.

 

EF 300mm f4L IS

Very sharp, peaking up one stop closed but falling off fast from f11. Good CA. Tested with best of 3 focus locations.

 

EF 300mm f4L IS + EF 1.4X II

Very usable peaking up one stop closed but falling off after that. Increased CA. Tested with best of 3 focus locations.

 

EF 300mm f4L IS + EF 2X II

Reduced but still usable sharpness, good one and two stops closed. Similar CA to with 1.4X. Note tested with manual focus on ground glass screen as auto-focus is disabled on a non-pro body. Tested with best of 5 focus locations. (NB There is a trick to tape over 3 pins in the teleconverter to retain auto-focus, this did not work well in moderate to poor light and hunted constantly never obtaining a useful AF lock, it might work better in bright sunlight).

 

EF 300mm f4L IS + EF 1.4X II + EF 2X II

With the version II teleconverters it is possible to stack them, 1.4X first on the body. The body ignores the 1.4X and a max stop of f8 was reported although it was really f11. Manual focus was again required.

Performance wise as might be expected is disappointing with a very soft centre performance and extremely soft edge and corner. Overall the contrast was poor.

Zooms

EF-S 18-55mm f3.5-5.6

Centre performance is good and edge not bad. Performance in the corners is falling off. Performance drops of at the long end. Optimum performance around f8-f11 although the wide end sharpens up a little quicker. Chromatic aberration is fairly high but this easy to fix in Adobe RAW. In this example (just one aperture at 24mm) the corner CA has been corrected by eye. This has reduced CA to about half the original value with a very slight increase in MTF 50%.

 

EF-S 10-22mm f3.5-4.5 USM

Centre is fairly sharp wide open but corners need f8-f11 to perform. Chromatic aberration is not bad except for the corners where it gets a bit high. The corner sharpness at 12mm seems to be poor, this seems to be due to poor sharpness in the lower left corner. The left edge is also an outlier at this focal length. Checking all focal lengths we see this side of the lens is always the worst performer although the severity of the problem reduces at longer focal lengths. Surprisingly 17mm is better than 12mm. These problems were not obvious in normal usage even in retrospect. It may be that the very close target distance at these focal lengths is showing up an issue that may be more to do with flatness of field rather than sharpness.

 

EF 17-40 f4L USM

A sharp lens but of course not as sharp as the longer primes. Quite sharp wide open but improving considerably in the corners one stop closed. Almost as sharp as it gets one stop closed at 17mm, however progressive zooming to tele requires 2 to 3 stops of closure for ideal sharpness; this may be due to non-ideal focus with the f4 maximum aperture under these test conditions. Similar sharpness throughout the zoom range with CA well controlled for a zoom in this focal length range, again improving significantly one stop closed. Some indication of flatness of field issues in the multiple focus tests (not shown).

 

EF 24-105mm f4L IS USM

A sharp lens performing very well against other lenses covering this focal length but suffering from slightly high CA at the widest end of the zoom range. This copy is about where Canon specify at the wide end but seems to have better fine detail than specified at the long end allowing it to keep up with the excellent 100mm f2.8 Macro USM in all but the corners. 

 

Tamron SP 150-600mm f/5-6.3 Di VC

A very impressive lens, centre sharpness at 400mm and above is better than the EF 300mm f/4L IS + 1.4xII. Corner shapness on APS-C is good, softening slightly at 600m. CA about 0.5 to 1.5 pixels.

Performance is still good below 400mm, slightly suspect wide open (f/5.0) at and below 200mm but recoveing well by f/5.6. Corners a bit soft on APC-C but recovering to reasonable by f/8. CA < 1pixel at 300mm and < 0.25 pixel at and below 200mm.

 

 

Groups

Primes

All the primes are very sharp. The 200mm and 100mm macro have the highest peak sharpness. This confirms experience in real life with these lenses. The 300mm has similar sharpness to the 50mm but over a limited range of stops, additionally sharpness drops off fast for the 300mm above f8.

The EF 28mm f1.8 centre sharpness is very close to the 50mm until f4 after which the 50mm takes over. For edge sharpness the 28mm catches up to the 50mm at f2.8 but after this point again the 50mm is sharpest. The 28mm corner sharpness never catches up with any of the other lenses.

 

17mm Region

The 10-22 and 18-55 are quite close, the 10-22 being better stopped down except for the corners. However the chromatic aberration of the 18-55 lets it down. The 17-40mm offers a clear improvement in general over both with much better resolution and about half the chromatic aberration. However, the maximum centre sharpness is very similar between all three lenses, although the 17-40 takes the day otherwise. 

 

24mm Region

Although slightly short of the 24mm mark the 10-22 is clearly better than the 18-55 even in the corners. The 17-40mm offers a clear improvement in general over both with much better resolution and about half the chromatic aberration. However, the maximum centre sharpness is very similar between the 17-40 and the 10-22. However the 24-105 takes the day otherwise in the centre and ties at the edge with the 17-40, the 17-40 pulls back the pole position for the corners. 

 

28mm Region

The 17-40mm is much better than the 18-55mm. The EF 28mm f1.8 is slightly sharper in the centre than the 17-40mm up to f11. The 28mm matches the 17-40mm edge sharpness to f8 with CA correction. In the corners the 28mm never catches the 17-40mm but is generally superior to the 18-55mm. In the centre the 24-105mm is only slightly ahead on the 28mm f1.8 until f8 but in the edge and corners the 25-105mm is a clear winner.

 

35mm Region

The 24-105mm wins over the whole frame.

 

50mm Region

As expected the 50mm prime wipes the floor with the 18-55. The 17-40mm does not catch up to the prime until f11 but is much better than the 18-55. The 17-40 CA levels are close to that of the prime. After f5.6 the 24-105mm outperforms the 50mm f1.4 in the centre and edge and beats the 50mm in the corner. 

 

100mm Region

The 24-105mm ties with the 100mm macro in the centre, in the edge the 100mm macro leads until f8, in the corners the 100mm macro leads all the time. This is a first rate result for the 24-105mm, comparing the measured performance to Canon's MTF data shows the zoom to be performing above specification.

 

300mm Region

The 300mm f4L IS is much sharper at wide apertures than the 200mm f2.8 + 1.4X. However, at f11 and below the 200mm f2.8 + 1.4X is sharper. This is a very unexpected result and seems to be due to the rapid fall off in sharpness of the 300mm f4L IS at smaller apertures. 

 

400mm Region

The 300mm f4L IS + 1.4X is much sharper than the 200mm f2.8 + 2X although the latter is good.

 

600mm Region

This compares the 200mm f2.8L II with double stacked TCs (EF 1.4X II and EF 2X II) and the EF 300mm f4L IS with the EF 2X II. The double stacked 200mm clearly can't substitute even in an emergency for the 300mm with 2X.

 

EF 200mm f2.8L II + TCs

The main effect of the 1.4X is to increase chromatic aberration, as far as sharpness goes the lens clearly has a lot to give showing as almost as sharp as the basic lens. The 2X strains the lens rather more although performance is probably still good on the centre and not too bad at the edges. However the corners are rather a little soft. The difference between centre sharpness of the 1.4X and 2X is about 10% and 15-20% in the corners at optimum apertures suggesting the 2X is better than using the 1.4X and cropping. Double stacking the TCs is surprising not too rubbish in the centre but dreadful at the edges and corners.

 

EF 300mm f4L IS USM + TCs

Both TCs are optimally sharp about a stop closed. Both the 1.4X and 2X cause loss of sharpness but remain usable. This suggests the lens' inherent sharpness is not quite as good as the 200mm f2.8 which was showed no real loss of image quality with the 1.4X and is still good with the 2X. Double stacking the TCs is dreadful at the centre, edges and corners showing considerable drop off from the 2X only performance..

 

Specials

An investigation into Filter Effect on Sharpness and AF; conclusion is none.

Test on the Image Stabilizer function of the EF 300mm f4L IS USM and EF 24-105mm f4L IS USM.

Looking at TC Sharpness relative to just up sampling the image.

See this page for tests with the Kenko 1.4X Pro,

See this page for tests at macro distances.

 

Summary Table

The following table summarizes the lenses at various focal ranges, the relative sharpness at different focal lengths can be compared with the Best of the Best plots. It is interesting to note that the sharpness of the 10-22mm @ 10/14mm is similar to the sharpness of the 300mm f4L IS with the EF 1.4X II and EF 2X II teleconvertors.

Focal Length Centre Edge Corner
17mm 18-55 @18, 10-22 @17 and 17-40 @17 are close but the 17-40 is best 17-40 best particularly from f5.6 17-40 best particularly from f5.6
24mm 24-105 24-105 and 17-40 tie 17-40 but 24-105 close
28mm 24-105 and 28/f1.8 very close to f11 with zoom in lead 24-105 24-105 with 17-40 close
35mm 24-105 24-105 24-105
50mm 50/1.4 until f8 then 24-105 50/1.4 until f8 then 24-105 50/1.4 until f8 then 24-105
100mm 24-105 and 100 macro tie 100 macro best until f8 but 24-105 close 100 macro best
200mm Only 200mm f2.8L available Only 200mm f2.8L available Only 200mm f2.8L available
300mm Amazingly 200mm f2.8L II + EF 1.4X better after f8 but before 300mm f4L IS Amazingly 200mm f2.8L II + EF 1.4X better after f8 but before 300mm f4L IS Amazingly 200mm f2.8L II + EF 1.4X better after f5.6 but before 300mm f4L IS
400mm 300mm f4L IS + 1.4X but 200mm f2.8L II not far behind particularly from f8 300mm f4L IS + 1.4X but 200mm f2.8L II not far behind particularly from f8 300mm f4L IS + 1.4X and 200mm f2.8L II more or less tie

 

Last Updated 23/08/2014

All Content 2005-16 Lester Wareham All Rights Reserved     

All material is supplied  as is and without warranty, use at your own risk.  

All opinions stated are the authors own.

All quoted information remains the copyright of the respective authors.

Any trade marks mentioned are the property of their owners.

  

Contact: 

Note this email gets spammed a lot so if you want to get your message through please prefix "Holding Tank" in the title

holdingtank@ware.myzen.co.uk