Canon EF 70-300mm f/4.5-5.6 DO IS USM comparative test

I am a firm believer that the real test of lenses is to take photographs of the subjects that you love. If you are happy with the photos, that is all you need to know about a lens. Taking photographs of test charts has never appealed to me. Life is too short. However, after taking a lot of real world photos with this lens, I began to wonder just how it compared to my other lenses. I had hoped it would be a lot better than the Canon EF 75-300 IS and maybe nearly as good as the Canon EF 100-400 IS but the photos didn't always look like that. Bob Atkins put out an appeal on photo.net for anyone who could do a comparison test. I had the lenses. I had the time, so why not.

I set about devising fair tests for the lens, comparing it to my 75-300 IS, 100-400 IS and 200 F2.8 lenses. The questions that were raised on discussion groups after my first review included bokeh, dreaminess, flare, and distortion, as well as resolution in dim light. All these issues are examined in the following tests.

Low contrast resolution

The first test was of a low contrast subject, a detailed colored map 22 inches by 15 inches, taped to a flat wall in a room lit by dim diffuse daylight. The camera, a 10D, was mounted on a sturdy Slik tripod 24 feet away in a dark corridor, filling the frame with the subject at 300mm. Exposures were taken with each lens, using the self-timer, at full aperture and F11. Focus was manual throughout and for the wide aperture shots, focus bracketing was used with the sharpest picture from each lens being used for the comparison to avoid mis-focussing biasing the results. No Unsharp Masking was used in any of the tests.

The major comparison, and the one shown here, was of all the lenses at 300mm, at both F5.6 and 11. Tests were also done filling the frame at 70/75mm for the Canon 70-300 DO and Canon 75-300 IS and at 100mm for all three lenses again at full aperture and F11.

250x500 pixel crops, corner of frame f5.6

250x500 pixel crops, centre of frame f5.6

250x500 pixel crops, centre of frame f11

Crops were taken from the center and lower left corner of all the images. Matching crops from all three lenses were tiled in groups on the screen ( Mitsubishi Diamond Pro) and magnified for visual comparison.

In the low contrast detail test the 100-400 IS was the clear winner at 300mm. No surprise there. It is an L-series lens and was not fully zoomed. The comparison between the Canon EF 75-300 IS and Canon EF 70-300 DO was less clear. At F11 both showed a little less detail than the 100-400 IS but it was not possible to choose a winner. Essentially they had similar resolution. At f5.6 however the 75-300 IS had a clear advantage over the 70-300 DO.

After compressing the files for web use I noticed that the file sizes were different. When there is less detail in an image, the compressed file will be smaller, so did these measurements confirm my own visual observations?

For the center crop the file sizes were as follows :-

High contrast resolution

The second test was of a high contrast set of lens test charts photographed indoors as described above. I mounted three test charts on a board 33 in x 23in. The portion shown is a cropped part of a Norman Koren test chart printed at 2880dpi positioned in the top right corner of the frame. Manually focussed shots were taken with the 70-300 DO, 75-300 IS, 100-400 IS and 200 F2.8 all at 200mm, at apertures of F5.6, 6.7 and 8. This target was also photographed outdoors with the camera handheld in bright light.

The high contrast test showed the lack of of chromatic aberration in the Canon EF 70-300mm DO which gave cleaner boundaries between black and white areas. Before compression for the web the 75-300 IS shots showed substantial chromatic aberration along these edges but this is difficult to see in the compressed image. The 70-300 DO improved dramatically as soon as it was stopped down, even by half a stop. At F6.7 and F8 it left the 75-300 IS behind and produced resolution similar to the 100-400 IS.

The comparison with the 200 F2.8 was included to show how close to maximum performance these zoom lenses could get. All three zooms showed less contrast than the prime, as is shown by the brighter white in the prime lens shots. This is to be expected because of the larger number of elements in zoom lenses. The lower contrast of zooms can sometimes be beneficial in digital photography as it reduces the risk of highlight burnout.

Bokeh

The third test examined bokeh, the quality and smoothness of the out-of-focus image. I compared the 70-300 DO and 75-300 IS with my 200 F2.8, which I have always thought gives excellent bokeh. The camera was set on a tripod outdoors on a sunny day and I manually focussed each lens on the left petal of a pink poppy. Three sets of pictures were taken at different angles and distances to give a selection of out-of-focus backgrounds.

Half frame crops, f5.6 200mm

The set shown is a crop of one of the backgrounds and is typical of all the results. Bearing in mind that I showed target-like out-of-focus highlights from water and described the bokeh as 'broken' and wiry in the previous review, I was surprised that under normal conditions, all 3 lenses, when set at 200mm and F5.6, gave very similar bokeh. The 200 lens opened up to F2.8 gave me the smooth bokeh I know and love but that is not possible with the slower lenses.

Interesting bokeh, half frame crop, f8 110mm, sea pinks out-of-focus in foreground, sparking sea in background

The test for bokeh suggests that most of the time there will be little noticeable difference in the quality of the more distant out-of-focus areas with the 70-300 DO. However, as demonstrated before, areas that are just out of focus can appear 'broken' and when photographing into the light, especially when there are specular highlights such as from sparkling water, the bokeh will look very different.

Flare

When I sent the first review to Bob Atkins, he asked about the flare pictures I had included. He pointed out that the lens hood is much less effective at 300 than at 70 and asked if the flare was worse than the 75-300 IS. The reason I included the flare shots in the first review was not just because it was worse than the 75-300 IS but because of the nature of the flare. I had never seen that type of flare in any other lens. It had a prismatic quality. When I moved my eye from side to side across the viewfinder, the flare changed color, like a rainbow. I set out to try and demonstrate this effect and compare it with other lenses in identical circumstances.

Full frame, f5.6 200mm, bottom row after auto-contrast applied

For this test I set up my 10D on a tripod and photographed elder flowers and leaves, into the light, with three lenses, the 70-300 DO, 75-300 IS and 200 F2.8, all at 200mm. All three lenses, even the prime lens, showed substantial veiling flare which lowered the contrast in the pictures. I have found that this loss of contrast is usually retrievable using Auto-contrast in PhotoShop Elements. The lower three pictures in the flare comparison show the improvement when this was used. With the 75-300 IS and 200 F2.8 lenses, usable results were obtained. With the 70-300 DO, the strong colored flare meant it was not possible to retrieve a usable picture. In the 70-300 DO picture the concentric circles of the Diffractive Optic are clearly visible in the colored area of flare.

Above is a 200x200 pixel crop from the shot with the Canon EF 70-300, showing the concentric circles from the diffractive optics visible in the flare pattern.

The flare test shows that the Diffractive Optics can cause substantial flare when shooting into the light. Care has to be taken to avoid this but good technique, such as always using the lens hood and shading the front of the lens can help to minimize it.

Distortion

Pincushion distortion was measured at 300mm. Both 75-300 and 70-300 were very similar.

Verdict on the Canon EF 70-300 DO

Pros

Excellent focus and stabilizer performance.

The small size means that you are more likely to carry it with you all the time and get shots of unexpected subjects.

It would make a good travel lens because of the substantial reduction in length made possible by the DO optics. An ideal travel lens set, in my view, would be a 17-40, a 50mm F2.5 (my choice) and the 70-300 DO. On a 10D or Digiral Rebel that would be equivalent to 27-480mm.

For candid street photography, of the type I have done in India and China, the 70-300 DO would be a great companion. The small size, black color and fast, silent focus would make it a winner. Even at 300mm the size of the lens gives little clue to the magnification being used. The stabilizer should allow sharp results even in low light situations such as indoor markets.

The 70-300 DO is also good for moving subjects. I have got sequences of birds in flight which are among the best I have taken.

It might also be good for romantic shots. Wedding portraits and high-key nudes would all benefit from the 'dreamy' qualities when used at full aperture.

Cons

There is an unpredictability about the results from this lens. I have had great results, sharp punchy pictures, and also soft hazy pictures. These tests show that the lens is weakest at full aperture at 300mm, especially with white subjects, and also into the light.

In the comparison shot of the white flowers, although the 70-300 DO lacks the chromatic aberration of the 75-300 IS, the results still look less sharp because of a white haze around the subject. The effect looks like a mild diffusion filter has been used, overlaying the sharp image with a soft one. This 'dreamy' quality appears when scattered (diffracted?) light submerges the normal definition of the lens. The 75-300 IS is one of the lower grade lenses produced by Canon. It is basically the same optical design as the much cheaper non-IS version, with an IS group added. There are reports that recent copies of this lens are better performers than the earlier ones but mine is one of the earliest ones. If full aperture performance at 300mm is really important to you, save money and buy the cheaper 75-300 IS.

With such an effective image stabilizer, and working in aperture priority, it would be easy to avoid using the 70-300 DO at full aperture. Even half a stop down from full aperture produces a noticeable jump in quality. The results available from the 10D at ISO 200 and even 400 are so good that using slow lenses stopped down is quite practical in most lighting situations but should you have to avoid full aperture on œ1,000 lens to get good photos?

The magnification of the 70-300 DO is noticeably less than the 75-300 IS. For example you need to zoom the 70-300 DO to 90mm to get the same image size as the 75-300 IS at 75mm at minimum focus. This is because the 75-300 IS can exceed the quoted focal length because of front cell focussing. The 75-300 IS can reach the equivalent of 330mm at maximum focal length depending on the focus position.You also lose half a stop at the minimum focal length, F4.5 instead of F4 compared to the 75-300 IS.

Conclusion

In real world photography, with normal subjects taken in good light, most people will be very pleased with the results from this lens. The resolution can be close to that of the 100-400 IS but with a much smaller, easier to handle lens. The excellent image stabilizer performance, coupled with the good balance of this lens allows sharp photographs to be taken hand-held at very low shutter speeds.

I have a photo taken with this lens as the wallpaper on my computer at the moment and every time I switch on the computer I marvel at the feather detail in the side-lit bird.

The major advantage of this lens, the small size, is only possible because of the DO optics. The disadvantages of this lens, prismatic flare, dreamy quality at full aperture and different bokeh are all caused, in my view, by the DO optics. Only you can decide on the balance of the trade-offs. I hope these tests will allow you to make a better informed choice.

Remember, photographs taken by a lens you have with you are ALWAYS better than photographs taken by a lens you have left at home.

(As ever, the findings only apply to a single example of each lens.)

Where to buy the Canon EF 70-300mm f/4.5-5.6 DO IS USM lens

Purchasing the Canon EF 70-300mm f/4.5-5.6 DO IS USM lens through the following links helps to support photo.net.
In our experience Amazon.com offers great service, is trustworthy and has a generous 30 day return policy. They also have excellent partners that offer competitive prices.

FEATURED AMAZON.COM MERCHANTS
Product: Canon EF 70-300mm f/4.5-5.6 DO IS USM Lens for Canon EOS Cameras		List Price: $2,000.00
Vendor	Availability	Price	Buy via Amazon
Amazon.com	Usually ships in 24 hours	$1,178.92	More info or
Adorama Camera	Usually ships in 1-2 business days	$1,190.00	More info or
17th Street Photo	Usually ships in 1-2 business days	$1,190.00	More info or
J&R Music and Computer World	Usually ships in 1-2 business days	$1,190.00	More info or
Product: Canon EF 100-400mm f4.5-5.6L IS USM Telephoto Zoom Lens for Canon SLR Cameras		List Price: $2,400.00
Vendor	Availability	Price	Buy via Amazon
Amazon.com	Usually ships in 24 hours	$1,360.00	More info or
Adorama Camera	Usually ships in 1-2 business days	$1,360.00	More info or
Calumet	Usually ships in 1-2 business days	$1,459.99	More info or
J&R Music and Computer World	Usually ships in 1-2 business days	$1,360.00	More info or
RitzCamera	Usually ships in 1-2 business days	$1,549.00	More info or