New 70-200 f2.8L IS compared to my old 28-135 IS - 2nd Batch

[Radtech1]

Again, these are full resolution crops without any manipulation, shot with my 10D, in Large JPG mode.

The camera was tripod mounted. I used a wired remote release to eliminate any camera shake from pressing the shutter release. Focal lengths were verified to be identical through the EXIF data, specifically, comparing 75mm to 75mm and 135mm to 135mm.

All exposures were done using only the center focus point, aiming at the window. Yes, I know that there is a fence there, but both the fence and the window were close enough to "infinity" that I used that anyway. All exposures were first set "out of focus" and the AF was allowed to find the focus itself.

To me, it looks as though that at 70mm focal length the images are just the same. My new $1700 lens shot just as well as the $350 one. ARRGH!At 135mm, the much cheaper 28-135 is clearly sharper. Note the patterns in the stucco and the white pipecap on the roof for the clearest example. DOUBLE ARRGH!!!

I asked the guy at Samys if there is any return if I am not happy with the results. He said there is no "test shooting" time allowed. Well, I am not happy. If it were you would you demand a refund, in spite to being told there was none? This is a VERY BIG bullet to bite.

28-135 at 70mm 125 at 6.7
http://users.adelphia.net/~radtech1/Photos/70test2X.jpg

70-200 L at 70mm 125 at 6.7
http://users.adelphia.net/~radtech1/Photos/70test2Y.jpg

28-135 at 135mm 125 at 6.7
http://users.adelphia.net/~radtech1/Photos/135test2X.jpg

70-200 L at 135mm 125 at 5.6
http://users.adelphia.net/~radtech1/Photos/135test2Y.jpg

[Tom W]

I can't dispute your hypothesis at this point - at 70 mm, f/6.7, the lenses are very much alike with a slight color difference (I can't tell which is more accurate since I haven't seen the actual scene).

At 135 mm, I noticed that the 70-200 was shot at f/5.6 while the 28-135 was shot at f/6.7. This may account for the slight change in sharpness, as well as the slight increase in brightness on the longer lens.

A few points or observations or whatever - First, the 28-135 is one of the best consumer-grade lenses out there. I've seen some call it the best "non-L" lens around. I don't have it, but if your copy is anything like my 28-105/3.5-4.5, its a keeper.

Second, the differences among lenses isn't all that great when stopped down a few stops, but become more apparent at the extremes, I.E., wide open or at f/4 or so. Considering the 28-135's good reputation, I'd expect it to perform very well stopped down a bit, though I wouldn't expect it to exceed the 70-200.

Third, don't use a filter for these tests on either lens. Depending on the quality of filter, some can cause flare on high-contrast subject matter due to their high reflectivity.

You've got some poor contrast on these crops - It might just be an exceptionally bright area of an otherwise more moderate scene, it might be flare, or it might be a hazy day. Its hard to tell here.

Just out of curiousity, where is the focus point in these shots?

[Radtech1]

No filters were used. The first shot of this series was the one at 5.6 - the reason that I chose that was because that 5.6 IS full open on the 28-135 and I wanted an "apples to apples" comparison. For whatever reason, the camera shot the remainder of the images at 6.7.

I am just really crestfallen. $1800 for the Legendary Canon Pro glass, and I was expecting sharper than a serpant's tooth. This is far from it.

I suspect it will be on EBay next week if Samy's does not let me take it back.

Rad

OH - focus point was the window under the AC unit on the roof.

[Tom W]

Yes, that's not encouraging. I would expect at least a modest improvement at f/6.7 and a larger improvement at wider apertures.

I think you ought to experiment a little further though - the subject matter in this test is rather hazy, which serves to mask the quality.

Will Samy's let you exchange for another lens? That one may just be a poor example.

[petiot]

Hi RadTech.

I agree. there is a clear advatage for the 28-135. And i understand that you feel a bit cheated on this purchase. I was very interested with your result. I am myself considering buying L lens, and i was looking everywhere for comparison between L and non L lenses. and to my surprise, there are non to my knowledge!!! nobody dared !! ;)

Your story throw a shadow of doubt, even though i have read so many "oh L Lenses are the best, no comparison possible with consumer lens" etc etc. I am also surprised by the lack of feedaback on this thread!! the explanation given are nowhere near to justify the poor perf of the 70-200 IS.

It confort me a litle bit more in my opinion that L lens are mainly worth for the built quality and the aperture they offer. but their price is by no means by the quality they offer and is only a pro "fashion". especially considering that anyway photo will be sharpen in PS! who do not ??? who dare saying they dont sharpen their photo!! ;-)

anyway, 200, f2.8, IS, you are a lucky man!! and i am sure oyu will overcome this first desillusion!

Dan

In fact i think that the lack of feedabck from people on this thread clearly shows (to me at least) that L lens are a

[Tom W]

Hi RadTech.

I agree. there is a clear advatage for the 28-135. And i understand that you feel a bit cheated on this purchase. I was very interested with your result. I am myself considering buying L lens, and i was looking everywhere for comparison between L and non L lenses. and to my surprise, there are non to my knowledge!!! nobody dared !! ;)

Your story throw a shadow of doubt, even though i have read so many "oh L Lenses are the best, no comparison possible with consumer lens" etc etc. I am also surprised by the lack of feedaback on this thread!! the explanation given are nowhere near to justify the poor perf of the 70-200 IS.

It confort me a litle bit more in my opinion that L lens are mainly worth for the built quality and the aperture they offer. but their price is by no means by the quality they offer and is only a pro "fashion". especially considering that anyway photo will be sharpen in PS! who do not ??? who dare saying they dont sharpen their photo!! ;-)

anyway, 200, f2.8, IS, you are a lucky man!! and i am sure oyu will overcome this first desillusion!

Dan

In fact i think that the lack of feedabck from people on this thread clearly shows (to me at least) that L lens are a

The lack of feedback is more likely related to the fact that the images take forever to show up - its hard to make a judgement on something that one cannot see. I had to refresh several times and save all 4 images to get to a point where I could make true side-by-side comparisons.

[SDK^]

Like others have said - at small Apertures (higher F numbers) the gap between normal lenses and L lenses is less.
If you could use the 28-135 at 70mm F3.5 and you were to compare that image to one shot on the 70-200L lens you would see the difference.

Remember - it's not just lens (image) quality you're paying for in the 70-200L 2.8, you're also getting a very fast lens that can be used in dark rooms.

[Pekka]

Most likely your focus points are not same in all tests.

Your test is not scientific but it still says what Canon's MTF curves say, too: near f8 both lenses are about the same in sharpness. If you value sharpness at 5.6 then you're bying the wrong lens. 70-200 f/4 would be equally good "daylight zoom" for less money.

What 70-200 f/2.8L (IS) is designed for is low light use - at 2.8 it can do very sharp images from edge to egde. This is what people pay for.

Other value in it is bokeh (out-of-focus quality), where is 28-135 is quite mediocre.

Then there are factors like focusing speed, focusing accuracy, flare control, same aperture in both ends of zoom, IS speed, color accuracy not to mention reliability and weather protection.

But I do think that your copy of 70-200 may not be the best one out there, or then it is true that IS version is not as sharp as non IS. My 70-200 (non IS) it tack sharp, even with 1.4X extender.

[maderito]

I don't get it. Why spend $1000+ for a fast lens and then worry about its performance when stopped down?

You can get the 70-200 f/4 L for a lot less and virtually the same optics. As Pekka points out, the 70-200/f2.8 was meant to meet the needs of those who need fast focusing, image-stabilized, large aperture zoom lenses that have good to excellent quality across the entire image frame, at all apertures, and at all focal lengths. That's asking a lot - and that's why you pay a lot.

I am happy with my decsion based on the costs/benefits: I have both the 28-135 IS and the 70-200/f4. They serve different needs very well.

[drisley]

SOmebody else noted the same thing when comparing the 300d kit lens to the much vaunted 50mm F1.8 prime.
However, like your pictures, the shots were taken stopped down.
When comparisons were taken with lenses wide open, the difference was obvious:

http://photography-on-the.net/forum/showthread.php?t=33438&highlight=

[Tom W]

That is true, drisley - I did a quick comparison between an "L" zoom and my trusty 28-105/f3.5-4.5 zoom at f/4.0, 50 mm:

http://home.comcast.net/~trwilk3/Images/28_105_f4_topcrop.jpg

http://home.comcast.net/~trwilk3/Images/24_70_f4_topcrop.jpg

I don't think that its hard to spot which one is which. And this is a relatively easy, low-contrast subject.

As you stop down, the differences become less apparent.

As manderito said, shooting at middle-apertures won't really show the differences between the lenses, at least as far as sharpness is concerned. But when there is a need for low-light shooting, extreme contrast, or the potential for flare, the "L" zooms rise to the occasion.

[tannoy]

One other possibility is that your 70-200 may not be in calibration. I recvently took my 10 D and all of my lenses to Canon for service and I was told by them that every lens has "electronic trim" meaning that the ideal focus point can be adjusted. I am sure this is of little help after plunking down the $ for such a lens but every company puts out bad product. I also own the 28-135 is and a 70-200 F4L. I really enjoy both lenses for different reasons but my 70-200 has much better edge to edge sharpness than the 28-135, the 28-135 is however one hell of a cheap lens...

Cheers,

Darrin

[karusel]

Wait, wait, wait, wait... you guys are saying that fast L glass is in terms of sharpness inferior or equal to consumer grade glass when stopped down more than 5,6???? Well I'll be damned...

[IanD]

Tom,
John Coltrane.......kool!

[Tom W]

Wait, wait, wait, wait... you guys are saying that fast L glass is in terms of sharpness inferior or equal to consumer grade glass when stopped down more than 5,6???? Well I'll be damned...

Ummm... no. Equal to or better, but the differences, at least in sharpness, dimiinish as you stop down. This is highly dependent on which consumer grade lens is being discussed, though.

[Malaxos1]

I was considering getting the Sigma 70-200 2.8 but found that the sweetspot on those lenses are at around f8. So I figured that if the lens was not what I wanted at 2.8 then I should get the 70-200L f4 being that it was sweet throughtout. I don't think the quality is going down as the apeture is being dcreased, I just think that the apetures that the test is being done at is the sweet spot of the the non L lens. However said that people buy the 70-200L IS for the sake of speed is correct. While I don't have the 2.8, I do shoot with my f4 wide open in most situations...Dean

[defordphoto]

Anyone see a ruler around here?

[msvadi]

Tom,
John Coltrane.......kool!

The first thing I noticed ;) Where is a copy of Love Supreme? ;)

[maderito]

Anyone see a ruler around here?
Jim - Actually, you need a scale. . .

Lbs., Price, Lens Model
0.82, $230, 28-105
1.10, $410, 28-135 IS
1.56, $580, 70-200 f/4
2.80, $1,140, 70-200 f/2.8
3.20, $1,649, 70-200 f/2.8 IS

Price = 323*Wt^(1.33)

:shock:

(B&H specs and prices)

[Tom W]

Anyone see a ruler around here?

Don't make me get out my ruler picture!! ;)

[defordphoto]

Ahhhh...found one...

http://psephos.adam-carr.net/libya/gaddafi.gif

[Leszek]

I remember doing some MTF comparison (from Photodo data) between Canon 28-135, Sigma 70-200/2.8 HSM and Canon 70-200/2.8.
If memory serves me well, 28-135 peak performance was about 70 mm at f/8. It was as good (or better) as the Canon and Sigma, so no surprise here if the pics are similiar.
A bit of surprise at 135 mm, where 28-135 gets softer than at 70 mm. Also, Canon 70-200 improves as it gets longer (again getting a bit softer at 200 mm).
So, IMHO at 135 mm the 70-200 should be a clear winner.
Well, look at it this way: if you are happy with the pictures from 70-200, then maybe your 28-135 is just a very good, sharp example of the lens ?

[Tom W]

Ahhhh...found one...

http://psephos.adam-carr.net/libya/gaddafi.gif

ROFL! I can't touch that one. :D

[karusel]

Leszek, this might be the answer to the mystery... the 28-135 is at 70mm at its best, and 70-200 at it's worst. Makes sense to me.