Canon EF-S 18-135 IS vs. EF 28-135 IS USM

[ErectedGryphon]

The EF-S will not work on a 5D mark 1 or 2, the lens body usually goes deeper in the cavity, causing the mirror to strike it and breaking the camera. In addition the care in producing the glass on the EF-S is not as stringent as the EF, the smaller sensor only uses the center of the lens, full frames use more of the glass so they have to be more diligent on the glass production. Thats one reason for the increased prices.

I don't agree with that. The EF-S 17-55mm and the EF-S 10-22mm lenses are very good. Many people say that they are on par with L zoom lenses, in terms of image quality.

Point, I never said they didn't make great images, what I meant was that the outer edges of the glass isn't used on a crop sensor camera. Therefore manufacturing isn't as stringent, which in no way effects the image quality, and it actually brings down the price of the lens.

Another feature of USM vs. most (not all) Non-USM, that we overlooked, is the internal focus. The front lens of USM lenses do not spin while focusing. With most Non-USM lenses, the front lens spins, which will require you to hold your Polarizing and Gradient filters during focus, that if not careful can cause resistance an wear out the focusing motor or strip gears.

EDIT: According to the specification on the Canon sight this particular lens has internal focusing, so not really an issue

 

[DB83]

What lens are you talking about?

 

[chip]

I would noy buy a Canon EF 28-135 IS USM on a crop frame Canon 7D. It is not wide enough 28x1.6=45mm - a normal lens. You will not like it and wish you could zoom out more.

The Canon EF-S 18-135 IS is better for a 7D, but wait, the new EF-S 15-85mm f3.5-5.6 is the best stay on lens for a crop frame Canon.

 

[usayit]

If we were talking Full Frame Sensor, I would say 72mm Diameter (28-135) is better than 67mm (18-135), but were talking EF-S so it may not be that big of a difference, since all the glass wont be used anyway.
...
Point, I never said they didn't make great images, what I meant was that the outer edges of the glass isn't used on a crop sensor camera.

I do believe the EF-S lenses have a smaller image circle... so I don't believe there is any correlation between either statement and overall image quality. I don't think filter diameter and element sizes are a correlation in any format/brand/mount. I think Mike was pointing that out. There are external and internal focusing designs in the canon line equipped with both USM and non-USM.

You can talk terminology and specs all day long... in terms of image quality the only decisions to made are done looking over samples.

 

[DB83]

I would noy buy a Canon EF 28-135 IS USM on a crop frame Canon 7D. It is not wide enough 28x1.6=45mm - a normal lens. You will not like it and wish you could zoom out more.

The Canon EF-S 18-135 IS is better for a 7D, but wait, the new EF-S 15-85mm f3.5-5.6 is the best stay on lens for a crop frame Canon.

The 15-85 is a 800 dollar lens which from what I've read isn't quite worth the money. If I'm going to spend that much on a lens, I may as well either get the 17-55 f/2.8L IS USM or the 24-105L. Still would lean towards the 24-105L though since I'm looking for more of a walk around lens. The telephoto lenses come later.

 

[ReoFlex]

I posted 4 comparison photos on a Flickr site that I set up. I am NO fan of Flickr. The site is confusing as hell to get going on. Smugmug is far and away a better site! But go here and hopefully you'll find the pictures of the test that I posted. There should be 4 double pics with the New lens on the left and the old lens on the right. You'll be amazed. Be sure to click on the photos and use the magnifier button above the pic in order to enlarge it for a better look.
Hopefully, this will get you there:
Flickr: reoneill1's Photostream

 

[ReoFlex]

The 28-135mm IS USM is a lackluster lens at best. It should be retired. If you're using it, you're not getting the photo quality you will get with any number of other newer Canon lenses, especially the newest ones. This lens is also NOT one you will use on a full frame camera if you purchase one. It is simply NOT good enough. You'll want better lenses for those. Same is true with a 7D. Here's a tip. Purchase what you need now, NOW. In the future purchase what you'll need then, THEN. Simple. Don't buy a lens today thinking about tomorrow. Things are changing so fast that you can never be sure that what you get today will even be applicable tomorrow anyway. Live and photograph for today. Address your present needs. Hell, you may not even be alive tomorrow. =)

 

[usayit]

The 28-135mm IS USM is a lackluster lens at best.

I think we heard you the first and second time. :er:

I am usually skeptical of anyone raving about any product.
I am usually skeptical of anyone ranting about any product.

I've shot with the 28-135mm IS and it wasn't "lackluster" as bad as you make it out to be. Please post your comparisons so the OP can see for themselves.

 

[DB83]

It's ok. I won't make my decision based on any one person's opinions.

 

[Derrel]

I went to Canon's official USA web site, to look at the MTF graphs for the old 28-135 and the new 18-135 zoom....but they have jiggered with the results on the new lens. I find this disingenuous on Canon's part, at best. I've been considering adding a 7D to my kit since I first heard about it, but you know,
the sensor's ultra-dense high-megapixel count with exceedingly small pixels means that lenses are going to have to be able to deliver truly *EXCELLENT* performance, or the sensor's high MP count will be absolutely wasted.

The second issue is that with the high MP count of the 7D, diffraction will probably set in at apertures LARGER than f/5.6. At 12 MP on 1.5x, diffraction starts cutting sharpness at f/5.6; apertures like f/8,using the 24-105L are looking not very good to me, based on this Canon 5D and 7D owner's side-by-side comparisons.

Last night I spent like an hour at Fred Miranda, trying to find out the honest truth about the 7D from actual owners. I have also been searching the web for full-sized images, and honestly, I think the 7D's performance is craptastic with anything less than Canon's very-best lenses, and then ONLY at larger apertures like f/4 to f/5.6. With zoom lenses in the consumer price, the wide-open apertures are not going to be optically good enough,and by the time the cheaper lenses are stopped down to f/5.6, the tiny pixel size will already be causing diffraction and a loss of image quality.

Here is the best comparison I found of a head-to-head lens for lens shot
showing the 7D's issues that concern me the most"
Image Quality: 7D vs 5D original ? - FM Forums
Excellent example of the 5D being sharper than the 7D on the same scene,both shots made using the 24-105L at f/8

Landcape with 24-105L at f/8 http://farm3.static.flickr.com/2531/3985223796_6e954783d7_o.jpg
ISO 1600 side by side http://farm4.static.flickr.com/3530/3984515563_76df58c0f1_o.jpg

 

[usayit]

Kinda sounds like a issue between IQ of camera bodies not IQ between two lenses as being currently discussed.

 

[Derrel]

Usayit wrote, "Kinda sounds like a issue between IQ of camera bodies not IQ between two lenses as being currently discussed."

No, go to Canon's web site and look at the truncated MTF graphs for the 18-135,and then look at the full MTF graphs for the 28-135 lens, and then maybe you'll understand that all Canon is showing at the MTF figures for the VERY CENTER of the new lens, while the older lens has a full MTF graph which shows itsgood center performance as well as its lousy performance out to the edges. What is Canon hiding? Looks to me like potentially atrociously poor edge performance in the newer 18-135 lens. Look at the MTF info they provide on the new lens...center figures only....hmmm. Why?

That was my first point about the 18-135 vs 28-135 debate.

The second point I made is about how demanding the 7D is,and how its performance on the 24-105L is not very good at f/8, and about the way the 7D's ultra-dense sensor will show deleterious diffraction effects at smaller apertures. The OP asked about two specific zoom lenses for use with a specific body, the 7D,and my comments ought to be construed to mean, " both the 18-135 and 28-135 are consumer-level zoom lenses,and will produce craptastic images on the 7D at small aperture like f/8, and also at wider apertures because said lenses are not very good wide open."

My third point was that consumer-grade zoom lenses are *unlikely* to look good on the camera in question, which is the 7D. And both the 18-135 and 28-135 must be considered "consumer grade" lenses. Looking at the poster above's three sample scenes on Flickr, it seems like the older 28-135 he owns has a serious problem with flare or veiling glare whenever shot toward even moderately strong light, but that the degree of resolution in both the 18-135 and 28-135 are reasonably close; it's hard to tell in such small samples, but one thing we need to keep in mind is that the newest cameras can be programmed by their manufacturer's to read the information from each,specific lens, and in the case of Nikon's higher-end cameras, and having the camera be aware of *specific* lens characteristics, the camera itself can remove chromatic aberration the lens has when it converts the data from the sensor. Since the 7D is Canon's first color-aware light metering camera, and it has Canon's very latest Digic image processor, I expect that it will be able to utilize every last bit of lens ability from the very newest lenses,and probably many older lenses.

I do not suspect the poster's motives or methodology, but I wonder if the camera itself might be optimizing the output of the 18-135 thru sophisticated image processing in its new Digic image processing engine,and Lightroom is merely accurately opening the file up almost optimized perfectly. What I see is lower contrast from the 28-135, and more veiling; I could boost the contrast up with a slight tweak if I wanted, so they two would look identical contrast-wise. It's hard to tell from such small samples as those above, but how the camera's own JPEG engine and other software interprets the sensor data has become pretty sophisticated these days.

When a camera "knows" the RGB components of the light, like the 7D does, and when the camera is fully aware of every lens's particular characteristics, the images ought to look really good. I'm not doubting the poster above and his side-by-side backyard test results: it is clear that *his* 28-135 has a flare/veiling problem when even weak backlighting is present against a predominantly dark field, like the ivy on his fence. Maybe the 28-135's rear element coatings are slightly hazy, or maybe the newer lens really has a much more flare/veiling-resistant optical design.

I'm still considering a 7D as a good value at $1600; I just want to see how it performs,and I've been searching all over for good FULL-sized samples,and what I am seeing is that this camera looks,well, like it demands lenses that have high MTF scores,and I don't considered the 28-135 to look good on that parameter, and Canon is hiding the 18-135's full-field potential on its web site. Still 28x 1.6 = 44.8 FOV, and I'm not too down with that. 18>28.8:thumbup:

 

[ReoFlex]

Derrel: Sounds like you're a more advanced user than I. Interesting what you said about the lenses. Maybe so. Perhaps my test was flawed as well, or just specific to MY 28-135 lens? Dunno for sure.
Usayit: I posted a link for you to see for yourself. Go check it out. Also, I've owned two 28-135 lenses, and borrowed my son-in-laws for this particular test. ALL of them have been "lackluster." IMHO. Three "flukes" maybe??? And talk about a "lackluster" macro lens? The 28-135mm IS USM (Macro) is the very definition of "lackluster" Give it a try and see for yourself. =)

 

[ReoFlex]

Derrel: I like what you said about the color aware light metering: "Since the 7D is Canon's first color-aware light metering camera, and it has Canon's very latest Digic image processor, I expect that it will be able to utilize every last bit of lens ability from the very newest lenses,and probably many older lenses." - Then perhaps this is why the colors were so very superior in the new lens images??? This really surprised me. Could the lens be tweaked to capitalize on this new camera ability? Or vice-versa? The images I put up were scaled down by Flickr, however, if you click on the image itself when it opens in a new window you can click on the little magnifier icon above the picture and it opens it in a larger rendition. Was only the second time I've posted images to Flickr, and am not sure of the best way to do it. Your stuff is good, I like what you have to say, your experience shows. Very helpful

USAYit: Sorry bud, I do tend often to drive a point I'm trying to make home a little "hard." Best I think to leave the final verdict to the observer perhaps?

 

[Derrel]

ReoFlex asked, "Could the lens be tweaked to capitalize on this new camera ability? Or vice-versa?" The answer is YES,YES,YES. Panasonic is doing simply amazing things with lenses in their high-end compacts, eliminating optical aberrations that exist within the lens, but which are processed OUT when the image data from the lens is handled by the image processor in the camera. DXO software can do the same thing in post on your computer with specific lenses in its database (dozens of lenses). Nikon's newer, better cameras can remove chromatic aberration from the CPU-equipped AF Nikkor lenses, based on the EXIF information and the lens ID information transferred electronically. Nikon Capture software can remove residual CA from phots, allowing lenses like the 10.5mm fisheye to retail for $699, not $3,500 with ALL the CA isssue addressed in the lens itself. Leica's new lenses also have a coding system that tells the M8 and M9 which specific lens is on the camera,for similar reasons, so the sensor info can be processed as good as is possible based on boatloads of data that is lens-model specific.

Every lens has its own,specific design and every lens of that type performs *almost* identically, unless an element is decentered, or the coatings get cleaning marks, or the lens is dirty. If you say three different examples of one zoom model performed badly,I believe it. Nikons 35-135mm f/3.5~4.5 AiS was one of the world's worst zoom lenses--just a horrific design. Making a wide to short tele zoom is fairly tricky--it crosses boundaries of wide-normal-tele. Kinda like a Swiss army knife-jack of all,master of none...

Your test showed me the flare and veiling glare performance clearly. Cameras are computers, and lenses feed data to processors and to software.