17-40L vs 28 or 24 primes

I was just wondering if the 17-40L lens still beats out Canons primes in 28 or

24mm? This can be either the cheap or more expensive 28 or 24. I woudl guess

the 24L would beat it, but I've heard the 17-40 still may be better. I'm

thinking in terms of sharpness and contrast. Obviously the primes can open up

more. I was just debating whether it was worth getting one of these primes for

landscape while also having the 17-40.

As a note I use 35mm still, but have plans down the road to go full frame digital. So that is important as well. I already have a Rebel XT.

I have the 17-40 4L and 24 2.8. Although the zoom is very good--save slowness and plenty

of barrel distortion--I use the prime when I shoot indoors or need to go light. So I find it

useful. I've heard so many reports of greatly varying copies of both these lenses I' not sure

what the norm is. However both of mine are very sharp and nearly the same at 24mm F4.

Surprisingly, flare resistance is about the same as well.

 

To me the biggest disadvantage of the zoom is upgrading my polarizer. A nice 77mm one is

$150! I have plenty of 58mm ones.

I went with the 17-40 as a wide angle solution for the 10D. I prefer primes for their speed and weight, and strongly considered the 24 or 28 (the f/2.8 versions). But, I like the 17-40 a lot. The zoom is high quality, with great build, and it's capable of producing very sharp and contrasty images. It compares favorably with my other primes. I use it for landscapes and in a variety of other situations. The zoom can also be used effectively indoors with bounced flash, but I don't like using flash. Even if the images don't look "flashed", a flash means more weight on a camera.

 

I have a 35/2 that I often use for indoor/low light situations, but I think I would find a 24/2.8, or perhaps a 28/1.8 (USM!) more useful. That said, I'll likely stick with the 35/2; I'm certainly not going to get rid of the 17-40.

 

--tom

I have the 17-40 (and love it), but I'm toying with the notion of buying a prime in this range as well, for low-light use. I also have two primes in the range of my 24-105 (50 and 85).

 

I don't use primes enough to warrant the purchase of a 24- or 35mm f/1.4L, but I'm sorely tempted by the 28/1.8. I'm attracted to this lens because I don't want to buy another non-USM model, but the 28/1.8 seems overpriced for the average reported quality.

 

I suppose that if/when an opportunity arises where I *need* it, I'll go ahead and buy one. In the meantime, I think I'll wait, and hope (against hope?) that Canon will update/improve their non-L normal and wide angle prime line-up (24-50mm).

I'll join the club and say that I've also considered acquiring one or more of the inexpensive

wide primes to supplement my 17-40mm.

 

The 17-40 is an excellent lens and does, as others pointed out, produce good color and

contrast and is capable of very sharp images. Of course, sometimes at certain focal

lengths it can seem a bit soft in the very corners and (though it doesn't often affect me

much) there can be some barrel distortion.

 

I've thought about getting the 35mm f/2 and the 24mm f/2.8. The main question in my

mind is whether they would (or would not) provide any advantage in sharpness overall and

in the corners. I'm mainly interested in the answer to this at f/8-ish apertures (for

landscapes) but a comparison wide open would also be interesting.

 

I own the 50mm f/1.4 prime and use it when it is close to the right focal length for a given

shot because it is noticably sharper than my zooms. Would this hold true (even if to a

lesser degree) with the 24mm or the 35mm? (Oh, heck, tell me about your 28mm f/1.8

while your at it, too.)

 

Thanks.

I previously owned the 17-40L, and now own and use the 24mm f2.8, 28mm f2.8, and 35mm f2.0 on a Canon 5D. I liked the zoom. However, I noticed when printing some images taken around 28mm, f8, tripod mounted, at 13" x 19" that they lacked the microcontrast and "wow" factor that should have been there. So I compared it to my 35mm f2.0. For distant subjects (> 15 feet from the camera), the 35mm resolves substantially more detail than the 17-40L. So I then investigated it further, and purchased the 24mm, and found that the primes all out performed the zoom for distant subjects, but not necessarily close subjects. The contrast and color of the zoom may have been better, but I think it's actually just that the zoom causes underexposure for some reason. My copy of the zoom was a fine performer, but all of the primes (the 28mm f2.8 only by a small margin) beat it, at least in some circumstances. I think it's because the zooms are very complicated designs so that if they aren't performing perfectly, the image will be fine, but it will be missing microcontrast, but that's just my theory. Anyways, I plan on buying a 28mm f1.8 soon to replace my f2.8 (I prefer the bokeh of the f1.8). BTW, Castleman, on his website, states that the 17-40mm f4 outperforms various primes, and his tests are far more scientific than mine. Interestingly, Castleman states that his 20mm prime outperforms the 17-40mm f4, but most people regularly "dis" the 20mm prime. I think there's a tremendous amount of sample variation in the zooms, and I also strongly suspect that at least some samples are prone to inconsistent (mis)performance. There's apparently also sample variation in the 20mm prime, though I've see very little sample variation in the primes that I've tested (they're all excellent, if not superb). . . I've read of some people who complain for instance about the corners of the 24mm f2.8, but I tested two copies and, though there were some differences, the corners of both are/were very good by f4 to 5.6, and excellent by f8. I am very demanding, having come from a view camera background, and I can't figure out why people aren't raving about the 24mm f2.8. Maybe the complainers tested it on a 1DS II, where its limitations are more apparent? I don't especially like the bokeh of the 24mm f2.8, but it's usually decent at f2.8 - f4, and it's not really a high priority for a 24mm lens.

Howard,

 

I agree - the 17-40mm gave about 1/2 stop less exposure with flash on my 20D when compared to my 35mm/2.0 - I put it down to the difference between f-stops and t-stops (actual light transmission - as given on cine lenses).

 

Didn't think much more about it because I've now changed to the 5D, and don't use flash anyway.

 

I think the 17-40mm may have been optimised for close distances - it seems to equal the 35mm prime at close distances, but the prime is clearly better for distant detail.

You'll have to cross-reference these two multi-lens reviews, but Castleman tests all three lenses you ask about here:

 

http://www.wlcastleman.com/equip/reviews/17-40/index.htm

 

http://www.wlcastleman.com/equip/reviews/24_70/index.htm

 

(Click on his "Equipment review list" at the bottom of the linked pages to see tests of the 24L, the 28/1.8, the 24-105, the 24-70, etc.)

Roger -- I agree that it's probably the difference between T stops and F stops that accounts for the exposure difference. I was hoping to short circuit that discussion. I've often wondered whether Canon programs exposure information (i.e., T stops) into the CPUs for the cameras for its various lenses. I doubt it, but I think I've read somewhere that it may. . . Of course, our 17-40mm underexposure theory suggests not

Hey Scott, I own both a 17-40 and an EF 24/2.8, a 5D and a 350D, I'd be happy to do some tests of both on either body at varying apertures if you like - just let me know here.

 

I only recently got my 17-40 as an upgrade from a 10-22 EF-S, but from experience I've found the 24/28 to outperform that lens - particularly in low light situations, where small apertures and long exposures were used, I'm not sure why but I think my EF24 is a particualrly good copy.

Ben wrote: "Hey Scott, I own both a 17-40 and an EF 24/2.8, a 5D and a 350D, I'd be happy

to do some tests of both on either body at varying apertures if you like - just let me know

here."

 

Hey, Ben, that sounds very interesting. I'd sure be interested to see the results of such a test

since teh 24mm f/2.8 is a lens I've been interested in - and because I currently own the

17-40mm.

 

Dan

<cite>I agree - the 17-40mm gave about 1/2 stop less exposure with flash on my 20D when compared to my 35mm/2.0 - I put it down to the difference between f-stops and t-stops (actual light transmission - as given on cine lenses).</cite>

 

<p>I'd be more inclined to put it down to the difference in flash metering. The 17-40 returns distance information; the 35/2 does not. <a href="http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00GPLS">This can make a heck of a lot more than just half a stop's difference to E-TTL II flash metering (which your 20D has).</a></p>

The answer is "it depends". On what? On a lot of things. If you have a good 24/28 f/2.8 they would beat the 17-40L. I love the quality of the pics from the 17-40L, the color and contrast are very nice. My 24/2.8 was a weird lens. It was ok on film, but on DSLRs it was as good (bad?) as the cheap kit 18-55 lens. I wouldn't spend money on the 24L or 35L - the cheaper non-L's are very good, if you can get a good copy.

I had the same itch some time ago, and bought a 28 f/2.8 to complement my 17-40, and did some detailed comparisons. I found the 28 to be slightly brighter, noisier, less reliable in focusing, and at best equal to the 17-40 -- disappointing overall, and so I returned the 28.

 

My tests were done mostly if not entirely at close range and considered sharpness and contrast. I didn't notice distortion to be significant for my needs. I used a 20D, semi-live monitoring, an angle-finder for manual and assisted focusing, tripod, checked for focusing accuracy, etc..

 

Since then, I've also noticed that my 17-40's auto-focus at wide angles isn't always satisfactory (although it's difficult to notice this using just the little viewfinder).

 

I'm afraid the 28 f/2.8 is no Distagon 21... but it might be worth your trouble to do a test for yourself with option to return; that way you'll get results you can personally rely upon, and perhaps your findings will be more positive than mine. But if I did this again, I'd probably try the 24 first.