By the comparisons, the 24-70 tamron isn't that much "worse", and for the price, offering 2.8 may be better.

Just picked one up. A very nice lens allbeit pricey. It is very sharp at 4.0 getting just a tad better at f8 but you have to looking at 100% samples to see it. I had the 24-105 and it was an OK lens. I was never really happy with the distortion or the darkening in the corners. This lens seems to have fixed both of those issues while being sharper and in a smaller package. I wanted a zoom to travel with and my 24-70 2.8MK1 was just too heavy (even though mine is a great copy) so it was always left at home. I can vacation with this and a fast prime and be more than happy. I never considered 2.8 fast enough for use without a flash indoors anyway so this should fit.

Just a head's up for those of you checking out your new 24-70 f/4 IS lenses: having done some further testing the weak point of the lens is actually around 50mm, rather than 70mm. That's kind of unusual, most zooms are at their weakest at one extreme or the other, but pretty consistently this one is at 50mm.

It's not bad, by any means, but a bit less resolution than at 70mm.

RCicala wrote: Just a head's up for those of you checking out your new 24-70 f/4 IS lenses: having done some further testing the weak point of the lens is actually around 50mm, rather than 70mm. That's kind of unusual, most zooms are at their weakest at one extreme or the other, but pretty consistently this one is at 50mm.

I think it is purposely optimized that way. Canon has realized that most people shoot zooms at their extremes most of the time. Further, most testing and MTF charts are done at the extremes.

It is a shift in fad/fashion. In the old days, photographers knew the worth of the "normal" focal length and were more desiring of high performance there, especially since they were comparing against primes. These days many photographers have not learned the usefulness of "normal" focal lengths and depend on gimmicks like extremes and Instagramming.

These days, most photographhers are comparing zooms versus zooms and not zooms vs. primes. The 18-55 crop factor shooter moves up to 24-70 or 24-105 full-frame. Hence the big push in 2012 by Canon on the L 'kit' lens: 24-105 f/4 L.

A 'landscape' zoom for me (like this would be) has to be decent at 50mm too, which is where I shoot a lot of panos at.

Monito wrote:
I think it is purposely optimized that way. Canon has realized that most people shoot zooms at their extremes most of the time.

Aah. You masters at Canon have carefully designed the lens this way.
But make no mistake: this is not a flaw/limitation of the lens.
Just a masterfully crafted feature for the benefits of Canon users.

RCicala wrote:
Just a head's up for those of you checking out your new 24-70 f/4 IS lenses: having done some further testing the weak point of the lens is actually around 50mm, rather than 70mm. That's kind of unusual, most zooms are at their weakest at one extreme or the other, but pretty consistently this one is at 50mm.

As always, thanks for your efforts.

I've seen similar behavior with other lenses, where there's a dip in sharpness mid-zoom.
It depends on the optical formula and zoom design, I guess.

For reference, here's the PhotoZone test of the 16-35L II. It has a minor dip at 26mm:
http://www.photozone.de/canon_eos_ff/435-canon_1635_28_5d?start=1

I think the 17-40L does a similar thing at ~30mm - before becoming very sharp again at 40mm.

jorkata wrote: Aah. You masters at Canon have carefully designed the lens this way.
But make no mistake: this is not a flaw/limitation of the lens.
Just a masterfully crafted feature for the benefits of Canon users.

You have nothing to say. You are unable to discuss the facts or the implications, so you descend to attacks.

Try discussing the facts and their implications like I did. I'm sure you can. Give it a try.

Monito wrote:
You are unable to discuss the facts ...

And how do you know what Canon has realized and why they designed this lens the way they did?
You are not presenting any facts here.

Lens design involves compromises - and we are seeing one such compromise here.
Why do you have to spin it an 'optimization' .

RCicala wrote:
Just a head's up for those of you checking out your new 24-70 f/4 IS lenses: having done some further testing the weak point of the lens is actually around 50mm, rather than 70mm. That's kind of unusual, most zooms are at their weakest at one extreme or the other, but pretty consistently this one is at 50mm.

It's not bad, by any means, but a bit less resolution than at 70mm.

Probably a fair trade for many (although obviously it is a poor one for some). You get your crisp edge to edge landscapes and your sharp wide open 70mm portraits out of it (http://www.fredmiranda.com/forum/topic/1181595 it looks a lot sharper f/4 70mm in this comparison) and the middle is a bonus. A standard zoom that does 24mm well has been rarer than hen's teeth and is more impressive than on that does 50mm OK of which such already existed. But it obviously depends how you tend to shoot.

RCicala wrote:
Just a head's up for those of you checking out your new 24-70 f/4 IS lenses: having done some further testing the weak point of the lens is actually around 50mm, rather than 70mm. That's kind of unusual, most zooms are at their weakest at one extreme or the other, but pretty consistently this one is at 50mm.

It's not bad, by any means, but a bit less resolution than at 70mm.

Disappointing to hear that is the expected behavior for a normal copy since it's a rather important focal length for me (based on EXIF data of my past photos). Checked results from TDP and it appears resolution at 50 mm remains poor even at f/8 (particularly at the center) when compared to performance of 24-105 lens at 50 mm.

I think I am leaning towards the 24-105 f/4 lens now. I am sure Canon could have done better if they were not so ambitious and tried to incorporate some macro feature into their 24-70 f/4 lens.

thw2 wrote:
RCicala wrote:
Just a head's up for those of you checking out your new 24-70 f/4 IS lenses: having done some further testing the weak point of the lens is actually around 50mm, rather than 70mm. That's kind of unusual, most zooms are at their weakest at one extreme or the other, but pretty consistently this one is at 50mm.

It's not bad, by any means, but a bit less resolution than at 70mm.

Disappointing to hear that is the expected behavior for a normal copy since it's a rather important focal length for me (based on EXIF data of my past photos). Checked results from TDP and it appears resolution at 50 mm remains poor even at f/8 (particularly at the center) when compared to performance of 24-105 lens at 50 mm.

I think I am leaning towards the 24-105 f/4 lens now. I am sure Canon could have done better if they were not so ambitious and tried to incorporate some macro feature into their 24-70 f/4 lens.

I think you should give the 24-70 f4 a try and not focus so much on the results of Digital Picture. By the samples they have, yes there is a fall off at 50 mm compared to the 24-105 but that is really the only focal lenght I see where this happens on that test of those 2 copies. The new lens seems better pretty much everywhere else. Mine seems very good all the way through the focal lengths. If you get the 24-105 be prepared for a lot of barrel distortion at 24mm (the worst I have seen) and some pretty severe corner darkening at 24 and f4. The barrel distortion was the one reason I could really warm up to the 24-105. It can be corrected but you will give up resolution when you do that.

I will try to take some photos today with my 50 1.4 and 24-70 4.0 today to compare.

twistedlim wrote:
I will try to take some photos today with my 50 1.4 and 24-70 4.0 today to compare.

Please do. I am really hoping the weakness at 50 mm is not a normal characteristic of this lens.

There is a comparison between the Canon 24-70 f/4L IS and the Tamron 24-70 f/2.8 VC on youtube: http://www.youtube.com/watch?v=4nMmX7_ua_w

Looks like the weak 50mm is noticeable there as well. Tamron seems to look better at 50 & 70mm. Add the 1 stop advantage and $200 price break and the Tamron has me sold. Though it was slower at focusing and doesn't have the macro ability but it still looks like a much better lens in many ways to the new Canon

RobDickinson wrote:
A 'landscape' zoom for me (like this would be) has to be decent at 50mm too, which is where I shoot a lot of panos at.

It is reportedly considerably better than decent at around 50mm. The point wasn't, I think, that it is not good at 50mm, but that it is stronger at the shortest and longest focal lengths. Perhaps RCicala could clarify this?

It is sort of sad that the observation that the lens is unusual in providing excellent performance at the long and short ends of its range, which is somewhat special for a zoom, has now begun to morph into the new, unintended, and apparently quite wrong notion that it is poor at 50mm!

This is how these oddball intermet memes often get started, I'm afraid. It is unusual to be present at the very start of the silliness!

Dan

Dan, I agree. I saw it coming, but I decided to get out of the way...

A 50 f1.4 weighs next to nothing; f1.8 even less, either can be carried in a shirt pocket if necessary, I doubt; as Dan says, the 50mm on this zoom is bad at all, just not optimal at 50mm like it is wide-open at 24/70...stop it down one stop and it probably kicks arse.

Heck, carry the 50 f2.5 macro, it's super sharp, small, light weight, would do great for panos...

Roger...help

twistedlim wrote:
If you get the 24-105 be prepared for a lot of barrel distortion at 24mm (the worst I have seen) and some pretty severe corner darkening at 24 and f4. The barrel distortion was the one reason I could really warm up to the 24-105. It can be corrected but you will give up resolution when you do that.

Why, oh, why does the discussion of some lens always have to lead to the denigration of some other lens?

It is quite possible that more than one lens can be quite good. In fact, among the 24-105mm and the two 24-70mm lenses we have three very fine lenses with different complements of strengths and weaknesses - and any one of the three could be the "best" choice for a particular photographer based on a range of factors.

The 24-105 does, indeed, exhibit barrel distortion and vignetting at 24mm (the latter when wide open), but it is easily corrected in post and the supposed distortion effect of the correction is essentially invisible in a rather large print.

Dan

The sample crops from TDP suggest that the performance at 50 mm (center of image) is worse than all other zoom lenses, including cheap ones, and that it does not improve substantially by stopping down, and that the copy variation is significant.

It is not about being less excellent than at the extremes, but about being the worst lens you can choose for 50 mm stopped down work. The sample #1 at TDP is that bad.

I find it strange that someone who has not used the lens, can dismiss several reports that support the results from TDP, based on measured numbers from a single source. I trust the numbers from Roger Cicala, but I also found some of the other sources credible.

alundeb wrote:
The sample crops from TDP suggest that the performance at 50 mm (center of image) is worse than all other zoom lenses, including cheap ones, and that it does not improve substantially by stopping down, and that the copy variation is significant.

It is not about being less excellent than at the extremes, but about being the worst lens you can choose for 50 mm stopped down work. The sample #1 at TDP is that bad.

I find it strange that someone who has not used the lens, can dismiss several reports that support the results from TDP, based on measured numbers from a single source. I trust the numbers from Roger Cicala, but I also found some of the other sources credible.


Not sure if you are referring to my post or to something else, but since I responded a bit negatively to the new "bad at 50mm" business I'll reply a bit.

The familiar "you haven't used the lens" criticism is useful as far as it goes, but at this point few of us have extensively used all three of the current 24mm-(more than 24mm) Canon L options. I have not used it yet, and I don't think I made any claims to have concrete personal experience with the 24-70mm f/4. I think I kept my comments in the realm of things that a person could reasonably observe both about the specific lens and about the general question of how lens performance is evaluated - at least that was my intent.

It is reasonable and possible to have something useful to say about a lens performance discussion from that perspective - though I'll freely acknowledge that long and intense use of a lens can inform one's opinion in other useful ways. (I have that experience with the 24-105, and I now also have some experience with the f/2.8 24-70 II.) My comments were more about how we judge lenses than about the specific performance of the f/4 24-70.

My points were basically two:

1. Forum discussions often seem to morph the observation that a good thing is best in condition A into an assumption that it must therefore be bad in condition B. I think that something like that may be happening here with the f/4 24-70 lens, where there doesn't seem to be much evidence that the lens performs in a manner that could objectively be described as "poor" at 50mm but just that this isn't its best FL. The real question isn't whether it is better or worse at 24mm, 50mm, or 70mm, but rather how it performs as a photographic tool for our purposes. We can say something similar about almost any zoom lens - I have yet to meet a high quality zoom that performs equally well at all focal lengths and apertures.

2. When comparing things like lenses, forum discussions (and similar comparisons) often can lead towards a false and simplistic assumption that one thing is "best" and that other things are "deficient" - e.g. we try to pick a winner and losers. My point is that there are so many factors in play - the photographer's specific needs, functionality, etc. - that this can be a false way to evaluate gear. In many, many cases we are actually comparing a number of things that are quite good. I believe that is almost certainly the case here, based on my actual experience with the 24-105 and the 24-70 f/2.8 II and on reading a lot of evaluations and commentary on a lot of lenses.

I went back to the original link that started this thread to see what the author actually said about the 50mm performance. Here is a quote:

"We did find that 50mm resolution was slightly lower than 70mm for every copy. The center / weighted average at 50mm for the 24-70 f/4 IS was 875 / 700, compared to 920 / 750 at 70mm. Not a huge drop, but it was consistent. This is a bit surprising, but not a total shock. Some wide angle zooms exhibit similar behavior and the dip in resolution isn’t extreme."

(Emphasis added to original text.)

That is in line with the point I was trying to make. I hope this clarifies things a bit.

Take care,

Dan

(edited after original post)