Photozone spanks the 24-70 F4 USM L IS

[Quasimodo]

While I trust their methods and unbiasedness at Photozone, as several have pointed out, the sheer number of lenses not tested is also an important aspect to consider when reading the different reviews.

How many lenses does TDP, DPReview, DXO use for tetsing lenses?

You make a fair point.  From the reviews I've read, it appears that each site has strengths and weaknesses:

Roger at LR is king of sharpness data from my perspective, b/c he tests all of LRs stock of the same lens (in some cases, dozens of the same lens).  That trumps most everyone out there for data, but he's not the thoughtful tips/feedback/insights guy that Bryan Carnathan is.  Roger also has (IMHO) massive street cred on being a camera nerd first and a fanboy last.  He'll blow any design out of the water if it underperforms.

TDP -- Carnathan does test multiple copies but not a huge number.  He brings up copy to copy variation when something odd occurs and obtains new glass to verify odd findings.  See his Tamron 24-70 IS and new Mk II 24-70 reviews -- fishy stuff came up and he itemized the differences.

Further, his site is more of a broad read of useability and his personal (though admittedly expert) take on things.  His site excels at fun image-level comparisons of all the lenses taking the same shot (or test page) from a tripod, and then allowing you to mouseover to see what's going on:  L vs. non-L vs. Sigma, aperture, ISO, etc.

I read his site fairly religiously for all-around reasons.  It's thoughtful, useful commentary from a knowledgable user.  But his is a world of insight and adjectives -- it's not a lab-like mountain of details.

Photozone has the most nicely broken down sharpness data that I can read easily.  They have middle/border/extreme resolution data for many focal lengths on a zoom, and they report it for most common apertures selected.  Very helpful.  This is a site opens your mind a bit as you tend to comparison shop.  At a site like this, you see non-L glass punching its weight brilliantly against L glass if you stop it down just a shade.  Big limitation?  I believe it's only one lens they test.

DXOs interesting angle is having combinatorial data of bodies + lenses.  No idea how many lenses they actually test.   I don't read this site much as DXO has (groan) their own metrics on their own hardware that I haven't bothered to learn how to read.  I still don't understand their data fully.

I don't read DPreview reviews as they require a dozen page turns to get ad views.  Fail.  They are wasting my time versus their competitors.

I don't read Ken Rockwell much, but my goodness, if you want a fine detail about metal vs. plastic, if the front element rotates during focusing, how many points on the sunstar created when stopping down, etc. then he's your guy.  Absurd little details for the nerdy camera lover.  TDP has some of this, too, but it's inconsistently reported from review to review.

For me, I trust:

• TDP for everything but data -- it's just there, easy to find, and thoughtfully explained.
• Roger at LR for hard sharpness numbers.  End of story.
• Photozone for quick, easy comparison of two lenses I might be considering

- A

Thank you for answering the question, and in such an informed manner.

G

[curby]

See his Tamron 24-70 IS ... review

Do you have a link?  This page doesn't seem to have many details.  Thanks!

[ahsanford]

See his Tamron 24-70 IS ... review

Do you have a link?  This page doesn't seem to have many details.  Thanks!

I forgot -- he didn't post it as a formal review because of his quirky findings.  He posted it under news:

Initial story in May of 2012: http://www.the-digital-picture.com/News/News-Post.aspx?News=2357

Follow-up in Jan 2013:  http://www.the-digital-picture.com/News/News-Post.aspx?News=4053

He has a pre-loaded mouseover compare chart here:  http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=786&Camera=453&FLI=0&API=0&LensComp=786&Sample=0&SampleComp=1&FLIComp=0&APIComp=0

- A

[psolberg]

FINALLY reviewers are starting to take note of focus shift, a topic which had been buried as a footnote as all everybody cares these days is 100% crops of some stupid black and white chart.

it is however surprising to see such absurd amount of focus shift on a slow zoom lens. I kind of expect that behavior on f/1.8 glass or faster.

[Albi86]

I don't know about focus shift, but what supporters of this lens are terribly overlooking is that the 24-105 has a range and a price that leave room for compromise, whereas a 24-70 f/4 sold for 1600$ has to be epitome of optical perfection itself - and it's not.

[Zlatko]

... whereas a 24-70 f/4 sold for 1600$ has to be epitome of optical perfection itself - and it's not.

It was introduced at $1499.  It is now $1449.  It will likely be less in the months ahead, and even less if included as part of a kit.

No one has ever made a perfect zoom for $1449.  The epitome of optical perfection itself would cost a lot more than $1,449.  For $1449 you get an improved lens, not the epitome of optical perfection.

[Albi86]

... whereas a 24-70 f/4 sold for 1600$ has to be epitome of optical perfection itself - and it's not.

It was introduced at $1499.  It is now $1449.  It will likely be less in the months ahead, and even less if included as part of a kit.

No one has ever made a perfect zoom for $1449.  The epitome of optical perfection itself would cost a lot more than $1,449.  For $1449 you get an improved lens, not the epitome of optical perfection.

The fact is that it's not an improvement because it left as much as 35mm of range behind its predecessor. It's likely that most of the optical advantages derive from this alone.

What you get is a lens whose specs and performance hardly justify its price. And in fact I don't see many owners of a 24-105 throwing it out and running to buy this 24-70.

So yes, when you charge 1450$ for such a lens it'd better be flat-out amazing.

[sagittariansrock]

I don't know about focus shift, but what supporters of this lens are terribly overlooking is that the 24-105 has a range and a price that leave room for compromise, whereas a 24-70 f/4 sold for 1600$ has to be epitome of optical perfection itself - and it's not.

I am sure many if not most people who own a 24-105 wouldn't feel the need to upgrade to a 24-70 IS. I am also sure that most people looking for a standard zoom for FF will still prefer a 24-105. I don't think that a comparison between the two lenses is the big question here.
The so-called "supporters" of the 24-70 are merely trying to reason that it doesn't make sense to "spank" a lens without basis. However, a justifiable basis would be if the lens consistently demonstrates the problem of focus shift (something that I recently learned about) or if Canon decides to replace the 24-105 with the 24-70 without lowering the price or fixing the issue (assuming it really is an issue across many samples).

[Zlatko]

... whereas a 24-70 f/4 sold for 1600$ has to be epitome of optical perfection itself - and it's not.

It was introduced at $1499.  It is now $1449.  It will likely be less in the months ahead, and even less if included as part of a kit.

No one has ever made a perfect zoom for $1449.  The epitome of optical perfection itself would cost a lot more than $1,449.  For $1449 you get an improved lens, not the epitome of optical perfection.

The fact is that it's not an improvement because it left as much as 35mm of range behind its predecessor. It's likely that most of the optical advantages derive from this alone.

What you get is a lens whose specs and performance hardly justify its price. And in fact I don't see many owners of a 24-105 throwing it out and running to buy this 24-70.

So yes, when you charge 1450$ for such a lens it'd better be flat-out amazing.

Sure, it's not an improvement because of the reduced 35mm of range ... if focal length range is your measure of improvement.  In that case, 10X or 20X super-zoom would be the ultimate improvement.  But focal range is not the usual measure of improvement.  The facts are as shown by the numbers in ahsanford's post above.  Those numbers clearly show an overall improvement in IQ.

Of course, owners of the 24-105 aren't "throwing it out" to buy this 24-70.  That doesn't reflect on the 24-70 at all.  The 24-105 is a fine lens in its own right and remains a good value.  So no one is crazy enough to throw it out.  The 24-70 is an alternative with some improvements and differences, not an obvious replacement.  Many people will find the 24-105 meets their needs and budget better.  That doesn't mean that other people won't find that this new 24-70 meets their needs and budget better.  That's the beauty of alternatives.  They meet the needs of different people with different needs and applications.

Optical perfection is a very high standard.  You don't get it for an extra $400-$600 from any manufacturer.

[mrsfotografie]

I don't know if I would call that a spanking. The review is mostly very positive with one big gotcha. That big gotcha plus the high cost of this lens means I would expect few to choose it over the 24-105. Maybe that does add up to a spanking. 

Well... Residual Spherical abborations, focus shift, soft @ F 4, and a not so compelling price.. I'll hold on to my 24-105 for a while longer. A lens that continue to deliver.

+1 I don't see the point of the 24-70 F/4

[ahsanford]

BTW ... I'd still go for the 24-70 2.8 over the f/4. It's not that much extra.

Crystal ball: Canon releases a 24-70 2.8L IS at $2400, and the non-IS version settles in at $1600 - $1800

I don't see it.  If and when that pickle jar of a 2.8 IS lens is offered, those jerks will want extortion money for it -- I'd say $3k.  It's a holy grail sort of lens.

That said, if Nikon offers one first, Canon will only punish us with a nasty markup, and then your prediction will be right in line.

- A

[Zlatko]

FINALLY reviewers are starting to take note of focus shift, a topic which had been buried as a footnote as all everybody cares these days is 100% crops of some stupid black and white chart.

it is however surprising to see such absurd amount of focus shift on a slow zoom lens. I kind of expect that behavior on f/1.8 glass or faster.

FWIW, this poster on the FM forum says he tried to replicate the focus shift with his new 24-70/4L and could not find it:
http://www.fredmiranda.com/forum/topic/1194688/1#11395622

... but of course, that's not the last word on the topic, as others may get different results.

[Zlatko]

BTW ... I'd still go for the 24-70 2.8 over the f/4. It's not that much extra.

Crystal ball: Canon releases a 24-70 2.8L IS at $2400, and the non-IS version settles in at $1600 - $1800

I don't see it.  If and when that pickle jar of a 2.8 IS lens is offered, those jerks will want extortion money for it -- I'd say $3k.  It's a holy grail sort of lens.

That said, if Nikon offers one first, Canon will only punish us with a nasty markup, and then your prediction will be right in line.

- A

Sure, everyone one wants optical excellence and great durability at bargain prices.  Who wouldn't?  But is that realistic?  Are they "jerks" for not offering that?  Are photographers who charge high prices for excellent photography and service also "jerks"? 

Try buying the real estate for a lens factory, building your own lens factory and hiring your own highly skilled workforce, and paying for all of the materials, training, benefits, upkeep, taxes, etc. that go with that, and then see how cheaply you'll want to sell your professional grade lenses.  My guess:  not very cheaply.

[sagittariansrock]

It is sometimes worrisome that people maintain negative feelings towards a brand they are loyal to, but it's not a really bad thing altogether. We should be able to criticize and praise at the same time. So, while I generally think Canon makes excellent photography equipment, it is refreshing to see people criticize genuine mistakes. Take a look at the Apple forum, and the excess of fanboyish behavior will disturb you. You say one negative word, and ten people will pounce on you advising you to switch to PC (or Android) immediately or just deal with it.
I think there are both good and bad things about the criticism of the 24-70 IS- comparing it with the 24-105 or making a big deal out of a yet unproven (except for one flying pig till now) issue is not so good. Criticizing people who have bought, used and liked the lens- also not good. Criticizing Canon's pricing policy- not so bad, actually. Not just Canon- Nikon came up with a very expensive 80-400 VR (II?) as well. Are these companies trying to mark their items up to make up for their losses due to natural disasters, bad economy, etc.? That is bad, considering they are taking advantage of brand commitment rather than charging a product based on its performance alone. And that is unfair, as it is unfair for Apple to charge $ 30 for a lightning connector adapter, taking advantage of people who have already purchased Apple-compatible products that are suddenly no longer compatible.

[curby]

I agree with the above, with two caveats/clarifications:

1) Apple has more than its share of rabid fans, but it's not just the media that has insanely high expectations of them; their users howl about every real and perceived fault.  I'm a long-time Mac user, and have caused many friends and family to purchase them too, but boy do I hate their mice, cloud services, and every laptop screen they've ever put out before the retina.

2) The market will bear what the market will bear.  Going back to Apple, everyone said the iPad mini was overpriced, but they still couldn't keep them in stock.  Lowering the initial price would have just been a bad business decision that would have left money on the table.  $30 for a Lightning cable is a lot, but now there are cheaper options available.  $1500 for this lens is a lot, and will either plummet in price if the naysayers here are right, or will stay high if it does turn out to be as popular as Canon bet on.

Ultimately, I hope the performance problems Photozone found are limited in nature, as supported by other reviews not mentioning the problem and other users who can't reproduce the problem.  If not, then Canon truly made a misstep not only in the pricing, but in releasing a lens that simultaneously offers high-magnification performance as a headlining feature yet suffers from critical issues when using that feature.