Tamron 24-70 f/2.8 Di VC USD Announced!!! Stabilized 24-70!

Marsu42 wrote in post #14410775
But you can very well compare them on how they overlap: Is the 70-105 on the 24-105L just "extra" without drawbacks (answer: yes, it even makes the lens sharper @70mm because it's mid-range),

Now that is funny right there! By that logic the 28-300L should be really sharp at 200mm since that is "mid-range". Sorry, but no. Making the lens have a 4.3x zoom range is what keeps it from being as sharp as it could be at 70mm if it had a 2.9x zoom range... Much like giving the 24-70L an f/2.8 aperture is what keeps it from being as sharp as it could be at f/4.

This is basically the same thing I just said above about the 24-70, only difference is that you are turning it into a negative rather than recognizing it as a tradeoff just like you did with the 24-105. You can not have it both ways. If the extra 35mm reach is "bonus" on the 24-105 at the cost of overall IQ at all focal lengths (which it is), then the extra stop of aperture most also be accepted as "bonus" on the 24-70 VC. Otherwise you are just being blatantly biased.

As to which is the sharper lens between the 24-70 VC and 24-105. I already told you I have both and I believe my 24-70 VC to be sharper at F/4. It's not like it would be hard to be sharper than the 24-105 either.... As I showed already even the Tamron 28-75 is sharper at 40mm than the 24-105 according to Photozone.de's testing (as it is at 28 vs 24, and 75 vs 70), and the 24-70VC is sharper than the 28-75.

XTC1 wrote in post #14411251
What are the three lenses that are better than the 35L at f/4?. 24L II... and the other two?

Center weighted averages, F/4, top 10, Canon's in Red:

Canon EF 70-200mm F/2.8 USM L IS II @70mm only = 3555
Zeiss Makro-Planar T* 100mm F/2 ZF (ZE) = 3550
Canon EF 24mm F/1.4 USM L II = 3540
Sigma AF 70mm F/2.8 EX DG Macro = 3533
Voigtlander Nokton 58mm F/1.4 SL II = 3518
Samyang 35mm F/1.4 AS UMC = 3511
Canon EF 50mm F/1.4 USM = 3493
Canon EF 35mm F/1.4 L = 3482
Samyang 14mm F/2.8 IF ED AS = 3467
Zeiss Distagon T* 28mm F/2 ZF (ZE) = 3455

The 35L is only 13th on the list if you look at just the normal F/4 average because of its corners being a good bit softer. But it is number one at center sharpness at f/2.8 and all stops below.

Also Photozone lacks long-lens results, and I imagine the 400 f/2.8L IS II and maybe a couple others could give some of these lenses a run for their money.

I have a spreadsheet created laboriously by hand with every MTF result they have posted for Full Frame Canon testing. It also have various averages calculated that I can sort by.

arentol wrote in post #14411527
Now that is funny right there! By that logic the 28-300L should be really sharp at 200mm since that is "mid-range". Sorry, but no.

I was stating the fact that many zooms are designed so that there's a little iq drop on one or both ends if due to distortion alone - and of course this makes sense, no one is forcing you to use the extremes on zooms at all times. Looking at the Tamron, I guess there's a reason why it's more like 65mm than 70mm on the long end, maybe the iq drawback was more than they could justify. That doesn't say anything about absolute sharpness.

arentol wrote in post #14411527
As to which is the sharper lens between the 24-70 VC and 24-105. I already told you I have both and I believe my 24-70 VC to be sharper at F/4.

That's the keyword - your 24-70vs vs your 24-105, and why would I doubt that? It's just that with the lenses I tested, it was the other way around, and the digital picture review says so, too. I hope we'll get more reviews on more lens samples to know what to reasonably expect from the Tamron, meaning if there are few "golden samples" among the mediocre bulk or the other way round.

Charlie wrote in post #14411462
I've posted 100% crops from both the Tamron and the Canon 24-70L in this thread... the tamron had the canon beat in every scenario, so if you think the tamron is bad, you must also believe the canon is bad. In that case, you're a 24-105 type of guy. Different strokes for different folks.

Same here - I guess arguing about the Tamron is rather pointless unless we see more reviews on more lens samples.

arentol wrote in post #14411527
This is basically the same thing I just said above about the 24-70, only difference is that you are turning it into a negative rather than recognizing it as a tradeoff just like you did with the 24-105. You can not have it both ways. [...] Otherwise you are just being blatantly biased.

You are aware that I'm not trying to defend or bash either lens, are you? I'll probably buy the Tamron myself if the current price drops persist, actually I hope the Tamron turns out better than some reviews say because I'd very much like to have an affordable f2.8 with vc. And usually I'm the one ridiculing red ring religion. Your experiences and the reviews you cite are just exactly the opposite I read and tried for myself.

Marsu42 wrote in post #14411867
You are aware that I'm not trying to defend or bash either lens, are you? I'll probably buy the Tamron myself if the current price drops persist, actually I hope the Tamron turns out better than some reviews say because I'd very much like to have an affordable f2.8 with vc. And usually I'm the one ridiculing red ring religion.

It sounded like bashing to me when you claimed that wider apertures make a lens worse at narrower apertures while in the same breath saying increased zoom range does not compromise the overall sharpness of the lens at all focal lengths. A few of your other posts recently only reinforced the idea you were bashing. I am sorry if I misunderstood.

Nope, still wrong. There is no way for someone to say with authority that the longer zoom range doesn't affect the sharpness of the lens at any given focal length, because a version of that lens without the additional zoom range doesn't exist to test against, and because that statement runs contrary to the known "rules" of lens design.

The reality is that YOU are assuming that if the Tamron were f/4 only and otherwise the same size and build quality it would be sharper at f/4 than it is now. Your assumption is based on years of collective experience and knowledge among the photography community, and your assumption is correct. However, that same collective experience and knowledge tells us that the longer a zoom range is the lower its overall quality throughout the range, and therefore a shorter zoom range lens of otherwise the same general size, build quality, and aperture, will have better sharpness and IQ. You simply can not accept one of these assumptions while denying the other in the same breath and maintain any claim to being unbiased.

The simple fact is that if Canon set out to make a top quality $1100 24-70 f/4L IS lens it would be sharper than the current 24-105L, or any other 24-105 F/4 IS lens they could possibly design today at a similar price point. This is simply the nature of the optical beast.

arentol wrote in post #14411988
The simple fact is that if Canon set out to make a top quality $1100 24-70 f/4L IS lens it would be sharper than the current 24-105L, or any other 24-105 F/4 IS lens they could possibly design today at a similar price point. This is simply the nature of the optical beast.

Actually I wish they would. 24-105 isn't great at 105 anyway. And we have absolute best 70-200 II to cover 70-105 (and even at 2.8 it blows 24-105 out of the water).

No, the math is not that simple.

Shorter zoom range does not always equal better IQ. It matters a heck of a lot where that range actually is. For instance, a lens that is 2:1 but wholly under 70mm is much, much harder to design than a 3:1 lens that is wholly above 70mm. That's why the first really good, professional-IQ zooms were all in the telephoto range.

So even though 24-100 is a wider range than 24-70, the extra range is at the "easy" end...so it may well be easier to design--other factors being equal. In the case of the 24-100, that was also intended to be a "kit" lens, albeit a high-end kit lens. Cost was definitely a factor, and that's where design compromises have been made.

I suspect Tamron's older 28-75mm was considerably simpler to design than their (or Canon's) 24-70...yes four millimeters can make that much difference when they're all below 70.

RDKirk wrote in post #14412341
No, the math is not that simple.

Shorter zoom range does not always equal better IQ. It matters a heck of a lot where that range actually is. For instance, a lens that is 2:1 but wholly under 70mm is much, much harder to design than a 3:1 lens that is wholly above 70mm. That's why the first really good, professional-IQ zooms were all in the telephoto range.

So even though 24-100 is a wider range than 24-70, the extra range is at the "easy" end...so it may well be easier to design--other factors being equal. In the case of the 24-100, that was also intended to be a "kit" lens, albeit a high-end kit lens. Cost was definitely a factor, and that's where design compromises have been made.

I suspect Tamron's older 28-75mm was considerably simpler to design than their (or Canon's) 24-70...yes four millimeters can make that much difference when they're all below 70.

I find this hard to believe.. as a blanket-statement. If any zoom "below 70mm" becomes so much harder to design, how do we get high-quality UWAs at equivalent or lower prices? ie. EF-S 10-22(yes, APS-C lens, but still), 16-35, 17-40 to name a few. All fall well under the 3x zoom range.

1Tanker wrote in post #14412487
I find this hard to believe.. as a blanket-statement. If any zoom "below 70mm" becomes so much harder to design, how do we get high-quality UWAs at equivalent or lower prices? ie. EF-S 10-22(yes, APS-C lens, but still), 16-35, 17-40 to name a few. All fall well under the 3x zoom range.

There isn't anything cheap about the 16-35, and it's barely over 2:1. The compromise of the 10-22 is that it's only covering 15x22mm (do you see a rectilinear 10-22 designed for 24x36mm at any price or aperture? Nope, you don't). The compromises of the 17-40 are that it's longer at both ends than the 16-35 and it's slower. Those are significant compromises.

My point, and what I said up front, is that the math is not simple. Meeting design points is a complex game, and what seems to an observer to be a small compromise--such as designing for 15x22 instead of 24x36--may make all the difference of meeting all the major design points.

But if there is one statement that can be made as a blanket statement, it's that lenses significantly below the diagonal of the format are harder to design for an SLR than lenses above it.

That's because the SLR needs a focus distance farther out than the depth of the mirror box (usually around 45mm for the 24x36mm format). Any focal length shorter than the depth of the mirror box requires a retrofocus design, which is always more complex than a simple straightforward design.

Plus the fact that any lens with a focal length shorter than the diagonal of the format is also more difficult to design because even if it's not retrofocus (such as with a non-SLR), it must still contend with the increased aberrations that are caused by the light rays striking the sensor at more oblique angles as well as having been bent to greater angles.

Lenses shorter than the depth of the mirror box or the diagonal of the format are harder to design than those which are not, and that's a blanket statement that you can depend on.

Do that in a zoom and it's even more complex.

RDKirk wrote in post #14412579
There isn't anything cheap about the 16-35, and it's barely over 2:1. The compromise of the 10-22 is that it's only covering 15x22mm (do you see a rectilinear 10-22 designed for 24x36mm at any price or aperture? Nope, you don't). The compromises of the 17-40 are that it's longer at both ends than the 16-35 and it's slower. Those are significant compromises.

My point, and what I said up front, is that the math is not simple. Meeting design points is a complex game, and what seems to an observer to be a small compromise--such as designing for 15x22 instead of 24x36--may make all the difference of meeting all the major design points.

But if there is one statement that can be made as a blanket statement, it's that lenses significantly below the diagonal of the format are harder to design for an SLR than lenses above it.

That's because the SLR needs a focus distance farther out than the depth of the mirror box (usually around 45mm for the 24x36mm format). Any focal length shorter than the depth of the mirror box requires a retrofocus design, which is always more complex than a simple straightforward design.

Plus the fact that any lens with a focal length shorter than the diagonal of the format is also more difficult to design because even if it's not retrofocus (such as with a non-SLR), it must still contend with the increased aberrations that are caused by the light rays striking the sensor at more oblique angles as well as having been bent to greater angles.

Lenses shorter than the depth of the mirror box or the diagonal of the format are harder to design than those which are not, and that's a blanket statement that you can depend on.

Do that in a zoom and it's even more complex.

Ok..thanks. I still believe Arentol's statement, that smaller zoom ranges are easier to design, and "usually" offer better results. You can look at the Tokina 11-16 /2.8. Yes, it's also for APS-C, but it is f/2.8. The theory goes, that Tokina limited the zoom-range to just 1.5x. to keep IQ and cost and size in check.. which seems to be a perfect example of smaller zoom ranges being preferable.

RDKirk wrote in post #14412341
No, the math is not that simple.

It is if you don't make it more complicated than it has to be.... Oh look, you do make it more complicated:

Yes they do if the shorter one is contained entirely within the range of the longer one, which is what we are talking about here.

Yes, but a 5:1 that is 2/5ths below 70mm is harder to design than a 3:1 that is wholly above 70mm and similarly, a 3:1 which is 2/3rds under 70 is harder to design than a 2:1 which is wholly under 70mm... Which is basically the exact situation at hand.

So, to be clear, you think that it would be easier to make a high quality 3:1 33-100 than a 3:1 24-72 (I agree), but you also think it would be easier to make a high quality 4:1 24-100 than a 3:1 24-72? . Just tack on an extra 9mm on the "Hard" end, but it is EASIER because there is an extra 28mm on the long end? I could understand and might not entirely disagree if you said a near 3.5:1 35-120 MIGHT be easier. After all your argument is essentially that each mm closer to 0 makes it harder to make a longer zoom range, and I would not disagree with that general idea. But to argue that two lenses both starting at the same EXACT focal length, but one being much longer in total, would result in the longer one being easier and higher IQ.... That is just ridiculous.

Heck, if that is the case then please explain why we aren't drowning in a sea of 24-100 f/2.8's right now if they would actually have better IQ than 24-70 f/2.8's? Sure, they would be big and weigh a lot, but probably not any more than a 70-200 f/2.8 and people would love to have the superior IQ.

Not sure what compromises you are talking about, nor why that would be relevant to the point at hand.

I am sure it was simpler, not considerably, but definitely simpler... Curiously though nobody has ever made a 28-120 f/2.8 that blows the 28-75 away like you are suggesting would be trivially easy to do. Would pair fantastically with my 120-300 f/2.8.

Sorry, but your entire argument is silly beyond belief.

This thread is getting excited now since everyone in here has his own right to defend what you like or what you do not like. IMO I would judge for myself of using what lenses I like the most by comparing between the one I had owned and the one I own now.

Why do you have to bring in of someone's reviews called themselves a PRO and believe what they wrote (?) Why don't you buy yourself if you can afford one then do yourself the comparison or write your own reviews based on your experiences of using the two lenses you have (Tamron 24-70 VC vs Canon 24-70L or 24-105L). If you could not do that please should not have disagreements with people who have spent times to do a favor for you by reviewing them. All these would help you deciding to buy the lens that fit your budget or make you purchase the so-called "L" or red ring lens.

To be honest I like very much all the L lenses and would buy them within my budget. I myself was not open-minded and thought all third party lenses could not compared to even the EF-S lenses. Since I read this thread and other reviews about Tamron 24-70 VC I have decided to sell my 24-70L and pulled a trigger to buy it. I own this lens just about more than a week and already fall in love with this lens on my 5D3.

Thanks to dedicated members like KenjiS, Charlie and Arentol who have brought more excitation by reviewing and comparing between Tamron and Canon lenses to this thread. I have made my decision according to this thread to cancel the 24-70L II and buy the Tamron 24-70 VC and would not regret.

tancanon58 wrote in post #14412834
Since I read this thread and other reviews about Tamron 24-70 VC I have decided to sell my 24-70L and pulled a trigger to buy it. I own this lens just about more than a week and already fall in love with this lens on my 5D3.

Now if only I could get a copy like yours or Charlie's... *crosses fingers for second copy carefully handled by UPS employees somewhere around Michigan*

So I just spent a few hours with the lens.

Brief thoughts?

1. I do NOT think I received a GREAT copy.
2. Center sharpness is not as good as I was expecting.
3. Edge sharpness is poor, but I expected this.
4. Vignetting not as bad as expected.
5. Onion bokeh. I think I got it in one shot but otherwise seems under control?
6. Focus speed right where I expected it to be/a hair slower than my 85/1.8
7. Quieter than expected in regards to focus motor BUT ESPECIALLY in regards to VC. It's quieter than Canon's IS mechanism.
8. Build quality is excellent. Weight and balance on my 5DII is also excellent.

I am going to play with it over the weekend, but it will likely go back.

A few notes on the shots below:

-Taken with a 5DII
-Picture mode: Neutral, 0 in-camera sharpening
-No USM used in Photoshop!
-Center Point AF, AWB
-Hand-held
-Most taken with flash
-All taken after sunset

Remember, center sharpness does not seem ideal. I couldn't correct for it using MA but then again these were hand-held. I will bust out the tripod over the weekend for more accurate testings.

Straight-from-camera, reduced in size: