TDP Image Quality posted -- 70-200 f/4L IS II: it's not much better

[ahsanford]

Holy cow, and not a good holy cow:

https://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=1198&Camera=979&Sample=0&FLI=0&API=0&LensComp=404&CameraComp=0&FLIComp=0&APIComp=0

The new optical design doesn't look that new at all. I know it's just one copy, but:

• At 200, the Mk II looks a hair better in midframe. It's not a resounding win, but it's better.
  
  
• At 135, the Mk II is a step forward.
  
  
• At 70, I'll take the Mk I thank you very much.

We expected this from the f/2.8L IS III, not from the new optical design f/4L IS II.

[Checks TDP specs]

The IS improved, MFD got smaller, one more blade was added and the filter thread got bigger... but the element count is the same and the IQ looks very similar. Just how new and improved is this optical design? Is this another 24-105L II sort of situation?

- A

 

[ahsanford]

Re: TDP Image Quality posted: 70-200 f/4L IS II, and it's not much better

A 9th blade, shorter MFD and better IS warranted a new optical design?

Why didn't Canon just 'paint job + coating refresh' this one as well?

- A

 

[Larsskv]

Compare it to the 70-200 f2.8 LII, and I think one should be happy with the new f4 LII.

The difference between I and II isn’t mind blowing, but given the quality of the I, I don’t think that would be possible. The II seems very good too me, but I don’t think people will upgrade from the I because of IQ differences.

 

[AlanF]

Compare it to the 70-200 f2.8 LII, and I think one should be happy with the new f4 LII.

The difference between I and II isn’t mind blowing, but given the quality of the I, I don’t think that would be possible. The II seems very good too me, but I don’t think people will upgrade from the I because of IQ differences.

+1
The published MTFs on the Canon site are pretty similar so not much difference is expected. It's also best not to stare at the TDP charts and look for minor differences in one pair of lenses. TDP is a fine site but the image comparisons aren't set up to be a fine statistically satisfying analysis.

 

[ahsanford]

Compare it to the 70-200 f2.8 LII, and I think one should be happy with the new f4 LII.

The difference between I and II isn’t mind blowing, but given the quality of the I, I don’t think that would be possible. The II seems very good too me, but I don’t think people will upgrade from the I because of IQ differences.

+1
The published MTFs on the Canon site are pretty similar so not much difference is expected. It's also best not to stare at the TDP charts and look for minor differences in one pair of lenses. TDP is a fine site but the image comparisons aren't set up to be a fine statistically satisfying analysis.

Bryan usually gets a few more copies of the newer lenses. Let's see if he gets one here.

- A

 

[AlanF]

If you really want a proper comparison, wait for lensrentals. There you will get a series of controlled measurements under highly reproducible conditions on usually 10 copies and MTFs at various frequencies systematically. The TDP charts tell you something about low frequency MTFs (acutance) on the one or sometimes few copies of lenses tested. Annoyingly in the newer charts using the 5DSR, the converging lines don't go close enough to see resolution differences as they are too separated to blur together.

I have made mistakes in the past buying lenses based on looking at TDP charts.

 

[Nitroman]

Well i was certainly disappointed with the Canon 24-105mm F4 IS Mark II. It was almost no better than the Mark 1 - apart from a slightly better IS.

Why do Canon bother to release new versions when there is little improvement?!

With a new Canon 5Dsr in the pipeline, we need better lenses.

 

[Don Haines]

And once again, you have to face the possibility that this is as good as you can make a 3X zoom lens.....

 

[ahsanford]

And once again, you have to face the possibility that this is as good as you can make a 3X zoom lens.....

...for $1299, yes. It's possible.

But if this is the case, you can only squeeze so much added performance out of a lens with more pixels behind it. At some point, the lens will cry uncle and you are only realizing X% of your sensor's resolution.

So Canon needs to offer better lens designs (and ask for more money if that's what they cost to make) or Canon should stop climbing sensor resolution mountain for a few cycles until they can.

- A

 

[Don Haines]

And once again, you have to face the possibility that this is as good as you can make a 3X zoom lens.....

...for $1299, yes. It's possible.

But if this is the case, you can only squeeze so much added performance out of a lens with more pixels behind it. At some point, the lens will cry uncle and you are only realizing X% of your sensor's resolution.

So Canon needs to offer better lens designs (and ask for more money if that's what they cost to make) or Canon should stop climbing sensor resolution mountain for a few cycles until they can.

- A

For many of us, we want the quality of primes and the versatility of zooms..... without acknowledging that at higher levels the two are mutually exclusive. If you really want to be shooting a high megapixel camera and want to get as much out of it as possible, you need a bag (a LARGE bag) full of primes. You look at the series 2 big whites and that shows you what you can do when you abandon zooms..... unfortunately for those of us who are financially challenged, these are very expensive lenses and there is not the market for slower versions, as the zooms have captured the bulk of that market....

For example, the 400F5.6. WAY! More affordable than the F2.8 version, and despite being a 20+ year old design with antique materials, is still comparable in sharpness to the latest 100-400 zoom. A version 2 of that lens would beat the pants off of the 100-400 series 2 zoom, but the market numbers just are not there.... people like zooms!

 

[takesome1]

For example, the 400F5.6. WAY! More affordable than the F2.8 version, and despite being a 20+ year old design with antique materials, is still comparable in sharpness to the latest 100-400 zoom. A version 2 of that lens would beat the pants off of the 100-400 series 2 zoom, but the market numbers just are not there.... people like zooms!

Despite all the rumors over the years the 400 f5.6 is getting an upgrade.Even IS maybe?
Maybe it would take away from the 100-400 sales, the series I 100-400 we saw the question from budding wildlife photographers whether they go with the zoom or the prime. It always seemed IS was the tipping point.
Personally I wouldn't mind having a 400mm light weight prime with IS for those time I want to go compact.

 

[takesome1]

The original 70-200 f/4 is a great lens.
When they upgraded the big whites the largest improvements were in IS and weight.
The improvement in IQ was marginal.
Perhaps it is the same here, the improvement will be with other features.

 

[Don Haines]

For example, the 400F5.6. WAY! More affordable than the F2.8 version, and despite being a 20+ year old design with antique materials, is still comparable in sharpness to the latest 100-400 zoom. A version 2 of that lens would beat the pants off of the 100-400 series 2 zoom, but the market numbers just are not there.... people like zooms!

Despite all the rumors over the years the 400 f5.6 is getting an upgrade.Even IS maybe?
Maybe it would take away from the 100-400 sales, the series I 100-400 we saw the question from budding wildlife photographers whether they go with the zoom or the prime. It always seemed IS was the tipping point.
Personally I wouldn't mind having a 400mm light weight prime with IS for those time I want to go compact.

This is a lens that I doubt we will see, but if it came out I would be the first to buy it.

As to the 70-200 F4 IS, I doubt that I would ever get one. This lens is not an upgrade from the previous version, it is a replacement. New buyers get the new lens, previous owners are happy with the old one, and three or four people with money to burn buy the new one because it is marginally better.

 

[Keith_Reeder]

Why do Canon bother to release new versions when there is little improvement?!

It's naive and unrealistic to expect dramatic improvements to lenses which are already as good as they are; but people will complain about Canon being lazy and resting on their laurels if they don't update.

They can't win...

 

[Don Haines]

Well i was certainly disappointed with the Canon 24-105mm F4 IS Mark II. It was almost no better than the Mark 1 - apart from a slightly better IS.

Why do Canon bother to release new versions when there is little improvement?!

With a new Canon 5Dsr in the pipeline, we need better lenses.

If you want better lenses, get primes.

Realistically, most zooms are now as sharp as they can be made. Improvements are limited to coatings, IS, quality of machining, and the electronics. This improves copy variation and AF performance..... things that do not show up on a MTF chart, yet are solid reasons for a new version.

 

[BillB]

The original 70-200 f/4 is a great lens.
When they upgraded the big whites the largest improvements were in IS and weight.
The improvement in IQ was marginal.
Perhaps it is the same here, the improvement will be with other features.

Starting about 2012, new Canon lenses and cameras have included iterative focussing capabilities that significantly improved the AF consistency of mirror based PDAF. Don't know whether it has any effect on dual pixel PDAF. Presumably, all new lens updates including the 24-105 and the 70-200s include this iterative capability.

Roger Cicala reported this development in a series of Lensrental blog posts in the summer of 2012.

 

[sanj]

Compare it to the 70-200 f2.8 LII, and I think one should be happy with the new f4 LII.

The difference between I and II isn’t mind blowing, but given the quality of the I, I don’t think that would be possible. The II seems very good too me, but I don’t think people will upgrade from the I because of IQ differences.

+1
The published MTFs on the Canon site are pretty similar so not much difference is expected. It's also best not to stare at the TDP charts and look for minor differences in one pair of lenses. TDP is a fine site but the image comparisons aren't set up to be a fine statistically satisfying analysis.

What does this mean? Would be helpful to know. Thanks.

 

[takesome1]

Compare it to the 70-200 f2.8 LII, and I think one should be happy with the new f4 LII.

The difference between I and II isn’t mind blowing, but given the quality of the I, I don’t think that would be possible. The II seems very good too me, but I don’t think people will upgrade from the I because of IQ differences.

+1
The published MTFs on the Canon site are pretty similar so not much difference is expected. It's also best not to stare at the TDP charts and look for minor differences in one pair of lenses. TDP is a fine site but the image comparisons aren't set up to be a fine statistically satisfying analysis.

What does this mean? Would be helpful to know. Thanks.

AlanF can explain his meaning.

TDP only samples a few lenses on a population of thousands and thousands of lenses, it is best to read the TDP review to understand what you are looking at.
Notice the review of the 24-70mm f2.8L II shot with the 1Ds III.
There are actually 4 different copies used, partially because of copy variations and problems with new lenses.

https://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=787&Camera=453&FLI=0&API=0&LensComp=0&CameraComp=0&SampleComp=0&FLIComp=0&APIComp=0

While AlanF is correct, TDP does however puts out the effort to find a copy of the lens tested that at least performs as you expect it should.

 

[sanj]

Thank you!

 

[AlanF]

What do I mean about image testing on TDP? Well, Bryan takes shots of a standard chart and posts crops of a few of the coarse regions on it. It's OK for a quick look and a rough estimate of what the lens is like. If it's a lens that is very weak in regions, like the corners, or isn't very sharp in general, you will see it. But, if the lens is reasonably sharp, it's difficult to see how sharp it really is because he doesn't show the parts of the chart with very finely spaced lines. So, it's difficult to compare lens that are pretty good and say that one is better than the other.

Where he has tested more than one lens, he does show the results of multiple copies, but that is rare.

So, usually only one copy is tested and it is at a coarse level of just the thick lines on charts.