24-70 GM ii vs 28-70 f2 GM vs 20-70 f4 G

Brett8883

Leading Member
Messages
944
Reaction score
346
Ahh I’m in a conundrum. I have the Sigma 24-70 DG DN Art and I just don’t like it so I want to replace with a first party lens but which one.

Professionally, I do architectural photos and real estate and so what I usually do is have my 12-24 GM on one camera and my sigma 24-70 on another. The two camera set up with those two focal length ranges works great but I just don’t like the sigma.

The 24-70 GM ii is the obvious contender to replace the Sigma but it’s just kind of a boring focal range or there’s a part of me that has trouble rationalizing replacing a like for like lens I already have.

The 28-70 f/2 is much more interesting and I can use it for portraits and other things besides work. Though it is big and heavy and I worry a bit if I’m going to miss that 25-27mm since sometimes I do take the 24-70 as my only lens for some things. Seems irrational I know.

The 20-70 f4 is also interesting. It’s very small and light and that extra 4mm on the short end might come in handy sometimes when I take it as an only lens. For my architectural/real estate work I don’t need anything faster than f/4 anyway but how does the image quality compare to the GMs? I normally wouldn’t consider a non-GM Sony but I got the 20 1.8 G and that is an excellent lens, as good or better than any GM for what I do so not sure if that extends to the zooms in the G line.

Anyway, just hoping to get some opinions or thoughts that might help me to make a decision. You guys always think of things I hadn’t.

Thanks
 
These are all super nice lenses, but VERY different in size and price. Only you can decide. But if u r willing to take along the wide zoom too, you might as well go with the 24-70GMii. If you really need an extra 2/3 stop of light, then the f2 zoom might be worth the extra size, but you sacrifice 24 (vs 28mm)
 
Yeah, think it's hard to do much besides a quick reaction. I only have the 20-70/4 of these several lenses and only for a short time. I don't have a need for critically high end "best" image quality and could find the premium pricing hard to justify. I don't have a business "need" nor does my own viewing or sharing require the highest performance levels, either.
Ahh I’m in a conundrum. I have the Sigma 24-70 DG DN Art and I just don’t like it so I want to replace with a first party lens but which one.

Professionally, I do architectural photos and real estate and so what I usually do is have my 12-24 GM on one camera and my sigma 24-70 on another. The two camera set up with those two focal length ranges works great but I just don’t like the sigma.
Two camera vs one camera one or two lenses depends on your interests, needs. I'd likely prefer lighter as "better" for my casual uses.
The 24-70 GM ii is the obvious contender to replace the Sigma but it’s just kind of a boring focal range or there’s a part of me that has trouble rationalizing replacing a like for like lens I already have.

The 28-70 f/2 is much more interesting and I can use it for portraits and other things besides work. Though it is big and heavy and I worry a bit if I’m going to miss that 25-27mm since sometimes I do take the 24-70 as my only lens for some things. Seems irrational I know.
It may be more different than the 24-70/2.8 gmii in thinking of a change from the other 24-70.

My "usual" lenses for one lens carry have been the 18-135 and more recently, 28-200. You didn't mention which body/bodies in use? Or where you might be shooting? With the 12-24 available as second lens or second body I don't think the 24mm to 28mm gap would be an issue. I seldom go wider so my impression, for me, would be even as a single lens approach, probably wouldn't be a problem.

Having said that, if in a condensed "Old City" environment, cathedrals, interiors congested twisty streets, etc. - a not usual environment for me, then the wide as a second lens could be much more useful and swapping lens more necessary - and a second camera more convenient also heavier and bulkier and all that, too.

I really spent a lot of time looking and enjoying the sights over pushing photography on a recent trip. In Switzerland, the 28-200 never came off the A7Riv. On shore excursions along the Rhine, I got a lot more use than usual of my 17-28 and likely would expect a lot of use of the 12-24 had that been my wide. A lot of swapping, two lens/two cameras would have been heavy, maybe overall more convenient. That's in the "what you do" vs "what I do" side of things so may not apply?

Maybe run a sort of used focal lengths for some trips, etc.?
The 20-70 f4 is also interesting. It’s very small and light and that extra 4mm on the short end might come in handy sometimes when I take it as an only lens. For my architectural/real estate work I don’t need anything faster than f/4 anyway but how does the image quality compare to the GMs? I normally wouldn’t consider a non-GM Sony but I got the 20 1.8 G and that is an excellent lens, as good or better than any GM for what I do so not sure if that extends to the zooms in the G line.
Post that trip, I added the 20-70/4. Very convenient, considering wider than 24mm and 28mm. I happen to get a fair amount of longer use at time but that wouldn't differ in considering 20-70, 24-70 or 28-70. I like it, haven't done critical testing. f4 works for me for travel with noise reduction, I don't typically look for shallow depth of field, perhaps might be more concerned at times to get deeper.
Anyway, just hoping to get some opinions or thoughts that might help me to make a decision. You guys always think of things I hadn’t.

Thanks
 
You didn't mention which body/bodies in use? Or where you might be shooting? With the 12-24 available as second lens or second body I don't think the 24mm to 28mm gap would be an issue. I seldom go wider so my impression, for me, would be even as a single lens approach, probably wouldn't be a problem.
I have an A1 and an A7RIV. I am thinking you are right that the gap from 24-28 would be no problem. There’s nothing that can fill the role of the 12-24 so I’m always going to have that with me. I can always crop a little bit on the 24 if I really need to.



So I think it’s between the 28-70 f2 and 20-70. If I could be sure the image quality of the 20-70 was on par with the 28-70 it would probably be an easy decision. My biggest thing is I shoot in harsh light a lot and that one of my problems with the sigma. It just folds under less than ideal lighting. I actually have the 20 G, 35GM, 50 1.2 GM, and 135 GM so really my only holes in the prime arsenal is at like 24ish mm and 85mm and then of course I need the zoom for work. I was thinking the 28-70 could fill all 3 holes decently well.

The 20-70 wouldn’t fill any holes in my prime arsenal but it’s relatively inexpensive fills my need for a workhorse zoom and is small. But I’d need to know the image quality is on par with the GMs just a bit slower.

Decisions decisions decisions…
 
These are all super nice lenses, but VERY different in size and price. Only you can decide. But if u r willing to take along the wide zoom too, you might as well go with the 24-70GMii. If you really need an extra 2/3 stop of light, then the f2 zoom might be worth the extra size, but you sacrifice 24 (vs 28mm)
 
Besides some local trips, I've used the 20-70/4 in a couple of my usual spots, one of the Spanish colonial era missions and a public garden, but these aren't ideal test conditions, the garden is seasonally different and the mission isn't "crisp." Plastered adobe so irregular surfaces, the interiors are dim so higher isos needed, etc. I've been pleased with it in comparison to several other lenses I've used in the same place. But not having the 28-70/2, can't say how they compare.
 
These are all super nice lenses, but VERY different in size and price. Only you can decide. But if u r willing to take along the wide zoom too, you might as well go with the 24-70GMii. If you really need an extra 2/3 stop of light, then the f2 zoom might be worth the extra size, but you sacrifice 24 (vs 28mm)
Sums it up well. I guess I’m looking for somebody to say the 28-70 is one of those special lenses or the 20-70 has the same IQ as the GMs it’s just slower. Otherwise, the 24-70 is a tried and true workhorse that I know works.
I have the 20-70, not had it that long. Generally pretty good but not very happy with it at around 20mm, maybe l got a bad copy?

There is a lot of distortion, it is about 9% barrel at 20mm, but l think it is more complex than that, my lens corrections don't fully work. Also in high contrast situations l get a lot CA, again the software corrections don't fully correct it. I'm also getting some white lines, think this is from CA correction, not fully resolved it yet. Also some faint black lines which seem to be from the CA edges. These are only noticeable if you zoom in, ie 50% or more.

Another thing field curvature is quite strong, so the corners are not that sharp and they don't even improve when stopping down.

l'm looking at buying the Sony 16-35 PZ, which in tests seems much better at 20mm.

As you have a 12-24, you don't need 20mm. I don't know about the other lenses you mention. I would look at some lens tests.



lf you are in the USA, why not rent one for a day and do some testing.
 
These are all super nice lenses, but VERY different in size and price. Only you can decide. But if u r willing to take along the wide zoom too, you might as well go with the 24-70GMii. If you really need an extra 2/3 stop of light, then the f2 zoom might be worth the extra size, but you sacrifice 24 (vs 28mm)
Sums it up well. I guess I’m looking for somebody to say the 28-70 is one of those special lenses or the 20-70 has the same IQ as the GMs it’s just slower. Otherwise, the 24-70 is a tried and true workhorse that I know works.
I have the 20-70, not had it that long. Generally pretty good but not very happy with it at around 20mm, maybe l got a bad copy?
There is a lot of distortion, it is about 9% barrel at 20mm, but l think it is more complex than that, my lens corrections don't fully work. Also in high contrast situations l get a lot CA, again the software corrections don't fully correct it. I'm also getting some white lines, think this is from CA correction, not fully resolved it yet. Also some faint black lines which seem to be from the CA edges. These are only noticeable if you zoom in, ie 50% or more.

Another thing field curvature is quite strong, so the corners are not that sharp and they don't even improve when stopping down.
l'm looking at buying the Sony 16-35 PZ, which in tests seems much better at 20mm.

As you have a 12-24, you don't need 20mm. I don't know about the other lenses you mention. I would look at some lens tests.

lf you are in the USA, why not rent one for a day and do some testing.
Not seeing that with my copy processing with Capture One and the embedded Sony lens corrections.

64618099393b46f99c60ba134cd8e42f.jpg

4f017b30edc54279b5d8271fe0e979e9.jpg

4154635c391c45aa852801c5c75022a3.jpg

225b8027c676412fa16c5ef1846f85c3.jpg

I have done some brick wall tests. There is definitely field curvature but the distortion isn’t an issue for me with the Sony corrections. Even at f10, focussing on your partner puts the flowerbeds out of focus a bit, just enough to draw your eye to the main subject.

Andrew

--
Infinite are the arguments of mages. Truth is a jewel with many facets. Ursula K LeGuin
Please feel free to edit any images that I post
 
Last edited:
These are all super nice lenses, but VERY different in size and price. Only you can decide. But if u r willing to take along the wide zoom too, you might as well go with the 24-70GMii. If you really need an extra 2/3 stop of light, then the f2 zoom might be worth the extra size, but you sacrifice 24 (vs 28mm)
Sums it up well. I guess I’m looking for somebody to say the 28-70 is one of those special lenses or the 20-70 has the same IQ as the GMs it’s just slower. Otherwise, the 24-70 is a tried and true workhorse that I know works.
I have the 20-70, not had it that long. Generally pretty good but not very happy with it at around 20mm, maybe l got a bad copy?
There is a lot of distortion, it is about 9% barrel at 20mm, but l think it is more complex than that, my lens corrections don't fully work. Also in high contrast situations l get a lot CA, again the software corrections don't fully correct it. I'm also getting some white lines, think this is from CA correction, not fully resolved it yet. Also some faint black lines which seem to be from the CA edges. These are only noticeable if you zoom in, ie 50% or more.

Another thing field curvature is quite strong, so the corners are not that sharp and they don't even improve when stopping down.
l'm looking at buying the Sony 16-35 PZ, which in tests seems much better at 20mm.

As you have a 12-24, you don't need 20mm. I don't know about the other lenses you mention. I would look at some lens tests.

lf you are in the USA, why not rent one for a day and do some testing.
Not seeing that with my copy processing with Capture One and the embedded Sony lens corrections.

64618099393b46f99c60ba134cd8e42f.jpg

4f017b30edc54279b5d8271fe0e979e9.jpg

4154635c391c45aa852801c5c75022a3.jpg

225b8027c676412fa16c5ef1846f85c3.jpg

I have done some brick wall tests. There is definitely field curvature but the distortion isn’t an issue for me with the Sony corrections. Even at f10, focussing on your partner puts the flowerbeds out of focus a bit, just enough to draw your eye to the main subject.

Andrew
In the lens tests the CA is quite high, certainly higher than on my Canon lenses. In some of my images, which are higher contrast than yours above, the CA is fairly extreme.

The distortion is more noticeable if you have straight lines, near to the edge of the frame, of course, especially if you tip the lens up. If l was doing architecture l would want to test one out first before buying.

l don't have Capture One but l tried DxO Pure Raw, ACR and even the Sony RAW converter.

The JPEG's we're also producing the same problems.

Crosby beach l was there last year. It was sunny all the time, for 2 days.
 
Last edited:
These are all super nice lenses, but VERY different in size and price. Only you can decide. But if u r willing to take along the wide zoom too, you might as well go with the 24-70GMii. If you really need an extra 2/3 stop of light, then the f2 zoom might be worth the extra size, but you sacrifice 24 (vs 28mm)
Sums it up well. I guess I’m looking for somebody to say the 28-70 is one of those special lenses or the 20-70 has the same IQ as the GMs it’s just slower. Otherwise, the 24-70 is a tried and true workhorse that I know works.
I have the 20-70, not had it that long. Generally pretty good but not very happy with it at around 20mm, maybe l got a bad copy?
There is a lot of distortion, it is about 9% barrel at 20mm, but l think it is more complex than that, my lens corrections don't fully work. Also in high contrast situations l get a lot CA, again the software corrections don't fully correct it. I'm also getting some white lines, think this is from CA correction, not fully resolved it yet. Also some faint black lines which seem to be from the CA edges. These are only noticeable if you zoom in, ie 50% or more.

Another thing field curvature is quite strong, so the corners are not that sharp and they don't even improve when stopping down.
l'm looking at buying the Sony 16-35 PZ, which in tests seems much better at 20mm.

As you have a 12-24, you don't need 20mm. I don't know about the other lenses you mention. I would look at some lens tests.

lf you are in the USA, why not rent one for a day and do some testing.
Not seeing that with my copy processing with Capture One and the embedded Sony lens corrections.

64618099393b46f99c60ba134cd8e42f.jpg

4f017b30edc54279b5d8271fe0e979e9.jpg

4154635c391c45aa852801c5c75022a3.jpg

225b8027c676412fa16c5ef1846f85c3.jpg

I have done some brick wall tests. There is definitely field curvature but the distortion isn’t an issue for me with the Sony corrections. Even at f10, focussing on your partner puts the flowerbeds out of focus a bit, just enough to draw your eye to the main subject.

Andrew
In the lens tests the CA is quite high, certainly higher than on my Canon lenses. In some of my images, which are higher contrast than yours above, the CA is fairly extreme.

The distortion is more noticeable if you have straight lines, near to the edge of the frame, of course, especially if you tip the lens up. If l was doing architecture l would want to test one out first before buying.

Crosby beach l was there last year. It was sunny all the time, for 2 days.
https://www.lenstip.com/643.5-Lens_...f_4_G_Chromatic_and_spherical_aberration.html

My experience is consistent with the Lenstip review. LaCA should be almost fully corrected by the Sony profile. Maybe you have a bad copy?

If you post an example of a problem image, I could try and match it with my copy and see how it processes in C1.

For what I shoot, the residual distortion is acceptable. For architecture, I’d probably use the 15/4.5, 21/2.8 or adapted EF 24mm TSE II in combination with a Tamron 28-75.

If architecture was my main interest, I’d look at other lenses.

If you tip an UWA lens up, you get perspective distortion, which is a property of the FL, not the lens. Even level, it looks odd.



00e84cfa6e154df9a0072bf3448130e0.jpg

Guess the beach and for a bonus.



6231ed0f8a574ab7aaa0bc1990914cc3.jpg

A

--
Infinite are the arguments of mages. Truth is a jewel with many facets. Ursula K LeGuin
Please feel free to edit any images that I post
 
Last edited:
These are all super nice lenses, but VERY different in size and price. Only you can decide. But if u r willing to take along the wide zoom too, you might as well go with the 24-70GMii. If you really need an extra 2/3 stop of light, then the f2 zoom might be worth the extra size, but you sacrifice 24 (vs 28mm)
Sums it up well. I guess I’m looking for somebody to say the 28-70 is one of those special lenses or the 20-70 has the same IQ as the GMs it’s just slower. Otherwise, the 24-70 is a tried and true workhorse that I know works.
I have the 20-70, not had it that long. Generally pretty good but not very happy with it at around 20mm, maybe l got a bad copy?
There is a lot of distortion, it is about 9% barrel at 20mm, but l think it is more complex than that, my lens corrections don't fully work. Also in high contrast situations l get a lot CA, again the software corrections don't fully correct it. I'm also getting some white lines, think this is from CA correction, not fully resolved it yet. Also some faint black lines which seem to be from the CA edges. These are only noticeable if you zoom in, ie 50% or more.

Another thing field curvature is quite strong, so the corners are not that sharp and they don't even improve when stopping down.
l'm looking at buying the Sony 16-35 PZ, which in tests seems much better at 20mm.

As you have a 12-24, you don't need 20mm. I don't know about the other lenses you mention. I would look at some lens tests.

lf you are in the USA, why not rent one for a day and do some testing.
Not seeing that with my copy processing with Capture One and the embedded Sony lens corrections.

64618099393b46f99c60ba134cd8e42f.jpg

4f017b30edc54279b5d8271fe0e979e9.jpg

4154635c391c45aa852801c5c75022a3.jpg

225b8027c676412fa16c5ef1846f85c3.jpg

I have done some brick wall tests. There is definitely field curvature but the distortion isn’t an issue for me with the Sony corrections. Even at f10, focussing on your partner puts the flowerbeds out of focus a bit, just enough to draw your eye to the main subject.

Andrew
In the lens tests the CA is quite high, certainly higher than on my Canon lenses. In some of my images, which are higher contrast than yours above, the CA is fairly extreme.

The distortion is more noticeable if you have straight lines, near to the edge of the frame, of course, especially if you tip the lens up. If l was doing architecture l would want to test one out first before buying.

Crosby beach l was there last year. It was sunny all the time, for 2 days.
https://www.lenstip.com/643.5-Lens_...f_4_G_Chromatic_and_spherical_aberration.html

My experience is consistent with the Lenstip review. LaCA should be almost fully corrected by the Sony profile. Maybe you have a bad copy?

If you post an example of a problem image, I could try and match it with my copy and see how it processes in C1.

For what I shoot, the residual distortion is acceptable. For architecture, I’d probably use the 15/4.5, 21/2.8 or adapted EF 24mm TSE II in combination with a Tamron 28-75.

If architecture was my main interest, I’d look at other lenses.

If you tip an UWA lens up, you get perspective distortion, which is a property of the FL, not the lens. Even level, it looks odd.

00e84cfa6e154df9a0072bf3448130e0.jpg

Guess the beach and for a bonus.

6231ed0f8a574ab7aaa0bc1990914cc3.jpg

A
I'm not keen on Lenstip testing, that is just flat charts in a studio, hardly real world in high contrast. Optical Limits have 1.2 pixel at 20mm, the Canon 16-35/4 is 0.55 pixel at 20mm.

ln high contrast images these are likely to be higher.

l don't put any of my images online,

I may do a trial of Capture 1 but too busy at the moment.
 
These are all super nice lenses, but VERY different in size and price. Only you can decide. But if u r willing to take along the wide zoom too, you might as well go with the 24-70GMii. If you really need an extra 2/3 stop of light, then the f2 zoom might be worth the extra size, but you sacrifice 24 (vs 28mm)
Sums it up well. I guess I’m looking for somebody to say the 28-70 is one of those special lenses or the 20-70 has the same IQ as the GMs it’s just slower. Otherwise, the 24-70 is a tried and true workhorse that I know works.
I bought the 20-70 when it came out, and that extra 4mm at the wide end is quite useful. It's a fairly light lens, too.

I preordered the 28-70 f/2, and it is most definitely a special lens. It's noticeably heavier, but very sharp, even wide open. I use it when photographing people. I wouldn't use it as a general purpose / walk-around lens, but when I'm photographing people, yes, definitely. You could avoid buying it if you bought the amazing 50-150 f/2, but that's a rather expensive lens.

One thing I should say is that Rashid7 is mistaken - f/2.8 is a full stop slower than f/2, not 2/3 of a stop. And the 20-70 is another full stop slower again.

The 24-70mm f/2.8 GM II is the classic choice, and it's surprisingly light for a 24-70 F/2.8, too.

Fundamentally, none of these is a bad choice :-)
 
Optical Limits have 1.2 pixel at 20mm,
Here is the full quote from them on CA:
Lateral CAs (color shadows in the outer image field) are quite well-corrected. Unsurprisingly, they are most pronounced at 20mm with an average pixel width of around 1.2px at the image borders. The CAs decrease a bit when zooming toward longer focal lengths.
the Canon 16-35/4 is 0.55 pixel at 20mm.
That is a lot less challenging design WRT FL range covered, 2,18 x (16-35) vs 3,5 x (20-70). Cherry-picking a comparison lens? Many Canon RF lenses at Opticallimts have the same CA data as the Sony you obviously are unhappy with, given the repetition rate of posts about your observations.
l don't put any of my images online,
Not even a test pic showing the issue?
 
Optical Limits have 1.2 pixel at 20mm,
Here is the full quote from them on CA:
Lateral CAs (color shadows in the outer image field) are quite well-corrected. Unsurprisingly, they are most pronounced at 20mm with an average pixel width of around 1.2px at the image borders. The CAs decrease a bit when zooming toward longer focal lengths.
the Canon 16-35/4 is 0.55 pixel at 20mm.
That is a lot less challenging design WRT FL range covered, 2,18 x (16-35) vs 3,5 x (20-70). Cherry-picking a comparison lens? Many Canon RF lenses at Opticallimts have the same CA data as the Sony you obviously are unhappy with, given the repetition rate of posts about your observations.
l don't put any of my images online,
Not even a test pic showing the issue?
I"m not cherry picking, that is what l used for 6 years before changing to Sony, it was not a RF lens.

interesting that LensTips has the distortion at 10.3% while Optical Limits is just under 9%.

l'm unhappy with some of my images at 20mm.
 
The 20-70/4 seems to be a no brainer for your work photography. My guess is you could probably take that one lens to most of your sites. It's not a GM and maybe not exciting but probably plenty of image quality for what you do.

However for your passion it should be the 28-70/2 GM. I have it and what an incredible lens. Not cheap but wow. The images look like they shot with GM primes
 
Here are a couple of shots with the 20-70. First day I had it, picked it up at Samy's in Pasadena. these are from the edge of the Eaton fire area. Raws when initially viewed in File Explorer, display some vignetting and distortion (depends some on focal length). I picked a few, reset them to original aspect ratio and only did "Auto" in Lightroom, I did export at 3840x2160. This gives corners and edges, in a consistent default approach.

I picked some 20s as that seems to be a focal length getting some concern. Not test chart controlled and maybe not optimal aperture, etc.

3641cbf09dc245d9bd7dd6228b6a845c.jpg

20mm, not a perfect dead-on shot so some angle/perspective, too.

05c86d0e95864a2092b2dc2fa508ed1b.jpg

20mm longer shot

fb362bbdcffb4a3aa0a41c6a28254f67.jpg

20mm

275bee9838de44bea1a57db2f1b58a12.jpg

70mm about same spot.
 
Last edited:
These are all super nice lenses, but VERY different in size and price. Only you can decide. But if u r willing to take along the wide zoom too, you might as well go with the 24-70GMii. If you really need an extra 2/3 stop of light, then the f2 zoom might be worth the extra size, but you sacrifice 24 (vs 28mm)
Sums it up well. I guess I’m looking for somebody to say the 28-70 is one of those special lenses or the 20-70 has the same IQ as the GMs it’s just slower. Otherwise, the 24-70 is a tried and true workhorse that I know works.
I have the 20-70, not had it that long. Generally pretty good but not very happy with it at around 20mm, maybe l got a bad copy?
There is a lot of distortion, it is about 9% barrel at 20mm, but l think it is more complex than that, my lens corrections don't fully work. Also in high contrast situations l get a lot CA, again the software corrections don't fully correct it. I'm also getting some white lines, think this is from CA correction, not fully resolved it yet. Also some faint black lines which seem to be from the CA edges. These are only noticeable if you zoom in, ie 50% or more.

Another thing field curvature is quite strong, so the corners are not that sharp and they don't even improve when stopping down.
l'm looking at buying the Sony 16-35 PZ, which in tests seems much better at 20mm.

As you have a 12-24, you don't need 20mm. I don't know about the other lenses you mention. I would look at some lens tests.

lf you are in the USA, why not rent one for a day and do some testing.
Not seeing that with my copy processing with Capture One and the embedded Sony lens corrections.

64618099393b46f99c60ba134cd8e42f.jpg

4f017b30edc54279b5d8271fe0e979e9.jpg

4154635c391c45aa852801c5c75022a3.jpg

225b8027c676412fa16c5ef1846f85c3.jpg

I have done some brick wall tests. There is definitely field curvature but the distortion isn’t an issue for me with the Sony corrections. Even at f10, focussing on your partner puts the flowerbeds out of focus a bit, just enough to draw your eye to the main subject.

Andrew
In the lens tests the CA is quite high, certainly higher than on my Canon lenses. In some of my images, which are higher contrast than yours above, the CA is fairly extreme.
Whose tests? Based on trustworthy reviews it seems there's mostly lateral CA which is easier to correct than LoCA, but yes all the correction does (on any lens) is desaturate the purple outline, leaving less visible grey behind, you'd really have to pixel peep a lot to notice though...



A 20mm prime will certainly do better and rely less on distortion correction to get there, but then you'd certainly hope so. Which Canon 20-70 lens does better? Comparing it against a 24-70 seems moot...
The distortion is more noticeable if you have straight lines, near to the edge of the frame, of course, especially if you tip the lens up. If l was doing architecture l would want to test one out first before buying.
If you tip the lens up the distortion is no longer the lens' fault, that's perspective distortion inherent to any lens this wide or wider...
l don't have Capture One but l tried DxO Pure Raw, ACR and even the Sony RAW converter.

The JPEG's we're also producing the same problems.

Crosby beach l was there last year. It was sunny all the time, for 2 days.
 
Optical Limits have 1.2 pixel at 20mm,
Here is the full quote from them on CA:
Lateral CAs (color shadows in the outer image field) are quite well-corrected. Unsurprisingly, they are most pronounced at 20mm with an average pixel width of around 1.2px at the image borders. The CAs decrease a bit when zooming toward longer focal lengths.
the Canon 16-35/4 is 0.55 pixel at 20mm.
That is a lot less challenging design WRT FL range covered, 2,18 x (16-35) vs 3,5 x (20-70). Cherry-picking a comparison lens? Many Canon RF lenses at Opticallimts have the same CA data as the Sony you obviously are unhappy with, given the repetition rate of posts about your observations.
l don't put any of my images online,
Not even a test pic showing the issue?
I"m not cherry picking, that is what l used for 6 years before changing to Sony, it was not a RF lens.

interesting that LensTips has the distortion at 10.3% while Optical Limits is just under 9%.

l'm unhappy with some of my images at 20mm.
Sounds like you might be better off with a prime or an UWA zoom then, but there's no 20-70 zoom better than the Sony, in fact there's few other 20-xx at all and there's none from Canon so...
 
These are all super nice lenses, but VERY different in size and price. Only you can decide. But if u r willing to take along the wide zoom too, you might as well go with the 24-70GMii. If you really need an extra 2/3 stop of light, then the f2 zoom might be worth the extra size, but you sacrifice 24 (vs 28mm)
Sums it up well. I guess I’m looking for somebody to say the 28-70 is one of those special lenses or the 20-70 has the same IQ as the GMs it’s just slower. Otherwise, the 24-70 is a tried and true workhorse that I know works.
I have the 20-70, not had it that long. Generally pretty good but not very happy with it at around 20mm, maybe l got a bad copy?
There is a lot of distortion, it is about 9% barrel at 20mm, but l think it is more complex than that, my lens corrections don't fully work. Also in high contrast situations l get a lot CA, again the software corrections don't fully correct it. I'm also getting some white lines, think this is from CA correction, not fully resolved it yet. Also some faint black lines which seem to be from the CA edges. These are only noticeable if you zoom in, ie 50% or more.

Another thing field curvature is quite strong, so the corners are not that sharp and they don't even improve when stopping down.
l'm looking at buying the Sony 16-35 PZ, which in tests seems much better at 20mm.

As you have a 12-24, you don't need 20mm. I don't know about the other lenses you mention. I would look at some lens tests.

lf you are in the USA, why not rent one for a day and do some testing.
Not seeing that with my copy processing with Capture One and the embedded Sony lens corrections.

64618099393b46f99c60ba134cd8e42f.jpg

4f017b30edc54279b5d8271fe0e979e9.jpg

4154635c391c45aa852801c5c75022a3.jpg

225b8027c676412fa16c5ef1846f85c3.jpg

I have done some brick wall tests. There is definitely field curvature but the distortion isn’t an issue for me with the Sony corrections. Even at f10, focussing on your partner puts the flowerbeds out of focus a bit, just enough to draw your eye to the main subject.

Andrew
In the lens tests the CA is quite high, certainly higher than on my Canon lenses. In some of my images, which are higher contrast than yours above, the CA is fairly extreme.
Whose tests? Based on trustworthy reviews it seems there's mostly lateral CA which is easier to correct than LoCA, but yes all the correction does (on any lens) is desaturate the purple outline, leaving less visible grey behind, you'd really have to pixel peep a lot to notice though...

https://phillipreeve.net/blog/review-sony-fe-20-70mm-4-0-g/#Chromatic_aberrations

https://www.lenstip.com/643.5-Lens_...f_4_G_Chromatic_and_spherical_aberration.html

A 20mm prime will certainly do better and rely less on distortion correction to get there, but then you'd certainly hope so. Which Canon 20-70 lens does better? Comparing it against a 24-70 seems moot...
The distortion is more noticeable if you have straight lines, near to the edge of the frame, of course, especially if you tip the lens up. If l was doing architecture l would want to test one out first before buying.
If you tip the lens up the distortion is no longer the lens' fault, that's perspective distortion inherent to any lens this wide or wider...
l don't have Capture One but l tried DxO Pure Raw, ACR and even the Sony RAW converter.

The JPEG's we're also producing the same problems.

Crosby beach l was there last year. It was sunny all the time, for 2 days.
I did not mention a 24-70?

The CAs l'm getting on certain images is higher than in the Phillips Reeve review. I do see he says "very high uncorrected distortion (profiles available that don’t work perfectly at 20mm)"

that is what lm finding. I note the Lenstip distortion is over 10%. Optical Limits is just 9%.

Corrections are leaving halos and faint black lines also. Sharpening of course makes this worse. Trying to remove them is difficult.

Looks like l made a mistake getting the 20-70, it was a toss up between that a 16-35/4.
 
Every lens is a compromise!
Those termed “perfect” are usually BIG and expensive.

I used the 20-70 in Andalusia last year. Great lens. But next time I would leave it behind, and replace with 16-35G, paired with 28-200

Those three weigh about 350g, 480, & 570. They are great “compromises”!

Recently I traveled through Yosemite and Mammoth Lakes with my 50-400. But when I went backpacking for 4 days, I took a D5600 with two kit lenses (18-55&70-300). I took 2 batteries, but one was enough! = great compromises

Keep it fun!
 

Keyboard shortcuts

Back
Top