FAQ FAQ  Forum Search   Events   Register Register  Login Login

Sony 28-75 2.8 vs CZ 24-70&others - artaphot test

Page  <1 78910>
Author
edwardkaraa View Drop Down
Senior Member
Senior Member

Joined: 05 February 2009
Country: Lebanon
Location: Thailand
Status: Offline
Posts: 552
Post Options Post Options   Quote edwardkaraa Quote  Post ReplyReply Direct Link To This Post Posted: 07 December 2009 at 19:54
It seems you agree with me completely, since I am talking strictly IQ, and you have confirmed my statement in your post.

I also fully agree with you about the points you are raising, namely size, weight and no filters

Originally posted by hornblade hornblade wrote:

Originally posted by edwardkaraa edwardkaraa wrote:

In any case, except from maybe the Nikon 14-24 (and the old Contax 35-70), no zoom lenses are good enough for landscape photography.
.


I don't really agree with this statement Edward. I'm not familiar with the contax, but the nikon 14-24 is hardly an ideal landscape lens. The deal breaker for me was its lack of support for filters. Polarizers are important even on ultrawides, and I use both graduated and solid ND filters a fair amount as well. Additionally, the lens is very large and heavy.

Optically, the 14-24 is a knockout, and definitely sharper in the extreme corners than the cz 16-35. Ultrawide zooms are notorious for being difficult to get right, and Nikon did do an amazing job with this lens. That said, I'm not sure the compromises they made (large bulbous front element, heavy size, etc.)   are worth it for very many landscape photographers. If they were, I think we would see more landscape photogs using nikon & the 14-24, but the lens seems to get most use in reportage.

I realize most of this discussion has been about optical performance, but I just wanted to mention that it's only one component of how well suited a lens is for a particular area of photography.
   
A900 ZA 24/2 85/1.4 135/1.8 24-70/2.8
 



Back to Top
jerome View Drop Down
Senior Member
Senior Member

Joined: 03 June 2009
Location: Germany
Status: Offline
Posts: 805
Post Options Post Options   Quote jerome Quote  Post ReplyReply Direct Link To This Post Posted: 07 December 2009 at 20:39
Originally posted by edwardkaraa edwardkaraa wrote:

The Leica 28/2.8 (latest version) is very sharp in the corners wide open. The Zeiss 28/2 (ZF/ZE) has a strong field curvature, but it is better than the ZA zoom in most of the frame. Zeiss 21/2.8 and Nikon 14-24 are both stellar wide open. Zeiss 25/2.8 and 35/2 are also excellent. Just to name a few.


The Leica 28 indeed is a counter-example, and probably the only one. As was to be expected, it uses asphericals to correct from coma, which degrades the outside areas of the frame in spherical design.

I suggest that you compare the published MTF for the Zeiss 28mm prime and the Sony CZ zooms to see which one is better (here). This being said, of all the Zeiss wide-angle, the 21mm is the most impressive. The old Minolta 20mm is a clever design, but the Zeiss is better.

The Nikon 14-24 is not a prime and does not do 28mm. Still, it is an optical tour-de-force and, as noted, not that convenient in the field.
9000-9xi-5D-A900A850-20f28-CZ24-28f20-35f14-50f17-50f14-50mf28-85f14-100mf28-100SF-135f28-STF-CZ135f18-200f28-x1.4-∑12/24-CZ16/35-CZ24/70-70/200G-70/300G-500f8-F58
Back to Top
edwardkaraa View Drop Down
Senior Member
Senior Member

Joined: 05 February 2009
Country: Lebanon
Location: Thailand
Status: Offline
Posts: 552
Post Options Post Options   Quote edwardkaraa Quote  Post ReplyReply Direct Link To This Post Posted: 07 December 2009 at 21:04
Jerome,

The problem is that you are still insisting on comparing Zeiss to Sony's MTF.

Sony's MTF are not only theoretical, they are also ridiculous. Have you ever seen a lens resolving close to 100% at 40 line pairs per millimeter (and on 80% of the frame)? Zeiss does not recognize Sony's MTF and Sony forbids Zeiss from posting the real MTF of the ZA lenses on Zeiss website.

Besides, supposing the MTF are comparable, those of the 28 are at f/2 and f/5.6, while those of the 24-70 are at f/2.8 and f/8. How can you compare that?

And, in my post that started this discussion, I was asking Sony to develop Zeiss primes which are better than the zoom. If Zeiss can make a zoom which is so good, it can certainly make a prime which is better. What do you have against this idea? It's not like you are forced to buy it if such lens(es) was ever introduced
A900 ZA 24/2 85/1.4 135/1.8 24-70/2.8
Back to Top
Vivec View Drop Down
Senior Member
Senior Member
Knowledge Base Contributor

Joined: 17 October 2008
Location: United States
Status: Offline
Posts: 957
Post Options Post Options   Quote Vivec Quote  Post ReplyReply Direct Link To This Post Posted: 07 December 2009 at 21:23
First of all, thanks to Stephan and Jerome for their extensive testing.

Nevertheless, I have to admit that I am quite skeptical. For example, I feel that in all of the tests here there is often a corner comparison that is at a different distance than the center (ie. the dome test, and corner on the hill).

In my own experience and testing, I have not found the 'bad' corners nor the field curvature (see here). Sure, there is probably some field curvature (as all lenses have it) but as one can see from my test, it is not nearly as pronounced as suggested here and will usually fall within the depth-of-field.

Moreover, if I look at the SLR gear figures, there nothing that suggests anything as bad as suggested here. The lens actually does really good with about 4 blur units in the corner at F2.8/24:
http://www.slrgear.com/reviews/zproducts/sony24-70f28za/ff/tloader.htm

I agree with Edward, it would be nice to have an even better prime that has less distortion and an even flatter field to get even sharper corners -- but is also light weight and small.
Back to Top
Vivec View Drop Down
Senior Member
Senior Member
Knowledge Base Contributor

Joined: 17 October 2008
Location: United States
Status: Offline
Posts: 957
Post Options Post Options   Quote Vivec Quote  Post ReplyReply Direct Link To This Post Posted: 07 December 2009 at 21:27
Originally posted by jerome jerome wrote:


To a certain degree, all wide-angles exhibit some amount of field curvature. Uncorrected, it shows on the MTF as a somewhat parabolic curve (the interested reader my refer themselves to the published MTFs for the Sigma 20mm f/1.8 for example).


Hmm, I don't think you can conclude this from an MTF figure. When a figure drops down at the end in a parabolic fashion, it just means that the sharpness/contrast is less further away from the center. Now, this might be field curvature but that can only be verified by showing that the sharpness is increased if measuring at a distance in front or beyond the center distance. This must be done separately and cannot be concluded from the MTF alone. for example, it can also be that there is no field curvature but the lens is just bad away from the center.
Back to Top
jerome View Drop Down
Senior Member
Senior Member

Joined: 03 June 2009
Location: Germany
Status: Offline
Posts: 805
Post Options Post Options   Quote jerome Quote  Post ReplyReply Direct Link To This Post Posted: 07 December 2009 at 22:07
Originally posted by edwardkaraa edwardkaraa wrote:

Sony's MTF are not only theoretical, they are also ridiculous. Have you ever seen a lens resolving close to 100% at 40 line pairs per millimeter (and on 80% of the frame)? Zeiss does not recognize Sony's MTF and Sony forbids Zeiss from posting the real MTF of the ZA lenses on Zeiss website.


I don't think that Sony would take the legal risks to post blatantly false data on their official web site.
9000-9xi-5D-A900A850-20f28-CZ24-28f20-35f14-50f17-50f14-50mf28-85f14-100mf28-100SF-135f28-STF-CZ135f18-200f28-x1.4-∑12/24-CZ16/35-CZ24/70-70/200G-70/300G-500f8-F58
 



Back to Top
jerome View Drop Down
Senior Member
Senior Member

Joined: 03 June 2009
Location: Germany
Status: Offline
Posts: 805
Post Options Post Options   Quote jerome Quote  Post ReplyReply Direct Link To This Post Posted: 07 December 2009 at 22:09
Originally posted by Vivec Vivec wrote:

Originally posted by jerome jerome wrote:


To a certain degree, all wide-angles exhibit some amount of field curvature. Uncorrected, it shows on the MTF as a somewhat parabolic curve (the interested reader my refer themselves to the published MTFs for the Sigma 20mm f/1.8 for example).


Hmm, I don't think you can conclude this from an MTF figure


I am not saying that. I say that if a wide-angle lens would not be corrected at all for field curvature, the MTF would be a parabolic curve, not the other way around.
9000-9xi-5D-A900A850-20f28-CZ24-28f20-35f14-50f17-50f14-50mf28-85f14-100mf28-100SF-135f28-STF-CZ135f18-200f28-x1.4-∑12/24-CZ16/35-CZ24/70-70/200G-70/300G-500f8-F58
Back to Top
Vivec View Drop Down
Senior Member
Senior Member
Knowledge Base Contributor

Joined: 17 October 2008
Location: United States
Status: Offline
Posts: 957
Post Options Post Options   Quote Vivec Quote  Post ReplyReply Direct Link To This Post Posted: 07 December 2009 at 22:16
Originally posted by jerome jerome wrote:

Originally posted by edwardkaraa edwardkaraa wrote:

Sony's MTF are not only theoretical, they are also ridiculous. Have you ever seen a lens resolving close to 100% at 40 line pairs per millimeter (and on 80% of the frame)? Zeiss does not recognize Sony's MTF and Sony forbids Zeiss from posting the real MTF of the ZA lenses on Zeiss website.


I don't think that Sony would take the legal risks to post blatantly false data on their official web site.


Well, it seems Canon, Nikon, and Sony all post 'theoretical' MTF figures, ie. calculated by a computer model of the lens. In that sense it is not 'false', it is just shows something different than you might have expected

It is still useful in my opinion to compare lenses across the line of a manufacturer -- just not that useful to compare to Zeiss MTF's.
Back to Top
Vivec View Drop Down
Senior Member
Senior Member
Knowledge Base Contributor

Joined: 17 October 2008
Location: United States
Status: Offline
Posts: 957
Post Options Post Options   Quote Vivec Quote  Post ReplyReply Direct Link To This Post Posted: 07 December 2009 at 22:17
Originally posted by jerome jerome wrote:

Originally posted by Vivec Vivec wrote:

Hmm, I don't think you can conclude this from an MTF figure


I am not saying that. I say that if a wide-angle lens would not be corrected at all for field curvature, the MTF would be a parabolic curve, not the other way around.


Right -- I misunderstood your first post.

Edited by Vivec - 07 December 2009 at 22:18
Back to Top
redmalloc View Drop Down
Senior Member
Senior Member

Joined: 16 July 2006
Country: United States
Location: United States
Status: Offline
Posts: 855
Post Options Post Options   Quote redmalloc Quote  Post ReplyReply Direct Link To This Post Posted: 07 December 2009 at 22:35
Originally posted by rockhound rockhound wrote:

I don't think it is correct to say that the old Tamron and new Sony are sharing the same optical design, unless there is info I am not aware of.


As indicated on Stephan's site, the Sony and Tamron share the same number of elements and groups. More to the point, in the test he also states, "The Sony AL 2.8/28-75mm SAM (obviously a clone of the Tamron / Minolta 2.8/28-75mm)" would indicate that he also see the performance as the same.

So if the lens has the same design and performs the same, it seems like a safe bet that it is the same.
T: 17-50 70-200 100-300

M: 7D 16 28-135 50M 85 70-210

S: A700
Back to Top
zoidberg View Drop Down
Groupie
Groupie

Joined: 14 January 2009
Location: Sweden
Status: Offline
Posts: 46
Post Options Post Options   Quote zoidberg Quote  Post ReplyReply Direct Link To This Post Posted: 08 December 2009 at 17:52
Originally posted by redmalloc redmalloc wrote:

Originally posted by rockhound rockhound wrote:

I don't think it is correct to say that the old Tamron and new Sony are sharing the same optical design, unless there is info I am not aware of.


As indicated on Stephan's site, the Sony and Tamron share the same number of elements and groups. More to the point, in the test he also states, "The Sony AL 2.8/28-75mm SAM (obviously a clone of the Tamron / Minolta 2.8/28-75mm)" would indicate that he also see the performance as the same.

So if the lens has the same design and performs the same, it seems like a safe bet that it is the same.

Also, Sony's and Tamron's brochures both have pictures of the optical design. The lens elements look exactly the same, so it is fairly obvious that the Sony is a derivative of the Tamron.
Back to Top
Vivec View Drop Down
Senior Member
Senior Member
Knowledge Base Contributor

Joined: 17 October 2008
Location: United States
Status: Offline
Posts: 957
Post Options Post Options   Quote Vivec Quote  Post ReplyReply Direct Link To This Post Posted: 09 December 2009 at 07:51
Ok, I am still not really buying the whole field curvature story.
Since a few days, I happen to be the happy owner of a Minolta 35mm F2 lens, so I thought I repeat my experiment at 35mm. Shame this is not at 24mm, but I don't have any other lens at that focal length -- but, on the good side, the 35mm F2 is one the best lenses from the Minolta line.

About the test: first I calibrated both my Zeiss and Minolta lens for back/front focus. Ouch, the Minolta 35mm is the first lens that needs a correction (of -10) on my A900. I took all photos at 3 meters (9') from a wall with an elinchrom light at 1/160s at iso 100. (except at F8, where I used iso 200). All shots are at standard lightroom import settings with equal white balance (so we can look for the Minolta colors ). Everything on tripod + MLU.

Here is the scene:


center at F2.8:


corner at F2.8:


center at F8:


corner at F8:


What can I say, it seems that both lenses are equally sharp in the center (perhaps the Minolta has the edge?), and that the Zeiss zoom is sharper in the corners at wide apertures (both are very good though). I see no evidence at all of field curvature -- in contrast, it seems both have a very flat field at this distance.

Personally, to me it seems that the Zeiss has amazing corner sharpness and I think that some of the weak corners found may be due to different distance than the focus distance..
Back to Top
jerome View Drop Down
Senior Member
Senior Member

Joined: 03 June 2009
Location: Germany
Status: Offline
Posts: 805
Post Options Post Options   Quote jerome Quote  Post ReplyReply Direct Link To This Post Posted: 09 December 2009 at 08:03
Even if I did not publish them, I also took pictures with the 24-70 at 24, 35 and 50mm. The zone of "corner unsharpness" is biggest at 24mm (*) and gradually goes smaller (that is not surprising, this is how zoom lenses work). At 35mm, it is almost entirely gone.

As to field curvature, it is quite possible from you test to infer that the prime may have some. It is not possible to be sure, for that you would need to take a few pictures at f/2.8 with varying focus and see wether the corner sharpness increases when the center is a bit defocussed.


(* corner sharpness is poor at 24mm, but still better than on the 16-35 at the same focal length; corner sharpness is catastrophic at 24mm on the 24-105)

Edited by jerome - 09 December 2009 at 08:07
9000-9xi-5D-A900A850-20f28-CZ24-28f20-35f14-50f17-50f14-50mf28-85f14-100mf28-100SF-135f28-STF-CZ135f18-200f28-x1.4-∑12/24-CZ16/35-CZ24/70-70/200G-70/300G-500f8-F58
Back to Top
ficofico View Drop Down
Groupie
Groupie

Joined: 07 December 2009
Location: Italy
Status: Offline
Posts: 55
Post Options Post Options   Quote ficofico Quote  Post ReplyReply Direct Link To This Post Posted: 09 December 2009 at 08:03
Because it's for testing, can you post the entire jpeg photo of the zeiss 24-70 at 2,8 at full resolution? Thanks.
Back to Top
Dyxum main page >  Forum Home > Equipment forums > Lens Talk > A-mount lenses Page  <1 78910>

Forum Jump Forum Permissions View Drop Down



This page was generated in 0.109 seconds.

Monitor calibration strip

Dyxum.com - Home of the alpha system photographer

Find us on Google+

Feel free to contact us if needed.

Links monetized by viglink VigLink