85/2.8 35/1.8 24/2 MTF curves on sony Japan

[Old Hydro]

[quote="David Kilpatrick"]OK, don't feel bad about not being able to afford the 24mm f/2 now!

David,

I don't know how to read these charts either. What is that you see about the 24mm F2 that you would not buy it now?

[David Kilpatrick]

It's not very good in terms of even coverage across the frame. Looks very good for APS-C but frankly a bit disappointing for full frame. High contrast however. The MTF chart shown is typical of lenses with compensated geometry (moustache shape distortion), slightly less than flat fields, and visible vignetting.

David

[dynax800si]

Here looks nice vs 24-70



A900: Zeiss ZA 2/24mm, 16-35mm, 24-70mm, MinAF 2.8/24mm, 3.5/17-35mm G

[agorabasta]

The tester should have used a manually focused series for the corner samples just in order to evaluate the effects of field curvature for both lenses tested. Then the centre/corner crop pairs should have been presented for both the centred focus and best match corner focus.
(We really need an LV FF body just to have the ability to MF precisely when needed...)

[Old Hydro]

It's not very good in terms of even coverage across the frame. Looks very good for APS-C but frankly a bit disappointing for full frame. High contrast however. The MTF chart shown is typical of lenses with compensated geometry (moustache shape distortion), slightly less than flat fields, and visible vignetting.

Thank you for explaining the chart. Do you think there will be improvement with the production version, or is the chart it?

I have to say it looks a lot sharper then the Zeiss 24-70 photos posted in this thread. I have wondered if it would be better, and to my eye is seems to be.

I also have an old Minolta 24mm from around 1987, and was thinking of the new 24mm as a good replacement. Now I'm not so sure the cost is worth it.



David

[David Kilpatrick]

First of all, the 24-70mm looks terrible. I think my 24-85mm old Minolta is better, though of course it does not do f/2.8. The MTF chart shown for any lens is the theoretical computed perfect performance. That's what an absolutely perfect one does.



This is the MTF for the Sigma 24mm f/1.8, and as you can see, it's miles worse than the CZ lens but with a very simple fall off in sharpness - meaning it's not all that wonderful even for APS-C wide open. The Sigma chart shows 10 and 30 lppmm wide open. Of course, the Sigma can be stopped down to f/2 and may well show a dramatic clear-up at that point. I have used it and actually found it not too bad (but not good enough to buy!).



The Sony CZ is very odd, the up and down wavering of the contrast indicates it has a field flatness like a pond with ripples, and you will get the same sort of effect as the Canon 24-105mm f/4 L series lens where patches of a shot seem less sharp than others, especially is the focus distance works against the lens field.

The lens should be superb stopped down, and Sony's chart shows 10/20/40 lppmm. 40 lppmm is a very fine detail to measure contrast at, and indicates the Zeiss will be very good indeed at showing textures.

David

[David Kilpatrick]

To explain one thing further - the lines shown as dotted and solid represent (Sony terminology) Radial and Tangential resolution; Sigma use the terms Saggittal and Meridional, which mean exact the same thing.

A lens with excellent bokeh, producing an image which does not seem have any 'directional' quality to the blur or loss of sharpness, has the R and T/M and S lines as close to being together as possible. If they coincide perfectly, the lens nearly always shows a perfect bokeh and circular defocus points. The Sigma is only tested wide open, but you can see that at 30 lppmm (the green lines) the two directions of test target lines are tracking each other closely.

This also happens with the red line (10 lppmm - web or enprint resolution) on the CZ at f/8 but the highest resolution 40 lppmm shows Radial lines (those passing through the centre point of the image) to have 40% more contrast than Tangential (those at 90 degrees to any radial line). This will give very fine detail a strong directional quality at f/8, more so even than wide-open.

What normally happens is that the focus plane of the radial and tangential lines differs. Since radial lines naturally tend to have more definition, as they use more of the centre of the outermost lens elements than tangential lines, focus always prefers this tiny adjustment relative to the average - and edge or corner sensors will tend to favour radial even more than the centre focus sensor.

It's complex but means the Zeiss may tend to produce a harsher bokeh regardless of aperture circularity, even compared to the 'inferior' Sigma.

It is the balancing of these aspects which Leica has traditionally excelled in, and which for a period Minolta learned to master as well.

David

[Edgars L]

85mm bokeh on sample image as well as MTF resolution/contrast seems to be really amazing. Taking in to account small size and price it should be really excellent lens, unfortunately only for APS-C.

Zeiss claims that they publishing truly measured MTF charts instead of other (Sigma etc) who giving us computer simulated (not practically measured) MTF charts. Therefore Zeiss charts always seems to worse than others.

Regarding to Zeiss 24-70 photos posted in this thread. I noticed it some time ago, he has really bad 24-70mm lens copy. Of course Zeiss extreme corners are much worse than central part of frame, but not that much, at least for my copy. I tested it several ways and never found such a bad corners even at f2.8.

[David Kilpatrick]

No, the 85mm is full frame. Its performance over APS-C is potentially amazing.

David

[Dr. Harout]

No, the 85mm is full frame. Its performance over APS-C is potentially amazing.

David

... and how about on FF?

[David Kilpatrick]

Excellent on FF, close to perfect on APS-C if the curves are typical of the final product. It's a pity it does not have an extended focusing action which would improve its use for close-ups.

David

[catalytic]

Too bad they didn't make the 85 mm faster at say f/2. They could have made a 90/2 which would evoke memories of classic 90 mm lenses. That would represent a perfect alternative to the bigger and more expensive CZ 85/1.4 with most of the attributes people want the 85/1.4 for. But obviously, they didn't want to produce a lens that could cannibalize sales from the premium 85/1.4.

Similarly, they could have produced a 35/2 designed for FF to provide a modern version of the old 35/2. That would have been very well received, i'm sure, but took the easy way... er, easy choice approach on this one and restricted it to an APS-C design. Maybe they were protecting the 35/1.4G too? Too bad, cause it would have been a hit!

[Birma]

It is great to have some further affordable options in Sony primes, but I would have liked something wider for APS-C, e.g. 16mm. That would give us a true WA prime for the smaller sensor cameras. Don't mind if it is f/4 or f/3.5, and give it a plastic mount and SAM. Would this be a lot more expensive to make than the 35/1.8?

Perhaps next time .

[Dr. Harout]

It is great to have some further affordable options in Sony primes, but I would have liked something wider for APS-C, e.g. 16mm. That would give us a true WA prime for the smaller sensor cameras. Don't mind if it is f/4 or f/3.5, and give it a plastic mount and SAM. Would this be a lot more expensive to make than the 35/1.8?

Perhaps next time .

If the 16mm pancake lens for the NEX proves to sell quite well, I guess they will be tempted to manufacture it for APS-C too.

[UrsaMajor]

If the 16mm pancake lens for the NEX proves to sell quite well, I guess they will be tempted to manufacture it for APS-C too.

Wouldn't that have to be a completely new design?

The registration distance for the NEX is much, much shorter than for the Alpha cameras, so it appears to me that the only way that a NEX lens would focus when used on an Alpha would be with the use of an adapter, or the addition of at least one new element in the lens itself - unless I am missing something.

With best wishes,
- Tom -