New Sigma 12-24?

[deathvalleydave]

Hi,

Can anyone, by chance, tell me when the new Sigma 12-24 lens will be available? And, having the "old" Sigma 12-24, does anyone know if this new version will be optically better?

Thanks for any input here,
Dave

[David Kilpatrick]

We discussed this with Sigma at Focus, on Monday. The new lens is recomputed, based on the 8-16mm which is an improved revision of the original 12-24mm. It's two steps removed, in other words. It is going to be available in Nikon fit first, maybe 2 months, Alpha in the summer they think.

David

[deathvalleydave]

Hi David,

By "improved revision" and "two steps removed", could you please explain what this means in image quality? For instance, the thing I would be most interested in is the resolution figures across the different apertures. I remember the original Sigma 12-24 getting 3 1/2 stars for optical quality at Photozone - and that's pretty good, but will the resolution figures also be better on the new revision? I'm sure they won't be anything like Nikon 14-24 good, but my main question is in practical terms, will the resolution be any better? I'm not quite sure what chromatic aberration is, and if I even notice it - but I think everybody notices what resolution is, i.e. how sharp images look - for instance, with the new Sigma 12-24, would you still use it mainly at F11? Or, will it be as sharp edge to edge at F8 now?

Thanks,
Dave

[David Kilpatrick]

Judging by the 8-16mm, which we have, the new 12-24mm will be perfect at f/8 rather than f/11, and will not have the same decentering issues (most 12-24mms are sharper one side than the other). As with most aspherical corrections, it's a complex trade off between flatness of field and straight-line geometry. If you keep the line rendering dead straight, it is at the expense of planar (flat field) focus.

When a curved focus plane is symmetrical, you can average the focus at normal apertures and it works. But if the lens is not mounted dead parallel to the film, or the camera mount has an error, a curved focus plane will exaggerate loss of sharpness at one side of the image or the other. The old 12-24mm has very good geometry but a curved focus field, and suffers from this effect unless absolutely perfectly centered and mounted.

The 8-16mm has less perfect geometry, but that can be corrected using ACR/LR lens profiles to be perfect. In exchange, it has a flatter focus field, and also appears to be better engineered.

Assuming the neew 12-24mm follows this pattern, it will probably not be quite as awesome for straight line rendering at 12mm, but it will be sharper and the sharpness will also be even over the frame - not biased to one side or the other of the shot.

David

[deathvalleydave]

David, many thanks for your reply, Dave.

[deathvalleydave]

Hi David,

Just had one more question on this topic. Could you by chance please describe how I can test the "decentering" issue that my Sigma 12-24 has? The only thing that I've noticed is the soft corners, distortion at 12mm, vignetting, and a tendency to underexpose on my A900. I'd be really curious if there's a simple test to perform, that will show the decentering effect, for instance is it most prominent at 24mm vs. 12mm? But, for starters, how can I even see what this is, in practice?

Thanks!
Dave

[David Kilpatrick]

Not all Sigma 12-24mms have decentering, but most show it. Basically if you shoot something at full aperture, and either the left (normally this side) or right (unusual) side of the shot looks less well focused, maybe a bit smeared, that's the issue. If any loss of focus or sharpness is perfectly balanced to all four corners, then the lens has no centering issued and is a good example - probably.

David

[deathvalleydave]

Hi David,

I performed a test with my Sigma 12-24 and A900:

I tested at 12, 15, 17, 20, and 24mm at F4.5, 5.6, 6.3, 7.1, 8, 9, 10, and 11. I turned Image Stabilization off. I used 2 sec. timer drive mode with tripod. I put three pieces of paper on a wall. The pieces of paper had large text written on them. I then looked at all the images and saw the normal softer and more out of focus images from F4.5 through F8. I did not seem to see that one side (left) was more out of focus than the other - it's possible, but if so, it would be hard to measure. I also noticed that from 17 - 24mm there was less distortion - so, these seem to be all the standard issues with a 12-24mm lens.

So, based on the results of my test, would you say that when the new Sigma 12-24 lens comes out, it would be worth it to trade in my current 12-24 for it? I remember asking you if it was worth paying $1899 for the Zeiss 16-35 - and at the time, I made the decision to keep my Sigma 12-24. Now, the new Sigma might be under $1000, but will the performance of the new Sigma be worth it to trade up? Maybe this is a "dumb" question to ask, but I was hoping you would help me out with this question.

Thanks,
Dave

[David Kilpatrick]

You really can't test the Sigma 12-24mm with a chart on a wall. It is not performing optimally at that distance, and the distortion and flatness of field will both be worse than at infinity. The 12-24mm should be tested on distant subjects, not a chart.

We will have a copy of the new 12-24mm as soon as it is available. It will take me ten minutes to find out whether the HSM and the new design make changing from the old model worthwhile. I've used four or five different 12-24mms on three different camera systems over the years, and not one has been perfect - but I had equally bad results from, for example, the Canon 16-35mm MkII on 5D MkII (almost exactly the same type of loss of sharpness at one side of the image, wide open, distant subjects). Therefore I don't panic about my 12-24mm being poor, I just stop it down as I am used to seeing mediocre full aperture results from even the best and most expensive such lenses. Exception: Nikon 14-24mm, used several copies, all wonderful.

Tokina has a new 16-28mm f/2.8 coming out but so far it is Nikon and Canon only, at £1060. They tell me it has been built for quality, without regard to keeping the size or weight down. If this appears for Sony it will be a third option for full frame.

David

[deathvalleydave]

I understand that my wall chart test was an example of an extreme stress test on the Sigma 12-24, and it really did show the flaws at the corners, but what it did show, was that based on what I saw, I can pretty much deduce that shooting at F11 on distant subjects will be pretty rock solid for me. Maybe my question for you will turn out being, "is it worth trading in for the new Sigma 12-24, if you're really just using it at F11?" I know you mentioned it might be better at F8 - so, that will be interesting, but at any rate, I look forward to your findings with the new lens!

[deathvalleydave]

Hi David, I had another question related to this lens/body combo - for correcting lens distortion, does LR3 provide good support? Or, do you really need to get DXO for the best support for distortion at 12mm on the A900?

Thanks,
Dave

[David Kilpatrick]

Sigma's profiles are comprehensive and fairly good. LR3 does an excellent job. Some profiles are not very good and over-correct vignetting, but the 12-24mm is not one of these.

David

[deathvalleydave]

Hi David,

Looks like the new Sigma 12-24 was reviewed on Photozone, with the 5DMkII, and it only got 2 1/2 stars for optical quality! I have the old Sigma 12-24, and based on this review over at Photozone, would you agree it would be pointless for me to "newer" Sigma 12-24? The earlier version of the Sigma 12-24 was reviewed on an A700 body, so maybe the score would have been lower, if reviewed on the camera I have, the A900.

I guess the thing that would interest me most is whether or not the extreme borders are sharper on the new Sigma 12-24. For instance, I believe you said best overall edge to edge results were at F11 on the old Sigma 12-24 - is it any better on the new one? Is F8 on the new one any better?

Thanks,
David

[David Kilpatrick]

I am still waiting for the new one to review. It has HSM, which the old one lacks. I am surprised that it did so badly - the new 8-16mm is an amazingly good lens, we have had three different samples to review and I bought one ex-demo for a good price (not from Sigma, one just turned up in Alpha mount at a dealer).

I have been assuming the advance in quality made in the 8-16mm would be carried through to the new version 12-24mm. If not, then that's another nail in the coffin for full frame working, for me.

David