Views on Sigma 18-125mm f/3.8-5.6 DC OS HSM for Sony

[alphaomega]

Briefly viewed a review in UK Amateur Photographer of the above Sigma 18-125 against Nikon's 18-105. The Sigma was rated best at the longer end and a competent performer as well as being of a better construction than the Nikon.

Has any member of this forum used the Sigma and have you any views on its optical performance? Theoretically it ought to perform better than the Sigma 18-250mm f/3.5-6.3 DC OS HSM as well as the Tamron of equivalent length but theory does not always play out in practice.

I ask the question as this lens might suit some of my requirements, but if it does not measure up I am not interested. I noted that the Sony model is not "loaded" with the Sigma anti-shake mechanism. That is a distinct advantage IMHO.

[bossel]

I remember long time ago I also looked at this lens but found many negative views (sorry don't remember where). I now often use the Sony 18-250 as walkaround lens.

[David Kilpatrick]

The 18-125mm OS lens may well be good, but the 18-125mm lens sold for Sony is the original - the very first design of this type ever to appear, before even the 18-200mm and longer lenses. I had one in 2005 for my KM 7D. It's a very sharp lens, but the distortion is overwhelming, maybe the worst pincushion distortion of all such lenses at medium to tele range. Only for this one reason I can't recommend it.

I'm prepared to bet that the new design has much lower distortion, like the Sigma 18-250mm OS, and also that AP have not taken distortion into account as heavily as sharpness when judging the lenses. The Nikon 18-105mm VR is a 2008 design and surprisingly good.

David

[alphaPDX]

The 18-125mm OS lens may well be good, but the 18-125mm lens sold for Sony is the original - the very first design of this type ever to appear, before even the 18-200mm and longer lenses. I had one in 2005 for my KM 7D. It's a very sharp lens, but the distortion is overwhelming, maybe the worst pincushion distortion of all such lenses at medium to tele range. Only for this one reason I can't recommend it.

I owned the original briefly, but sent it back - vignetting and edge sharpness were worse than the 18-70 kit, so why should I bother with a slightly longer mediocre lens? Later I picked up their 18-200 and found it quite satisfactory.

[bakubo]

It's a very sharp lens, but the distortion is overwhelming, maybe the worst pincushion distortion of all such lenses at medium to tele range. Only for this one reason I can't recommend it.

I'm prepared to bet that the new design has much lower distortion, like the Sigma 18-250mm OS, and also that AP have not taken distortion into account as heavily as sharpness when judging the lenses. The Nikon 18-105mm VR is a 2008 design and surprisingly good.

These days with PTLens and other distortion correction software I am less concerned about distortion since it is so easy to correct -- if the lens is in the database. Sharpness and CA seem to be more important. Of course, having it all is great, but if they have to compromise something then distortion is the easiest to deal with. I realize that the correction involves some interpolation, but I haven't noticed any big deleterious effects.

David, I would like to hear your thinking on this though. I know that you have often mentioned distortion as being one of your very important considerations.

[David Kilpatrick]

I tend to be a very precisely aligned sort of photographer - you will find me correcting 0.25° tilts in pictures with distinct horizons, and studying the balance of converging verticals very carefully to ensure that the final result is what I want. Distortion tends to screw up my ability to judge horizon level, in the viewfinder from the start. So the less the better.

Also, my normal workflow would call for distortion correction in PP (I do not use DxO, the database does not contain many of my regular combinations and it is extremely slow and tedious). It's more time. A good geometry saves time.

David

[alphaomega]

Thanks for the comments so far.

The 18-125mm OS lens may well be good, but the 18-125mm lens sold for Sony is the original - the very first design of this type ever to appear, before even the 18-200mm and longer lenses.

Looking at http://www.sigma-imaging-uk.com/lenses/ ... 25mmOS.htm there is no evidence that the Sony model is of an earlier design. From the technical specification you get the impression that the Sony model is identical to CAN/NIK except for OS. Could that be wrong and are we Sony users being fobbed off with an "old timer"? Maybe I should simply ask Sigma UK to clarify. Anyone with a name and e-mail address I could use?

[KevinBarrett]

I fixed your quote there, alphaomega. I think what is meant is that while the Nikon and Canon variants were later revised to add OS, and likely to add other improvements as well, the Sony did not receive this revision and remains an older design.

[David Kilpatrick]

I am wrong here - if the lens being sold for Sony is the new HSM, then it's an OS with the OS omitted, and quite a bit larger than the original version, 67mm thread as opposed to 62mm, 500g versus 385g, additional couple of elements in the design. The price is extremely low for an OS zoom at only a little over £300.

Sigma has different types of HSM, the cheapest does not allow manual focus over-ride (even with DMF enabled in the camera) and the best, even on Sony, allows Canon-style full time manual focus without setting anything - just touch the focus ring and you can immediately over-ride AF.

It would certainly be worth checking the lens out if both Sony and Pentax, as shown, are HSM equipped and just have disabled OS units. When the lens was first introduced, the Sony and Pentax versions remained non-HSM and were presumably the older model. But it's an early Sigma OS as far as I know - one of the first - and pre-dated their discovery that Sony had HSM compatibility. They now seem to be updating that aspect. I just wish they would do a new 12-24mm with HSM.

David

[Fotogeorge]

I believe this is the correct lens in question: http://www.bhphotovideo.com/c/product/5 ... l#features

The features impress me, because of the SLD (Special Low Dispersion) glass, a molded glass aspherical lens and two hybrid aspherical lenses, super multi-layer coating, and the lens has a minimum focusing distance of 13.8" at all focal lengths and has maximum magnification of 1:3.8. I was pondering the purchase of the Sigma 50-200 mm OS HSM lens, but maybe the 18-125mm is a better lens. I wonder if it has a better build and optical qualities. I wish it was a 2.8 lens. Maybe someone in the forum, who has any access to Sigma lenses, could check out the current lenses in a camera shop.

[bakubo]

Also, my normal workflow would call for distortion correction in PP (I do not use DxO, the database does not contain many of my regular combinations and it is extremely slow and tedious). It's more time. A good geometry saves time.

Fortunately, with PTLens I can do a pretty much automated correction in a second or two. All the lenses I use for both Sony and Canon are in the PTLens db so it works well for me. No disagreement about it being nice for the lens to not need any correction. Not knowing anything about lens design I am guessing that there are lots of trade-offs concerning things like sharpness, CA, various kinds of distortion, size, weight, and price. These days the kinds of distortion that can easily be corrected with software means that it isn't much of an issue for me anymore. I would rather have to correct it if it means the lens is smaller, lighter, cheaper, and sharpness/CA are better rather than have more compromise in some other area(s) to get better distortion correction.

[bakubo]

My wife has the original Sigma 18-125mm on her A100. I bought it in 2005 to use as a walkaround/travel lens for my KM 7D. It is nice and compact. No complaints about it. I also used to have one in Canon mount that I bought in 2004.

[stevecim]

there is a review of this lens on a canon body here

http://www.photozone.de/canon-eos/407-s ... 3856os_50d, center looks good the whole range, corners looks a bit soft at the 125mm end.

maybe someone with better understanding can comment on the test results ?

[David Kilpatrick]

First of all it looks as if the distortion problem with the original design has been fixed at the long end, but there is more present at 18mm.

For an example of the earlier lens, see:


http://www.pbase.com/davidkilpatrick/image/42373496

Secondly, the Canon 1.6X format apparently shows soft edges/corners at 125mm, this will only be worse on 1.5X.

David