Sigma 17-50mm 2.8 OS vs Sigma 17-70mm F/2.8-4 OS vs Sony 16-50mm F/2.8 SSM?

NAwlins Contrarian said:

Not sure, how it works on Sony, but 17-70mm has been sharper than 17-50mm F/2.8 on Nikon cameras. I have not tested the contemporary one's and I believe the contemporary one has the option to adjust focus at different distances, never useful for me.

If you are judging sharpness by the results you get with auto-focus, then one lens back-focusing or front-focusing, or doing so more than another, can easily create bigger differences in the resulting image sharpness than the lenses' optical differences would. If your 17-50 is back-focusing on your camera and your 17-70 is focusing accurately, that explains it. Have you tested your camera-and-lens combinations for back- and front-focus? If you really want to test lens sharpness, then you need to use a tripod and a still subject, and use magnified live view to manual focus for optimum accuracy.

Problems with back-focus and front-focus on any particular camera-and-lens combination are exactly why the the option to adjust focus at different distances (and focal lengths) may well be useful to anyone who really cares about sharpness. Sigma introducing its dock to calibrate some of its newer lenses this way was a great service to photographers. Of course, some cameras can make similar adjustments.

Click to expand...

This is the #1 reason I went mirrorless. No more focus accuracy issues like DSLRs are plagued with. As far as the 17-50, it's very sharp wide open on the LA-EA3.

[Back-focus /front-focus] is the #1 reason I went mirrorless. No more focus accuracy issues like DSLRs are plagued with.

Yeah I always wondered about cameras like the Hasselblad H5D with an 80 MP, CCD sensor: AFAIK, they relied entirely on their separate auto-focus systems, and even a very small amount of back-focus or front-focus would probably destroy the benefit of such a high-resolution sensor. (The newer medium format DSLRs like the Pentax 645Z and a Hasselblad H with a CMOS back presumably allow you to use magnified live view to nail focus, and/or can supplement PDAF with CDAF.) But I think the new top-end mirrorless cameras like the Hasselblad X1D and the Fuji GFX50S will have significant advantages in some applications.

All of that said, I have seen reports of back- and front-focus on mirrorless cameras with on-sensor phase-detection auto-focus. Seems strange to think it could happen, but I can't say conclusively that it can't.

NAwlins Contrarian said:

Not sure, how it works on Sony, but 17-70mm has been sharper than 17-50mm F/2.8 on Nikon cameras. I have not tested the contemporary one's and I believe the contemporary one has the option to adjust focus at different distances, never useful for me.

If you are judging sharpness by the results you get with auto-focus, then one lens back-focusing or front-focusing, or doing so more than another, can easily create bigger differences in the resulting image sharpness than the lenses' optical differences would. If your 17-50 is back-focusing on your camera and your 17-70 is focusing accurately, that explains it. Have you tested your camera-and-lens combinations for back- and front-focus? If you really want to test lens sharpness, then you need to use a tripod and a still subject, and use magnified live view to manual focus for optimum accuracy.

Problems with back-focus and front-focus on any particular camera-and-lens combination are exactly why the the option to adjust focus at different distances (and focal lengths) may well be useful to anyone who really cares about sharpness. Sigma introducing its dock to calibrate some of its newer lenses this way was a great service to photographers. Of course, some cameras can make similar adjustments.

Click to expand...

I am used to both front focussing and back focussing. For me it is just the matter of convenience, but NEVER noticed a difference in photo quality using two methods. I have set my focus AF-ON only (back focus) as I get annoyed by half shutter button focus. 85% of the time I shoot with single AF point, because of which I don't like my camera to change the focus every time I take pictures. Also, I like back-focus because it saves lot of battery avoiding multiple unnecessary focus for each shots.

Another advantage with 17-70mm has been close focus. From what I can remember with 17-50mm lens I had, it's not so impressive when it comes to close focussing comparing to 17-70mm. 17-70mm offers 8.6"(22cm) close focus distance, while 17-50mm offers only 11"(28cm) closest distance. Resulting which you can take greatest sharp bokeh images for affordable price. And the low light performance at F4 (at 50-70mm) is also great, very happy with it. This lens is on my camera most of the time, whatever maybe the lighting conditions, keeping 50mm 1.8, 18-140mm lenses in bag most of the times.

I am used to both front focussing and back focussing. For me it is just the matter of convenience, but NEVER noticed a difference in photo quality using two methods. I have set my focus AF-ON only (back focus) as I get annoyed by half shutter button focus.

Front-focus and back-focus don't mean what you think they mean. They refer to systematic focusing errors where the lens actually focuses in front of or behind the subject on which the camera purports to be focusing. Here's an article that explains the problem and Nikon's way to adjust for it with a semi-automated process:

https://www.dpreview.com/news/3468248279/nikons-automated-af-fine-tune-explained

If one of your lenses back-focuses or front-focuses to a significant degree and another focuses accurately, the one that focuses accurately will almost always seem sharper. This is why you cannot use auto-focus to test lens sharpness, at least unless and until you've determined that the camera-and-lens combination focus essentially perfectly together.

With most DSLRs and DSLR lenses, adjusting to fix back-focus and front-focus problems requires sending them to a service center. Some of the higher-end cameras allow you to adjust them to compensate for specific lenses. Sigma lenses that communicate with its USB dock can be adjusted:

https://www.dpreview.com/articles/4406844237/sigma-usb-dock-quick-review/2

And Tamron recently announced the TAP In Console to do the same thing with its newest lenses:

https://www.dpreview.com/news/69114...onsole-to-customize-lens-af-firmware-and-more

NAwlins Contrarian said:

I am used to both front focussing and back focussing. For me it is just the matter of convenience, but NEVER noticed a difference in photo quality using two methods. I have set my focus AF-ON only (back focus) as I get annoyed by half shutter button focus.

Front-focus and back-focus don't mean what you think they mean. They refer to systematic focusing errors where the lens actually focuses in front of or behind the subject on which the camera purports to be focusing. Here's an article that explains the problem and Nikon's way to adjust for it with a semi-automated process:

https://www.dpreview.com/news/3468248279/nikons-automated-af-fine-tune-explained

Click to expand...

Thanks for the detailed reply. This link is really helpful and I need to check all my lenses if any of them need any AF fine tuning, though I have not noticed any AF issues so far (from the tutorial, it looks easier to do on my D7200). Live view focus is a good option but at the cost of doubling the shutter count and battery drain. If I can fix such back-focussing issues on my camera, can you please fill me the need to focus adjustment in the lens (like Sigma contemporary)? This topic is really interesting, but also have the potential to upset me with my new camera and lens collection lol

Thanks for the detailed reply.... If I can fix such back-focussing issues on my camera, can you please fill me the need to focus adjustment in the lens (like Sigma contemporary)?

You're welcome. Chronic back-focus and/or front-focus errors are something you can fix by adjustments you can perform yourself on some cameras (typically newer, higher-end lenses) and some lenses (the Sigmas plus a few Tamrons that work with their manufacturers' respective lens docks). If you have both a camera and a lens that let you make these adjustments, you could use either, but do not need to use both. Which adjustment method is better depends on whether you use the lens on other cameras (in which case adjusting it for one camera might not fix it, and could actually make it worse, on another) and/or whether you camera body properly recognized your Sigma or Tamron lens for what it is (many do not), in which case it could possibly get confused about what lens is mounted.

Man those lenses look unbalanced on those small bodies. I would say Sony needs to beef up the bodies but truthfully an A7 + 24-70/4 will do the job. By the time you price out an A6300, an adapter and any of these lenses, you're only $500 or so away from FF.

I am really liking the A6300 though. Waiting for the price to come down. It should complement my A7II nicely. Will definitely fit in my wife's purse with the 30 1.4, and will track dogs and action much better than the A7 will, especially with 8 FPS with no blackout. I just don't know if I can live with the awful SEL55210. But I would really miss having lock on AF with any of the significantly better Canon teles.......

Does anyone know how well Canon teles work in AF-C on the A6000 + Techart?

....not sure you can really have issues with back focus etc on an mirrorless camera with on sensor focus. The MC-11 emulates native focus and as such the camera sensor is actually the focusing array.

[N]ot sure you can really have issues with back focus etc on an mirrorless camera with on sensor focus. The MC-11 emulates native focus and as such the camera sensor is actually the focusing array.

Maybe so, but IIRC the discussion above was about how well these lenses perform based on the posters' experience with them on SLRs. If you judge a lens based on an SLR's auto-focus system, you need to make quite sure that it is focusing accurately.

Also: there are certainly reports of back-focus and front-focus errors on mirrorless cameras. Although I'm skeptical of such reports, I'm not ready to state categorically that it cannot happen.

One way that back-focus or front-focus on a camera using on-sensor PDAF could happen is if the camera uses a one-shot 'focus to' approach instead of an iterative / feedback approach, and the lens does not accurately focus to the focus the camera calls for. IOW, if the camera detects the phase difference and calculates a focus based on that, then merely powers the lens's focus motor, but does not verify (through PD or CD) that the lens has reached the precise point of desired focus, errors could occur. Do any current cameras focus this way? I don't know. But I can imagine that a camera designed might plausibly choose to have a focus mode work this way to increase focus speed at the expense of focus accuracy--and for some situations, it might even be a reasonably design choice.

I'm almost certain Sony PDAF uses phase detect for the big movement and contrast detect for the final tune