Sigma Art Lens and Dock

Pistnbroke

Just purchased at a good price a Sigma 24-105 f4 Art intending to sell it on for a profit.
Testing it out on the D810 the pictures were rubbish even at F8. I set the Fine Focus adjust at 105mm but was not impressed. then I investigated the dock and watching u tube it looks like the required fine focus setting (16 of them) are all over the place from -17 to 0 through the focus range.
So anybody got an ART lens and was it rubbish until you used the dock?
Not sure weather to send it back ..the seller knows jack shit about cameras.
You would think it would come out the factory so you could put it on zero and it would work but it seems not ..am I right?
MORE THOUGHTS.
When you have set a series of Fine focus settings using a dock is the lens computer causing the camera to swing its FFA settings or is it done in the lens and the camera sits at 0. This would be important if you wanted to use a D5300.

Will ring Sigma on Monday and ask wtf is going on.

donaldejose

Let us know, sounds like some error inside the lens. It has received many good reviews so I doubt it has bad optics.

Rx4Photo

Sounds like you might have bought somebody else's headache. The two Sigma Arts that I've owned have been phenomenal albeit both primes. Consider whether it's worth the effort.

Ironheart

I'd write down the current settings, and zero it and and give a look see..

Pistnbroke

Lets just wind back a bit ....those of you who have them did you use the dock? did it work out the box? I bought it on ebay for £400 (auction against 26 others ) ..the grey price here is £550.
The immigrant seller knows nothing of lenses/ has poor english. The lens is perfect but the bar code on the box is ripped ( suspicious) so cannot tell if its with warrantee from sigma.The image stabilization is fantastic ,the picture is horrible. I bouht it to sell on and I dont have a dock (£40) Ebay/seller has agreed a return..dont want to spend on a dock if its faulty/no warrantee. Will phone Sigma in next few hours ..
Question.....should it work out the box ?
When set is it movable from nikon to nikon without re docking?

snakebunk

My Art lense and my new Sports lense were perfect out of the box, I haven't bothered to do any fine tuning yet. No new Sigma lense should be rubbish, as no Nikon or Tamron lense should be. Furthermore I think Art lenses are tested individually before they leave Sigma. I am afraid something has happened to your lense.

How is the image if you focus manually?

Pistnbroke

Spoke to sigma ...
Should work out the box ..slight FFA on camera may help
Should work with any Nikon (if you have more than one body) slight FFA for each body may be required.
Moving the focus point around the screen should be fine.
Warrantee only applies to european lenses and original purchaser.
Its going back as its not working for whatever reason /dropped/fiddled with on a dock.
Thanks everyone for there help.

MegapixelSchnitzel

Stop down to f/11.

decentrist

Sigma? Sigma is a wonderful brand. Af issues are rare!

ericktessier

I never saw or heard somebody owning an Art lens with perfect AF out of the box.
I do own Art 24, 35, 50 and 85mm. They're perfect when you have the dock and know how to adjust them. The dock really worth it.

spraynpray

@ericktessier:

Hey erick, If you want to post your experiences and (especially) your technique for setting up the AF using the dock, I'm sure some members would appreciate it.

ericktessier

Sure. Here are some things I learned while calibrating my Art lenses :

- It will take time and you shouldn't take shortcuts.

- The type of light in the scene and especially on the target point is important. In my case, standard house LED lights doesn't leads to perfect results with my D800. I'll have a pretty near perfect focus but often once outside focus won't be perfect.

- The type of target also influence the quality of the process. I had bad results with LensAlign and now focus on more "real" targets usually with text.

- IR lighting of the target might lead to more consistent results afterward with my D800 when I want to do it indoor with my standard house LED.

- The use of flash/speedlite/strobe is not recommanded since the problem is to focus, not only to light the target properly or enough during the shot.

- Tripod, tripod, tripod (which doesn't mean you need 9 legs or an enneapod)

- Position the trio camera/lens/tripod to each distance for which you have adjustments in the Sigma calibration program. The more precise the positionnement will be the more precise the adjustment will be.

- It will take time and you shouldn't take shortcuts (worth repeating here).

- Start with a rough adjustment for all the focus distances (for 1 focal length) before doing the fine. In some cases the "in between" focus distances will be impacted by 2 adjustments (those in between which they fall).

- Always do at least 2 shots before deciding if you need to give a + or a -. The first shot with the focus at the nearest focus distance the lens can go and the second shot with the focus at infinity. If you don't know if you should give a + or a - for this point because of the behavior explained in the next point, skip this focus distance and after having adjusted all the other focus distances you should see a trend and give an average value on that point.

- It is normal to have a front (or back) focusing if the focus was starting at infinity and going to near distance AND back (or front focusing) if the focus was starting at near distance and going to infinity. The adjustment will solve this.

- Once the fine adjustment is done and you think you are finished with this task do a final checkup of ALL the points to make sure they're perfect.

- Go shoot and keep in mind it might take you some other fine tune later.

- Remember that it will take time and you shouldn't take shortcuts.

- If you don't want to spend required time or invest in the dock you can always send me your lens and camera and I'll do it but don't expect to have the calibration made forever. There'll always be fine tune (even giving a +1) for different reasons.

- While I'll be calibrating my Art 135mm (once they're available) I'll try to find a mathematical link between the AF Fine Tune of MY D800 and the calibration of the Sigma program to accelerate the process. This is not considered as a shortcut.

- Always use the largest aperture (smallest f-number).

- Never judge AF problems of an Art lens if you don't have the dock and you are taking the lens out of the box. (Or if you bought it from somebody with the dock that didn't knew what he was doing and is selling the lens because of AF problems).

- Do not expect the lens to focus perfectly out of the box.

- Make sure you AF fine tune in your camera is set to 0 at the beginning. If you need to reach -20/+20 in the Sigma program then it's time to think at the AF Fine Tune setting but remember that this will affect all the focus distances.

- If you have a -20 and a +20 in the same lens for a same focal distance consider returning the lens.

When I want a challenge I do it handheld on a moving subject of unpredictable displacement under LED lighting, indoor and with my distance to the target in between the distances in the program.

Questions? Feel free to ask.

Pistnbroke

Tamron rules

Ironheart

@ericktessier Given that PDAF is a mechanical process, and will yield a statistical range of focus, wouldn't it be better to do several shots and average the results at each FL and distance?

See the following for what I'm referring to:
https://www.lensrentals.com/blog/2012/07/autofocus-reality-part-1-center-point-single-shot-accuracy/

ericktessier

@Ironheart Sure this is logically the way to do it and statistically the most precise way if you can, when you take your (let's say 5) 5 shots, for every shot, precisely know what amount of correction you would need to give to have a perfect shot. Then the average would be easy to do although the process would be a little bit longer and I'm not sure it would be more precise since when the focus is not perfect the comportment of the lens might not be the one of a predictable unit.

The question is : can you really know what amount of correction you have to give to make a perfect shot? I can't and I will not try because that would require too much brain cells for me. I'm very good to approximate it though and I prefer this approximation and the multiple iterations method to have a perfect result at the end. And because of the next question, there would necessarily have to be multiple iterations of this average method.

The other question is : can you rely on statistics and assume the lens will have a predictable behavior if the AF is not calibrated with your body? There must be a mathematical explanation but I don't need to find it and I was proven with the Art 85mm that if the adjustments are not precise, the comportment is not predictable enough to proceed to an average (of guessed values).

The major point is : is the lens perfect after the calibration? If yes then you will be able to thrust this extraordinary tool. If not then do it again. If you keep playing between -1/+1 for the same value and on several occasions then you might start to consider you're either doing something wrong or you should return the lens or you need to use an enneapod.

Pistnbroke

I am told the adjustment takes about 80 fittings of the lens to the dock/camera and takes about 2 hours for a zoom lens. So from comments by others (50% of shots in focus so sold three ) and my own experience .....no thanks. Seems Sigma technicians are at odds with what you are saying ie not good out the box

ericktessier

Oh don't get me wrong, beside the fact that I don't like using a zoom lens and prefer primes I wouldn't invest in a Sigma (Art or Sport) zoom because of the 16 points required and the smaller aperture at which the measure can be made.

If you are good and the viewfinder is well calibrated with the results you don't have to remove the lens more than 20 times and take more than 20 pictures. You don't have to take a picture every time, just see which way to go and estimate by how much. Do the 4 distances before removing the lens.

Pistnbroke

I just watched a review of the Tamron Dock and one comment passed was that it was better than the sigma system whose lenses even when calibrated failed to focus reliably.

ericktessier

@Pistnbroke Are you affiliated with Tamron (or the basic question : why do you put hyperlink on your Tamron word?)?

I have a couple of minutes to kill. Can you send the link to this video?

Edit : (Sorry I just saw the forum do add it automatically and I have, being ignorant, created 2 links to Tamron. Damn... now 3)

Pistnbroke

No links to tamron other than owning a 150-600... you will note I started this thread and so far am not convinced from what other owners have said that the Art series works 100% 100% of the time.

spraynpray

I've had a 50 and 35 and sold them both due to inconsistent focusing in low light. My 24-35 is good, but that's because I manually focus with that lens.

snakebunk

I don't question anyones experience but I still wish I could understand how a lens can be bad at focusing in low light. I have read some articles about auto focus but found no indication that light has any impact on the lens performance.

ericktessier

Maybe if the transmission % of the specific lens is lower than standard than it could affect the low light behavior of the AF but I don't think it's the case with the Art lenses.

Low light is more an AF problem from the camera but maybe I'm wrong assuming this. Maybe in low light, the communication with the lens implies more signals (amplitude or ...) or if the camera struggle to react as fast as in good light it could lead to erratic results.

snakebunk

Good points! I guess it could be that a low bandwidth on the lens causes trouble when auto focus gets challenging. But given the speed of current electronics I find it hard to believe, it shouldn't be that much data compared to large images for example.

ericktessier

I don't think it's the speed of electronics but maybe a compatibility between the signals?

Pistnbroke

Its Nikon "put the knife in" special software