FS7: Sigma 50-100mm DC Backfocus Changes with Zoom on Metabones EF Adapter

LennyLevy said:
Jacek, I would never doubt your expertise, My guess is that because of your position working with professional clients and equipment you just aren't coming in contact with the the more varifocal zooms out there these days.
I do agree with you that the culprit in many cases is the camera mount itself.
My experience with Nikon's on my F3 was that once I adjusted my back focus correctly all my old Nikon's were par focal but new ones weren't. on a collimater though I found my 17-35 held focus at the wide and telephoto end but drifted in the middle.

I am pretty curious about the Sigma 24-105. I have tried 2 of them on a few different Canon DSLR's and my FS7 with metabones adapter and they were wildly out , but some people on this forum have claimed their's were par focal. Poor quality control, operator error? I'd buy one i it was even close to par focal .
Click to expand...
It's hard to draw conclusions from very limited data. That's why I said I have checked many lenses and draw my conclusions based on that. As far as your F3 ordeal there are still many question marks:
The body flange is only half of the information. You need to check the lens back focus as well.
The 17-35 you had could be out of adjustment, worn out, broken or just a bad copy. That's why drawing conclusions from limited data is not really a good idea.
There are couple of important notes to consider (that's why I made the video that deals with back focus with emphasis on still lens/motion cam setup with adapter)
1. I worked for years as a lens tech in a rental house. It took back then a full time position to just mainly clean and maintain lenses . There was lot of shimming and tightening loose elements in lenses.
FULL TIME JOB guys!!! for a pro.
With the explosion of owner operators in last 10 years or so somehow everyone decided that lenses don't need maintenance any more.
The truth is that they actually need more precise maintenance:
with film we worked our tolerances for .001" circle of confusion. That lets you get away with back focus MURDER compared with todays 4K cameras. .001" covers 4 pixels on a 4k S35 sensor. Yes 4. If you want to make out single pixel you need to do better than that. Like 3-4 times better.
Still lenses don't need to be set to that precision because of AF. All they need to be is always able to reach infinity aka short back focus. And that's how manufacturers tend to set them.
I'm amazed how well these lenses are set out of the box given that level of precision needed. (of course I'm simplifying things there is sensor MTF, debayer quality, OLFP etc, etc)
IMO the manufacturers are doing a tremendous job with how consistent and precisely they set lenses
Hope that helps....
 
Two days ago, I was out shooting with my other Sigma lens, the 150-600. It struck me that that lens held the focus throughout that 4:1 zoom range.

Then I got to thinking, 'why is it that a lens with half the zoom range, the Sigma 50-100, has trouble holding focus?'

It can't be the flange back distance, because both lenses are using the same Metabones adapter.

So now I'm really wondering about my 50-100, if it's out of tolerance. Because it goes noticeably out of focus, noticed even on a small low resolution VF, when pulling out from a full tele setting. Seeing as my 150-600 can maintain focus over that much longer focal range, I am suspicious of the 50-100 being defective. It doesn't seem logical that a simpler lens would not be able to hold focus as a more complex lens in the same price range.
 
The more telephoto a lens the greater the Depth of Focus. Depth of Focus is the measure of how accurate the Flange Focal Distance must be. Depth of Focus range behind the lens, Depth of Field is range in front of the lens. This is why a wide angle iens is the most critical measure for setting backfocus.

It may be that your lens is out of spec, but it may also be that the Metabones adapter is out but the range available from the more telephoto lens allows for it.
 
It would be interesting if everyone here who owns a 50-100 Sigma did a backfocus test and reported the results. Then we could compare numbers and determine of our lens is an outlier or normal.
 
I also have a 12-28 Tokina, but backfocus is pretty good on it. My Sigma 150-600 seems to have near perfect backfocus.
It's the Sigma 50-100 that is way off. But it seems like this process of shimming the lens requires a clean room environment and special tools.
I have to treat it as a 'variable prime' lens, because any zooming while rolling results in ruined footage due to loss of focus.
 
Mark, My guess is - first as Mitch said BF on a telephoto is simply not critical and you would be unlikely to see it unless it was way off. Likewise I've noticed that with a wide angle zoom like the 12-28 with a limited range there is so much depth of field in front of the lens at both ends of the zoom, that we often don't notice back focus issues because nearly everything is in focus anyway. But when you get into the 50-100 range then you can easily see problems because both ends of the zoom are critical enough you'll see if they're in focus and unlike with a telephoto you are wide enough that the BF matters.
Hope that makes some sense.
 
Lenny sorry, that's not correct.
It's just the opposite.
Depth of field is reverse of depth of focus (sharp area behind the lens) Depth of focus is what the sensor records because that's where it's at. Wide lenses have great depth of field but minuscule depth of focus. That is simple physics of the angle of view.
 
Jacek,
I understand that quite well, but the other factor in play is that while a wide angle lens has very little depth of focus it may have tremendous depth of field. My experience is that if you aren't using a collimator to judge whether you're in focus or not , on a wide angle lens its often hard to tell whether you've nailed the focus precisely and it may not matter that much to the viewer.
If I'm using a 50 or especially a 100 mm lens at f2.8 the focus change between 10' and say 12' might be quite noticeable on screen. But if I'm on a 12mm lens chances are I'll never notice the difference at all unless its a very large screen. My Tokina 11-16 looks par focal to me but it may not be - I just know that looking through the viewfinder I'll never be that exact anyway.

That's why as a shooter even with a par focal lens, you never focus on the wide end of the zoom and then expect to hold focus when you zoom in to telephoto because you'll probably have missed the proper focus point. Instead you zoom in to tele where its critical and then you know that it will be OK when you zoom out to wide . Of course if you're back focus is off - yes the wide shot will be soft but its harder to notice say in a viewfinder or even on many 7" screens. That's why its dangerous.

In Basspig's case he was saying his 12-28 was parfocal and I guess I'm suggesting that if it was slightly off he might not even notice it unless he looked very carefully on a big screen. (A collimator would show it up right away.) But on a 50-100 you might catch a small error more easily. On the 150-600 it's the opposite.

When I said I used to have a 17-35 that got soft in the middle - I never saw that until I looked on a collimator - It always looked fine on a 17" monitor.

Does that make any sense?

Lenny
 
Last edited:
Lenny have you tried to throw off 50mm lens by 2 feet (the back focus amount will vary with distance) equivalent back focus and mount on 12mm lens?
You won't see an image just a blurr.. try it
 
Haha - Jacek i think you have caught me pushing my limited knowledge a little too far. You are probably right. Mea Culpa.

I guess my point is simple and perhaps I'm wrong about this also, but I notice that with shorter throw wide angle zooms, the exact focus point is just hard to be precise about unless you look very carefully at a large monitor. My experience with that is on my 11-16 Tokina which seems par focal enough to me for government work but I have no idea whether it really is. I was just speculating that might apply to a 12-28 but heck it was just a wild thought and may be completely wrong.
 
You must log in or register to reply here.
Back
Top