Any one adapted a follow focus system for focus stacking?

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

austrokiwi1
Posts: 350
Joined: Sun Sep 14, 2014 10:53 am

Any one adapted a follow focus system for focus stacking?

Post by austrokiwi1 »

I keep thinking about the merits of a focus rail( I use a stack-shot system) and using the focusing helicoid, and I wonder if any one had looked at adapting a follow focus system so that a stepper motor could control the movements of the helicoid? I can see some advantages with such a system with a lens like the Voigtlander APO 125mm Lanthar
Still learning,
Cameras' Sony A7rII, OLympus OMD-EM10II
Macro lenses: Printing nikkor 105mm, Sony FE 90mm F2.8 Macro G, Schneider Kreuznach Makro Iris 50mm , 2.8, Schnieder Kreuznach APO Componon HM 40mm F2.8 , Mamiya 645 120mm F4 Macro ( used with mirex tilt shift adapter), Olympus 135mm 4.5 bellows lens, Oly 80mm bellows lens, Olympus 60mm F2.8

elf
Posts: 1416
Joined: Sun Nov 18, 2007 12:10 pm

Post by elf »

I can think of more cons than pros

Cons:
  • Each individual lens would need its own program/algorithm and possibly for each fstop.

    Adding an extension tube or bellows would require new algorithms.

    Most new Olympus lens are focus by wire, so focus changes depend on the rate the focus ring turns.

    Lens without changeable focus like microscope objectives and the JML 21mm couldn't be used.

    Lens where the entrance pupil changes with focus changes (focal length changes) wouldn't work for panoramas.
Pros:
:?:

mjkzz
Posts: 1696
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

Actually, why not just rotate the focus ring? This is something I have been tinkering with for a while, though not very happy with results -- it is tricky to build such system, but controlling lens focus ring electronically is another way to go and seem to work great.

elf
Posts: 1416
Joined: Sun Nov 18, 2007 12:10 pm

Post by elf »

mjkzz wrote:Actually, why not just rotate the focus ring? This is something I have been tinkering with for a while, though not very happy with results -- it is tricky to build such system, but controlling lens focus ring electronically is another way to go and seem to work great.
My previous post assumed a stepper or encoder based dc motor attached with a belt to the focus ring.

If you want to control a lens electronically, then you need the APIs. Olympus no longer releases their API. Canon and Nikon as well as Sony have public APIs, but I haven't looked at them for several years. Legacy lens would be left out of this solution.

Micromotors with encoders are dirt cheap, but it would probably be better to get a 4000 pulse/revolution encoder in order to have enough resolution.

Or just get one of these and add the micromotor.

mjkzz
Posts: 1696
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

Hi elf,

Yes, I meant the same in my first sentence -- using stepper or servo motor to drive focus system, it is kind of tricky to build because I can not get it compact enough, need 15mm rod rail to mount it.

The second half sentence (or meaning) refers to controlling Canon lens electronically, I do not intend to do it to other brand lenses because a Canon lens can be used on other camera with proper adapter, well, maybe not able to focus to infinity. I have written code already, will shoot a video to show it.

Of course, the electronic method will not be able to control a great lens like MP-E 65 while motorized focus follower might work (does the focus ring move as you change magnification on MP-E65?)

austrokiwi1
Posts: 350
Joined: Sun Sep 14, 2014 10:53 am

Post by austrokiwi1 »

Thanks for the comments. I think for my purposes Follow focus would perhaps be best when using manual adjustment
Still learning,
Cameras' Sony A7rII, OLympus OMD-EM10II
Macro lenses: Printing nikkor 105mm, Sony FE 90mm F2.8 Macro G, Schneider Kreuznach Makro Iris 50mm , 2.8, Schnieder Kreuznach APO Componon HM 40mm F2.8 , Mamiya 645 120mm F4 Macro ( used with mirex tilt shift adapter), Olympus 135mm 4.5 bellows lens, Oly 80mm bellows lens, Olympus 60mm F2.8

rjlittlefield
Site Admin
Posts: 24057
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

astrokiwi1 wrote:I wonder if any one had looked at adapting a follow focus system so that a stepper motor could control the movements of the helicoid?
The ZenFocus system (http://zenslider.com/zenfocus/) has this functionality.

A year ago I had some experience with the unit. Unfortunately that experience was painful and ultimately fruitless except for lessons learned.

The biggest problem turned out to be an apparently simple but in fact quite challenging matter of materials choice. The belt that they provided was an ordinary timing belt. It was not sticky, not stretchable, and had teeth that did not mesh with anything on the structure of the lenses that I tested it with. As a result it slipped, even when tensioned to a point that placed what seemed like too much sideways force on the lens barrel. I eventually figured out that it was possible to solve the slipping problem by wrapping a piece of non-skid drawer liner material around the focus ring, under the timing belt. That was at best a fiddly solution. Worse, it exposed a second problem in that the motor could then put what seemed like a dangerous amount of twisting force on the focus ring, when the ring would hit its near/far limits. The final difficulty was that when I tried to make very small movement of the focus ring, the focus steps became irregular apparently due to chaos (in the formal sense) introduced by the combination of friction in the focus ring and springiness in various other parts of the system.

There were an assortment of other startup problems that were eventually solved, but in the end, because of the belt problem I was not able to get the unit to work safely and well with any of the 5 different lenses that I tried it with, including the MP-E 65. The company seemed not interested in developing a production quality solution to the problems, so finally I took the (for me) very unusual step of simply returning the equipment.

I continue to have faith that the motorized follow-focus approach could work fine for focus stepping, given careful control of friction, springiness, and torque control, or perhaps even just a better match between the follow-focus equipment and the lens to be controlled. But I cannot recommend the ZenFocus system for this task.
mjkzz wrote:MP-E65 ... does the focus ring move as you change magnification
The MP-E 65 does not really have a "focus" ring. The ring changes magnification. In most parts of the range the ring also changes focus distance, but that's a side effect and not a designed feature. The change in focus depending on ring rotation is highly nonlinear and there is a critical magnification around 1.5X where the focusing effect slows down, stops, and changes direction as the ring is turned. See http://www.photomacrography.net/forum/v ... 327#193327 for more discussion of that effect.

--Rik

mjkzz
Posts: 1696
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

belt driven follow focus sounds great until it is implemented (believe me, I did it). There are gear driven system where you attach a larger gear around the lens focus ring, either made of hard rubber or plastic. But it requires 15mm rod rail (popular for video shooters) to mount the whole system, kind of clumsy.

Thanks Rik, what I meant was when you change magnification on MP-E 65, does the ring move, if it does not then it should be fine to be driven by belt or gear. If the ring moves, then it is hard to do.

I did see an "auto" focus system designed for manual lens, meaning, even if you have a manual focus lens, you can still do auto focus (on Sony camera). The construction of that system is to move the lens back and forth as a whole assembly on a motorized mount adapter, kind of cool and focus speed is fast.

mjkzz
Posts: 1696
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

FYI, here is the link about the auto focus system for manual lenses on Sony camera

http://petapixel.com/2016/02/11/techart ... us-lenses/

ChrisR
Site Admin
Posts: 8671
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

what I meant was when you change magnification on MP-E 65, does the ring move, if it does not then it should be fine to be driven by belt or gear. If the ring moves, then it is hard to do.
Hard!
It's just simple, in that the ring changes the magnification - but the focus changes at the same time, in an unspecified way. So you have to move the camera - the distance is different. You have to have a little experience of it to know which direction the camera has to be moved, because the lens length changes so much, and at first surprisingly.
As Rik says, things reverse, "somewhere in the middle". I daresay one could determine (a) polynomial(s), to describe the movements. I could have tried it when I was younger!
Chris R

mjkzz
Posts: 1696
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

Thanks ChrisR.

It is very possible to build a translation table for the zigzag instead of polynomial. A translation table seems faster (no floating point calculation, which is slow on a lowly MCU) and it can address non-linearity easily -- basically you turn the polynomial into segments of lines to approximate it. For each segment, linear interpolation can be used. The added benefit is that, if you are using a wheel (type control device) to adjust the focus (magnification for that matter), you will not experience the zigzag by rotating the wheel. The table could be such:

1.0X 0
1.1X 200
1.2X 400
1.3X 500
1.4X 600
2.0X 1000
2.1X 950 // turning point
2.2X 900
.
.
.
.

The hard part is to initialize the motor for the two ends so that the table can be applied, that process might take some manual intervention, hence some inaccuracy, but I think it is manageable vs a moving turning ring where a gear or belt is set on.

ChrisR
Site Admin
Posts: 8671
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

That should work - a couple of LUTs probably.
Reminds me of linearising a thermocouple unit with 6502 assembler. Memory cost was the problem then :)

Something else to deal with is the growth of the lens.

The ring moves an equal distance from the camera per integer increase in magnification, = about 10mm between 1x and 2x, or 2x and 3x, etc.
The front end moves about double that amount.
That means that the subject position at 1:1 (WD 101mm)is within the body of the lens when it's adjusted to 5x (WD 41m) - unless you move the camera backwards as you increase magnification.

It's also heavy to turn. It's currently sitting next to one of their Internal Focus STM lenses - more than a little different!
Chris R

Post Reply Previous topicNext topic