Archive 2019 · Improved focus stacking MF accuracy with Sony's integer distance readout

rvh23
Offline
• • •
Upload & Sell: Off

p.1 #1 · p.1 #1 · Improved focus stacking MF accuracy with Sony's integer distance readout

Having recently evaluated the new Sigma 14-24 DG DN zoom, I have been looking at how best to manual focus this focus-by-wire lens for focus stacking. Unfortunately it has a non-linear focus mechanism that changes the lens throw according to how fast you turn the focus ring. So you can't easily use angular increments to set your stack focal points. And even more unfortunately the distance readout in Sony cameras (to date) switches to integers beyond 1m. Importantly though, the exif data retains far more accuracy, and from my limited testing with this particular lens seems to correspond reasonably well to real-world best focus distance. Certainly more so than the integers in the display.

In trying to find a more reliable way to use Sony's distance readout, I think I may have found something that helps.It turns out that with this lens, and I suspect others also, that although the integer readout is very inaccurate due to rounding, the distance at which the integer increments is highly repeatable according to the exif data.

I did three repeats of finding the display toggle point between "1m" and "2m", both at 14mm and 24mm focal lengths on the zoom, and took a shot at each adjustment. The exif data showed it was very close to 1.4m in all cases. Similarly, the toggle-point between "2m" and "3m" in the display was always very close to 2.5m in the exifs, and the one between 0.9m and 1m was always at 0.9m.

In other words, I can know the lens is focused very close to 0.9m, 1.4m, and 2.5m if I carefully find the associated toggle-points at 0.9-1m, 1m-2m, and 2m-3m, respectively.

You can find other toggle points according to your typical interests, but for me I think these distances in particular will be very helpful in stacking typical landscape scenes.

It might be interesting to know, if anyone has the inclination, whether the same method can be used with other lenses that send distance information to the camera.

Aug 20, 2019 at 03:19 AM

rvh23
Offline
• • •
Upload & Sell: Off

p.1 #2 · p.1 #2 · Improved focus stacking MF accuracy with Sony's integer distance readout

For verification, I just tested my CV 21/1.4 which also provides distance information to the camera, although I wouldn't normally use the electronic distance readout with this lens unless I couldn't use the lens markings for some reason. The 0.9-1m toggle point was identical to the Sigma zoom. Because it has a much smaller focus throw, the 1m-2m and particularly 2m-3m toggle-points are hard to determine accurately with this lens, but were about 1.5 and 2.5m - very similar to the zoom.

Although this is still just two lenses, it seems the integer distance value displayed in-camera increments when the higher-resolution information from the lens is about half way between the two integers, so you can use that point to get a higher resolution focus distance indication.

Aug 20, 2019 at 06:32 AM

Fred Miranda
Offline
Admin
Upload & Sell: On

p.1 #3 · p.1 #3 · Improved focus stacking MF accuracy with Sony's integer distance readout

I guess this would only work with lenses that report distance. So, Loxia lenses are out?

Aug 20, 2019 at 01:30 PM

realVivek
Offline
• • • •
Upload & Sell: Off

p.1 #4 · p.1 #4 · Improved focus stacking MF accuracy with Sony's integer distance readout

How so? Are the Loxias focus by wire? (I would not think they are)
I think the OP’s problem is with manually focusing an AF (fly by wire) lens.

Fred Miranda wrote:
I guess this would only work with lenses that report distance. So, Loxia lenses are out?

Aug 20, 2019 at 03:28 PM

Fred Miranda
Offline
Admin
Upload & Sell: On

p.1 #5 · p.1 #5 · Improved focus stacking MF accuracy with Sony's integer distance readout

realVivek wrote:
How so? Are the Loxias focus by wire? (I would not think they are)
I think the OP’s problem is with manually focusing an AF (fly by wire) lens.

I just mentioned because he was testing the CV 21/1.4 (with helicoid) but that would not work with Loxia lenses because they don't report distance.
(Also the reason they only support 3-axis IBIS instead of 5-axis)

Aug 20, 2019 at 04:05 PM

samosh
Offline
• •
Upload & Sell: Off

p.1 #6 · p.1 #6 · Improved focus stacking MF accuracy with Sony's integer distance readout

Anybody tried B8stack software?

Aug 20, 2019 at 04:44 PM

rvh23
Offline
• • •
Upload & Sell: Off

p.1 #7 · p.1 #7 · Improved focus stacking MF accuracy with Sony's integer distance readout

Fred is correct, my suggested method will work with any lens that reports distance to the camera, not just focus-by-wire or AF lenses. And therefore not Loxias, which don't. But it is most useful for lenses that don't have distance scales marked on them, where you have to depend on the camera-distance readout.

It turns out to be rather fortunate that the three toggle-points at 1m, 2m, and 3m, which correspond to actual distances of 0.9, 1.5, 2.5m, are well suited to focus stacking at about F8. If you add one or two more shots further away to that sequence, it covers a lot of landscape needs with wide angle lenses.

Aug 20, 2019 at 04:50 PM

rvh23
Offline
• • •
Upload & Sell: Off

p.1 #8 · p.1 #8 · Improved focus stacking MF accuracy with Sony's integer distance readout

I obtained another verification of my hypothesis that the toggle point between consecutive display values is a good estimate of a (lens-reported) distance setting that is halfway between the two.

In this case the lens was a CV 12. With that lens the display steps directly from 0.7m to 1m, and the toggle point shows 0.85m in the exif. Similarly, as with my other lenses, the toggle from 1m to 2m produces 1.5m in the exif. Beyond 2m the display goes directly to infinity, so not numerically useful, and in that case the toggle produces 3.4m instead.

Aug 20, 2019 at 06:14 PM


FM Forums \| Sony Forum \| Join Upload & Sell

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.