2024-03-11 •
SEND FEEDBACK |
Related:
focus stacking, focusing, Glenn K, reader comment, Sony, Sony A7R, Sony A7R II, Sony A7R III, Sony A7R IV, Sony A7R V, Sony mirrorless
The idea (well, one idea) is that by focus bracketing, a better overall image can be had in 3D scenes where it can be difficult to determine the best point of focus. Shoot the bracket, then later assess which is best.
Reader Glenn K writes:
Sony A7R V can do a 3-shot focus bracket, one front and one back focused… but I haven’t played with the spacing
DIGLLOYD: somehow I hadn’t noticed that!
With , the Sony A7R V fixes the shots at 3.
is applicable and can be configured also.
Useful to be sure, but not as useful as Hasselblad’s Symmetric mode, which places no restriction on the number of shots.
Sony A7R V: 3-shot focus bracket
Reader Michael Erlewine writes:
I am trying to understand and utilize what your wrote of the various kinds of focus stacking.
I understand the first (to infinity) and the second (from far to near), yet I still don’t understand the use of the Symmetric mode.
I would tend to use the second mode, and with my technical camera I lock the most distant part of the subject I want in focus and make my focus rail have a hard stop there, leaving the rest for bokeh. It’s great that you did the work up you did and I appreciate it. Now to fully understand it and put it to work.
DIGLLOYD: as I am thinking of it, Symmetric mode would be best used for a focus bracket for the purpose of selecting the best sharpness “landing zone” for the subject.
Make your best guess at optimal focus placement, then shoot with Symmetric with at least 4 total frames. One of them is likely to be significantly better for sharpness cutting through the subject in a better way.
The bonus of doing this is you could decide to do a “short stack” if conditions allow, or you could simply select one single frame as the best of the lot. You can’t lose.
Example
Consider this image from Earth Shadow Rises over Aspen.
Where exactly should you focus in order to have maximum foreground sharpness without giving up sharpness to the storage shed? Depth of field is asymmetric (more to the distance), and the lens could have a little focus shift and/of field curvature, and the subject spans a significant neaer/far distance.
All these things together make it difficult to be sure of the zone of sharp focus on such a complex scenes, including the visual impact of what is unsharp and by how much.
Indeed, the lens here *does* have a slight forward focus shift. I did pretty well with my focus placement (see the full aperture series), but it is likely that some slightly different focus placement would have produced a better overall result.
Earth Shadow Rises over Aspen
f1.7 @ 1/10 sec EFC shutter, ISO 80; 2023-10-18 17:37:24
Fujifilm GFX100 II + Fujifilm GF 55mm f/1.7 WR @ 45.3mm equiv (55mm)
ENV: Colorado, altitude 8400 ft / 2560 m, 42°F / 5°C
RAW: Camera VELVIA, Enhance Details, LACA corrected, WB 5000°K tint 18, +30 Shadows, +20 Whites, +10 Clarity, AI Denoise 10
[low-res image for bot]
CLICK TO VIEW: Hasselblad X2D System
2024-03-11 •
SEND FEEDBACK |
Related:
digital sensor, frame averaging, Hasselblad, Hasselblad medium format, Hasselblad mirrorless, Hasselblad X2D 100C, image stabilization, In-Body Image Stabilization, medium format, pixel shift, shutter, true-color sensor or image
Hasselblad X2D 100C
Back in October 2022 I wondered why the Hasselblad X2D did not include a pixel shift feature, ideally one like the Panasonic S1R.
The IBIS technology enables pixel shift and/or multi-shot, and with 0.1 pixel sensitivity, it ought to work great:.
The X2D 100C is equipped with an in-house-developed IBIS system for medium format cameras that currently is the most compact in the industry.* It can detect camera movements down to 0.1-pixel. These movements can then be compensated for by employing the 5-axis 7-stop image stabilisation to make shooting more effortless in low-shutter-speed scenarios.
Back in 2022 I asked these question:
DIGLLOYD: Does the X2D support any version of pixel shift or multi-shot?
Hasselblad: No. Multi-shot and/or pixel shift are not planned for the X2D 100C.
DIGLLOYD: Does the X2D support any form of frame averaging?
Hasselblad:The X2D 100C does not support frame averaging.
Here in 2024, there is no pixel shift and there is no frame averaging support. Why?
Hasselblad did not promise, but still it’s disappointing.
The frame averaging thing is particularly baffling—so easy to implement, yet only PhaseOne offers it. At least for some cases, frame averaging is a 'killer'feature, making the whole ETTR thing moot.
Hasselblad X2D 100C
Built-in 1TB SSD — insanely fast and very nice if your work allows for plugging the camera in for connectivity. And with USB 3.0, transfers should be very fast. If it allows also writing to a card as a backup (or alternative download solution), it's A-OK with me, but the specs say that a CF Express Type B card is limited to 512GB... why the limit on capacity?
OLED EVF with 5.76 megadots. Not as nice as the Sony A1, but on par with the Leica SL2 and superior to the 3.69m-Dot OLED EVF of the Fujifilm GFX100S. The 1.0X magnification means extraordinarily nice viewing.
WiFi support great for some though of no value in the field to me.
Flash sync speed up to 1/2000 second via the lens leaf shutters or 1/4000 with electronic shutter, compatible with various Nikon flashes.
Startup time — the CPU in the X2D needs to be 2X faster to just to support twice the megapixels. But the X1D and X1D II were already slow as molasses, making them a frustrating choice for quick shots: those cameras had to boot-up like some old Windows 3 PC. The desired image was often gone by the time the camera was ready! With high battery drain, it was not feasible to leave the camera turned-on in the field. This single issue is IMO a major consideration. Has Hasselblad made the camera startup in 1 second or less? IMO a camera MUST be ready within 1 second once turned on, or it is a marginal choice for many types of photography.
Controls— no 4-way controller, no direct AF control, paucity of useful buttons, buttons that are troublesome in the dark or with gloves, etc. IMO, the emphasis on attractive design is all good, but a serious flaw when it means the lack of crucial operational controls. I despise camera design that makes simple things harder and slower.
Hasselblad X2D 100C and Lenses
|
|
|
|
|
|
|
|
|
|
|
|
Hasselblad X2D 100C
2024-03-11 •
SEND FEEDBACK |
Related:
Airy Disc, circle of confusion, depth of field, diffraction, digital sensor, focus stacking, focusing, Hasselblad, Hasselblad medium format, Hasselblad mirrorless, Hasselblad X2D 100C, medium format, optics
Approximation of Airy Disc
vs sensor photosite size
The Hasselblad X2D user manual apparently contains an error regarding the step size: the figures are the same as for the 50-megapixel Hasselblad X1D II 50C. Of course, it could be that the the manual is correct and the firmware is buggy. Presumably not.
CLICK TO VIEW: Hasselblad X2D System
Hasselblad X2D 100C: Step Size
Most cameras offer 10 step sizes. Hasselblad went with 5 steps. This is probably OK.
Discussion in the X2D manual is really good with respect to its pictures/diagrams, in terms of helping photographers understand the general idea. But in terms of actionable choices vs aperture choice, it is not helpful.
Discussion ignores a key issue: Airy Disc size (see Resolving to the Sensor Resolution), and talks only to Circle of Confusion (CoC). The Airy Disc is the minimum resolvable spot size whereas CoC as used in the manual apparently defines a cutoff for how much out of focus blur is acceptable, eg depth of field.
Both Airy Disc and CoC must be considered for optimal images. Were that not true, f/22 would be the aperture to use for every landscape photo. But at f/22, you get a massive loss of contrast and resolution: a very deep zone of mashed detail (very large Airy Disc eg no fine detail) and grayed-out contrast... garbage.
There is a sweet spot for depth of field vs Airy Disc size vs sensor pixel pitch which boils down to f/5 to f/9 range for the X2D and similar and 35mm cameras. And only that range.
Below, the table is defining the cutoff criteria is for the Circle of Confusion used to determine the scaling factor based on aperture for the step size. But it is ignoring Airy Disc.
For any Step Size, use of f/11 vs f/5.6 means steps 2X as deep with 1/2 the frames. Given that, these numbers have no practical utility for the photographer. They are useful only in the context of the Airy Disc, which Hasselblad does not discuss.
Cutoff values as per Hasselblad user manuals
Step Size |
X1D II 50C |
X2D 100C |
Extra Small |
1 * PP = 5.3 |
1 * PP = 3.76 |
Small |
4/3 * PP= 7.1 |
4/3 * PP = 5.0 |
Medium |
2 * PP = 10.6 |
2 * PP = 7.5 |
Large |
4 * PP = 21.2 |
4 X PP = 15.0 |
Extra Large |
6 * PP = 31.8 |
6 X PP = 22.6 |
Recommendations for Step Size for Hasseblad X2D
Choose shooting aperture, then choose the Step Size that keeps the CoC + Airy Disc consistent with each other.
For example, it rarely if ever makes sense to use Extra Small at f/11 because the Airy Disc is huge at ~14 microns, so an cutoff off 3.76 makes little sense—the Airy Disc is 14X larger in area than the cutoff size.
However, it’s not always just numbers. Additional overlap in real-world images can be important when focus stacking. For example, two things at very different distances can cause blur haloes which make it nearly impossible to achieve a quality result. Tighter spacing can help in such cases. Another case is a scene with motion, where subject matter in one frame is more acceptable than another, and both are reasonably sharp.
Symmetric mode (single-shot intent for best actual focus): go with Extra Small to achieve fine-tuned focus positions, along with 4 to 6 frames. One of those should be just right for the 3D subject you are capturing.
All of this assumes perfection in execution by the camera. Accordingly, these trecommendations are tentative pending my field use of the X2D. The best way to be sure is to shoot the same stack once at with Small and another time with Medium, then stack them and check all resultings aspects.
Recommended Step Size vs Aperture
f/# |
Airy Disc Size |
Recommended Step Size |
Comment |
2 |
2.54 |
Extra Small, CoC = 3.76 |
Not advised for focus stacking |
2.8 |
3.55 |
Extra Small, CoC = 3.76 |
Not advised for focus stacking |
4 |
5.08 |
Small, CoC = 5.0 |
Not advised for focus stacking |
5.6 |
7.11 |
Small, CoC = 5.0
Medium, CoC = 7.5 |
Peak brilliance but more frames at f/5.6. |
8 |
10.15 |
Medium, CoC = 7.5 |
Brilliance is dulling, but fewer frames. |
11 |
13.96 |
Large, CoC = 15.0 |
F/11 impacts resolution and contrast noticeably. |
16 |
20.3 |
Large, CoC = 15.0 |
Not advised for any situation. |
Airy Disc size by aperture
Helps explain the above recommendations.
The Airy Disc size rapidly overwhelms the pixel pitch. For example, at f/8, the Airy Disc size is around 10 microns, so a Step Size of Medium makes sense. But at f/5.6, you’d perhaps want to use Small, and at f/4, Extra Small. This idea is NOT explained well.
Airy Disc size rapidly overwhelms the pixel pitch of the X2D . Thus the choice of Step Size really comes down not to depth of field as stated, but resolvable detail. And that depends only on aperture. Thus it becomes an exercise of choosing a step size that is symbiotic with the shooting aperture.
Airy Disc vs Aperture
f/# |
Airy Disk Diameter at 520nm (microns)
[Edmund Optics] |
2 |
2.54 |
2.8 |
3.55 |
4 |
5.08 |
5.6 |
7.11 |
8 |
10.15 |
11 |
13.96 |
16 |
20.30 |
Hasselblad X2D: focus bracketing support
Hasselblad also does an excellent job of explaining step size.
Hasselblad X2D: focus bracketing support