Focus Stacking

[mikeinpa]

Has anyone tried the Focus Stacking feature and can recommend settings?
I tried it last night both towards Infinity and from Infinity. Certainly not the best of trials but it certainly worked out better than a single shot. This I think was 17 shots and stacked in Helicon Focus.

[grotte]

Somebody here once suggested that the correct method is to stack towards the infinity. I am not sure why: perhaps to assure that at least the foreground is in focus? But my preference is to shoot from infinity to near. Step wise is much lens/aperture/subject dependent: here is a Helicon stack of 13 images in small step increments with XCD30 f/4. Still I rarely focus stack which will never turn X1D into a view camera.

[mikeinpa]

I was wondering about the question of towards or from infinity. Nikon use towards Infinity and on a tutorial from Steve Perry on focus stacking he recommended focusing on the nearest point then backing off a little manually and starting the sequence from that point. I think having the foreground in sharp focus is preferable so I'll stick to that in further trials and experiment with the combination of aperture, number of shots and the increment setting for landscapes in particular. I suppose I could always RTFM and see what I can learn from that.

[JCM-Photos]

I use focus stacking for landscapes since it was introduced on the X1D.

A step of 2 pixels is sufficient for my landscapes even on large A2 prints (at f:5.6 wit the XCD21)

beginning at close focus never gives a perfect sharp infinity because it mostly falls between 2 steps

beginning at infinity gives always perfect sharpness at infinity and it is also perfect at the nearest point if you set a sufficient number of frames.
Number of frames should be in excess, eliminating unnecessary frames before stacking

[bmikiten]

Somebody here once suggested that the correct method is to stack towards the infinity. I am not sure why: perhaps to assure that at least the foreground is in focus? But my preference is to shoot from infinity to near. Step wise is much lens/aperture/subject dependent: here is a Helicon stack of 13 images in small step increments with XCD30 f/4. Still I rarely focus stack which will never turn X1D into a view camera.

I still shoot 4x5 and 8x10 in the field. The idea of focus stacking is one of the reasons I purchased the X2D. For some reason, it doesn't look like a view camera shot. I can't put my finger on it....The image is beautiful but something is different.

[JCM-Photos]

Large format film looks different than digital medium format with modern lenses, even with my 50 Mpix X1D and 907X.
Large format has very fine details but these are never as clinical as digital MF. And as the format is huge the contrast softness is not seen as a lack of sharpness.

In MF I am a step less clinical with older Zeiss CF glass for Hasselblad V with the CFV II 50c back from the 907x, and appreciate it a lot.

You can even try very old C glass non T* with single coating, that's on step softer.

[mikeinpa]

I use focus stacking for landscapes since it was introduced on the X1D.

A step of 2 pixels is sufficient for my landscapes even on large A2 prints (at f:5.6 wit the XCD21)

beginning at close focus never gives a perfect sharp infinity because it mostly falls between 2 steps

beginning at infinity gives always perfect sharpness at infinity and it is also perfect at the nearest point if you set a sufficient number of frames.
Number of frames should be in excess, eliminating unnecessary frames before stacking

Thanks for your insight as I really want to get to grips with this on the XD2. Can you explain about the step of 2 pixels please? I only have the option of Extra Small, Small, Medium, Large and Extra Large.

[MGrayson]

I use focus stacking for landscapes since it was introduced on the X1D.

A step of 2 pixels is sufficient for my landscapes even on large A2 prints (at f:5.6 wit the XCD21)

beginning at close focus never gives a perfect sharp infinity because it mostly falls between 2 steps

beginning at infinity gives always perfect sharpness at infinity and it is also perfect at the nearest point if you set a sufficient number of frames.
Number of frames should be in excess, eliminating unnecessary frames before stacking

Thanks for your insight as I really want to get to grips with this on the XD2. Can you explain about the step of 2 pixels please? I only have the option of Extra Small, Small, Medium, Large and Extra Large.

There is a chart in the manual giving equivalents. The number of pixels refers to the size of the circle of confusion. That, and the lens settings, will determine the actual step size. I think medium is 2 pixels, but you should check.

[mikeinpa]

I use focus stacking for landscapes since it was introduced on the X1D.

A step of 2 pixels is sufficient for my landscapes even on large A2 prints (at f:5.6 wit the XCD21)

beginning at close focus never gives a perfect sharp infinity because it mostly falls between 2 steps

beginning at infinity gives always perfect sharpness at infinity and it is also perfect at the nearest point if you set a sufficient number of frames.
Number of frames should be in excess, eliminating unnecessary frames before stacking

Thanks for your insight as I really want to get to grips with this on the XD2. Can you explain about the step of 2 pixels please? I only have the option of Extra Small, Small, Medium, Large and Extra Large.

There is a chart in the manual giving equivalents. The number of pixels refers to the size of the circle of confusion. That, and the lens settings, will determine the actual step size. I think medium is 2 pixels, but you should check.

Thanks, I guess I didn't get that far in the manual.

[mikeinpa]

I watched this which was helpful

https://youtu.be/RvVhHXGSm3w

Many thanks for this Patrick.

Cheers

Mike

[Andy Miller Photo UK]

OK so I have tried this a bit and still have the following comment -- of all the ways to explain this feature Hasselblad chose the most confusing (they expanded my circle of confusion) !!!

Thank H, but the guidance it to try it and find out "perform multiple tests"  -- not guidance helpful at all.

I get it the step size is linked to the DOF given by the camera which depends on lens, aperture and distance to focal plane. OK -- and it says it is better to pick a larger number of exposures in a sequence than less "the camera will automatically make all the calculations for the user" "the camera will automatically stop when the lens cannot be focussed further or closer" --  well if this is infinity and not the hyperfocal distance there will be a lot of wasted shots.

why can't the X2D stop the sequence when the hyperfocal distance has been reached? If it can please provide details. Otherwise one has to OVER-ESTIMATE the number of shots needed and delete them later.

Most macro shots do not need this but most landscape shots do.

Take my 120mm MACRO LENS - at f/8 - min focusing distance is 430mm / ~17" and hyperfocal distance 4.36m away -- what step interval size and how many images to shoot. Trial and error does not seem to be a clever way to work.

Shooting say a 6" long Honey Mushroom about 20"/50cm away and then stack the images.
How to work out the interval size to use. DOF at 50cm is +/- 0.4CM, at 59cm +/-0.58cm and +/- 0.8 cm at 67cm what step size and how many shots are needed to cover the subject - but not reach infinity?

I allow a 50% overlap so for a 6cm "deep" subject  0.5cm *2 = 24 images -- each one is effectively 0.25cm steps - that is WHAT in terms of CoC and the step size I should use -- I simply have NO IDEA.

One tool https://www.pointsinfocus.com/tools/focus-stack-calculator/ - seeks step intervales of 1/8th the DOF - so 0.58*2/8 = 0.145  -- which means for a 7 cm subject  48-50 exposures would be required -- but which step size gives this?  The guidance provided is simply not helpful to me.

Is there a simple calculator or look up table one can use?  Inputs focal length, aperture and distance to front and rear of the subject

OR "dear H" -- why not give users the choice to set the nearest and furthest focal planes and then the camera can work out the settings needed to cover the distance in various step sizes -- OTHER cameras do it this way OR you can stop when the camera focusses at the hyperfocal distance.

[mikeinpa]

I've got to say that after watching the mathphotographer video that 800 images to form a stack does seem rather excessive.

[MGrayson]

why can't the X2D stop the sequence when the hyperfocal distance has been reached? If it can please provide details. Otherwise one has to OVER-ESTIMATE the number of shots needed and delete them later.

Just now, I focused at 50 feet, told the X2D to take 900 images towards infinity, and it stopped after 5 displaying "Lens Limit Exceeded."

The approximate formula for DoF is linear in the CoC size (sorry for the acronyms). So one COULD do a quick series with Extra Large steps, find out how many images are needed, and then scale up for any other CoC size.

While the table of microns is wrong for the X2D, the multipliers of pixel pitch are (I think) correct, so since CoC sizes are
Extra Small = 1 pp
Small = 4/3 pp
Medium = 2 pp
Large = 4 pp
Extra Large = 6 pp

If you needed X images for your Extra Large stack, you'd need this many images in your stack at different step sizes.
Extra Small = 6 X
Small = 4.5 X
Medium = 3 X
Large = 1.5 X
Extra Large = X (obviously)

For example, if you needed 30 images for your Extra Large stack, you'd need 30 x 4.5 = 135 images at Small. Safety margins are, of course, at your discretion.

[Andy Miller Photo UK]

Thanks but it really does not help me at all.
To start it is only helpful if you provide lens, aperture and subject distance information. Then ne can understand what the camera is seeing. 2nd your example is not remotely similar to mine.

I'm a simple 60+ fellow who does not care about coc or pp - I want the camera to do the job as efficiently as possible AND the X2D MISSES the mark by a mile.
The best/simplest approach is to set the near and far distance and have the camera work out how many shots it needs to take — the user can set a % overlap or this can be left to and auto setting.
This is 2023 not the 1970s -it is not like the processing power is absent from the camera.
The XF + IQ4 does something very similar to this but with a wee bit more manual input — no coc or pp knowledge required. So do almost every mirrorless body.

[MSuser]

Thanks but it really does not help me at all.
I'm a simple 60+ fellow who does not care about coc or pp - I want the camera to do the job as efficiently as possible AND the X2D MISSES the mark by a mile.u
The best/simplest approach is to set the near and far distance and have the camera work out how many shots it needs to take — the user can set a % overlap or this can be left to and auto setting.
This is 2023 not the 1970s -it is not like the processing power is absent from the camera.
The XF + IQ4 does something very similar to this but with a wee bit more manual input — no coc or pp knowledge required

You seem to be consistently unhappy in a variety of ways with what you have. Why not get a different camera which works the way you think it should?