Fun with X1D

[JoshMerwin]

Some pics from walking around yesterday. And two from the focus chart thingy. You can see the AF one latched on closer to the #4, but I can't really complain too much with that. If It was flat vertically, it would have locked on exactly where I wanted.

 

[Shashin]

View attachment 124736

I found a test I did long ago with the Zeiss CY 135f2 when I for the first time saw this phenomenon (I thought the lens is defect, but it isn't ;-)

I did a pic of books at an angle to the shelf.
You see how the depth of field becomes larger closing the aperture but the actual focus point shifts backwards.

But how are you determining focus shift? Midpoint between the extremes of sharpness in DoF? DoF does not increase in front of the focal point as much as behind it, which means if you are measuring the midpoint, you will see the "shifts" you are illustrating. Focus shift is related to spherical aberration. Your lens does not exhibit SA.

 

[mkerouac]

I have been doing longer exposures around my house and I find the dynamic range to be very excellent...not a problem at all. With this said I have only been shooting at iso 100.......also haven't done much experimenting with clipped highlights so can't answer to that yet. What I don't like is the almost constant green cast I get in Auto WB. Wish Hasselblad would fix this in a firmware update. Even when I look at the menus on the LCD there is a green cast naturally inherent in the screen display itself. Eleanor

I'm not sure the green cast is as bad in every camera. My X1D has a very slight green cast on the LCD. The AWB is actually OK outdoors, not as accurate indoors. As a point of reference, my H6D50c had a horrible green cast on the LCD. My friends sample was OK. My H6D100c has a great image on the LCD and it's the same LCD as the 50c. So I don't know if its lot to lot variation on the LCD's or the way they are color calibrated at the factory. But I'm skeptical that a firmware change will fix it. I think Hasselblad would need to add a tuning feature to the firmware, so the user could adjust the color balance. It would almost need to be like calibrating a monitor. The green in your AWB, I hope would be corrected with the new firmware.

 

[jerome_m]

Good to know. The reviewer I am discussing doesn't mention, in his Lensalign series, if he focussed AF or MF but it is clear that the wide open shot was nailed and therefore the shift does seem to be there, at least at close range. However, if as you say the AF system 'knows' about shift and applies corrections

Just to be sure we understand each others: I meant that the AF system applies focus shift corrections in the H series cameras (H6D, H5D and earlier), not in the X1D. But I suppose that Hasselblad will implement that feature in the H series, as they did in the H series.

 

[tashley]

Just to be sure we understand each others: I meant that the AF system applies focus shift corrections in the H series cameras (H6D, H5D and earlier), not in the X1D. But I suppose that Hasselblad will implement that feature in the H series, as they did in the H series.

I've not had a Hassy before but I thought that TrueFocus did something when you focussed and recomposed so as to shift focus slight according to the angles involved. I didn't realised it handled focus shift too. Looks like the X1D might need that with this lens in some circumstances.

 

[tashley]

Pursuant to the original thread title: I had the call nearly two weeks ago to say I'd have my unit by Friday 3rd and it didn't arrive. I've been on a trip the last week and today I had a call to say that it'll be 'end of Feb' - so that seems to be the latest ETA for the UK. At least that gives me time to hear if the FW updates expected this week (as I understand it) do anything to relieve the more pressing problems with the camera - then I can make up my mind.

 

[hcubell]

I've not had a Hassy before but I thought that TrueFocus did something when you focussed and recomposed so as to shift focus slight according to the angles involved. I didn't realised it handled focus shift too. Looks like the X1D might need that with this lens in some circumstances.

I am not aware that the X1D has the internal hardware required to implement True Focus. The H6 has a so-called yaw rate sensor in it.

 

[jdphoto]

There's more talk here then photos for considering it's a "fun with X1D" thread. Granted, the delays in shipping might be obvious, but I'd like to view some OOC Raw files. Perhaps even shoot with a color/WB target to see how Hassy handles manual/auto WB, auto ISO. or even high ISO. Maybe a DB link with no editing or tone mapping, just straight outta camera. It sounds very "beta" to me and I hope those that print big will get great results, but those just posting web sized images in Srgb it's not overwhelmingly convincing yet. The rendering, detail and color are similar to my A7R2. No, this is not sour grapes as I'm waiting until the kinks are worked out before considering any MFD options.

 

[masud]

I've not had a Hassy before but I thought that TrueFocus did something when you focussed and recomposed so as to shift focus slight according to the angles involved. I didn't realised it handled focus shift too. Looks like the X1D might need that with this lens in some circumstances.

The focus shift correction was introduced with the H3DII—it was called Ultra-focus—the lens communicated the focal length and the aperture with the body so that the body to make the correction in autofocus when the aperture changed after focusing at the moment just before taking the shot. When True-Focus was introduced with the H4D, it allowed focus/recompose using the yaw sensor (as you mention above), but it also kept the Ultra-focus feature. Hence it does both.

The X1D should at least be able to implement Ultra-focus (correct for focus shift) in a future firmware because the lens and the body do communicate. I do not know if the X1D has a yaw sensor for a full True-focus implementation. Anybody know the answer to this?

Cheers,
Masud

 

[jerome_m]

The focus shift correction was introduced with the H3DII—it was called Ultra-focus—the lens communicated the focal length and the aperture with the body so that the body to make the correction in autofocus when the aperture changed after focusing at the moment just before taking the shot. When True-Focus was introduced with the H4D, it allowed focus/recompose using the yaw sensor (as you mention above), but it also kept the Ultra-focus feature. Hence it does both.

Exactly.

The X1D should at least be able to implement Ultra-focus (correct for focus shift) in a future firmware because the lens and the body do communicate. I do not know if the X1D has a yaw sensor for a full True-focus implementation. Anybody know the answer to this?

The X1D does not really need true focus, since its measurement point can be put anywhere in the frame.

 

[hcubell]

The focus shift correction was introduced with the H3DII—it was called Ultra-focus—the lens communicated the focal length and the aperture with the body so that the body to make the correction in autofocus when the aperture changed after focusing at the moment just before taking the shot. When True-Focus was introduced with the H4D, it allowed focus/recompose using the yaw sensor (as you mention above), but it also kept the Ultra-focus feature. Hence it does both.

The X1D should at least be able to implement Ultra-focus (correct for focus shift) in a future firmware because the lens and the body do communicate. I do not know if the X1D has a yaw sensor for a full True-focus implementation. Anybody know the answer to this?

Cheers,
Masud

Would the Ultra-Focus be implemented in firmware or the Phocus software? In the H bodies, I think it is done in Phocus. I Suppose it only matters if you want to use LR/ACR instead of Phocus.

 

[masud]

The X1D does not really need true focus, since its measurement point can be put anywhere in the frame.

Good point [no pun intended].

 

[masud]

Would the Ultra-Focus be implemented in firmware or the Phocus software? In the H bodies, I think it is done in Phocus. I Suppose it only matters if you want to use LR/ACR instead of Phocus.

I think it needs to be in firmware; in the H it is in the firmware. I don't think you can change focus after the shot is taken (unless you had a light field camera like Lytro). Good for LR/ACR users.

 

[Arjuna]

I would think that Ultra-focus would have to be done in the camera, not in post-processing software like Phocus.

The H cameras are (D)SLR's where, AFAIK, the auto-focus is phase-detection, with the lens wide open. The lens then automatically stops down to the selected aperture when the exposure is taken. This introduces the possibility of focus shift, which Ultra-focus is intended to correct. The X1D, where the auto-focus is contrast detection on the sensor, and which I think is done at the taking aperture, is an entirely different situation, and Ultra-focus may not really be applicable, at least not in the same manner as for the H cameras.

 

[Nick-T]

Ultra focus is indeed done in camera, and no you won't see True Focus in the X1D (not really needed).

 

[ErikKaffehr]

Hi,

You have to take Diglloyd with a grain of salt. That said, many of his observations are good, but he sort of makes a bit to much noise about his findings and does not really invite to a constructive discussion.

Regarding focus shift, it has to do with spherical aberration. Spherical aberration means that rays going trough the periphery of a lens have a different focus than those passing trough the center. Stopping down cuts away the peripheral rays and makes focus more consistent.

Let's say, when I am manually focusing my zoom lenses at 11X magnification on the Sony A7rII, focus is not very distinct. Stopping down focus will be more distinct and optimum may move a bit.

Quite often the optimum focus shifts a bit when stopping down from large to medium apertures.

Contrast sensing AF systems can focus well stopped down, after all the focusing system just tries to maximise fine detail contrast. Phase detection is often working fully open, so the phase error measured by the AF system is not masked by depth of field. But, that may lead to focus being optimised for open aperture instead of shooting aperture. With good medium aperture lenses this may not matter a lot.

If a lens is sharp at full aperture it means that aberrations will be well balanced. If the lens gains a lot of sharpness when stopping down it is reasonable to expect that it will have some shift of focus.

I got some interest in subscribing to Diglloyds pages again. But his terms are restrictive and he doesn't really invite to a discussion, so I guess i will pass on a subscription. That said, much of the info is good and solid. But, it is a part of the business that his reporting may be a bit out of proportions.

Just a few examples:

• Diglloyd was the first to report on shutter induced vibrations on the Sony A7r. I think he was proved right on the basic issue.
• Diglloyd reported on tilted focal plane on the otherwise excellent Sony 90/2.8 G Macro. I can see the same, and LensRentals may have made the same observation.
• Diglloyd has also reported on noise effects of the Sony raw compression scheme. Quite a few image processing experts had the opportunity to check out some of his files, but Diglloyd's restrictions forbade a constructive discussion.

My take is that both methodology and raw images should be shared for common benefits and also that an open discussion is a beneficial learning experience for all involved.

So, my take is really is:

• If you are happy with your stuff just stay happy with your stuff.
• Discussion may widen our perspective, potentially good…
• Good discussions may be a learning experience

Kind regards
Erik

I hate to admit this but I had no real idea of what "focus shift" was. Diglloyd has some really bad stuff to say about focus shift in the 90 lens that I used for this shot. I always just focused on what I wanted (many times manually) and that was that. Anyway this was shot with my 90 lens iso 100 at f4. Looks OK to me. I've never worked commercially so possibly that's why I never knew about stuff like focus shift. Now I know but don't think I'll be worried about this on my two X1D lenses. They'er way good enough. Eleanor

 

[jerome_m]

I would think that Ultra-focus would have to be done in the camera, not in post-processing software like Phocus.

The H cameras are (D)SLR's where, AFAIK, the auto-focus is phase-detection, with the lens wide open. The lens then automatically stops down to the selected aperture when the exposure is taken. This introduces the possibility of focus shift, which Ultra-focus is intended to correct.

This is true.

The X1D, where the auto-focus is contrast detection on the sensor, and which I think is done at the taking aperture, is an entirely different situation, and Ultra-focus may not really be applicable, at least not in the same manner as for the H cameras.

Ultra-focus, which is simply nudging the focus to compensate for focus shift, can be done with any focus detection system. It even works with manual focus, ask experienced view camera users. It just requires to know how focus varies with aperture, which is functionally equivalent to having a model of the lens spherical aberration. Hasselblad certainly has this model, since they have the lenses blueprints.

In short: there is no reason why Hasselblad could not implement that feature in a later firmware. Whether they will actually do it and when will they do it is another matter, however.

Edit: the Ultra-Focus feature was already discussed in this forum in 2012 (!):
http://www.getdpi.com/forum/medium-...-focusshift-compensation-hasselblad-h-mf.html

 

[hcubell]

Does the X1D focus wide open or stopped down at the shooting aperture? I assume the former, as logically there shouldn't be any focus shift if the focus is optimized for the shooting aperture. So, the potential solutions are (1) to have the focusing done with the lens stopped down, or (2) to adjust for the focus shift through firmware adjustments in the focus point. Is either alternative workable? In the real world of taking photographs, how much of an issue is the apparent level of focus shift in the XCD 90mm lens?

 

[jerome_m]

Does the X1D focus wide open or stopped down at the shooting aperture?

The X1D focusses wide open. As I said in message #220 above: in the Sony mirrorless cameras, focus is done with the aperture closed down, so there is never any focus shift. This would be a possible solution for a firmware upgrade.

 

[hcubell]

The X1D focusses wide open. As I said in message #220 above: in the Sony mirrorless cameras, focus is done with the aperture closed down, so there is never any focus shift. This would be a possible solution for a firmware upgrade.

Thanks.