Hasselblad X2D first impressions

[pflower]

However, not always, sometimes, and I seem to have developed a sixth sense for when, deciding that LRc is not working - then I go to Phocus, do as much as possible and then export TIFF to LRc for the final step. My observation is that Phocus can do a better job on the 3FR files - particularly on landscapes with a lot of subtle Earth tones and/or large deep shadow areas, even to this day, and I like the noise processing in the shadows better with Phocus.

I concur with pretty much everything you say. I have used LR with an H3D-39 (since 2010), CFV 50c and now with the X1D. In the early days LR struggled with the 3FR files a bit - especially with highly saturated colours. But mostly I never used Phocus until recently. After a lengthy shoot in Athens I decided on a whim to see what Phocus could do. To my surprise the Tiffs exported from Phocus were very significantly "better" than just using LR. "Better" is of course a purely subjective view, but I was surprised (pleasantly) at what Phocus was giving me as distinct from LR.

So my workflow is now to import everything into Phocus and export everything as 3fffs which I import into LR after deleting the 3FR files. I then sift in LR and process those files I intend to print in Phocus, do basic adjustments there and export as Tiffs which are then adjusted finally for printing in LR. I would say that about 80% of the time the Tiffs, after final adjustment in LR, are significantly better to my eyes than the original 3fff file adjusted in LR. It is a bit of a laborious workflow but I have the time to go through it.

One thing I have never understood is why the 3fff files are smaller than the 3FR files. WIth the H3D the opposite is true - the 3fff files are a good 20% larger than the 3FR files. I queried this with a Hasseblad rep when I test drove the H6D and was told that was just because compression had not been applied to the 3FR files but would be in due course. That has never happened and I still don't know what is being removed or changed from the 3FR file on conversion to 3fff to explain the size differential. But I also agree that there is difference between a 3fff file and the 3FR as far as image quality is concerned.

 

[SrMphoto]

<snip>
But I also agree that there is difference between a 3fff file and the 3FR as far as image quality is concerned.

When I import 3FR and FFF files into Lightroom I notice only a slight difference in the default white balance. I do not see any difference between images once I equalize WB.

 

[hcubell]

So my workflow is now to import everything into Phocus and export everything as 3fffs which I import into LR after deleting the 3FR files. I then sift in LR and process those files I intend to print in Phocus, do basic adjustments there and export as Tiffs which are then adjusted finally for printing in LR. I would say that about 80% of the time the Tiffs, after final adjustment in LR, are significantly better to my eyes than the original 3fff file adjusted in LR. It is a bit of a laborious workflow but I have the time to go through it.

I have always avoided making adjustments to a TIFF in LR, because I always do at least some additional work on a TIFF in PS using non-destructive editing techniques before printing. If you make LR adjustments to a TIFF, those adjustments are permanently baked into the TIFF when you send it to PS. You can't later change the LR adjustments made to the TIFF. I try to maintain a non-destructive workflow from start to finish, because I invariably make small changes to a file as I make test prints.

 

[Chris]

I received my Hasselblad X2D body together with the 38mm and 55mm f/2.5 V lenses a few weeks ago.

As a previous X1DII owner I’m glad to report that this beautiful tool holds up to its promises! Did a non-scientific compare of the Ibis with the Leica SL-2 with 55 (X2D) and 50mm (SL-2) lenses on. With the SL-2 I could go down to f/1 (1/50) to shoot shake-free from the hand (sitting down, camera at eye level, close to my body). Shockingly with the Hassi I could achieve an astounding 0.5 seconds! That’s 4-5 stops more than the SL-2 and on a much bigger (44x33mm/100MP) MF sensor! Terrific!
AF speed also greatly improved with the new V-lenses and phase detection AF. And the list goes on: superb EVF, IQ beyond everything else in my humble opinion, especially the color rendering.

So far so good. There’s one glitch though: the lenses do not sit tight on the body’s lens bajonett. It becomes quit noticeable whilst operating the controls (MF clutch, programmable lens ring) on the new lenses: the lenses rotate ever so slightly/a few milimeters in the bajonett, which I find quite disturbing and inappropriate. It doesn’t degrade image quality, but interferes with smooth operation of the camera.

My dealer returned the body to Hasselblad. It’s currently sitting in Sweden. Hasselblad reported back that their research department is working on a solution on improving the X2D’s lens mount and that it would take some time, probably a few months. So at least it seems that I’m not alone here.

I was wondering if the same is true on the X2D with the older X- lenses line up?

Best,
Chris

 

[hcubell]

I received my Hasselblad X2D body together with the 38mm and 55mm f/2.5 V lenses a few weeks ago.

As a previous X1DII owner I’m glad to report that this beautiful tool holds up to its promises! Did a non-scientific compare of the Ibis with the Leica SL-2 with 55 (X2D) and 50mm (SL-2) lenses on. With the SL-2 I could go down to f/1 (1/50) to shoot shake-free from the hand (sitting down, camera at eye level, close to my body). Shockingly with the Hassi I could achieve an astounding 0.5 seconds! That’s 4-5 stops more than the SL-2 and on a much bigger (44x33mm/100MP) MF sensor! Terrific!
AF speed also greatly improved with the new V-lenses and phase detection AF. And the list goes on: superb EVF, IQ beyond everything else in my humble opinion, especially the color rendering.

So far so good. There’s one glitch though: the lenses do not sit tight on the body’s lens bajonett. It becomes quit noticeable whilst operating the controls (MF clutch, programmable lens ring) on the new lenses: the lenses rotate ever so slightly/a few milimeters in the bajonett, which I find quite disturbing and inappropriate. It doesn’t degrade image quality, but interferes with smooth operation of the camera.

My dealer returned the body to Hasselblad. It’s currently sitting in Sweden. Hasselblad reported back that their research department is working on a solution on improving the X2D’s lens mount and that it would take some time, probably a few months. So at least it seems that I’m not alone here.

I was wondering if the same is true on the X2D with the older X- lenses line up?

Best,
Chris

No issue with my lenses. 30mm, 45mm, 90mm, 135mm, 1.7x TC, and the 35-75mm zoom. Completely solid connection. I gather it is the new V series lenses.

 

[glaiben]

...the lenses rotate ever so slightly/a few milimeters in the bajonett...

I just checked my X2D. The 120/3.5 Macro is rock solid without any axial rotation play. The 80/1.9 and the new 38/2.5 V have a very slight rotational play - perhaps 0.25mm or so at the most. I've already sold the X1D-II and 907X so cannot compare them directly. There is considerably more play when attaching a Fotodiox adapter for a third party lens. Of course, given their price point, I'm not surprised.

...gregg

 

[KlausJH]

No issue with my lenses. 30mm, 45mm, 90mm, 135mm, 1.7x TC, and the 35-75mm zoom. Completely solid connection. I gather it is the new V series lenses.

The 55V shows also rotational play on my 907x while all other XCD lenses (21/30/45p/90) sit tightly. Sounds like there is an issue with the new lenses (too?)

 

[SrMphoto]

I received my Hasselblad X2D body together with the 38mm and 55mm f/2.5 V lenses a few weeks ago.

As a previous X1DII owner I’m glad to report that this beautiful tool holds up to its promises! Did a non-scientific compare of the Ibis with the Leica SL-2 with 55 (X2D) and 50mm (SL-2) lenses on. With the SL-2 I could go down to f/1 (1/50) to shoot shake-free from the hand (sitting down, camera at eye level, close to my body). Shockingly with the Hassi I could achieve an astounding 0.5 seconds! That’s 4-5 stops more than the SL-2 and on a much bigger (44x33mm/100MP) MF sensor! Terrific!
AF speed also greatly improved with the new V-lenses and phase detection AF. And the list goes on: superb EVF, IQ beyond everything else in my humble opinion, especially the color rendering.

So far so good. There’s one glitch though: the lenses do not sit tight on the body’s lens bajonett. It becomes quit noticeable whilst operating the controls (MF clutch, programmable lens ring) on the new lenses: the lenses rotate ever so slightly/a few milimeters in the bajonett, which I find quite disturbing and inappropriate. It doesn’t degrade image quality, but interferes with smooth operation of the camera.

My dealer returned the body to Hasselblad. It’s currently sitting in Sweden. Hasselblad reported back that their research department is working on a solution on improving the X2D’s lens mount and that it would take some time, probably a few months. So at least it seems that I’m not alone here.

I was wondering if the same is true on the X2D with the older X- lenses line up?

Best,
Chris

There is a slight rotational play when mounting V lenses on X2D. There is less play when mounting V lenses on X1D and less play when mounting non-V lenses. The slight rotational movements have no influence on IQ and do not bother me.

 

[jng]

I find no rotational play with any of my lenses (21, 30, 45, 90, 55V) on the X2D, although the fit is maybe a bit less snug than on my older X1D where it took some effort to mount and dismount the older style lenses. The new 55V is perhaps a bit less snug all around, but once the lens is mounted, it's fully seated and locked in place. I'm guessing that the lens mount on the newer body is just a bit more "relaxed" and we're seeing unit-to-unit variation around that specification. My X-V adapter is a bit loose in the mount (as were just about all of my Nikkor lenses back when I was using the Nikon system), but no issues there as far as I can tell. As long as the lens is held flush to the mount, this shouldn't affect image quality (but not to discount the psychic effect on the photographer).

John

EDIT: OK, I checked again and found that there's maybe a few degrees of play with the 55-V on my X2D, but I had to grab the front of the barrel and give it a good twist to move it at all. It's a non-issue for me.

 

[buildbot]

My dealer returned the body to Hasselblad. It’s currently sitting in Sweden. Hasselblad reported back that their research department is working on a solution on improving the X2D’s lens mount and that it would take some time, probably a few months. So at least it seems that I’m not alone here.

How does this happen? I assume they have full CAD designs for all of their cameras, with tolerances, so why would that change over time? Why would they need to research it - just make the mount to the same tolerances and specs as before???
I know there are differences across the Mamiya/Phase One mount over time, and my older lens do have some play on the XF for example, but that's over ~2 decades of the mount and they literally added a row of contacts to it.

 

[MGrayson]

How does this happen? I assume they have full CAD designs for all of their cameras, with tolerances, so why would that change over time? Why would they need to research it - just make the mount to the same tolerances and specs as before???
I know there are differences across the Mamiya/Phase One mount over time, and my older lens do have some play on the XF for example, but that's over ~2 decades of the mount and they literally added a row of contacts to it.

In theory, there is no difference between theory and practice. In practice, there is.

 

[PeterA]

How does this happen? I assume they have full CAD designs for all of their cameras, with tolerances, so why would that change over time? Why would they need to research it - just make the mount to the same tolerances and specs as before???
I know there are differences across the Mamiya/Phase One mount over time, and my older lens do have some play on the XF for example, but that's over ~2 decades of the mount and they literally added a row of contacts to it.

A CAD drawing is converted into a CAM cycle - the CAM cycle is put into the CNC machine. The machine is fed round stock ( in this case). The round stock is then turned to the CAD / CAM dimensions. The actual machined finished product dimensioned 'part' is then (typically) deburred and any other processes required to complete to spec are performed.

Throughout this cycle from design to machining - there are many factors which can impact on the final finished part. Tolerances designed by the design engineer are just that 'tolerances' that are designed to provide for an acceptable part in use. However machine tools wear over time, so the difference between one batch with a new tool being used and another batch/ part where the machine tool is nearing its end use and not replaced yet can be significant and are still acceptable if the finished product is within design tolerances. Depending on the amount of parts being made- you also have different quality check regimes in place. Small numbers can allow for greater quality control larger numbers usually rely on statistical quality control testing - the quality of the quality control testing regime is also not a constant itself - many variable come into play in process control and measurement and testing...

The more parts that have to 'fit' together eg camera mount to lens mount(s) the more the cumulative/additive of at the margin in or out of spec parts can influence the mating performance of the parts.

I am not aware of the in spec tolerances that the mount parts from various generations of lens mounts are or were - what I can say is that there is little to zero chance of having a '''perfect" or even similar 'theoretical' fit between them because of the manufacturing processes involved - the difference in fit between one lens and one body is going to be different between a different lens and the same body etc etc...

the CNC machining available to Hasselblad contractors when V lenses were being made are orders of magnitude in machining and engineering terms below teh capability of machines today - lets not forget that the V lenses were designed with tolerances that woudl allow for acceptable focus to be achieved using film as teh capture medium - the tolerances used back then were able to be less than tolerances today. As a hobbyist I have ben using a Schaublin CNC lathe designed to be able to theoretically turn accurately to 0.001mm - in the real world that 0.001 ( one thousandth of a millimeter) is not easy to achieve over even a small batch of parts - because minor changes in tool wear make compensations almost impossible at that level of tolerance - even 0.01 ( one hundredth of a millimeter) is not something that can be achieved consistently over many many exactly the same parts.

TLDR? - what Matt said.

 

[peterm1]

Getting back to the post-processing discussion, I have tried using Phocus and sending TIFFs to both Lightroom and Capture One, but it's such a pain and time suck without a clear payoff in terms of superior file quality IMHO that I will likely go back to just using LR with it's amazing masking tools, good DAM capabilities, and a few plugins like the Topaz AI tools, Exposure X7, etc.

 

[davidrm]

Personally I see a pretty drastic difference in how Lr inteprets White Balance (auto or otherwise) compared with Phocus, and I much prefer Phocus. I would probably still use Lr, if only it had a Luma curve tool. Personally I cannot do without that. So for me, I do culling and base edits in Phocus, cleaning up in Photoshop, very occasionally some extra editing, then catalog in CaptureOne. Files from all other digital cameras are 99% pure CaptureOne. Works for, but it’s clearly not ideal.

 

[SylB]

A CAD drawing is converted into a CAM cycle - the CAM cycle is put into the CNC machine. The machine is fed round stock ( in this case). The round stock is then turned to the CAD / CAM dimensions. The actual machined finished product dimensioned 'part' is then (typically) deburred and any other processes required to complete to spec are performed.

Throughout this cycle from design to machining - there are many factors which can impact on the final finished part. Tolerances designed by the design engineer are just that 'tolerances' that are designed to provide for an acceptable part in use. However machine tools wear over time, so the difference between one batch with a new tool being used and another batch/ part where the machine tool is nearing its end use and not replaced yet can be significant and are still acceptable if the finished product is within design tolerances. Depending on the amount of parts being made- you also have different quality check regimes in place. Small numbers can allow for greater quality control larger numbers usually rely on statistical quality control testing - the quality of the quality control testing regime is also not a constant itself - many variable come into play in process control and measurement and testing...

The more parts that have to 'fit' together eg camera mount to lens mount(s) the more the cumulative/additive of at the margin in or out of spec parts can influence the mating performance of the parts.

I am not aware of the in spec tolerances that the mount parts from various generations of lens mounts are or were - what I can say is that there is little to zero chance of having a '''perfect" or even similar 'theoretical' fit between them because of the manufacturing processes involved - the difference in fit between one lens and one body is going to be different between a different lens and the same body etc etc...

the CNC machining available to Hasselblad contractors when V lenses were being made are orders of magnitude in machining and engineering terms below teh capability of machines today - lets not forget that the V lenses were designed with tolerances that woudl allow for acceptable focus to be achieved using film as teh capture medium - the tolerances used back then were able to be less than tolerances today. As a hobbyist I have ben using a Schaublin CNC lathe designed to be able to theoretically turn accurately to 0.001mm - in the real world that 0.001 ( one thousandth of a millimeter) is not easy to achieve over even a small batch of parts - because minor changes in tool wear make compensations almost impossible at that level of tolerance - even 0.01 ( one hundredth of a millimeter) is not something that can be achieved consistently over many many exactly the same parts.

TLDR? - what Matt said.

All this is true and well explained, but you're explaining the situation for old V lenses mounted on 500 or 2000 series, the problem appeared with new XCD V lenses, which is much less understandable (at least that's what I think, as the rotational play happens with 38mm and 55mm f/2,5 if I read correctly)
This V series will create a lot of confusion...

 

[PeterA]

All this is true and well explained, but you're explaining the situation for old V lenses mounted on 500 or 2000 series, the problem appeared with new XCD V lenses, which is much less understandable (at least that's what I think, as the rotational play happens with 38mm and 55mm f/2,5 if I read correctly)
This V series will create a lot of confusion...

yes there may be some confusion on my part and yes it is not understandable for me either ....I hope that the problem is confined to Chris' camera and explained by a faulty mount on the camera body....

 

[MGrayson]

Personally I see a pretty drastic difference in how Lr inteprets White Balance (auto or otherwise) compared with Phocus, and I much prefer Phocus. I would probably still use Lr, if only it had a Luma curve tool. Personally I cannot do without that. So for me, I do culling and base edits in Phocus, cleaning up in Photoshop, very occasionally some extra editing, then catalog in CaptureOne. Files from all other digital cameras are 99% pure CaptureOne. Works for, but it’s clearly not ideal.

L*a*b processing is a great thing and can be done in PS as well. It is too bad that LR doesn't go that far. The interplay of contrast and saturation is always a headache. BTW, I didn't even know that Phocus had a Luma curve. It's not the best documented software out there, alas. Do you know if it has ANY auto settings? I couldn't find any, and I find them useful for starting points, because at least I know what the machine is thinking.

 

[jduncan]

Some first impressions of the X2D, FWIW and FYI and YMMV, actually YMWV I ordered mine on announcement day and received it on November 7th, after a multi-day FedEx delay. I didn't order any of the new XCDV lenses as of yet. First impression are that I really like it so far.

The EVF is a very nice improvement, the IBIS is excellent, the PDAF focusing and overall operation is "snappy" compared to the X1Dii. A few small things that stood out:

• the brightness of the EVF can now be adjusted (yay) and it overcomes the super bright Santa Fe, NM Sun
• manual focusing is far easier for me with the higher resolution EVF (some of you know my eyesight is terrible)
• the tilting rear screen is great for tripod use
• the internal storage is a bigger change than I would have imagined - just pick the camera up and go (I bought a CFexpress card but haven't even put it in the camera yet)
• the top screen is way more useful than I originally thought would be the case (originally I thought I would miss the mode dial, haven't even thought about it once)
• AWB and colors are great (as is the X1Dii, 907x/CF50cii and H3D39ii)
• the upload time via USB-c connection is super fast, the camera charges when plugged into the MacPro even during uploading (not so on a PC for some reason) and the USB-c charging is very fast
• battery life seems about the same as the X1Dii (black band higher capacity batteries)
• the X2D is a bit bigger and heavier than the X1Dii, not really noticeable in use
• works well with the existing XCD lenses I tested it with (21, 30, 45p, 65, 90, 120, 135+1.7x and 35-75) - blackout time is marginally less than X1Dii, 120 is slow and it was slow before, 35-75 is great and noticeable quicker (I still have the X1Dii and have tested side-by-side)
• the IBIS is excellent and has brought new life to my array of old V lenses

I am still waiting for an L-bracket - ordered the Peipro L-bracket for X2D from Alibaba ($130) and it should arrive in two weeks.

A few things that didn't stand out, nothing new here, move along.....

• I really miss the they didn't include the embedded GPS
• focus peaking is sorely missing, the assist-indicator is okay but not great
• I really miss the cable release, sigh - hopefully a USB-c based release will arrive
• focus stacking will hopefully arrive via FW update
• blinkies are either on/off - it would be better to have it be one of the scrollable view modes on playback
• lossless compressed RAW would be nice
• the XY electronic spirit level is not a great user interface or I'm just not stable enough to get it reliably level handheld, works fine with tripod and either ball or geared head
• really miss the crop framing bars for various formats (especially square and Xpan)
• wish the focus distance indicator also showed DoF
• a joystick for focus point setting would have been very welcome (like the 907x grip)
• yeah, auto-ISO in manual and a live histogram would be welcome

I kept the GFX100s and lenses to compare side-by-side before making the final decision and for me it was quick, way quicker than I thought it would be - the GFX100s and lenses have found new homes with loving owners Yes the GFX100s has more advanced focus modes and is certainly more refined and feature-rich, lenses cost less ALTHOUGH I never particularly liked using it, plus it is heavier and bulkier to me than the X1Dii or X2D. GFX image quality is excellent although I'll easily give the nod to the X2D and in my style of use they are virtually the same speed of operation and focus (although blackout time is lower on the GFX) - I just didn't use the GFX beyond the basics. Haven't missed it once.

I used the Leica S007 heavily from 2015 to 2021 with 10,000's of images - so I can't compare it side-by-side with the X2D today - it is an incredible system with fantastic lenses. For me it became too heavy and bulky and the used value was still good in early 2021 so it was great timing for me to move to the X1Dii. I have looked at images taken in the same location, time of year and similar lighting (an estimate at best) and for my work the X2D is providing better results with less effort (PaulS and MGrayson have already discussed this). So while I loved the S I do not miss it at all.

The S was traded in for a full Leica SL system with zooms and a some incredible prime SL Summicrons (which are astonishing). So if I need fast operation with IBIS and tracking it's a nice compliment to the X2D.

Now I need to relearn how to scale and upload photos to getDPI to share a few samples

Kind regards, Glenn

Great news. Thanks for sharing.

Cable Release:
Some of the misses are the kind of stuff that makes me believe they did not ask any photographer. The camera has a primitive AF system, not even basic continuous AF like the only SLR, which means that the camera is particularly useful with stuff like product photography, landscape, deliberate portraits, and food: Ideal conditions for a cable release. If they have introduced a wireless remote, or a smartwatch trigger it won't be ideal but workable.

GPS:
They. removed the built-in GPS, again as if they don't know that the camera will be used for landscape and travel photography. Mi guess is that it was about costs but they could have written the software to use the x1d GPS module. It's not rocket science.
They simply choose not to do it.


Best regards.

 

[SrMphoto]

Great news. Thanks for sharing.

Cable Release:
Some of the misses are the kind of stuff that makes me believe they did not ask any photographer. The camera has a primitive AF system, not even basic continuous AF like the only SLR, which means that the camera is particularly useful with stuff like product photography, landscape, deliberate portraits, and food: Ideal conditions for a cable release. If they have introduced a wireless remote, or a smartwatch trigger it won't be ideal but workable.

GPS:
They. removed the built-in GPS, again as if they don't know that the camera will be used for landscape and travel photography. Mi guess is that it was about costs but they could have written the software to use the x1d GPS module. It's not rocket science.
They simply choose not to do it.


Best regards.

How many landscape/travel cameras do you know that have a GPS? Sony, Nikon, and Fuji do not have built-in GPS. Of the current models, only Pentax K1 II and Nikon Z 9 have GPS, AFAIK.
I stopped using cable release a long time ago, and use a timer instead. I understand that there are situations where a remote release is required.

 

[usm]

L*a*b processing is a great thing and can be done in PS as well. It is too bad that LR doesn't go that far. The interplay of contrast and saturation is always a headache. BTW, I didn't even know that Phocus had a Luma curve. It's not the best documented software out there, alas. Do you know if it has ANY auto settings? I couldn't find any, and I find them useful for starting points, because at least I know what the machine is thinking.

Well, there’s no auto-mode like doing magical things on every image individually. But you can setup some fixed corrections you like and apply them automatically while importing fff files. So you can create a setup for example for each lens or indoor an outdoor....

I don’t like it, because I want do start from a more or less uncorrected version.