Jump to content

Medium Format Rumors


Patrick FR
 Share

Recommended Posts

Advertisement (gone after registration)

In the interview posted today, you quote the Fuji Manager as saying, "But in future our goal is to satisfy everyone."  This is ridiculous.  You can never satisfy everyone.

 

The camera market is increasingly competitive and difficult survive in.  While it might be nice to make a Medium Format camera, is the market big enough and profitable enough to justify the investment in a new sensor, new body, and new lenses?  I think that this is a niche market which will not justify the diversion of resources.

 

It makes much more sense to try and make the X-series APS-C range the best it can be and further invest in the lenses available for this range.  There are still many people who criticise Fujifilm X cameras for things like colour fidelity and quality of images.  While we might love Fuji cameras for what they are, I think that the company should do everything to remove these objections and make their cameras the unquestioned best.

 

Nobody can be all things to all people, but Fuji should focus its resources on the X series and not dilute those resources with other ventures.

Link to post
Share on other sites

In the interview posted today, you quote the Fuji Manager as saying, "But in future our goal is to satisfy everyone."  This is ridiculous.  You can never satisfy everyone.

 

Why not? Mostly it's a question of price for the customer. In the future MF cameras might be affordable for everyone. I'd like it will.

Link to post
Share on other sites

well, it is , after all, a famous proverb:

 

“You can please some of the people all of the time, you can please all of the people some of the time, but you can’t please all of the people all of the time”.”
 
So it is not unlikely that “ satisfying everyone” is an hyperbole but not something that can be possibly done on a global scale.
 
So some people will be satisfied by Fuji and some won’t. Some will buy the Fuji to come and some won’t.
 
Nothing new under the sun.
Link to post
Share on other sites

Just my thought in light of the new mirrorless Hasselblad -- Fuji should start with something a little different but familiar.  Consider a MF analog to the X100.  Single lens, very compact (for MF), rangefinder form etc etc. Mirrorless in the tradition of GA645 and/or GW670 but with all mod cons.  Call it an X645.

 

Rumors said it would use a Beyer sensor -- this puzzels me - why not a use Xtrans technology??

Link to post
Share on other sites

BTW the more I think about it, the more I have to conclude that this analysis is probably bang on:

 

I agree with gdanmitchell that Fuji likes to zig when the market is zagging. The reason Fuji has been successful at APS-C is that Fuji APS-C ISN'T a crop system - it's a full frame system for a different frame size! The difference is that all the lenses are optimized for the actual frame size. If you look at Nikon DX or Canon EF-S, there are two distinct types of lenses - APS-C lenses (which are generally cheap zooms - there have been a few primes and better zooms, but they aren't the majority) and FF lenses, many of which are either needlessly bulky or odd focal lengths on APS-C sensors (a 24-70mm f2.8 is a classic example of both, not going usefully wide on APS-C as well as being 50% bigger and heavier than it has to be).

 

There are a few focal lengths that serve different roles in the two formats, but happen to work in both (35mm is the best example - it's a useful slight wide on FF, a normal lens on APS-C, and a 35mm FF lens is very compact, so there's no real disadvantage to using it as a normal lens). There are others that work both ways, but the lens designs are different (a 24mm wideangle on FF is similar to a 35mm if used on APS-C, BUT a good FF 24mm is quickly becoming a big, heavy lens; while it's easy to build a very compact one if you only need to cover APS-C). Long telephotos also work both ways - they inherently cover huge formats (most 300mm f2.8s cover 4x5", or would if they didn't have baffles to prevent reflection), and they tend to be "the longer the better", rather than an optimized focal length for a particular effect.

 

By designing an entire lens lineup for APS-C, Fuji has the right focal lengths at reasonable sizes (note that their "50" is actually a 56 - a regular 50 works as a portrait lens on APS-C, but it's always just a bit short -  56 is a better length)... There are some big lenses, but they are unavoidably big lenses, not needlessly bulky ones. A 100-400, especially a scary-sharp 100-400, will never be a small lens!

 

I'd expect (and hope for) them to do the same thing in medium format. If they embrace 33x44mm and design all their lenses for it, they will have significant size and quality advantages over manufacturers trying to repurpose 645 film lenses. On the other hand, I hope they go for a newer sensor, rather than that 3 year old 50MP one. It would be very easy for Sony to make a ~70-75 MP 33x44 mm sensor that uses the same technology generation as the X-Pro 2 sensor, the A7rII sensor and the big 100 MP sensor Phase One has. Fuji and Pentax are the logical customers...

 

I am now voting for a cropped 44 x 33mm MFD system. I think this is more future proof than a full frame MFD system. 

 

The problem with full frame MFD is that it is hard to make the lenses large enough to take full advantage of the larger size of the format. This in turn limits the maximum aperture of lenses. For example, comparing macro lenses, the Leica S (cropped MFD) 120mm macro has a maximum aperture of f/2.5, whereas that of the Hasselblad H (full frame) 120mm macro is only f/4. 

 

Another advantage of the X-system is that it is possible to make the optics "huge" relative to the sensor size. This is probably one of the reasons that X-system lenses are sharper shooting wide open than full frame lenses. The higher sensor resolution climbs the more it merely accentuates the hopeless softness of larger format lenses shooting wide open, and while you could say "well then just make the optics bigger and better to match", the lenses will be huge and ridiculously expensive. 

 

While larger format sensors have an advantage at the moment when it comes to light collecting ability, resulting in better S/N ratio and dynamic range, there will come a point when it will become increasingly less necessary to resort to something as crudely simplistic as a brute increase in sensor size to achieve that. Then increasing format size will result in ever diminishing returns, while only highlighting the disadvantages that come with increased format size. 

 

The next advantage of a 44 x 33mm format system is that if will make it easier for Fuji to design the mount to take IBIS. If what they say is true and an IBIS mount results in a larger lenses, then all the more reason to make it a 44 x 33mm sensor system to avoid an excessive blow-out in lens sizes. Once again, I certainly hope that Fuji will not release their system based on the current Sony 51MP 44 x 33mm sensor and that they will wait for the next generation of this sensor format to come out first. At which point, if that sensor has about 75MP resolution, then without IBIS it would start to become a tripod only studio camera, with this only getting worse in future as sensor resolutions increase. It's all very well creating a portable MFD system, but if it becomes almost impossible to shoot with it hand held, then it negates the point of the system.  

 

The last advantage of a cropped MFD sensor is that it makes it easier to spread focus points more widely. The last thing you want is to be limited by having too few focus points crowded into the centre of the viewfinder. It is all very well having shallower depth of field, but if you have only one focus point like a Hasselblad...

Link to post
Share on other sites

Rumors said it would use a Beyer sensor -- this puzzels me - why not a use Xtrans technology??

I guess the marketing idea is very simple: to share the same sensor between cameras of different prices. Such big sensors as known have very vide distribution of critical electrical parameters and just small part of sensor's yield satisfies the specs of Hassel.

 

However it would be not bad for us to get these not the top performed sensors in future Fuji MF camera if the price will be much reasonable in comparison to Hassel price. 

The same way it was done before: Hasselblad H5D-50C, Phase One IQ250 got the best performed DD/noise sensors and the same by process but lowest performed sensors were sold in cheap Pentax 645Z. 

The price of one x-trans die (sensor) would be too high for Fuji if not share semiconductors yield with somebody else.

Link to post
Share on other sites

If there is one thing I am certain about and that is if it doesn't have a X-trans sensor, I'm not going to have anything to do with it. Just having another Bayer sensor MFD camera out there is insufficient product differentiation.

 

But price will. If Fuji is cheaper then the blad, this will bring in more sales. Will hopefully see later on what Fuji brings to the table.

Link to post
Share on other sites

I guess the marketing idea is very simple: to share the same sensor between cameras of different prices. Such big sensors as known have very vide distribution of critical electrical parameters and just small part of sensor's yield satisfies the specs of Hassel.

 

However it would be not bad for us to get these not the top performed sensors in future Fuji MF camera if the price will be much reasonable in comparison to Hassel price. 

The same way it was done before: Hasselblad H5D-50C, Phase One IQ250 got the best performed DD/noise sensors and the same by process but lowest performed sensors were sold in cheap Pentax 645Z. 

The price of one x-trans die (sensor) would be too high for Fuji if not share semiconductors yield with somebody else.

I thought, perhaps incorrectly, that the basic semiconductor chip is the same for current Nikon and X-trans.  The difference being the RGB pattern (X vs Beyer) that is overlaid.  If that is correct, most of the Hassleblad and X-trans process would be the same - you would only need to allocate wafers (or die) to one product or the other until you were a number of process steps down stream.

Link to post
Share on other sites

  • 2 weeks later...
  • 4 weeks later...

Still no news on the price point though? A medium format camera could be great up sell (or even recapture) those who have looked at Sony a7riis. I'd love a medium format kit for landscapes and serious tripod stuff but the existing ASPC line up for travel etc.

 

Price will be key though!

Link to post
Share on other sites

My own desire would be to see a Medium Format Rangefinder aesthetic. I always wanted to see another Mamiya 7 type offering from somebody. A light weight (relatively speaking) field camera. A high resolution "monster" I could take on hikes (along with a tripod) for landscape photography. And yes, I do desire really large prints. I am excited to see what new technologies and developments are around the bend. 

Link to post
Share on other sites

Does it mean that Fuji admits that the X-Trans pattern is inferior to the Bayer pattern? - Just kidding.

 

But it would be strange if they praise X-Trans for APS-C format because it gives better IQ and then use Bayer on their medium format camera. Medium format buyers want the very best IQ and will compare the Fuji MF camera with the PhaseOne XF 100MP. If X-Trans can help to fight above its weight they should use it.

 

I personally do not care neither what color pattern is used in the MF nor if X-Trans is a few percent better or worse than Bayer.

Link to post
Share on other sites

Does it mean that Fuji admits that the X-Trans pattern is inferior to the Bayer pattern? - Just kidding.

 

[...]

 

Depends on how Fuji want to do it, either they go back to the R&D and works on the XF sensor to be converted into MF or use one that is already existing on the market and just adapt their technology to it.

 

Both solutions are perfectly viable on a production level, I guess it's mostly a choice that Fuji had/would made.

Link to post
Share on other sites

The decision to omit the X-trans configuration is almost certainly driven by cost. They want it to be cost competitive with the Pentax 645Z. 

 

Perhaps once the cost of manufacture of 4433 sensors comes down, Fuji might consider adding it to a future release. The original X100 had a Bayer sensor, and only later did the X100 acquire the X trans sensor. The days of Bayer sensors are numbered. They will be gradually replaced by non-Bayer sensors of all different varieties e.g. stacked sensors and organic X-trans sensors.

 

A Japanese magazine did a comparison between the Sigma 40MP Foveon sensor vs 16MP X-trans vs Canon/Nikon Bayer sensors. It suggested that the X-trans sensor does have the ability to resolve detail to a degree that outstripped its 16MP limitations:

 

https://plus.google.com/116458677975033889029/posts/Z39J4VuvYwA

Link to post
Share on other sites

Just another little word about the design of the upcoming Fuji medium format camera. I see we have a mock up suggesting that the Fuji is going to resemble the Hasselblad X1d. The problem is that Fuji have said this:

 

https://fujifilm-blog.com/2015/06/30/interview-with-mr-takashi-ueno-from-fujifilm-tokyo-why-dont-fujifilm-make-full-frame-dslr/

 

“Firstly, the angle of light that film and imaging sensors can receive differ from each other. Film can receive light at the slanted angle of up to 45 degrees without any problem, but in case of the digital camera, the light needs to be as perpendicular to the sensor as possible. Slanted angle light causes mixed colours and therefore the real colours sometimes cannot be reproduced. In order to receive the light perpendicular to the sensor, it is important to make the rear glass element on each lens as big as possible to put the light beams parallel from the outlet of the light to the sensor. Finally, the back-focus distance should be shortened as much as possible to eliminate the degradation in image quality.”

 

 
In other words, when light hits the corners of the sensor at too steep an angle there is degradation of image quality and vignetting. The shorter the flange distance, the steeper the angle of light incidence at the sensor corners:
 
flange-focal-distance-and-angle-of-incid
 
However, there are also advantages to having a short flange distance. On of these is improved acutance. Being able to eliminate the space taken up by the mirror box is also an advantage in lens design. However, make the flange distance too short and you run into the problem with degradation of corner performance.
 
One way to have a short flange distance and then reduce the angle of light incidence in the corner is to introduce telecentric design principles into your lens. You can see this on Sony FE mount lenses, which are often remarkably long. The way this lens elongation works is by creating extra distance between the sensor and the lens exit pupil position, which is moved further away from the sensor using an additional rear element. The disadvantage of this is that it makes the lenses really big, and the shorter you make the flange distance the more the lens size blows out. Because aperture size also affects the angle of light incidence in the corners, the wider the maximum aperture the more you need this telecentric lens extension i.e. when you increase the maximum aperture a lot it causes a disproportionate blow out in the lens size. So if you put these two elements together—short flange distance plus ultra wide maximum aperture—you end up with a massive lens.
 
With this you can understand the way lens designs work on mirrorless systems with extremely short flange distances. The best example of this is the Sony FE mount, a full frame mount with APS-C dimension. The reason why the full frame FE mount has an 18mm flange distance is because it was originally an APS-C mount ("NEX mount"). Compare that to the Fuji XF mount, which has a flange distance of 17.7mm, or the APS-C Canon EOS-M mount at 18mm. Now you should be able to understand why Sony FE mount lenses tend to avoid ultra wide maximum apertures, and when you do get these, the lens size blows out to become disproportionately large.
 
On the other hand, while increasing the flange distance reduces the size of the lens, you don't want to make it too long either, because that would negate the advantages of having a short flange distance. You can see the advantages of a longish flange distance in the M4/3 system, which has a 19.25mm flange distance, which is slightly longer than APS-C mounts despite being a smaller format. However, the lenses become really compact, even when they have fast apertures, and are much more affordable. It's all matter of give and take. Either you have a slightly bigger body with lots of compact lenses, or you can have a smaller body and lots of really big (i.e. expensive) lenses.
 
Now we have another example of a mirrorless mount that has followed the Sony example of setting an ultra short flange distance and that is the Hasselblad X mount, which has a flange distance of about 20mm. That is only a millimetre or so more than the M4/3 mount flange distance of 19.25mm. So we have a medium format mount with dimensions more like an APS-C or M4/3 mount. If you look at the Hasselblad X mount lenses, you will notice that they have quite slow maximum apertures. 
 
In Hasselblad's case the decision may be deliberate. They want a small, portable system. It is an enthusiast's entry level system designed for slow lenses with modestly short focal lengths, rather than a studio photographer's system. Short focal lengths permit the omission of a retrofocal element and this helps the compactness of the lens further. On the other hand, by having such an ultra short flange distance, they are giving up on being able to readily make fast lenses (at least not without a huge blow out in fast lens size to brick like proportions), as well in the performance of ultra wide focal lengths where the corner angle of incidence becomes particularly steep. Hasselblad have deliberately introduced certain strengths and limitations into their X system so that it doesn't complete with their H system. 
 
As for Fuji, they don't have multiple competing medium format systems. There is probably only going to be one system. And Fuji have expressed concern with excessively steep angle of light incidence in the corners. That can only mean that Fuji are going to take the flange distance of their XF system and increase it proportionately up to medium format dimensions. That will make the body larger, but it will keep the size, weight, and price of their lenses down. My prediction is that they will take the XF system flange distance and increase by the crop factor from the 4433 size sensor. This ought to give you an approximately 34mm flange distance, which is still a lot less than the 53mm of the Leica S medium format DSLR mount. That means the new Fuji will likely be nowhere near as slim as the Hasselblad X1d, but that is a good thing in many ways. We should welcome it, because it will control the blow out in lens size, especially when designing ultra wide aperture lenses. 
 
 
Link to post
Share on other sites

Just another little word about the design of the upcoming Fuji medium format camera. I see we have a mock up suggesting that the Fuji is going to resemble the Hasselblad X1d. The problem is that Fuji have said this:

 

https://fujifilm-blog.com/2015/06/30/interview-with-mr-takashi-ueno-from-fujifilm-tokyo-why-dont-fujifilm-make-full-frame-dslr/

 

 

 

In other words, when light hits the corners of the sensor at too steep an angle there is degradation of image quality and vignetting. The shorter the flange distance, the steeper the angle of light incidence at the sensor corners:

 

flange-focal-distance-and-angle-of-incid

 

However, there are also advantages to having a short flange distance. On of these is improved acutance. Being able to eliminate the space taken up by the mirror box is also an advantage in lens design. However, make the flange distance too short and you run into the problem with degradation of corner performance.

 

One way to have a short flange distance and then reduce the angle of light incidence in the corner is to introduce telecentric design principles into your lens. You can see this on Sony FE mount lenses, which are often remarkably long. The way this lens elongation works is by creating extra distance between the sensor and the lens exit pupil position, which is moved further away from the sensor using an additional rear element. The disadvantage of this is that it makes the lenses really big, and the shorter you make the flange distance the more the lens size blows out. Because aperture size also affects the angle of light incidence in the corners, the wider the maximum aperture the more you need this telecentric lens extension i.e. when you increase the maximum aperture a lot it causes a disproportionate blow out in the lens size. So if you put these two elements [/size]together—short flange distance plus ultra wide maximum aperture—you end up with a massive lens.

 

With this you can understand the way lens designs work on mirrorless systems with extremely short flange distances. The best example of this is the Sony FE mount, a full frame mount with APS-C dimension. The reason why the full frame FE mount has an 18mm flange distance is because it was originally an APS-C mount ("NEX mount"). Compare that to the Fuji XF mount, which has a flange distance of 17.7mm, or the APS-C Canon EOS-M mount at 18mm. Now you should be able to understand why Sony FE mount lenses tend to avoid ultra wide maximum apertures, and when you do get these, the lens size blows out to become disproportionately large.

 

On the other hand, while increasing the flange distance reduces the size of the lens, you don't want to make it too long either, because that would negate the advantages of having a short flange distance. You can see the advantages of a longish flange distance in the M4/3 system, which has a 19.25mm flange distance, which is slightly longer than APS-C mounts despite being a smaller format. However, the lenses become really compact, even when they have fast apertures, and are much more affordable. It's all matter of give and take. Either you have a slightly bigger body with lots of compact lenses, or you can have a smaller body and lots of really big (i.e. expensive) lenses.

 

Now we have another example of a mirrorless mount that has followed the Sony example of setting an ultra short flange distance and that is the Hasselblad X mount, which has a flange distance of about 20mm. That is only a millimetre or so more than the M4/3 mount flange distance of 19.25mm. So we have a medium format mount with dimensions more like an APS-C or M4/3 mount. If you look at the Hasselblad X mount lenses, you will notice that they have quite slow maximum apertures.

 

In Hasselblad's case the decision may be deliberate. They want a small, portable system. It is an enthusiast's entry level system designed for slow lenses with modestly short focal lengths, rather than a studio photographer's system. Short focal lengths permit the omission of a retrofocal element and this helps the compactness of the lens further. On the other hand, by having such an ultra short flange distance, they are giving up on being able to readily make fast lenses (at least not without a huge blow out in fast lens size to brick like proportions), as well in the performance of ultra wide focal lengths where the corner angle of incidence becomes particularly steep. Hasselblad have deliberately introduced certain strengths and limitations into their X system so that it doesn't complete with their H system.

 

As for Fuji, they don't have multiple competing medium format systems. There is probably only going to be one system. And Fuji have expressed concern with excessively steep angle of light incidence in the corners. That can only mean that Fuji are going to take the flange distance of their XF system and increase it proportionately up to medium format dimensions. That will make the body larger, but it will keep the size, weight, and price of their lenses down. My prediction is that they will take the XF system flange distance and increase by the crop factor from the 4433 size sensor. This ought to give you an approximately 34mm flange distance, which is still a lot less than the 53mm of the Leica S medium format DSLR mount. That means the new Fuji will likely be nowhere near as slim as the Hasselblad X1d, but that is a good thing in many ways. We should welcome it, because it will control the blow out in lens size, especially when designing ultra wide aperture lenses.

 

I do not pretend to be an expert on lens design, but I think you have misread Ueno's statement that you quoted. Ueno' states "In order to receive the light perpendicular to the sensor, it is important to make the rear glass element on each lens as big as possible to put the light beams parallel from the outlet of light to the sensor." He DOES NOT say that a way to reduce the angle of the light from the outlet of the lens to the sensor is to increase the flange distance. In fact, in his very next sentence, he says the opposite, "Finally, the back-focus distance should be SHORTENED as much as possible to eliminate the degradation in image quality. " [Emphasis mine] As I read that second sentence of Ueno's statement, this is a separate, independent image quality factor.

 

As I studied the diagrams in your post, it seems to me that an increase of the flange back distance will not affect the angle of light to the sensor. Draw a straight line from the point of Convergence to the sensor. The angle of that line to the sensor will be 90 degrees if the sensor is parallel to the back of the lens. As light rays leaving the lens will travel in a straight line, a ray emerging from the point of convergence which strikes the corner of the sensor will be the long side of a right triangle formed by the sensor and a line from the point of convergence to the middle of the sensor. Given that the angle of the line from the point of convergence to the sensor is 90 degrees, the other two angles of the triangle must add up to 90 degrees. So, as long as the line from point of convergence to the sensor edge and the sensor's midpoint are straight, the angle of light hitting the sensor edge will remain constant regardless of the distance between the back of the lens and the sensor.

 

To put this another way, as long as the back of the lens and the sensor are parallel, any line intersecting both parallel lines has to cut them at the same angle. It would make no difference how far apart those parallel lines are, the angle that the bisecting line creates won't change.

 

I noticed on your diagrams above, that in both Diagram A and Diagram C, the lines leading from the point of convergence to the sensor edge are not straight. Rather, they seem to be bent, thereby allowing the angle of the line to become more perpendicular to the sensor, but in Diagram B, the lines remain straight. Why is that? Obviously, the change in angle to the sensor seen in Diagrams A and C has nothing to do with the distance, and everything to do with whatever caused the change in direction of those lines.

 

Again, I'm not pretending to be an expert on this, but from the information you used in your argument, I'm not sure that the conclusion regarding the flange back distance is correct.

Edited by Jlrimages
Link to post
Share on other sites

All mirrorless lenses have a reduced back focus distance. That is integral to mirrorless lens design. That particular part of lens design is to optimise the acutance of the lens rather than to reduce incidence of light angle at the corners primarily. 

 

Yes, the concern about angle of light incidence reducing corner acutance is a basic principle of modern digital optical design. Yes, a shorter flange distance is part of the equation in mirrorless design that needs to be overcome and accounted for. This is the reason why most companies now have patents for curved sensors as this reduces the angle of corner incidence even when flange distance is much reduced. Sony also said that the corner vignetting resulting from a shorter flange distance in a mirrorless mount was the reason for them adding BSI to their sensor for the a7RII. Sony engineers explained that the reason for this is that sensors have a three dimensional architecture. If the light hits the architecture at an angle, it casts a shadow, just like the sun rays at sunset causes buildings in a city to cast a shadow. BSI is intended to partially counteract this. 

 

Also the diagram I quoted does not show the effect of the addition of telecentric design as this has the effect of effectively increasing the point of convergence from the sensor thus reducing the angle of corner induce. The diagram as shown clearly shows that the angle of light incidence on the sensor corners is increased by shortening the flange distance. It is mathematically IMPOSSIBLE to do otherwise. 

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
 Share


×
×
  • Create New...