I'd always assumed that a 48mm x 48mm sensor was unlikely. However, with the new Sinar/Leaf/Rollei camera, there's suddenly lots of 48mm x 48mm talk. How realistic is this bigger sensor?

i believe there will be for sure a larger sensor, the question is when. will it be the next generation or in the 2. generation from no on. i belive more in the last ...... but i am sure the sensors will grow.
personally i hope not in the 48x48 direction which will bring bigger files which mostly will be cropped. i would hope for a real 645 sensor .

What would be the manufacturing implications of a genuinely "full frame" 645 sensor (ie 55mm or 56mm wide) versus a 48mm x 48mm sensor, versus the current sensors?

Dalsa has the 48X48 sensor but so far not have been made into commercial use yet, or at least as I know it. I think the evolution of format moved from 6X6 then to 6X4.5 for some good reasons and may be in some time there will be a small support for such a sensor for the camera platforms use everyday. At least Hy6 offer such a possibility if such sensor does become commercially available but really it is still designed as camera use both digital and film rather than Hasselblad primary designed for digital and with film option.
I was more or less hoping for just the current or true 645 size sensor but in a phsical movment enclosure so it can take single 645 shor or automatically stitched piture by moving the sensor - of course this will not be useful for action shot but I guess for a lot of application this will be useful.

High K, hello.........
after reading your experiences of the training and after listening to thierry i think i start to be interested to see the HY6 ....... also very nice that there are plans to make again a rotatable back. i have that on my gottschalt and its such a practical detail.

I think the camera has some great potential, to be honest. The rotation mechanism, if finally made possible with acceptable tolerance, will be indeed a very good feature, what I understood is that is has been tested and will only be introduced when Sinar is satisfied with its performance.

Hi
I think in terms of marketing a true 6x4.5 would be best. A square chip you will be cropping 95 % of the time.

Denis

Unless you like the square for e.g. portraits (some even use it for landscapes ).
A Hy6, WLF and 48x48 sensor sounds great to me!

But what if the cost is rather high> I always hear people say " when that arrives i will buy it".... but then they say " lets wait for the right price " and " I do not need it at this moment ...) but the actually want to say "" I do not have the monney to buy that, but i really like to show off about it""
Sensors are getting better each year and when your business is profittable you just buy a new back each time your business feels the need for it.

For back manufacturers, 48x48 is a great way to release some new uber model with little technological investment.

Plus, it might come handy to help a new platform - like the Hy6 - making the headlines with a "differentiator".

The only thing that is unclear is whether photographers really needs the solution, besides those who just want to be seen with the latest stuff of course.

Regards,
Bernard

I'd love to see a big square sensor.

Before I went MF digital I used almost exclusively 6x6.

Embrace the square!

Square will not do much for me so I would pass on it, but come out with 6x7 and I will not hassitate even a second to pay $60K for this double size sensor.

http://andrenapier.com/

As far as I can tell, the new Sinar/Leaf/Rollei camera is based on 56x56mm square format for one simple reason: it is based on Rollei's 6x6 film format system including its lenses. Nothing about this "format inertia" is evidence that Kodak, Dalsa or any other sensor maker has told the Rollei team that larger, squarer sensors are coming.

A reversion to square from the now dominant 4:3 sensor shape is particularly unlikely. With industry leader Hasselblad-Imacon/Fuji along with Mamiya and Pentax having moved to 645 (56x42mm) and smaller formats, the market for sensors usable by none of them would be far too small.
Filling out more of the 56x42mm frame seems less unlikely, but there is no hint of that either: the latest generation of Dalsa and Kodak sensors did not increase at all in size.

Thanks for that. Over the years I've seen several comments that different sized sensors are all made from a standard size "chip wafer", I was wondering how a 48mm x 48mm or 56mm x 42mm sensor would fit within this standard wafer and what the attendant cost implications would be?

Wafers are easily big enough to fit larger sensors: diameters of 200mm are widely used, with 300mm coming in to use. The number of sensors per wafer does go down a bit faster than in proportion to sensor area, but that does not seem like a major cost factor.

The biggest cost factor with increasing sensor size seems to be the decreasing proportion of usable chips: the likelihood of a chip being rejected due to fatal defect goes up rapidly with increasing chip size, once chips are large. For example, if 1/4 of chips at one size are usable, only 1/16 will be usable at twice the area, giving about 1/8 as many usable chips per wafer. Another example: Kodak's 36x48mm sensors seem to cost about fifty or more times as much as its 4/3" sized ones, while the area is only eight times as great. And there is competitive pressure on those 36x48mm sensors, from Dalsa.

If you ask me, I'd rather have a 4:5 ratio sensor than the popular 3:4.

645 is 3:4 and all bigger sensors in 36-7 X 48-9 mm. sizes. One good thing about this ratio, it fits perfectly magazines pages.
The 4:5 ratio at current sensor sizes would be 39X49mm. If full frame, 44X55.

4:5 ratio equals 6X7, 4X5, etc, film formats. which in turn fit perfectly traditional darkroom paper sizes.
I know, we are in the digital era now, but 4:5 ratio works better (at least for me) for portraits and weddings. The extra space at the sides in verticals can help a lot in positioning bride veils, the couple and the works.
3:4 ratio is more constrained and 2:3 ratio is just plain unfitable for serious wedding work.

I always enjoyed the extra room I had with my Hasselblad 55X55mm negatives doing portrait shots. I would go for a full frame square sensor, but a 44X55 sensor would be cheaper and just as good. Of course only if provided with a revolving back. No re-attach funny thing.

Just my opinion.
Regards

The trick of the large Dalsa chips is to stitch several smaller ones together to one big surface. This way they can decrease the high outtake rate considerably. A modern CCD as it appears in the Leaf and Sinar backs is built up of six smaller pieces. This stitching technique was invented by Philips. The entire Philips lab/factory for large high quality imaging CCD's was sold to Dalsa. A little history from Philips Password Magazine, issue 19:

"Started in Philips Research in the early 70s, solid-state imaging based on CCD (Charge-Coupled Device) technology was a big success technically but less so commercially in the consumer field. The company developed one of the first sensors for HDTV and launched this in the HDTV cameras used in the 1992 Barcelona Olympics. Subsequently acquired by Philips Semiconductors for developing image sensors for consumer applications, the SSIS group expanded its activities into CMOS-based sensors. When Philips decided spin-off its professional image sensor activities, the professional part of the SSIS business was sold to the DALSA Corporation, Ontario, Canada in 2001 to form DALSA Professional Imaging. DALSA, which specializes in digital imaging for professional applications, was particularly interested in acquiring Philips’ CCD imaging technology and intellectual property which offers major advantages over competing CCD imaging technologies. This includes excellent dark-current performance and low power consumption as well as Philips’ proprietary ‘chip stitching’ technology that allows sensor chips to be stitched together into large-area sensors. Also interesting to DALSA was Philips’ advanced 0.5-micron wafer process for image sensor semiconductors."
Click to view attachment

Forgot to add: the separate smaller CCD parts that form a Dalsa chip are 12mm x 24mm. So stitching two more to the 36mm x 48mm CCD will already result in a CCD of 48mm x 48mm. Given the fact that the outtake rate is well under control due to this stitching technique, it is only a matter of willpower from Dalsa's side to create this economically feasible CCD. I am sure we will see it in the near future, especially since the back makers that are locked out of the H3D route will have an edge then. There are such huge amounts of Hassy V system bodies out there that could be used with this 48mm x 48mm CCD that it would be stupid not to recognise that fact. And on the high-end side we will have the Hy6 to work with this large square imager.

Very interesting!

That leaves me again with the question why the MF backs have such high prices. In the past it has been said by almost everyone who has discussed that topic that the large sensor surface results in very low yields. Apparently this argument was wrong, the actual surface of the parts is not even as large as a APS/ Dx sensor. In this case I guess that even larger (than 48x48mm) sensors will be made one day. 48x60mm as the digital version of a 6x7 film format comes to mind...

Could it also be that the centerfold issue of the Leaf backs is caused by stitching? This is what I always have guessed, I just believed that 36x48 was formed by 2 "FF" sensors, 24x36 each.

regards, bernie

this sensor already exists in Dalsa's drawer!

The question is if the backmakers decide to take it.

Thierry

Dear Alba,

I had explained in a tread some months ago, that the costs of a high-hend digital back are not longer due to the high costs of the sensor: this is past. 10 years ago the price of a sensor was about 50% to 60% of the total costs of a back.

Today these sensor costs are much less, less than 20% of the total (manufacturing) costs of a back.

The rest of the costs of a MFDB?: R&D and Software development represent a big part of it.

Centerfold: yes, this problem comes from the "stitching" process. It has as well been well explained a few months ago in another tread by member Brumbaer.

The 38x48mm sensor have six little and rectangular "stiched" parts: if you hold the sensor in an angle to the light you shall see it easily.

Best regards,
thierry

edited for addendum

I'm sure someone will correct me if I'm wrong, but I think the stitching happens in the exposure phase on a single wafer. It's not stitching 6 separate chips together. So, you still need a relatively large, relatively flawless continuous area of silicon.

Oh, backmakers are such pussies

Ed


The 38x48mm sensor have six little and rectangular "stiched" parts: if you hold the sensor in an angle to the light you shall see it easily.

Isn't it the other way around? By that I mean a back manufacturer had to have ASKED for it to be made. I doubt Dalsa or Kodak just make these expensive sensors and wait for them to sell. So who was it?

Although I'd like the Hy6 to be the 1st platform to utilize a 48*48 chip, I wouldn't be surprised if Hasselblad get there 1st with a square chip to attach to all the V series cameras out there, or even adapt the H series to it, let's face it, their platform is king of the hill right now even if the H series is less than ideal, a LOT of people are buying it and rentals are doing very well too.
We don't hear too much of them in terms of developments / rumors, I guess they have their heads down working on a few surprises themselves, maybe one of them is in terms of a larger sensor.

-Nik

And, Thierry, how about Sinar's willpower to use this 48mm x 48mm CCD?

I very much doubt that the V series will be resurrected in that way. It just can't compete with some of the newer platforms with its half f-stops, lack of metering and auto exposure modes, lack of autofocus, lack of digital integration, etc. The Hy6 would beat it in almost every possible respect.

The H series can't be adapted to 48x48 because the viewfinder can't be enlarged to show the whole image, and the lenses aren't designed with a sufficient image circle. Just makes no sense at all.

I can only see the 48x48mm back going on a Hy6/AFi or maybe even a Mamiya RZ67 via an adapter.

Personally I'd rather see a full-frame 645 chip (42x56mm) which I won't be cropping 99% of the time. If a chip has to be square then why not 56x56mm?

If a chip is 56mm x 56mm where does the area surrounding it go? In the case of the Hy6 there is sufficient space, but I'm not sure this is valid for all other 6x6 systems. Of course a 56mm x 56mm CCD would be the ultimate imager to have. Full digital coverage of a Zeiss 6x6 fisheye... a dream would come true.

Hassy probably won't support that one... if I am not mistaken, their latest 28 mm only covers 36 x 48 mm, right...

I don't see Kodak investing money in a format that one of their main customer isn't likely to use. Now Dalsa might be different...

Cheers,
Bernard

If Dalsa already makes this CCD (see Thierry's comment) and its main back making customers (Leaf and Sinar) have the camera under their own brand name to use it (the Hy6/AFi), why would they not make a back to max out the advantage of that camera? Isn't competition about exploiting your unique selling points?

The lens creates an image circle, so in THEORY it could produce the 48 x 48mm image. Things that may cause a problem include the rear lens baffle, the mirror, obstructions in the mount, etc. It's at least plausible that the 28mm lens could produce a 48x48 image circle.

Are you sure the Hassy 28mm will cover a 48mm x 48mm area without light fall-off in the corners? For a 48mm x 36mm CCD a smaller image circle is sufficient. (Same reason why you can't use a 4.5x6 lens on a 6x6 camera.)

No comment, as said in earlier posts: it makes no sense to speculate on something which might take years until available.

And to be honest, although discussions have surely taken place, I am not informed (yet).

Dear Nik,

Sometimes it is one way around, sometimes the other: this time Dalsa have developped this sensor (and even much bigger = see one of my earlier posts a few weeks ago) and is looking to be able to sell it.

But they certainly won't go in production stage unless they get firm order(s).

Best reggards,
Thierry

Dear Ed,

Yes, they are!

Such a decision has to be thought over: there are much more aspects than being the technological leader by having the biggest sensor and the possible resulting sales potential . The risk factors are at least as important, if not more.

Best regards,
Thierry

GBPhoto,

Yes, you are right.

Best regards,
Thierry

Sure, I agree 100% with you on this.

Regards,
Bernard

A square chip in a DB is an absolute certainty in my view. You don't place an order with a camera manufacturer for a square format camera that will accept lenses specifically made to cover square and then only produce 4:3 backs for it. It's a bit late don't you think? The thousands of man hours spent developing the camera's optimum shape, size and weight distribution for square would be wasted, not to mention the time spent on lens design to cover this format. Can anyone from Schneider offer any input or validate this?

No disrespect, but being a pussy is one thing, being stupid is another!

-Nik

Hi
If they do come out with a square chip being Dalsa for Leaf & Sinar. They will also have to make one for the Mamiya 6x7 format. Personally I don't see the advantage to square.

Dear Nik,

you are certainly 100% right, absolutely! No question about that.

My point is: such a decision has to be thought over carefully, and having all the aspects and risks in front of the eyes. All these considered, the decision can go in both directions.

And more over, even if the decision to go with it is taken, it is not for TOMORROW!

So why speak and speculate about something which is not here and ready, and which might take certainly more than a year to be in the market?

Others and myself have been much "criticized" here (no harm for me) for speaking about the Hy6 when it was launched during Photokina. And it was an officially announced product.

Best regards,
Thierry

For me personally a Hy6 with a rotating 36x48mm MFDB would be all I want.
For my work the file-size that comes out of my eMotion 75LV is more than large enough.
After I sold my Blad and went to the RZ and after that the Fuji GX680 I really enjoyed the rotating
backs.
Now I'm waiting to take delivery of the Hy6 and that's the camera for me.
I was forever cropping my Polaroids with the 6x6 Blads.
However,if Sinar,Phase or Leaf will come out with a square sensor there will be lots of users that
want a larger and square file.

Dear Thierry,

Could you please elaborate us on all the aspects and risks that are involved in a decision to use a larger and probably square CCD in a Sinar back? What puzzles me is that the CCD is readily available, it uses known interpolation technique since it is just an enlarged piece of silicon and has the same Bayer pattern as smaller ones, and it works with the same electrical technique as the current CCDs Sinar uses. What would make it take years to implement this larger CCD when the underlying technique is readily available? (This is not meant as a rhetorical question, but one I would like to get an answer to.) Also, Yair, feel free to chime in on this subject.

I can understand that you don't want to "speculate" here. Speaking about future plans of your employer could harm today's sales and we all understand that companies always try to stretch their customer's wallets to the max, preferably for a long period of time. That's just plain business. However, stretching wallets untill they snap would be the end of all business in this area. I have known several photographers who went broke only because of their very high investments in digital technology that became obsolete long before they had reached their return on investment. It's not just the digital back makers that take risks here. We are all in the same boat and some more solidarity from the back makers with the photographers would be most welcome, IMHO.

Regards,
EPd

Willem, didn't you think it was a nice luxury to be able to crop afterwards? And also: do you never want to use real wide angles anymore? A larger CCD does not nessesarily have to lead to larger files. A 45MP or 50MP 56mm x 56mm back would just be fantastic, I think. Lower noise, higher ISO values. Everything is already there: the lenses, the cameras, the routine. The waiting for is that full frame back to make it all work the way it did with film.

Because it's been a slow week work-wise and I'm bored. Thankfully next week looks better. Drinks are on you at Photokina '08 if my speculation was right.

-Nik

I believe Thierry is referring to the business risk, not the technical merits of the chip itself.

Epd,
I would be happy with a 50MP 56x56 full frame sensor.
At the moment I have a four year lease on my e-75LV and I won't be upgrading till then.
It will have to do for now.
Like I said before,I prefer to frame the image while I shoot and the images are always either
landscape or portrait.
Very rarely square.
WR.

I very much doubt that the V series will be resurrected in that way. It just can't compete with some of the newer platforms with its half f-stops, lack of metering and auto exposure modes, lack of autofocus, lack of digital integration, etc. The Hy6 would beat it in almost every possible respect.
-foto-z
I think there are so many V systems gear around the world waiting for the right dback at the right price, that it doesn’t need to be resurrected. Dback makers could easily profit from existing V system. Of course, Hy6 would be beat it, but at an extra $ 20k usd for body, lenses and accessories. I think shooting raw, half stops are not a real necesity.
Eduardo

---------

A square chip in a DB is an absolute certainty in my view. You don't place an order with a camera manufacturer for a square format camera that will accept lenses specifically made to cover square and then only produce 4:3 backs for it. It's a bit late don't you think? The thousands of man hours spent developing the camera's optimum shape, size and weight distribution for square would be wasted, not to mention the time spent on lens design to cover this format. Can anyone from Schneider offer any input or validate this?
No disrespect, but being a pussy is one thing, being stupid is another!
-Nik
I agree with you Nik. I see a 6X6 sensor for the near future. However
putting a rectangular sensor in it, it’s not a waste of resources. It just shows the versatilty of the original idea. And that was not to have to keep turning the camera from hor to vert, over and over. In the digital era, since sensor area is such an expensive comodity, rectangular sensors make economic sense. Film was (still is) so inexpensive, that having the extra room didn’t make any real harm in our pockets.
Eduardo

---------

If they do come out with a square chip being Dalsa for Leaf & Sinar. They will also have to make one for the Mamiya 6x7 format. Personally I don't see the advantage to square.
-mcphoto
I believe square is actually a convenience from the film era, not an advantage. Just, it caught up with many photographers and square format became favorite for many photographers doing portraits, weddings and whimsical landscapes among other things.
Eduardo

--------

Everything is already there: the lenses, the cameras, the routine. The waiting for is that full frame back to make it all work the way it did with film.
-Epd
Exactly. Its already out there (inside thousands of drawers). Now, we don’t need more megapixels, we need sensors that aproach or equal film sizes. I’m sure if Canon were a MF maker too, they already had a full-frame sensor. So, maybe it’s just a power war among sensor makers and mf gear makers.
Eduardo

You are right Nik, it’s been a long boring week.
Best regards all
Eduardo

Dear EPd,

See foto-z's comment: so it is, and I was thinking that I had been clear and precise in my explanation. What puzzles me, is that you seem so well aware about all and can't figure out the risks involved.

And I don't want to speculate simply because I prefer speaking about things I know (or think to know). This means that I simply don't know (as already said) and it has nothing to do with future plans of my employer or harming sales. You are suggesting that it is a simple matter, that one has just to drop the new sensor in its casing and then the new digital back is ready to hit the market. My experience of the past shows me that one has to be careful with such statements.

I cannot speak for other companies, but wish to clear your statement about "stretching the wallet of customers or potential customers", concerning Sinar: no, this is definitively not our philosophy, and our nearly 50 years of existence in this professional market is a little prove of it.

And as for myself and my suggestions of investing to a photographer ready or willing to invest (being it into a view camera or into a digital back), or even willing to upgrade his current system, my own philosophy has always been to know first his REAL needs and possibilities: there are many photographers out who know me and who would tell you that although ready to buy the top of the line of our analog cameras or digital backs, I have suggested them to start slowly by investing in either a f1 or f2 camera, or then to make their first steps into digital with a lower resolution instead of buying the top of the line and most expensive, when I was feeling they didn't need more or couldn't afford more, or that the technical possibilities of what I was suggesting was enough for them . My guess is, that there are others out, either colleagues or from other companies, sharing the same philosophy and acting the same way.

So yes, we are in the same boat and we strongly feel so, and that is what has always driven my acts.

Best regards,
Thierry

Understood, Nik!

As for the beer at PK '08: I already owe one to "eronald", that would definitively be a "though" PK for me!

Thierry

Dear Thierry,

Like you I prefer not to speculate whenever this is not needed. I could think of many both business risks as well as development risks for the making of DB's with larger sensors, but why dreaming it up? So I was asking you since I would expect you to be an expert in this specific field. If one has access to an expert, why not ask? Generally this is how I become so well informed.


I'm not suggesting anything but I am raising obvious questions. I was actually asking about specific experiences that could tell us why it is not so easy to develop a new version of a DB, when a lot of basic techniques seem to be there already. Your experience with past matters would be very interesting to hear about in detail. You seem to suggest that it is not just business risks involved here, rather technical risks as well. I would love to learn about those. Perhaps you will tell things that I already knew, but more likely you will have to add useful information to my knowledge base. And that of the others reading this as well.


To me it seems that this 50 years of existence is the very proof of Sinar's ability to stretch the customer's wallet. I have been watching Sinar's offerings for a loooong time, but I have never found anything that suggested Sinar tried to break down high prices of professional photo gear. On the contrary, they have always been in the forefront of highest priced companies. Now, before we go into the semantics of pricing: I do not want to suggest that anything would be wrong with that per se, or that Sinar's products would not be of high quality. Most of them are. But they come at a very high price, even for photographers who can afford them.


Now you are speculating, Thierry. I personally have very different experiences with representatives from both Sinar and other companies. No grudge here, but they surely aren't always angels.


Thanks, I'll remember that.

Best regards,
EPd

Well, I am may be an expert in my field, though never having pretended so, but even experts don't know it all (see my introduction post, when I joined LLF).




See above.

There no specific technical risks, rather uncertainties about timing and readiness for the market. I think everybody has experienced this, when companies are announcing a new product (being it a DB maker or other companies in other fields of high-tech) and how is the reality as opposed to the announcement. I am not an expert in this field, but I have experienced delays, due to issues which did not turn out the way they should have or have been thought they would. All I am saying is that it is dangerous to make any timing statement for a new product.




That is your opinion, which many do not share. You have the right to have this opinion and I respect it.




You probably know the costs of living in Switzerland, and what that means for a Swiss company, in the photographic field or else. You probably also know that a company's budget is calculated with a certain gross margin which ensures that the company is viable. And this gross margin is adapted to the market price realities.



It is simply too easy to "accuse" somebody of speculations by being anonymous: I have said it in an earlier post. I would be pleased to put a name on you, other than EPd, the same way you are putting my name Thierry on my posts. It is also too easy to badmouth about representatives (being it from Sinar or other companies). And you should read what I have written: "there are others out ... sharing the same philosophy...". I am well aware that not all are "angels".




Best regards,
Thierry