Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: Medium format technology question  (Read 3232 times)
larkis
Full Member
***
Offline Offline

Posts: 196


WWW
« on: November 05, 2012, 10:28:10 AM »
ReplyReply

I don't want to start a debate about the D800e vs medium format as this has been beat to death on this forum and others. My question is about the differences of format size in the digital age and about the companies that make sensors.

In the film days when a new emulsion came out and provided less grain, more resolution or dynamic range, bigger cameras that used bigger areas of that stock got more benefit. A fine grain film at the same print size from a bigger camera simply looked better (under normal circumstances).

What seems to be happening in the digital age is that the sensor technologies don't really propagate up and are kept within one format. When sony or whoever else comes up with a aps-c sensor that crams 24mp into a small area, could they not scale that same surface area to a 645 size sensor ? Sure it would cost more because of reduced yields and lenses would need to be very sharp but that's another matter. It almost seems that the medium format players are at the mercy of their sensor suppliers and the ones who sell to 3rd parties seem to be lagging behind.

If medium format is supposed to represent the highest quality image capture for specialized applications, what are the medium format players doing to give themselves the edge increased surface area of film used to provide ? Can you imagine the same sensor that is in the DP2/DP1 Merrill in a low ISO back at 645 size ?
Logged

My photography blog
http://blog.dominik.ca/
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7405


WWW
« Reply #1 on: November 05, 2012, 11:21:46 AM »
ReplyReply

Hi,

My guess is that all this depends very much on market volume. The firms who make MF sensors are opting for a small high margin market.

CMOS vendors work with much larger markets but probably operate with smaller margins.

There is a difference in APS-C and larger sensors as the production equipment used to expose the mask used in chip production can expose an APS-C frame and not a full frame. So "full frames" used to be stitched. Canon is said to have a 0.5 micron stepper that can expose a full frame sensor in a single exposure.

I would think that Sony or any other vendor could make larger sensors, but the market is probably not big enough.

Best regards
Erik


I don't want to start a debate about the D800e vs medium format as this has been beat to death on this forum and others. My question is about the differences of format size in the digital age and about the companies that make sensors.

In the film days when a new emulsion came out and provided less grain, more resolution or dynamic range, bigger cameras that used bigger areas of that stock got more benefit. A fine grain film at the same print size from a bigger camera simply looked better (under normal circumstances).

What seems to be happening in the digital age is that the sensor technologies don't really propagate up and are kept within one format. When sony or whoever else comes up with a aps-c sensor that crams 24mp into a small area, could they not scale that same surface area to a 645 size sensor ? Sure it would cost more because of reduced yields and lenses would need to be very sharp but that's another matter. It almost seems that the medium format players are at the mercy of their sensor suppliers and the ones who sell to 3rd parties seem to be lagging behind.

If medium format is supposed to represent the highest quality image capture for specialized applications, what are the medium format players doing to give themselves the edge increased surface area of film used to provide ? Can you imagine the same sensor that is in the DP2/DP1 Merrill in a low ISO back at 645 size ?
« Last Edit: November 12, 2012, 11:10:06 PM by ErikKaffehr » Logged

Don Libby
Sr. Member
****
Online Online

Posts: 726


Iron Creek Photography


WWW
« Reply #2 on: November 05, 2012, 11:41:38 AM »
ReplyReply

 I'm lucky in that I have 2-seperate systems to shoot and compare from.

First take in account we are landscape and nature photographers and our "studio" can often be found hanging near the side of a cliff or onto sand dune. While I shoot primarily with medium format using either a Cambo WRS or Phase DF with an IQ160 my wife uses a Canon 1DsIII.  I'm the guy that processes all our images. I know we're speaking a difference in megapixels between the 160 and 1DsIII however that difference hasn't always bee there as I also uses a P45+ and P30.

I'm attempting to lay some background/groundwork to my reply so bear with me.  If I don't do this I'm sure at least one-person will chime in on how great the D800 is over anything dealing with digital medium format and only film is good, etc etc etc.

Disregarding the difference in megapixels for a moment.  In 8 out of 10 images I process between MF and the 1DsIII I can see a clear difference in image quality.  Again these are our files.  Color is often times better; there's a richer "feel" to the image  Talking about sensor size I see a much broader area of coverage when using a similar focal length which helps in print size.  Then there's a sharpness that could be attributed to pixels however I feel it's also the sensor size as well.

Image quality is also effected with the ISO used and in some cases the 1DsIII wins out however we always attempt to use the lowest possible range that we can get away with.

My feelings are that no matter the resolution (megapixels) Canon or Nikon squeezes onto their sensors they will remain exactly what they are -  35mm.  They might be excellent 35mm but 35mm nevertheless.  For me - as a landscape/nature photographer given the choice of a 35mm and medium format with the same megapixels I'll always go for the medium format.

As far as film goes?  Unless you have your own wet darkroom no thanks.  Shooting film, having a second party develop it then scanning into a digital format defines why I gave up film years ago.  Shooting digital I have 100% control over the image from the moment I capture it through the printing process.

No certain if this answers your question but it was fun writing it.....

Don

Edited:  Also what Erick said. (It would be very helpful to have "Like" and "BS" buttons
« Last Edit: November 05, 2012, 11:43:31 AM by Don Libby » Logged

larkis
Full Member
***
Offline Offline

Posts: 196


WWW
« Reply #3 on: November 05, 2012, 01:59:43 PM »
ReplyReply

I do agree with medium format winning for me when I shoot landscapes over 35mm based solutions (heck, I even hate the 3:2 aspect ratio), but shooting with the foveon sensor on the Merrill has given me a bit of a pause when it comes to per pixel quality and especially color details that normally get killed unless they occupy a bigger chunk of the sensor. I guess the medium format people will need to do something to widen the gap even more.
Logged

My photography blog
http://blog.dominik.ca/
torger
Sr. Member
****
Offline Offline

Posts: 1474


« Reply #4 on: November 05, 2012, 03:56:53 PM »
ReplyReply

I'm rather format agnostic. As a landscape photographer with deep DoF I'm not interested in a "look", I just want neutrally rendered sharp images with reasonably accurate color. And I want movements.

Theoretically the 35mm format can do that for me, but I think we'll see problems over ~30 megapixel resolution concerning lens resolving power especially on the wides, and we'll not see many tilt-shift products, haven't seen any high resolution 35mm focal length yet for example. However, concerning resolving power I'm less sure today than I were say 6 months ago, I've now realised that MF lenses are not as good as I first believed, and sharper 35mm "high res" lenses are seen on the horizon, so maybe 35mm can become really high res too.

(Sony Exmor is way ahead of Canon technology in terms of base ISO, so I don't think the Canon comparison is so good these days if you want to see what the best 35mm can do.)

I also think 4:3 aspect ratio suits my shooting style much better than 3:2, but I don't particularly like to pay 5x the price for it :-).

I find it charming to work with my all-mechanical Linhof Techno, I like the workflow. It's primitive in a nice cosy way. What I want to happen is that MF continues to stay competitive, and 35mm don't get too good in the tech camera space. However, the day I can get say ~40 D800-like megapixels with 24, 35, 50, 90mm high resolution tilt-shift focal lenghs and have a Canon-like live view, I will find it very hard to motivate staying with the MF system. Except for that I probably won't find any buyer for it when that happens Wink.

For MF to survive in the long-term in traditional photo applications I think we need to see a change, lower prices and more flexible tech (higher ISO, efficient live view). The problem is that 35mm is getting too good so it becomes increasingly harder to motivate the 5x - 10x spending for MF. The are advantages, but the problem is the inbalance between subtle advantages (or just the joy of being a bit different) and the large difference in price.
« Last Edit: November 05, 2012, 04:03:27 PM by torger » Logged
MrSmith
Sr. Member
****
Online Online

Posts: 889



WWW
« Reply #5 on: November 05, 2012, 05:10:37 PM »
ReplyReply

MF has been short changed in the digital age, film emulsions were available in 35mm and 120 so 400asa was the same but you had a bigger area to play with at 6x6/6x7 now sensors that don't cover 6x45 are called MF and 400asa is barely usable.
It's a shame they are unlikely to reach parity again soon. :-(
Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7405


WWW
« Reply #6 on: November 05, 2012, 10:07:18 PM »
ReplyReply

Hi,

There is little doubt that there are advantages to size. The problem is really that most development is going on small sensors. If you take Sony, who seems to be the dominant vendor of CMOS sensors, the DSLR market (full frame + APS-C) market is very small in units but makes a very large part of the income. Sony migrates development from small sensors into DSLR size sensors.

I guess that the MFDB market is to small for a vendor like Sony.

Best regards
Erik


MF has been short changed in the digital age, film emulsions were available in 35mm and 120 so 400asa was the same but you had a bigger area to play with at 6x6/6x7 now sensors that don't cover 6x45 are called MF and 400asa is barely usable.
It's a shame they are unlikely to reach parity again soon. :-(
Logged

BJL
Sr. Member
****
Offline Offline

Posts: 5129


« Reply #7 on: November 06, 2012, 07:27:24 AM »
ReplyReply

There is a difference in APS-C and larger sensors as the production equipment used to expose the mask used in chip production can expose an APS-C frame and not a full frame. So "full frames" used to be stitched. Canon is said to have a 0.5 micron stepper that can expose a full frame sensor in a single exposure.
This ancient myth needs to be refuted yet again.
- Yes, Canon has been making for almost eleven years the FPA-5500ix, a stepper with field size of 50x50mm and 0.5micron minimum feature size, the only one still in production that could make a 36x24mm sensor without stitching (Nikon used to make one too, but discontinued it long ago.)
EDIT: Canon seems to have discontinued the FPA-5500ix; it is no longer on the stepper product page at http://www.usa.canon.com/cusa/semiconductor/products/semiconductor_equipment/steppers

- No, this stepper is useless for making a viable 36x24mm sensor, because its minimum feature size of 500nm is too large. Canon itself has indicated this in repeated explanations of the need for stitching in making its 36x24mm sensors, and these explanations came at a time when Canon was already making that "jumbo stepper".
As a guideline, good CMOS pixels seem to need to be fabricated with a minimum feature size about twenty or thirty times smaller than the pixel pitch, so that stepper would only be good for pixel pitch about 10 to 15mm or greater, limiting 36x24mm format to at best about 3600x2400 and more likely only 2400x1600. Though I expect a few posters might declare their desire for a sensor of such low resolution for the sake of huge per pixel DR, the DSLR marketplace clearly disagrees.

There seem to be just a couple of sensors made using that stepper: the Kodak/TrueSense KAF-4320 and KAF-4301 CCDs for X-rays and such that are about 50x50mm, 2084x2084 pixels, 24 micron pixel pitch.
« Last Edit: November 06, 2012, 08:00:27 AM by BJL » Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7405


WWW
« Reply #8 on: November 06, 2012, 08:11:14 AM »
ReplyReply

Hi,

According to chipworks all Canon APS-C or FF sensors they have analysed were made using 0.5 micron rules.
I don't know if that is relevant in the context.

Best regards Erik



This ancient myth needs to be refuted yet again.
- Yes, Canon has been making for almost eleven years the FPA-5500ix, a stepper with field size of 50x50mm and 0.5micron minimum feature size, the only one still in production that could make a 36x24mm sensor without stitching (Nikon used to make one too, but discontinued it long ago.)
EDIT: Canon seems to have discontinued the FPA-5500ix; it is no longer on the stepper product page at http://www.usa.canon.com/cusa/semiconductor/products/semiconductor_equipment/steppers

- No, this stepper is useless for making a viable 36x24mm sensor, because its minimum feature size of 500nm is too large. Canon itself has indicated this in repeated explanations of the need for stitching in making its 36x24mm sensors, and these explanations came at a time when Canon was already making that "jumbo stepper".
As a guideline, good CMOS pixels seem to need to be fabricated with a minimum feature size about twenty or thirty times smaller than the pixel pitch, so that stepper would only be good for pixel pitch about 10 to 15mm or greater, limiting 36x24mm format to at best about 3600x2400 and more likely only 2400x1600. Though I expect a few posters might declare their desire for a sensor of such low resolution for the sake of huge per pixel DR, the DSLR marketplace clearly disagrees.

There seem to be just a couple of sensors made using that stepper: the Kodak/TrueSense KAF-4320 and KAF-4301 CCDs for X-rays and such that are about 50x50mm, 2084x2084 pixels, 24 micron pixel pitch.
« Last Edit: November 06, 2012, 08:13:07 AM by ErikKaffehr » Logged

ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7405


WWW
« Reply #9 on: November 10, 2012, 12:29:42 AM »
ReplyReply

Hi,

A I wrote before I think the market is to small. Another problem may be there is no exact sensor size. Would the MF industry standardize around a single format, like full frame 645 the market may be more attractive. Now, it seems that MF manufacturers need different sensors they can sell at different prices. That would really force a CMOS sensor vendor to develop several MF-sensors for a very limited market.

I have also doubts that much of the existing MF equipment has the precision to take advantage of the resolution that the best sensors today can offer. Schneider and Rodenstock produce some impressive glass for MF digital and cameras like the Alpa probably have the mechanical precision.

Scaling up the Sony Alpha sensor to IQ180 size would yield 143 MPixels.

Leica has teamed up with an European vendor, CMOSIS, for the Leica M type 240. Here is some info from CMOSIS about that sensor: http://www.cmosis.com/news/press_releases/new_leica_m_uses_cmosis_24_mp_cmos_image_sensor

I would not rule out that CMOSIS could also develop a sensor for the Leica S.

The page sited above indicates that the sensor has similarities to modern CMOS technology:
"The new custom-designed sensor chip, counting 6,000 x 4,000 pixels on a 6 x 6 m grid across the active area of 36 x 24 mm, is made by STMicroelectronics (STM) in Grenoble, France, using 300mm wafers in their IMG175 CIS technology. STM's 110nm frontend and 90nm backend CIS technology with copper metallization was originally developed for CMOS image sensors with 1,75m pixels for mobile phones and other consumer applications. The large die size, larger than the reticle size, requires the use of one-dimensional stitching.
 The imager for the Leica M is based on a 6 x 6 m pixel size, yielding a linear full well capacity of ≥40,000 electrons and a linear dynamic range close to 76dB. Pixel data are digitized by patented low-power, high-speed 14-bit column AD converters. The sensor features an electronic rolling shutter with global reset and noise cancellation through both analog as well as digital correlated double sampling (CDS) resulting in low temporal and spatial noise and non-uniformities."

76 DB would mean a DR of 12.7 EV, a bit short of the DR measured for the Nikon D600 (13.4 EV) by DxO.

Best regards
Erik




I don't want to start a debate about the D800e vs medium format as this has been beat to death on this forum and others. My question is about the differences of format size in the digital age and about the companies that make sensors.

In the film days when a new emulsion came out and provided less grain, more resolution or dynamic range, bigger cameras that used bigger areas of that stock got more benefit. A fine grain film at the same print size from a bigger camera simply looked better (under normal circumstances).

What seems to be happening in the digital age is that the sensor technologies don't really propagate up and are kept within one format. When sony or whoever else comes up with a aps-c sensor that crams 24mp into a small area, could they not scale that same surface area to a 645 size sensor ? Sure it would cost more because of reduced yields and lenses would need to be very sharp but that's another matter. It almost seems that the medium format players are at the mercy of their sensor suppliers and the ones who sell to 3rd parties seem to be lagging behind.

If medium format is supposed to represent the highest quality image capture for specialized applications, what are the medium format players doing to give themselves the edge increased surface area of film used to provide ? Can you imagine the same sensor that is in the DP2/DP1 Merrill in a low ISO back at 645 size ?
« Last Edit: November 10, 2012, 01:27:52 AM by ErikKaffehr » Logged

BJL
Sr. Member
****
Offline Offline

Posts: 5129


« Reply #10 on: November 10, 2012, 09:01:17 AM »
ReplyReply

Hi,

According to chipworks all Canon APS-C or FF sensors they have analysed were made using 0.5 micron rules.
I would like a source for that, including clarifying if the 500nm is used exclusively or only for some early coarser stages: that big FPA-5500ix stepper is advertised in part for doing early rough stages in conjunction with other steppers. If true, it would seem that there is some other problem with using that big stepper to make sensors as opposed to purely digital chips, given that Canon has repeatedly said that it needs to use stitching even some years after introducing the FPA-5500ix. Then again, I do not know enough to say if the "design rule" dimension matches exactly with the "minimum feature size" spec. of the stepper used.

P. S. I found the source:
http://www.chipworks.com/blog/technologyblog/2012/10/24/full-frame-dslr-cameras-canon-stays-the-course/
Some interesting speculation there as to whether Canon is limited to 22MP and prevented from adopting column-parallel ADC by its continued use of 0.5 micron design rules.
« Last Edit: November 10, 2012, 09:42:01 AM by BJL » Logged
Justinr
Sr. Member
****
Offline Offline

Posts: 1009


WWW
« Reply #11 on: November 12, 2012, 08:17:02 AM »
ReplyReply


If medium format is supposed to represent the highest quality image capture for specialized applications, what are the medium format players doing to give themselves the edge increased surface area of film used to provide ? Can you imagine the same sensor that is in the DP2/DP1 Merrill in a low ISO back at 645 size ?

I distinctly remember BJP analysing Kodak's FF 14n and deciding that it was two smaller chips stuck together. It was an outstanding camera that was ahead of its time and I have wondered since if this approach has been utilised since by any mainstream manufacturers, if two can be assembled together then why not four? Frame rate might be compromised but I doubt that it would be a great drawback in MF.
Logged

ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7405


WWW
« Reply #12 on: November 12, 2012, 12:50:36 PM »
ReplyReply

Hi,

As far as I understand all sensors larger than APS-C are stitched. It would probably be possible to stitch larger sensors. The issue is probably more about profitability. Small series.

I'm not really sure it is about just sensors. You also need ASICs to process the data. Scales of economy are different if you produce hundreds of thousands or just hundreds.

Best regards
Erik

I distinctly remember BJP analysing Kodak's FF 14n and deciding that it was two smaller chips stuck together. It was an outstanding camera that was ahead of its time and I have wondered since if this approach has been utilised since by any mainstream manufacturers, if two can be assembled together then why not four? Frame rate might be compromised but I doubt that it would be a great drawback in MF.
Logged

Justinr
Sr. Member
****
Offline Offline

Posts: 1009


WWW
« Reply #13 on: November 12, 2012, 02:47:33 PM »
ReplyReply

Hi,

As far as I understand all sensors larger than APS-C are stitched. It would probably be possible to stitch larger sensors. The issue is probably more about profitability. Small series.

I'm not really sure it is about just sensors. You also need ASICs to process the data. Scales of economy are different if you produce hundreds of thousands or just hundreds.

Best regards
Erik



Excuse my innocence here Erik but by stitching do we mean the joining together of sensors to act as one sensor or the joining together of two images from two non connected sensors (mounted alongside each other) to give the appearance of one picture?

Cheers

Justin.
Logged

ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7405


WWW
« Reply #14 on: November 12, 2012, 04:11:36 PM »
ReplyReply

Hi Justin,

When the sensors are produced a set of masks are exposed on the silicon substrate, these masks and the chemical processes involved build the sensor, layer by layer. The equipment making these exposures is called a stepper. Steppers normally have an aperture slightly larger than APS-C. So any sensor that is larger than APS-C is essentially made making several exposures for generating a single chip. That is sometimes called stitching.

So it is not a question of two devices but a single device made with multiple exposures.

Best regards
Erik





Excuse my innocence here Erik but by stitching do we mean the joining together of sensors to act as one sensor or the joining together of two images from two non connected sensors (mounted alongside each other) to give the appearance of one picture?

Cheers

Justin.
Logged

FredBGG
Sr. Member
****
Offline Offline

Posts: 1651


« Reply #15 on: November 12, 2012, 08:20:22 PM »
ReplyReply

When sony or whoever else comes up with a aps-c sensor that crams 24mp into a small area, could they not scale that same surface area to a 645 size sensor ?

They probably could, but most likely are not interested in the dwindling market for various reasons.

The main thing is that the principal CMOS sensor makers make cameras and know what can be achieved with the formats they are already working with.
The very top end of CMOS cameras is so close today to what MF can obtain that they see no growth or stable market.
I'm sure that Sony is looking closely at what Zeiss is doing with the new line of DSLR lenses. Sony also has a good idea of what is going on in the MF world.
Just look at Hasselblad's desperate move to make money with the Lunar.
Sony would only make the required investments if they could get a guaranteed order quantity and if the company ordering were solid enough for the business
to not be too risky.


Regarding Sony, Canon or Nikon coming out with a medium format camera... that is unlikely.
One of the reaons is that the Spin type marketing required to justify MF to more than the users that really do need it is not in their style.

I think that if a new entry were to be made it would have to be a larger than 645 format to really be able to differentiate itself from 35mm DSLR.

While MF manufacturers like to talk about a sensor twice the size it actually is a sensor that is twice the area, however resolution and most visual characteristics are more of a linear function
so the size difference really is not that significant.

Sensor size difference between a 35mm ff DSLR and the biggest MF sensor is not that big if you look at it from a linear point of view that is more relevant.

36 vs 53.7 is a 49.167 % increase.

To summerize Sony, Canon, Nikon etc would simply look at these two images and then tell themselves why bother and why take the risk.






both are crops from this framing:



One is 35mm DSLR and one is MF digital.

(credit) http://www.photigy.com/nikon-d800e-test-review-vs-hasselblad-h4d40-35mm-against-medium-format/

The CMOS manufacturers are way more likely to focus on improving their camreas and sensor to compete against their more relevant
competition.

MF digital is really a sector for those companies that are stuck in that format and their per item margins have to grow in order to survive in a dwindling market.

Five or six years ago high end fashion photographers would be using MF nearly all the time, now most are using 35mm DSLRs



« Last Edit: November 12, 2012, 10:04:37 PM by FredBGG » Logged
jsiva
Jr. Member
**
Offline Offline

Posts: 99


« Reply #16 on: November 12, 2012, 09:45:13 PM »
ReplyReply

They probably could, but most likely are not interested in the dwindling market for various reasons.

The main thing is that the principal CMOS sensor makers make cameras and know what can be achieved with the formats they are already working with.
The very top end of CMOS cameras is so close today to what MF can obtain that they see no growth or stable market.
I'm sure that Sony is looking closely at what Zeiss is doing with the new line of DSLR lenses. Sony also has a good idea of what is going on in the MF world.
Just look at Hasselblad's desperate move to make money with the Lunar.
Sony would only make the required investments if they could get a guaranteed order quantity and if the company ordering were solid enough for the business
to not be too risky.


Regarding Sony, Canon or Nikon coming out with a medium format camera... that is unlikely.
One of the reaons is that the Spin type marketing required to justify MF to more than the users that really do need it is not in their style.


I think that if a new entry were to be made it would have to be a larger than 645 format to really be able to differentiate itself from 35mm DSLR.

While MF manufacturers like to talk about a sensor twice the size it actually is a sensor that is twice the area, however resolution and most visual characteristics are more of a linear function
so the size difference really is not that significant.

Sensor size difference between a 35mm ff DSLR and the biggest MF sensor is not that big if you look at it from a linear point of view that is more relevant.

36 vs 53.7 is a 49.167 % increase.

To summerize Sony, Canon, Nikon etc would simply look at these two images and then tell themselves why bother and why take the risk.

The CMOS manufacturers are way more likely to focus on improving their camreas and sensor to compete against their more relevant
competition.

MF digital is really a sector for those companies that are stuck in that format and their per item margins have to grow in order to survive in a dwindling market.

Five or six years ago high end fashion photographers would be using MF nearly all the time, now most are using 35mm DSLRs


Really?  Sony is a marketing machine and if they saw an opportunity to exploit a market, they would.  This was their main premise for entering the DSLR market in the first place.  Given the trouble they have had of late, do you really think they would leave out a viable market because of an unsubstantiated claim you make on their behalf about "Spin type marketing"? Your definition, not mine.

BTW, the second photo is the one with MF.

If you think Sony does not need to change fundamentally take a look at their charts over the past 5 years.  While you are at, it, please do the same with Fuji.

I am only responding here as I really enjoy reading this forum, and of late, I see this kind of egregious commentary just going on endlessly on multiple threads.

P.S.  The spell checker is free, please use it.
« Last Edit: November 12, 2012, 09:48:25 PM by jsiva » Logged
FredBGG
Sr. Member
****
Offline Offline

Posts: 1651


« Reply #17 on: November 12, 2012, 10:10:29 PM »
ReplyReply



If you think Sony does not need to change fundamentally take a look at their charts over the past 5 years.  While you are at, it, please do the same with Fuji.

.....

P.S.  The spell checker is free, please use it.

Did I say I think Sony needs to change or not..... No

Do you really think that making MF sensors in the small quantities MF makers sell will make any difference for a giant corporation like Sony?

I think that after the Hasselblad Lunar debacle I doubt Sony wants to go near an MF company for some time.

Spell checker..... really. You sound like my 5th grade English teacher. Please excuse the odd spelling mistake... it comes with speaking many languages Wink
« Last Edit: November 13, 2012, 10:34:29 PM by FredBGG » Logged
Justinr
Sr. Member
****
Offline Offline

Posts: 1009


WWW
« Reply #18 on: November 13, 2012, 05:48:12 AM »
ReplyReply

Hi Justin,

When the sensors are produced a set of masks are exposed on the silicon substrate, these masks and the chemical processes involved build the sensor, layer by layer. The equipment making these exposures is called a stepper. Steppers normally have an aperture slightly larger than APS-C. So any sensor that is larger than APS-C is essentially made making several exposures for generating a single chip. That is sometimes called stitching.

So it is not a question of two devices but a single device made with multiple exposures.

Best regards
Erik





So if I understand this right, half the chip is exposed  and then it is moved so the other half is exposed and we are talking about the accuracy of locating the chip so that no join lines appear on the image?

It's mind boggling to think that it can be done to 5 microns (if I've got that right).


Cheers

Justin.
Logged

ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7405


WWW
« Reply #19 on: November 13, 2012, 07:30:14 AM »
ReplyReply

Hi,

Yes, indeed. The only thing is the precision needs to be much higher. Sony uses 0.18 micron rules. Also keep in mind that they do this over several layers. So they mount, expose, unmount, go into chemical treatment, coat with new resist remount, reexpose and so on.

Best regards
Erik
Logged

Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad