Poll on sensors for Canon DSLRs. (II)

Started Jun 29, 2011 | Discussions
Great Bustard
Forum ProPosts: 21,954
Like?
Poll on sensors for Canon DSLRs. (II)
Jun 29, 2011

At the end of the last thread, the following comment was made:

The only thing that's a bit conforting is that all this is only usual Internet forum and doesn't really accomplish anything.

Right you are, sir!

So, for those that have not "switched to more useful things", I present the continuation of the Poll on Sensors for Canon DSLRs!

CameraCarl
Veteran MemberPosts: 3,176Gear list
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to Great Bustard, Jun 29, 2011

Have you tabulated the results yet?

 CameraCarl's gear list:CameraCarl's gear list
Sony Alpha NEX-6 Canon EF 70-300mm f/4-5.6L IS USM Canon EF 24-105mm f/4L IS USM Sony E 16-50mm F3.5-5.6 PZ OSS Sony Vario-Tessar T* E 16-70mm F4 ZA OSS +1 more
Reply   Reply with quote   Complain
Great Bustard
Forum ProPosts: 21,954
Like?
Threatening the complacency
In reply to Great Bustard, Jun 29, 2011

Don't really have anything to add -- well said!

zerozeronine wrote:

In your reasoning, you assumed "since Canon's management is not stupid," but I worry if there's some face-saving or inability to be nimble because they've already invested in building fabrication plants. There's also the egos/politics of the engineers who have been working on their sensors thus far.

I'm also hoping that Canon feels threatened, and will do their best to make changes for the better. But they may not be seeing things this way. Do we really know that they think their sensor inferiority is really a problem, from a business standpoint? It seems like something that mostly photo geeks or fine-art pros would care about. The general attitude of the masses with their gear is like the emperor without clothes. They wouldn't notice stuff even if you pointed it out, and they simply assume whatever brand they chose is the best.

So, the only way I see Canon feeling threatened is if many pros who use their 1 series cameras defect to Nikon because of the shadow noises. If that's not the case, and we still see rows of white lenses at sports events, Canon probably won't be worried. Only when pros defect and make the masses wonder about getting a Nikon instead will any of this matter.

Does anyone have a pulse on what the pros are saying and whether they're migrating to Nikon?

My worry is that this shadow-noise issue is rather still in the realm of aficionados, and even pros aren't raising a fuss over it, in which case Canon might be too complacent with their current sales to care.

Reply   Reply with quote   Complain
Great Bustard
Forum ProPosts: 21,954
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to CameraCarl, Jun 29, 2011

CameraCarl wrote:

Have you tabulated the results yet?

No, it wasn't a thread where people played by the rules by putting their answer in the subject line, as requested.

However, there seems to be a polarization with many feeling that Canon will provide, and others feeling that Canon simply doesn't have the capacity to keep up the juggernaught.

Reply   Reply with quote   Complain
zerozeronine
Contributing MemberPosts: 595
Like?
Can Canon do the right thing? (reply to bobn2's posting)
In reply to Great Bustard, Jun 29, 2011

In your reasoning, you assumed "since Canon's management is not stupid," but I worry if there's some face-saving or inability to be nimble because they've already invested in building fabrication plants. There's also the egos/politics of the engineers who have been working on their sensors thus far.

I'm also hoping that Canon feels threatened, and will do their best to make changes for the better. But they may not be seeing things this way. Do we really know that they think their sensor inferiority is really a problem, from a business standpoint? It seems like something that mostly photo geeks or fine-art people would care about. The general attitude of the masses with their gear is like the emperor without clothes. They wouldn't notice stuff even if you pointed it out, and they simply assume whatever brand they chose is the best.

So, the only way I see Canon feeling threatened is if many pros who use their 1 series cameras defect to Nikon because of the shadow noises. If that's not the case, and we still see rows of white lenses at sports events, Canon probably won't be worried. Only when pros defect and make the masses wonder about getting a Nikon instead will any of this matter.

Does anyone have a pulse on what the pros are saying and whether they're migrating to Nikon?

My worry is that this shadow-noise issue is rather still a concern of aficionados, and even pros aren't raising a fuss over it, in which case Canon might be too complacent with their current sales to care.

Also, I know your guess is that Canon probably won't be able to overcome Sony's ADC column advantage soon, but you don't think they haven't been working on some radical upgrades while they've been releasing their consumer models with easy evolutionary upgrades (I don't know...but am hoping). They've had about 2 years since the 7D and 3 since the 5D II. I can't imagine they've just been idle, while others have been making advances. There's no fundamental reason why Canon can't find their own solution (the way Sony did), unless you know Canon's engineers are lazy.

Kaz

Reply   Reply with quote   Complain
bobn2
Forum ProPosts: 27,829
Like?
Re: Can Canon do the right thing? (reply to bobn2's posting)
In reply to zerozeronine, Jun 29, 2011

zerozeronine wrote:

In your reasoning, you assumed "since Canon's management is not stupid," but I worry if there's some face-saving or inability to be nimble because they've already invested in building fabrication plants. There's also the egos/politics of the engineers who have been working on their sensors thus far.

All those things come into play in real companies run by real people. But competent management realise when the writings on the wall, when a previous decision was a bad one and when to stop pouring good money after bad. Even Olympus worked out when to call time on Four Thirds. The investment in fab plant is not necessarily a white elephant. One thing they could do, if the decided to Use Sony sensors, is to licence the designs and fabricate them themselves. Or, if they went into partnership with someone like Panasonic or Samsung (not impossible, Sony has a partnership with Samsung for making LCD's, in fact they make the LCD's you find on the back of canon DSLR's) the fab lines would still be in use. At the last resort, they could sell it on.

I'm also hoping that Canon feels threatened, and will do their best to make changes for the better. But they may not be seeing things this way. Do we really know that they think their sensor inferiority is really a problem, from a business standpoint? It seems like something that mostly photo geeks or fine-art people would care about. The general attitude of the masses with their gear is like the emperor without clothes. They wouldn't notice stuff even if you pointed it out, and they simply assume whatever brand they chose is the best.

I think the masses pick up an idea that some brand is 'best', it's not always clear what that is based on, but canon's success in DSLR's has at least been based on the perception that their sensors are 'better', I think. As they lose that, their sales will slump, and the popular photography press is full of how good the new Sony sensors are.

So, the only way I see Canon feeling threatened is if many pros who use their 1 series cameras defect to Nikon because of the shadow noises. If that's not the case, and we still see rows of white lenses at sports events, Canon probably won't be worried.

We don't still see rows of white lenses, we see a speckle of white and black lenses. What must be worrying Canon is that the 1DIV didn't make a dent in Nikon sales. Nikon more than countered it by upgrading the D3 with a new sensor far better in QE than anything Canon can do. It seems even Nikon is more agile at sensor development than Canon, possibly because they concentrate on a few, rather than attempting the range that Canon does.

Only when pros defect and make the masses wonder about getting a Nikon instead will any of this matter.

Does anyone have a pulse on what the pros are saying and whether they're migrating to Nikon?

Don't know. All the pros I know migrated to Nikon with the D3 and haven't been moved by the MkIV, but I see a fairly restricted set of pros.

My worry is that this shadow-noise issue is rather still a concern of aficionados, and even pros aren't raising a fuss over it, in which case Canon might be too complacent with their current sales to care.

I shouldn't think so. After the hit with the MkIII, Canon will be looking closely at what their customers are saying, the problem is, can they easily do something about it. They managed to get the best video in the business quickly, with some ingenious but simple sensor mods, which shows when they know how to do something they can do it quickly. I just think that catching up Sony is not easy for them, which is why they don't just do it.

Also, I know your guess is that Canon probably won't be able to overcome Sony's ADC column advantage soon, but you don't think they haven't been working on some radical upgrades while they've been releasing their consumer models with easy evolutionary upgrades (I don't know...but am hoping).

I think that they are working on something. I think that's why the 1DsIV is now nearly a year overdue. The D3X showed that high DR was what that section of the market wants, and Canon has to achieve it somehow. A I said somewhere, on chip column ADC's like Sony uses are not the only way of doing it, but other solutions might be more expensive and need debugging. It might well be that Canon is spending the effort on on-chip digitisation, which is why its taking so long, but its taking so long doesn't bode well for future ability to keep pace.

They've had about 2 years since the 7D and 3 since the 5D II. I can't imagine they've just been idle, while others have been making advances. There's no fundamental reason why Canon can't find their own solution (the way Sony did), unless you know Canon's engineers are lazy.

I don't think they are lazy. I just think that there are fewer of them, with a narrower skill set (particularly not in the digital side) and that their efforts have been wasted elaborating similar sensors (two different 18MP APS-C sensors, 2 different 12MP APS-C sensors, 2 different 21MP FF sensors and a 16MP APS-C sensor). I think the reason canon has had to do that is the economic necessity to balance production on two different fab lines. Nikon's sensor team is likely smaller than Canon's but in the time that they've produced three sensor designs, Canon's have had to produce seven.
--
Bob

Reply   Reply with quote   Complain
David Hull
Veteran MemberPosts: 5,330Gear list
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to Great Bustard, Jun 30, 2011

Poll on sensors for Canon DSLRs. (II)

I generally hate it when people do this but in this case, I was rather hoping you would -- it has been quite enlightening, thanks.

Dave
--
Flickr: http://www.flickr.com/photos/26158506@N07/

 David Hull's gear list:David Hull's gear list
Canon EOS 50D Canon EOS 5D Mark III
Reply   Reply with quote   Complain
Great Bustard
Forum ProPosts: 21,954
Like?
Glad to hear it!
In reply to David Hull, Jun 30, 2011

David Hull wrote:

Poll on sensors for Canon DSLRs. (II)

I generally hate it when people do this but in this case, I was rather hoping you would -- it has been quite enlightening, thanks.

Bob's post just above:

http://forums.dpreview.com/forums/read.asp?forum=1032&message=38783423

was really, really good.

Reply   Reply with quote   Complain
KLO82
Contributing MemberPosts: 663
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to Great Bustard, Jun 30, 2011

I wish I cud know:

(i) What is the percentage of sales/ market share Canon has lost due to less DR/ more read noise at lower ISOs?

(ii) What percentage of additional sales/ market share Nikon/ Sony has gained due to more DR/ less read noise at lower ISOs?

(iii) What percentage of users of Nikon/ Sony think that with high DR/ low read noise sensors of Sony, they r creating images which they never thought they wud b able to? (In other words, how many of the users r taking advantage of the high DR/ low read noise at base ISOs?)

(iv) What percentage of users bought cameras with latest Sony sensor to take advantage of it's "ISOlessness"? What percentage of users r actually taking this advantage?

(v) What percentage of the users of this latest sensors of Sony know that they r using a sensor which has more DR/ less read noise than competitors? What percentage of the users know that they r using a sensor which is "ISOless"?

(vi) What percentage of Canon users who r using sensors that came after 5DII feel that they r not being able to make images they wud like due to low DR?

Reply   Reply with quote   Complain
buellom
Regular MemberPosts: 457Gear list
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to Great Bustard, Jun 30, 2011

We don't know what Canon will deliver in short term, probably this year.

Sony, Panasonic and now Olympus came up with new sensors and Canon probably will respond this year. At least, this is to be expected. If then the new canon sensor are not state of the or best in class, the question gets relevant.

A few notes:

1. Canon is big enough to produce own sensor and to developp own sensor. BMW and Mercedes are smaller (revenue) than Volkswagen or Toyota - does that mean they won't be able to produce competetive cars in the future?

2. The differences between sensors does not seem that big in my eyes. There is a difference, sure. But to me it's more important to get the right quality lens infornt of the sensor. And this is an economic question too for the customer. In short: I'd chose a system rather based on available and payable lenses than sensors.

3. From a business point of view: I think the real thread for Canon is coming from mirrorless cameras. They eat sales at the low end and this is where Canon makes revenue (and profit). And cell phone cameras eating sales from compacts.

4. The market for cameras is pretty saturated (as far as I see). So you need really inovations to make people upgrade. A new e.g. 5DIII with just less shadow noise - would it make me investing 2000+ Euros? Probably not. I know about the limitations of the sensor (or whatever the cause is of the shadow noise) and can work around it in most cases. I want a new and better 17-40 lens rather than a new camera. Meanwhile there are interesting non Canon options.

5. My fear is that Canon thinks they have to improve video in the first place to get more sales. Sure, this might end with an improvement of still capability too. But not necessarily.

-- hide signature --

********************
http://www.freude-am-licht.de
********************

Reply   Reply with quote   Complain
bobn2
Forum ProPosts: 27,829
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to buellom, Jun 30, 2011

buellom wrote:

1. Canon is big enough to produce own sensor and to developp own sensor. BMW and Mercedes are smaller (revenue) than Volkswagen or Toyota - does that mean they won't be able to produce competetive cars in the future?

It means that they can't sink as much development resource into them as VW and Toyota, and on the whole cannot be as competitive. But the motor industry is a good example. If you compare the sensor to the engine, car companies routinely share engine development you'll find the same engines in cars of competitive companies. The reason is simple, economies of scale add up to a great deal in the economics of engine production. It's quite possible to design and build a low volume engine, just like it is possible to design and build a low volume sensor, but amortising the R&D costs (which with modern emission requirements can be considerable) and production setup costs across a small number of engines adds considerably to the final cost of the product, which in the mass markets produces an immediate competitive disadvantage. In the end, it makes business sense to share the engine in order to increase the volume. The same is true in aircraft engines, where joint developments between competitors are routine.

-- hide signature --

Bob

Reply   Reply with quote   Complain
sparkling elk
Contributing MemberPosts: 800Gear list
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to Great Bustard, Jun 30, 2011

practical question:
is there someone who owns or could compare both the D7000 and 1D4 ?
both sensors have 16mp resolution.

of course the 1D4 pixels are larger, but if the Sony-D7000 sensor is superior, it should show that it comes close to or exceeds the performance of the Canon 1D4 sensor.
--
Eckhard Aland
http://www.flickr.com/photos/proud-elk/collections/

 sparkling elk's gear list:sparkling elk's gear list
Ricoh GR Canon EOS-1D Mark III Canon EOS-1D Mark IV Canon EF 24mm f/1.4L II USM Canon EF 50mm f/1.2L USM +14 more
Reply   Reply with quote   Complain
buellom
Regular MemberPosts: 457Gear list
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to bobn2, Jun 30, 2011

Well, yes and no.

Your argument (economics of scale) is valid of course. But when a company has a certain size the argument looses weight. (Whatever this certain size might be, Canon certainly is large enough.)

A large company can sometimes be too large and R&D gets lazy or inflexible or bureaucratic or driven by to many other interests withhin the company.

Regarding automobile industry: Yes, there are cooperations. But they are not going as far as you suggested.

Another thought that might be relevant for future R&D budgets:

Financially Nikon is doing ok at the moment, but not really fine. They are operating at (relativly) high risk: How long will it take that they are in deep trouble if a new release doesn't sell as expected? Not very long.

and Sony: They are runing two systems at the moment (nex + SLT). What if they do not make the profit they want to? How long do you think will they keep the camera business? There were already rumors, that they might stop producing FF cameras. This seems wrong now, but as far as I know this was something that was discussed within the company. And will be dicussed again if returns are not as expected.

Sony made a relevant step ahead in sensors (and they probably invested a lot to do so), but this has to pay out now. And this is not only a question of iq the sensors can deliver. If the FF Sonys fail and Nikonn is the only customer for FF sensors, Sony might stop investing in this area. Thus, Nikon would have to set up it's own fabs or source sensors somewhere else. However, in this case it would be good for Canon to have its own sensors and not to rely on someone else. See the difficulties Olympus has with sensors.

-- hide signature --

********************
http://www.freude-am-licht.de
********************

Reply   Reply with quote   Complain
noirdesir
Forum ProPosts: 10,767
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to bobn2, Jun 30, 2011

bobn2 wrote:

buellom wrote:

1. Canon is big enough to produce own sensor and to developp own sensor. BMW and Mercedes are smaller (revenue) than Volkswagen or Toyota - does that mean they won't be able to produce competetive cars in the future?

It means that they can't sink as much development resource into them as VW and Toyota, and on the whole cannot be as competitive. But the motor industry is a good example. If you compare the sensor to the engine, car companies routinely share engine development you'll find the same engines in cars of competitive companies. The reason is simple, economies of scale add up to a great deal in the economics of engine production. It's quite possible to design and build a low volume engine, just like it is possible to design and build a low volume sensor, but amortising the R&D costs (which with modern emission requirements can be considerable) and production setup costs across a small number of engines adds considerably to the final cost of the product, which in the mass markets produces an immediate competitive disadvantage. In the end, it makes business sense to share the engine in order to increase the volume.

To give some examples:

  • BMW and Mercedes have been developing jointly the drivetrains for hybrid cars because they felt they were two small to warrant the investment

  • BMW uses Peugeot engines (though slightly modified) in its Mini cars

  • Porsche uses VW six cylinder engines in it Cayenne SUV and the whole Cayenne is very much based on the VW Touareg

  • VW and Ford for almost a decade sold only slightly different versions of the same car as their main minivan in Europe (essentially when minivans startet to take off in Europe, they weren't sure the volumes were high enough to warrant a car on their own, they now are high enough)

  • ditto for Fiat and Peugeot, same car as their minivan

Reply   Reply with quote   Complain
buellom
Regular MemberPosts: 457Gear list
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to noirdesir, Jun 30, 2011

But they all have the ability to design and to produce an engine by their own as one of their core competences ...

Canon uses (mainly?) Sony sensors in their Powershots. Guess these are the Minis of BMW. But Canon and BMW produce their own "engines" too and this is for both an area of core competence.

-- hide signature --

********************
http://www.freude-am-licht.de
********************

Reply   Reply with quote   Complain
xfz
xfz
Regular MemberPosts: 129
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to bobn2, Jun 30, 2011

bobn2 wrote:

buellom wrote:

1. Canon is big enough to produce own sensor and to developp own sensor. BMW and Mercedes are smaller (revenue) than Volkswagen or Toyota - does that mean they won't be able to produce competetive cars in the future?

It means that they can't sink as much development resource into them as VW and Toyota, and on the whole cannot be as competitive. But the motor industry is a good example. If you compare the sensor to the engine, car companies routinely share engine development you'll find the same engines in cars of competitive companies. The reason is simple, economies of scale add up to a great deal in the economics of engine production. It's quite possible to design and build a low volume engine, just like it is possible to design and build a low volume sensor, but amortising the R&D costs (which with modern emission requirements can be considerable) and production setup costs across a small number of engines adds considerably to the final cost of the product, which in the mass markets produces an immediate competitive disadvantage. In the end, it makes business sense to share the engine in order to increase the volume. The same is true in aircraft engines, where joint developments between competitors are routine.

Just checked, Sony's Revenue is 88B at market cap of 26B, 20B cash and 12B debt,
Canon's Revenue is 46B at market cap of 58B, 11B cash and 234M debt.

So Canon is a much better managed company, and has more net-cash (11B vs 8B) at disposal. And Sony is a beleaguered company trying to redefine itself. This is reflected in their market cap. Canon is the one with deeper pocket because it has more net-cash and more focused.

Also, how much resource do you think needed for R&D for a sensor? I guess it is not much for a company size of Canon. Once all the logistics are in place like Canon is, it at most at the level of 10M. The crucial part is to find the right people to do it right. I am surprised Canon has been using the 18MP sensor for their APS-C for such a long time. And from all the experiment chips Canon show now and then, Canon definitely can make many different sensors but somehow they think the 18mp is the best compromise for their mass production. But that doesn't mean they don't have anything in their arsenal, it is not uncommon they have 1-2 generation newer chips in the lab waiting.

In the end, the photography is more about system, not a single part. Canikon know it, so even in old bad days for Nikon, Nikon still survived fine. Because sensor is such an important piece for Canon's strategic market, I don't think Canon will abandon their sensor development, and the competition is good the consumers.

Bob

Reply   Reply with quote   Complain
bobn2
Forum ProPosts: 27,829
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to buellom, Jun 30, 2011

buellom wrote:

Well, yes and no.

Your argument (economics of scale) is valid of course. But when a company has a certain size the argument looses weight. (Whatever this certain size might be, Canon certainly is large enough.)

Not at all, however large the company, there is no point investing $1 billion to get $500 million worth of business. Canon's sensor business produces 3 million sensors a year, maybe with an average value of $30, that's $90 million turnover per year. Sony produces 100 million sensors per year, maybe with an average value of $10, that's $1 billion turnover per year. If both look at a 5 year return on investment, then $500 million is an infeasible investment for Canon, but thoroughly feasible for Sony, since the amortisation of that investment will double the cost of Canon sensors, while it will only add 10% to the cost of Sony's.

A large company can sometimes be too large and R&D gets lazy or inflexible or bureaucratic or driven by to many other interests withhin the company.

A large company will do detailed accounting on every part of their investment. Canon would be mad to spend $500M on sensor investment if that money would recoup more profit spent on, say, printer development. As I said above, it makes no sense to invest more than will be recouped in sales.

Regarding automobile industry: Yes, there are cooperations. But they are not going as far as you suggested.

You are wrong, they go further than I suggested. There are many joint developments and joint ventures all over the place.

Another thought that might be relevant for future R&D budgets:

Financially Nikon is doing ok at the moment, but not really fine. They are operating at (relativly) high risk: How long will it take that they are in deep trouble if a new release doesn't sell as expected? Not very long.

Irrelevant to Canon's business position.

and Sony: They are runing two systems at the moment (nex + SLT). What if they do not make the profit they want to? How long do you think will they keep the camera business? There were already rumors, that they might stop producing FF cameras. This seems wrong now, but as far as I know this was something that was discussed within the company. And will be dicussed again if returns are not as expected.

Again, how is this at all relevant to the discussion in hand. Sony runs two camera divisions, it might well merge them or close one (probably Minolta). I suspect they will not make any more FF cameras. I can't imagine they made a profit on the A900. They very possibly did on the sensor, since they sold a bunch to Nikon.

Sony made a relevant step ahead in sensors (and they probably invested a lot to do so), but this has to pay out now.

It is paying out, to the semiconductor business. They are selling sensors as fast as they can make them and taking on more and more fab facility.

And this is not only a question of iq the sensors can deliver. If the FF Sonys fail and Nikonn is the only customer for FF sensors, Sony might stop investing in this area.

Sony has invested in developing a technology that they can use to produce very many sensor variants very fast. Nikon will get their FF sensor, if they pay for it, whether or not Sony imaging has a FF camera. It will be a purely commercial transaction, just like the D2X sensor, or the several specialist sensors Sony Semiconductor makes for the film camera business.

Thus, Nikon would have to set up it's own fabs or source sensors somewhere else

Nikon already gets its own designed sensors at fabs other than Sony. There is a world oversupply of semiconductor fab at the moment, one reason that Sony is able to buy up plant at attractive prices. The market conditions favour fables companies, such as Nikon, at the moment. Interestingly, Toshiba's sensor business, the fourth largest in the world, is going fabless.

. However, in this case it would be good for Canon to have its own sensors

Its fab lines are dead capital, and look increasingly like a poor investment. I have no doubt Canon could get fab more cheaply on the open market. I've no doubt that it could get sensors more cheaply by buying Sony ones.

and not to rely on someone else.
See the difficulties Olympus has with sensors.

And see the difficulties Nikon doesn't. Olympus' difficulty is having chosen a non-standard size of sensor which limits their choice of suppliers, and also not having the resource to invest in their own sensors, even as much as Nikon.
--
Bob

Reply   Reply with quote   Complain
bobn2
Forum ProPosts: 27,829
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to xfz, Jun 30, 2011

xfz wrote:

bobn2 wrote:

buellom wrote:

1. Canon is big enough to produce own sensor and to developp own sensor. BMW and Mercedes are smaller (revenue) than Volkswagen or Toyota - does that mean they won't be able to produce competetive cars in the future?

It means that they can't sink as much development resource into them as VW and Toyota, and on the whole cannot be as competitive. But the motor industry is a good example. If you compare the sensor to the engine, car companies routinely share engine development you'll find the same engines in cars of competitive companies. The reason is simple, economies of scale add up to a great deal in the economics of engine production. It's quite possible to design and build a low volume engine, just like it is possible to design and build a low volume sensor, but amortising the R&D costs (which with modern emission requirements can be considerable) and production setup costs across a small number of engines adds considerably to the final cost of the product, which in the mass markets produces an immediate competitive disadvantage. In the end, it makes business sense to share the engine in order to increase the volume. The same is true in aircraft engines, where joint developments between competitors are routine.

Just checked, Sony's Revenue is 88B at market cap of 26B, 20B cash and 12B debt,
Canon's Revenue is 46B at market cap of 58B, 11B cash and 234M debt.

So Canon is a much better managed company, and has more net-cash (11B vs 8B) at disposal. And Sony is a beleaguered company trying to redefine itself. This is reflected in their market cap. Canon is the one with deeper pocket because it has more net-cash and more focused.

Canon would not be a well managed company if it put invested more in development of a product than it could recoup. The issue is not the overall size of the companies, nor even their spare cash, it is the size of their sensor businesses, and the relative revenue from those. Because Sony's is much larger, it can afford to put more resource into it. Likewise Canon will put much more resource into printing technology development than Sony, because Canon makes big revenues from printers and copiers while Sony doesn't. What you are suggesting is that Canon's 3M sensors per year can support the same development investment as Sony's 100M. That is absurd.

Also, how much resource do you think needed for R&D for a sensor? I guess it is not much for a company size of Canon.

Not much if canon feels like spending money it will never recoup. And it is not the cost of R&D for a sensor, it is investment in acquiring a technological capability. Sony's sensors are mixed signal, high quality analog and high speed digital integrated on the same chip. That is not a trivial thing to accomplish, and requires a dual expertise as well as semiconductor technology different from the simple digital or analog case. This capability, and its refinement is something that Sony has developed over time, with a great deal of investment. There is no reason to think that Canon could do the same without a similar level of investment.

Once all the logistics are in place like Canon is, it at most at the level of 10M.

Show me the calculations. Anyhow, even if Canon were to make only that sensor for a year, that would be $3 on every chip. And I'm sure that you are thinking of a variant of the existing tech, not a completely new technology.

The crucial part is to find the right people to do it right. I am surprised Canon has been using the 18MP sensor for their APS-C for such a long time. And from all the experiment chips Canon show now and then, Canon definitely can make many different sensors but somehow they think the 18mp is the best compromise for their mass production. But that doesn't mean they don't have anything in their arsenal, it is not uncommon they have 1-2 generation newer chips in the lab waiting.

But none of them is very different, or much better. The point is not simply to have a different sensor, it is to have one that performs as well as Sony's offerings, which needs a radical technology change.

In the end, the photography is more about system, not a single part. Canikon know it, so even in old bad days for Nikon, Nikon still survived fine. Because sensor is such an important piece for Canon's strategic market,

You are contradicting yourself. If the sensor wasn't 'an important piece' for Nikon's 'strategic market' then it isn't for Canon's. And if it is, then they need the best sensor that they can have, regardless of its source.

I don't think Canon will abandon their sensor development, and the competition is good the consumers.

Canon isn't a philanthropic organisation for the consumer's interest, it is there to maximise its profit, and if it's strategy leads to it lagging the competition in sensor capability, it needs to change it or risk losing market share.

-- hide signature --

Bob

Reply   Reply with quote   Complain
bobn2
Forum ProPosts: 27,829
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to buellom, Jun 30, 2011

buellom wrote:

But they all have the ability to design and to produce an engine by their own as one of their core competences ...

Canon uses (mainly?) Sony sensors in their Powershots. Guess these are the Minis of BMW. But Canon and BMW produce their own "engines" too and this is for both an area of core competence.

That sounds like what Nikon does. Retain the competence, use it strategically but in the end, select what will make the most money. Nikon seems to understand that their own sensors can only beat Sony's in very specific applications, where their superior QE is key (D3S) or where the IQ factor is less critical (D3100). And remember that the D3S sensor, while overall not as capable as what Sony is doing now, is better than anything Canon has.

The issue is business sense, not prestige. Move away from cars, note that Sony no longer makes its own LCD's for TV's (a core market). It has a joint venture with Samsung for some, others it buys from the likes of LG.
--
Bob

Reply   Reply with quote   Complain
x-vision
Senior MemberPosts: 1,035
Like?
Re: Poll on sensors for Canon DSLRs. (II)
In reply to Great Bustard, Jun 30, 2011

Very interesting discussion.

FYI, in the last two years Canon has filed & obtained multiple patents for what looks like a new generation of sensor design and manufacturing.

This year alone, for example, they have obtained at least three patents for reducing noise - including one for pattern noise (aka banding).

Canon's designs are not based on per-column ADCs (like Sony's) but Canon is obviously hard at work fighting noise in the sensor.

So, I fully expect that Canon will very soon match and maybe even exceed Sony's current performance.

Overall, though, I'm also not convinced about the long-term viability of Canon making its own sensors - or at least not all of them.

Sony is a diversified and higher volume sensor manufacturer, so they will always be more agile and quicker to implement new technologies than a lower-volume, specialized manufacturer like Canon.

FF sensors are likely to remain an in-house technology for Canon. Crop sensors, however, could very well be outsorced to another manufacturer. Don't know.

Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads