*UPDATED* Big Megapixel Camera Next Week.

[LetTheRightLensIn]

Not looking for megapixels -- 21 is already too many for FF. Huge files, slow transfers, slow processing.

Looking for better low light performance; lower noise; better EVF previewing for longer exposures; GPS; wireless downloading; wireless charging; programmable exposures longer than 30 seconds; Switchable IR filtering (or none... I'd be happy to slap an IR block filter on the lenses when shooting normally.) (C'mon Canon, don't you want to sell a boatload of highly profitable filters???)

You know... things that will actually improve the images and the regimes, and the shooting experience, as opposed to giving us, yet again, an even more detailed view of the flaws in the lenses.

But of course, what we'll GET is... more megapixels. Bigger files. Slower processing. Slower transfers. Oh yeah, and more effort wasted on trying to inappropriately multipurpose these beautiful still cameras into video cameras. And maybe a new battery style so as to obsolete the hugely expensive ones we've already bought.

I am *so* tired of "Hey, look more megapixels!" and frankly, I'm not going to buy unless the camera actually performs better. Increases in megapixels work *against* my buy-in. And no, throwing out half or 3/4ths of the pixels in the camera isn't the right way to go -- the less sensor area involved in taking the shot, the more noise there will be. With a nominally high megapixel camera, you have to take the entire image and then bin the results in order to approach the actual noise reducing effect of larger sensels. So modes like SRAW and SRAW2 don't really help (unless the cameras are modified to do the binning, that would be a useful change.)

I like the extra reach and detail more MP gives, although at it's getting to the point I'd rather have a least a couple more fps than more MP although ideally I'd have 6.3fps+ plus more MP.

And I love that they added in video and I really hope DIGIC V can do proper scaling from the entire sensor each frame. It will get rid of all the nasty moire and artifacts andgratly boost SNR. Video would be truly phenominal at that point. I think it will. They will get killed by the video competitors if not. If they make things this fast they should also be able to do some binned stills modes too.

And yeah it sure would be nice if they lowered read noise and fixed banding so we get more DR.

 

[LetTheRightLensIn]

His eyes lifted, "since we got the 5D2 - 3 years ago, we haven't sold a single 1DsIII. At 3 years the 5D3 would be more than a suitable release and more than able to match most of Nikons plans, It doesn't even need that much work to get them selling in volumes quicker than Canon can produce them, and certainly the equal of any D800 nikon might have to offer! This would leave Canon till Christmas to sort out their 1D stragey ready for the games..."

Whichis why I think they need to stop protecting the darn 1Ds series so much and worry about D800 and such and soup up the 5 series so it's not just a great sensor in a Rebel performance body.


You've actually raised an interesting point:

There is a rather limited market for the 1DsIII - most shops, with the exception of pro dealers would not have sold many 1DsIIIs anyhow. On the other hand the 5DII sells like hotcakes.

Even if the 5DII has cannibalised a lot of 1DsIII sales, the net result for Canon has probably been a good one.

It would not surprise me if the higher volumes make for better margins on the 5DII. Is there anyone who has specific knowledge of this? (I mean margins for Canon, not for distributors and retailers.) Canon cannot go to market without the 1DsIII though, as then they would give up ground to Nikon.

If however Canon appeal more to the medium format market with the 1DsIV, that may change things.

 

[LetTheRightLensIn]

Re: Big Megapixel Camera Next Week.

Consider an alternative line of thinking. Canon has already established the precedent that there would be a 5D3 which has the same resolution as a 1Ds4.

Please explain how they established that precedent. The original 5D had lower resolution than the then-current 1Ds model. The 5D Mark II has the same resolution as the current 1Ds model. So, if I flip a coin and it's heads, and the next time it's tails, tails is the precedent, and the third flip will surely be tails? Woah baby, Vegas here I come! The only 'precedent' I see is that the 5DIII will NOT have *more* resolution than the 1DsIV, and even then, n=2 makes for a pretty flimsy precedent.

Oh, c'mon, don't go all statistical on me! Will you at least admit that a lot of people EXPECT that the 5D3 will have the same number of pixels as the 1Ds4 and would make their buying decisions based on that expectation? (The 5D2 is the more recent precedent and therefore is more salient.) The poster I was responding to certainly seemed to assume that.

I kind of hope it won't. I could see a super high 40MP 1Ds studio cam and then a faster, more general say 28MP 5D3.

Then again Canon will probably toss us a 5D3 with 40MP and 1fps and 3pt AF (non-cross) and cheer when the Nikon D800 with 7fps and their top AF arrives with 'only' 24 MP....

 

[LetTheRightLensIn]

Re: Big Megapixel Camera Next Week.

Oh, c'mon, don't go all statistical on me! Will you at least admit that a lot of people EXPECT that the 5D3 will have the same number of pixels as the 1Ds4 and would make their buying decisions based on that expectation? (The 5D2 is the more recent precedent and therefore is more salient.) The poster I was responding to certainly seemed to assume that.

5DII and 1DsIII have the same sensor probably for cost saving reasons, Canon probably thought it could differentiate these 2 models using the other features of the 1 series, what happened is what we all know 5DII cannibalized so much of the 1DsIII that it almost killed it

the fact that the sensors were the same made studio photographers think: why should I pay 2X the price for faster autofocus and weather sealing and tank like build in my studio with no IQ improovement


the recent rumors about a high mpx 5DIIi and very high mpx 1dsIV indicates that canon will not repeat the same mistake and (hopefully) put some more advanced features (autofocus, metering, fps ...) in the 5DII and differentiate from 1 series with mpx

indeed, and one can only hope

 

[c-law]

Why does everyone go gaga over dual card slots?

Data protection. Memory cards do fail or get corrupted occasionally.

yeah but how often is that?
Over 6 years of sometimes extremely heavy usage I've had not a single failure.
So wouldn't something like say 6-8fps rather than 4fps be a MUCH bigger deal on it than dual card slots???

How often does that happen? How often does it need to happen to ruin your reputation?

Chris

 

[ursinus71]

: fyngyrz on August 19, 2011, 05:47:26 AM

But of course, what we'll GET is... more megapixels. Bigger files. Slower processing. Slower transfers. Oh yeah, and more effort wasted on trying to inappropriately multipurpose these beautiful still cameras into video cameras. And maybe a new battery style so as to obsolete the hugely expensive ones we've already bought.

Again fyngyrz this is ill-informed. Nothing is sacrificed to incorporate video. You're paying more for the overall camera, but it certainly doesn't mean that they cut corners in the still dept. This is another cliched criticism that has no basis in reality.

 

[neuroanatomist]

Nothing sacrificed? How do you know? R&D dollars (or yen, in this case) are finite. Engineering (soft- or hard-ware) effort devoted to video capability cannot be used for still capabilities. Granted, many features overlap, but some do not.

 

[epsiloneri]

Nothing sacrificed? How do you know?

Off course he doesn't know, but it's not an unreasonable assumption. It could be that the effort that goes into video results in more sales of the camera (fair assumption) and by doing so effectively decreases the price per unit (including R&D $$), compared to what would have been the case if there was no video. I'm not saying this is the case, but I think that it reasonably could be. As long as video features do not conflict with still features, I see no point in complaining.

 

[-zero-]

Nothing sacrificed? How do you know?

Off course he doesn't know, but it's not an unreasonable assumption. It could be that the effort that goes into video results in more sales of the camera (fair assumption) and by doing so effectively decreases the price per unit (including R&D $$), compared to what would have been the case if there was no video. I'm not saying this is the case, but I think that it reasonably could be. As long as video features do not conflict with still features, I see no point in complaining.

money spent on developing new technology for video on a dslr is money that cannot be spend developing new technology for stills on the same dslr

I think this is what neuro was saying

 

[epsiloneri]

money spent on developing new technology for video on a dslr is money that cannot be spend developing new technology for stills on the same dslr

And money spent on marketing is also money not spent on R&D. So would we get better and cheaper cameras if Canon spent no money on marketing but diverted it to R&D instead?

 

[Orangutan]

Nothing sacrificed? How do you know? R&D dollars (or yen, in this case) are finite. Engineering (soft- or hard-ware) effort devoted to video capability cannot be used for still capabilities. Granted, many features overlap, but some do not.

Which parts don't overlap?

• Rapid capture of frames from the sensor: already done to a great degree for Live Mode. The next question is whether it can be sped up/scaled-up for true video, but that's already done (due to marketing necessity) for P&S cameras, so it must be similar for higher-end sensors. There's no reason to believe this is any harder on a DSLR than a P&S. R&D needed: routine adaptation to the specs of a new sensor.
• Processing frame capture to video: this is just software and, again, they already do this in P&S cameras using same (or very similar) chips. It may be the same algorithm verbatim, or a family of algorithms that are adapted based on the specs. R&D needed: routine adaptation to the specs of the new sensor.
• Sensor heat dissipation: this is a bit trickier to judge, but I don't see how it would be worse on a larger, less-dense sensor than on a small, dense P&S. DSLR users, particularly pros, would demand longer capture times than P&S users, and it may be easier to cool through a small P&S body than a tank-like DSLR. Nevertheless, seeing how video appeared to be a near afterthought for the 5D2, I can't imagine that this is a huge barrier either. Furthermore, it seems entirely plausible that creating a sensor that dissipates heat better would add to IQ of still images, not detract. R&D needed: make sure the sensor can dissipate heat well.

The only thing I believe would be different about adapting a DSLR for video (vs. P&S) is the expectation of longer capture times, hence the need for heat dissipation. For a DSLR that already has live view, adapting it for video may not be a true freebie, but it's damned close. Furthermore, the heat dissipation R&D is only a bonus for still IQ. If you want to save significant R&D money on a still camera you'll need to get rid of Live View: that's where the R&D delta is likely found.

I invite critique of the above.

 

[-zero-]

money spent on developing new technology for video on a dslr is money that cannot be spend developing new technology for stills on the same dslr

And money spent on marketing is also money not spent on R&D. So would we get better and cheaper cameras if Canon spent no money on marketing but diverted it to R&D instead?

R&D and marketing department are independent from each other, companies allocate a specific amount for each (often the marketing is much cheaper than R&D) so more money on marketing has no impact on the R&D budget

but if you allocate a team of engineers to improve a video feature on a camera, the same team cannot work on another feature related to stills

money and manpower are often the same thing from the company point of view

 

[neuroanatomist]

companies allocate a specific amount for each (often the marketing is much cheaper than R&D) so more money on marketing has no impact on the R&D budget

Companies spend more on R&D than on Marketing? Thanks! With tropical storm Irene passing over my head, I really needed a laugh this afternoon...

Seriously, Canon often buys ad spots for the Super Bowl - and at a few million dollars for 30 seconds of air time, that would pay the salaries of a LOT of engineers. I can say than in pharma, there are about $2 spent on marketing for every $1 spent on R&D - if I had to guess, I suspect the ratio may be even higher in the electronics industries.

Let's look at the numbers. Canon's list of Marketing Subsidiaries and Affiliates sums up to around 17,000 employees - and that's just 'external' marketing, not internal staff. Do you really think they are anywhere close to that staffing level for R&D? I sure doubt it.

As for marketing spend not impacting R&D spend? A nice second chuckle. Of course it does - yes, they are separate lines on the ledger, but if more money is spent on one line, it has to come from somewhere...and most corporations are inclined to take first from R&D, because R&D spend doesn't provide an immediate return on investment.

 

[epsiloneri]

R&D and marketing department are independent from each other, companies allocate a specific amount for each (often the marketing is much cheaper than R&D) so more money on marketing has no impact on the R&D budget

Are you saying Canon cannot allocate funding between marketing and R&D as they wish?

but if you allocate a team of engineers to improve a video feature on a camera, the same team cannot work on another feature related to stills

If you had the money, you could hire another team of engineers, because:

money and manpower are often the same thing from the company point of view

Precisely. This is however besides the point. I'm sorry I did not explain it sufficiently, so let me repair my mistake by going into more detail:

The thesis is that there is a fixed amount of money available to R&D, and by spending money on video R&D, you are not spending it on still photography. Therefore, we would get better still photography cameras if Canon focused all their R&D towards still photography and forgot about video. This is your argument, correct?

But what if the assumption that there's a fixed amount of money available for R&D is wrong? What if Canon actually can afford to spend more money on R&D if there are more units sold and they make more profit? And then, what if developing video features contributes to more sold units and more profit for Canon? Then the R&D budget would be able to increase and not remain fixed. The total $$ spent on R&D would increase, and perhaps even the R&D spent on still photography would increase compared to the case when no video features were developed.

I took the example of marketing because it's so obvious (I thought) that they need marketing to sell products. If they took away all the marketing, they would sell very few units (in comparison), and because of that their total budget for R&D would not increase as you would perhaps naively think (because you directed funds from marketing) but would actually decrease (because the total $$ available would decrease).

 

[epsiloneri]

With tropical storm Irene passing over my head, I really needed a laugh this afternoon...

Take care and hold your camera in a firm grip Good thing you still have internet.