p.12 #1 · Hogan's article - The DSLR cash cow is alive for how long?
technic wrote:
Sony A9 has the lead in ML cameras for real action (fast AF, high framerate, minimal viewfinder blackout etc.) using stacked sensor, technology that is totally lacking at Canikon (although Nikon might use the Sony tech in future camera). Probably it is too expensive for more widespread use but it's essential to make a jump forward compared to top DSLRs that currently lead for action (if you ignore A9). And cameras like D500 offer excellent AF performance at a much lower price level than Sony A9 (not everyone needs FF ...).
Other factor that is holding mirrorless back from a real jump forward is the almost total lack of computational imaging. This is most visible compared to recent smartphones, but it is one of the features that would more easily separate ML cameras from DSLRs. ML requires more electronics and processing power than DSLR so it makes sense to use that as the basis for computational imaging. But like with very fast ML AF, it looks like the technology/cost is not quite there yet and will probably take at least a few years before we see something compelling. For now there is nothing more than a "promise" that ML is more suitable for this than DSLR platform.
While Canon has nice technology with DPAF especially for video it has its own issues, see recent discussions on Canon forum. I would not want a camera that fully relies on DPAF for focus.
Global shutter is another feature that could help some people make the jump, depending on implementation. There are still a lot of technological steps in the pipeline that can entice users to upgrade to ML (I'm assuming most of them won't be implemented on DSLR, but who knows); hopefully we won't have to buy a new camera (and wait several years) for every one of those steps. ...Show more →
Agreed, on pretty much all counts except that Canon does have the stacked sensor tech (they've got patents and have shown sensors, but not used it in production). I would not be surprised to find out that the 1DXIII is using some variant of that tech.
Otherwise I pretty much agree with you across the board.
p.12 #2 · Hogan's article - The DSLR cash cow is alive for how long?
technic wrote:
Sony A9 has the lead in ML cameras for real action (fast AF, high framerate, minimal viewfinder blackout etc.) using stacked sensor, technology that is totally lacking at Canikon (although Nikon might use the Sony tech in future camera). Probably it is too expensive for more widespread use but it's essential to make a jump forward compared to top DSLRs that currently lead for action (if you ignore A9). And cameras like D500 offer excellent AF performance at a much lower price level than Sony A9 (not everyone needs FF ...).
Other factor that is holding mirrorless back from a real jump forward is the almost total lack of computational imaging. This is most visible compared to recent smartphones, but it is one of the features that would more easily separate ML cameras from DSLRs. ML requires more electronics and processing power than DSLR so it makes sense to use that as the basis for computational imaging. But like with very fast ML AF, it looks like the technology/cost is not quite there yet and will probably take at least a few years before we see something compelling. For now there is nothing more than a "promise" that ML is more suitable for this than DSLR platform.
While Canon has nice technology with DPAF especially for video it has its own issues, see recent discussions on Canon forum. I would not want a camera that fully relies on DPAF for focus.
Global shutter is another feature that could help some people make the jump, depending on implementation. There are still a lot of technological steps in the pipeline that can entice users to upgrade to ML (I'm assuming most of them won't be implemented on DSLR, but who knows); hopefully we won't have to buy a new camera (and wait several years) for every one of those steps. ...Show more →
There's no reason why Canon and Nikon won't include as much mirrorless technology as possible in future DSLR bodies, like they are doing with the D780 and to some extent with the 1DX III. This is what sets them apart from Fuji and Sony who in reality now only offer mirrorless cameras. In addition, it will increase production volume of the sensors, processors etc. in question, possibly reducing the cost per unit. The Nikon D6 is already on its way, the next from Nikon will probably be a D860 and hopefully a D510 and D7600.
I'm not so sure about the low-end DSLR bodies. They are much bulkier than the corresponding mirrorless cameras, and cannot offer a viewfinder experience that competes with a camera like the Z50. But who knows... if they are profitable for Nikon, they won't go away anytime soon.
p.12 #3 · Hogan's article - The DSLR cash cow is alive for how long?
mawz wrote:
Agreed, on pretty much all counts except that Canon does have the stacked sensor tech (they've got patents and have shown sensors, but not used it in production). I would not be surprised to find out that the 1DXIII is using some variant of that tech.
Otherwise I pretty much agree with you across the board.
Canon may have stacked sensor tech patents, but production cost must be an even bigger problem than for Sony: you cannot run a profitable sensor production line (with dedicated technology) at very low numbers like for a Pro ML camera. Sony has the advantage that they have this tech already in production for various sensors, including several camera sensors. I'm sure Canon will get there at some point, but it will take years.
p.12 #4 · Hogan's article - The DSLR cash cow is alive for how long?
Jorgen Udvang wrote:
There's no reason why Canon and Nikon won't include as much mirrorless technology as possible in future DSLR bodies, like they are doing with the D780 and to some extent with the 1DX III. This is what sets them apart from Fuji and Sony who in reality now only offer mirrorless cameras. In addition, it will increase production volume of the sensors, processors etc. in question, possibly reducing the cost per unit. The Nikon D6 is already on its way, the next from Nikon will probably be a D860 and hopefully a D510 and D7600.
I'm not so sure about the low-end DSLR bodies. They are much bulkier than the corresponding mirrorless cameras, and cannot offer a viewfinder experience that competes with a camera like the Z50. But who knows... if they are profitable for Nikon, they won't go away anytime soon....Show more →
I think we will see more low-end DSLR bodies simply because they are cheaper to build than comparable ML bodies. Replacing that cheap pentamirror/mirrorbox assembly with electronics still costs a lot of money (plus extra power, memory, etc.). The low end will disappear when ML bodies finally get cheaper for the consumer, while providing at least similar margin to the producers.
p.12 #5 · Hogan's article - The DSLR cash cow is alive for how long?
technic wrote:
I think we will see more low-end DSLR bodies simply because they are cheaper to build than comparable ML bodies. Replacing that cheap pentamirror/mirrorbox assembly with electronics still costs a lot of money (plus extra power, memory, etc.). The low end will disappear when ML bodies finally get cheaper for the consumer, while providing at least similar margin to the producers.
I have read that mirrorless is less expensive to build because of lower part counts, and labor costs achieved by getting rid of the whole mirror assembly, focusing sensors and other things. A low end EVF should be very economical to build in quantity. Look at all the very low cost mirrorless point and shoots that are on the market.
p.12 #6 · Hogan's article - The DSLR cash cow is alive for how long?
technic wrote:
I think we will see more low-end DSLR bodies simply because they are cheaper to build than comparable ML bodies. Replacing that cheap pentamirror/mirrorbox assembly with electronics still costs a lot of money (plus extra power, memory, etc.). The low end will disappear when ML bodies finally get cheaper for the consumer, while providing at least similar margin to the producers.
I follow your logic, and that means that large numbers would be needed to make that happen. Electronics are only cheap when produced in large numbers, and with a declining camera market, that might not happen overnight. The mirror box assembly of the 3xxx and 5xxx cameras on the other hand probably doesn't change much from model to model, and can be produced dirt cheap on already existing equipment. Yes, the $500 DSLR may be hard to beat.
Jan 18, 2020 at 06:14 PM
Steve Spencer Offline • • • • • • • Upload & Sell: Off
p.12 #7 · Hogan's article - The DSLR cash cow is alive for how long?
Jorgen Udvang wrote:
I follow your logic, and that means that large numbers would be needed to make that happen. Electronics are only cheap when produced in large numbers, and with a declining camera market, that might not happen overnight. The mirror box assembly of the 3xxx and 5xxx cameras on the other hand probably doesn't change much from model to model, and can be produced dirt cheap on already existing equipment. Yes, the $500 DSLR may be hard to beat.
What seems to be beating the $500 DSLR is the phone that is already in people's pocket. I don't see much future even in the next year or two for a $500 DSLR.
p.12 #8 · Hogan's article - The DSLR cash cow is alive for how long?
Steve Spencer wrote:
What seems to be beating the $500 DSLR is the phone that is already in people's pocket. I don't see much future even in the next year or two for a $500 DSLR.
I think you're right and while there are still buyers for the $500 ILCs it remains to be seen whether there are enough buyers to justify producing many of them. Canon probably has an advantage with their distribution and name recognition so they might be able to keep a low end camera or two around longer but I'm not optimistic that we'll see very many in the future.
p.12 #9 · Hogan's article - The DSLR cash cow is alive for how long?
RobCD wrote:
I think you're right and while there are still buyers for the $500 ILCs it remains to be seen whether there are enough buyers to justify producing many of them. Canon probably has an advantage with their distribution and name recognition so they might be able to keep a low end camera or two around longer but I'm not optimistic that we'll see very many in the future.
Most smartphones still don't have decent tele/zoom lenses so a DSLR with extra lens or a more traditional superzoom compact are still superior - depending on ones style and subject, but certainly for travel, nature etc. This might change when we get affordable smartphones with both SWA and good quality 5-10x tele lenses, plus IS (suitable for good light conditions only, obviously) but I don't see this happening in the near future.
Some of my family members purchased a Canon DSLR in recent years assuming it would produce better vacation pictures than their iPhone - and it does, but only in some conditions while in other cases the iPhone does better ... I doubt most buyers will/can judge if the DSLR provides the better picture, marketing is a big factor. There is pervasive marketing suggesting that smartphones are BETTER than DSLR/ML cameras (especially from Apple), so maybe in time most buyers are going to believe that and prefer their smartphone for "serious" images.
p.12 #10 · Hogan's article - The DSLR cash cow is alive for how long?
1bwana1 wrote:
I have read that mirrorless is less expensive to build because of lower part counts, and labor costs achieved by getting rid of the whole mirror assembly, focusing sensors and other things. A low end EVF should be very economical to build in quantity. Look at all the very low cost mirrorless point and shoots that are on the market.
People frequently claim that but I have never seen real numbers to back it up.
I know some numbers from Canon DSLR production from years ago and I can assure you the actual production cost for the entry models is really low. I seriously doubt a decent ML camera (with an EVF that isn't clearly behind a pentamirror OVF) can be build for the same cost. Probably the cost per unit sold is more due to production numbers and development cost than actual parts cost; many recent DSLRs are mostly "from the parts bin" while ML cameras are constantly evolving and development cost per unit must be much higher, for now.
p.12 #11 · Hogan's article - The DSLR cash cow is alive for how long?
technic wrote:
People frequently claim that but I have never seen real numbers to back it up.
I know some numbers from Canon DSLR production from years ago and I can assure you the actual production cost for the entry models is really low. I seriously doubt a decent ML camera (with an EVF that isn't clearly behind a pentamirror OVF) can be build for the same cost. Probably the cost per unit sold is more due to production numbers and development cost than actual parts cost; many recent DSLRs are mostly "from the parts bin" while ML cameras are constantly evolving and development cost per unit must be much higher, for now....Show more →
Take a look at the parts count, mirrorless is a massive reduction in parts counts from any DSLR. Like a 2-3 times reduction in parts via the removal of all mechanical mechanisms except the shutter (And I expect to see the shutter go soon enough in the cheap ones). Sony's A6x00 series makes clear how little actual dev cost can be gotten away with (there's so much parts reuse that one of their big releases got out the door literally using parts already labelled for the predecessor model, the A6400 in this case, some shipped with A6300 shell parts, including model ID label)
And EVF panels can be REALLY cheap. Probably not cheaper than making a pentamirror find, but certainly cheaper than calibrating it.
Remember, EVF's have traditionally been present in superzoom P&S's, which generally came in cheaper than any DSLR.
p.12 #12 · Hogan's article - The DSLR cash cow is alive for how long?
Rational tells me new tech R&D at the beginning would be high, but less parts to assemble and calibrate would be far more cost effective in the long run. At a certain point in time, it is imaginable that low end current DSLR cost is lower than a MLC, but those products are discounted heavily, not bringing in as much revenue won't it ?
It is great for manufacturers turner if we think DSLRs are cheaper to build, then they are safe to price higher than DSLR counterpart. That pricing scheme, obviously failed with Z7 sales.
p.12 #13 · Hogan's article - The DSLR cash cow is alive for how long?
When did the first Nikon Coolpix appear? I bought my first in 1999. The digital camera industry has come a long way over 2+ decades minimizing costs, while the quality and prowess of the cameras (Nikon's at least) are excellent.
The MILC is the latest in a long vintage (the N1 series was also important in refining R&D and factory production).
Mass production of DSLRS after the SLR legacy must have been refined too, but they cannot escape a minimum of mechanical parts and fine hands on skills. The halcyon of the film SLR was probably the FM3a, and Nikon stopped this amazing camera because of high production costs, and the surviving F6 has a high price. A D5*00 is a cheapo product by comparison, but no escaping the mechanics at its core.
p.12 #14 · Hogan's article - The DSLR cash cow is alive for how long?
technic wrote:
Most smartphones still don't have decent tele/zoom lenses so a DSLR with extra lens or a more traditional superzoom compact are still superior - depending on ones style and subject, but certainly for travel, nature etc. This might change when we get affordable smartphones with both SWA and good quality 5-10x tele lenses, plus IS (suitable for good light conditions only, obviously) but I don't see this happening in the near future.
Some of my family members purchased a Canon DSLR in recent years assuming it would produce better vacation pictures than their iPhone - and it does, but only in some conditions while in other cases the iPhone does better ... I doubt most buyers will/can judge if the DSLR provides the better picture, marketing is a big factor. There is pervasive marketing suggesting that smartphones are BETTER than DSLR/ML cameras (especially from Apple), so maybe in time most buyers are going to believe that and prefer their smartphone for "serious" images....Show more →
You're confusing the issue here. You're trying to interject your reasoning for why someone would or should buy a DSLR and I'm talking about what's actually happening in the market. Smartphones are killing the compact and low end ILC market. You'd have to talk to all of those people that decided not to buy a DSLR to find out why they didn't care about the advantages a DSLR offers or why they cared more about the advantages their smartphone offers. I'm simply observing what's actually happening and I don't see it ending any time soon. In fact, I think smartphones will continue to move higher up market and in 2 or 3 years could be impacting even the mid-level ILCs like the Nikon Z50.
p.12 #15 · Hogan's article - The DSLR cash cow is alive for how long?
mawz wrote:
Take a look at the parts count, mirrorless is a massive reduction in parts counts from any DSLR. Like a 2-3 times reduction in parts via the removal of all mechanical mechanisms except the shutter (And I expect to see the shutter go soon enough in the cheap ones). Sony's A6x00 series makes clear how little actual dev cost can be gotten away with (there's so much parts reuse that one of their big releases got out the door literally using parts already labelled for the predecessor model, the A6400 in this case, some shipped with A6300 shell parts, including model ID label)
And EVF panels can be REALLY cheap. Probably not cheaper than making a pentamirror find, but certainly cheaper than calibrating it.
Remember, EVF's have traditionally been present in superzoom P&S's, which generally came in cheaper than any DSLR. ...Show more →
I agree that all this (less parts, declining electronics cost, more standardization) will make ML cameras cheaper in the end, but not NOW. As to EVF's, I really doubt that buyers of $500-2000 ML cameras will accept the quality of the cheap EVF's that are present in superzooms of a few years ago. The few fixed lens compacts that have nice EVF's are pretty expensive, way above the cost of an entry level DSLR.
p.12 #16 · Hogan's article - The DSLR cash cow is alive for how long?
Jorgen Udvang wrote:
The D810 is heavier than either of the cameras in my comparison, and the 24-70mm f/4 has no equivalent for Nikon DSLRs. The closest comparable lens would be the 24-120mm f/4 which is around 200g heavier than the 24-70, but offers nearly twice the reach.
I know this isn’t a direct comparison but I don’t care. It isn’t a contest. This lighter rig produces images as good or better for while being considerably more compact. Works for me.
p.12 #17 · Hogan's article - The DSLR cash cow is alive for how long?
Steve Spencer wrote:
What seems to be beating the $500 DSLR is the phone that is already in people's pocket. I don't see much future even in the next year or two for a $500 DSLR.
RobCD wrote:
I think you're right and while there are still buyers for the $500 ILCs it remains to be seen whether there are enough buyers to justify producing many of them. Canon probably has an advantage with their distribution and name recognition so they might be able to keep a low end camera or two around longer but I'm not optimistic that we'll see very many in the future.
I agree with both of you.
The only really important difference between smartphones and ILCs is the "specialty lenses" that can be attached to a MILCs/DSLRs. Such as: telephoto, TS/E, macro. Anybody dumping thousands of dollars into those lenses ain't gonna use a $500 ILC. Probably. Most likely a mid $2k+ camera.
Therefore...
The camera makers: Canon, Nikon, Sony, etc need to cater to the mid-up scale to wealthy to serious/real pro photographer. At least in the long(er) run. And the "long run" is not as long as the bean counters at said camera companies think. Sure Canon and Sony would be able to have one or two small/basic cameras in their lineup due to their size/scale but that's it.
Nikon might get one generation of camera below the Z 50 (say Z 30) but I think those will dry up also, unless they are uber cheap.
BuT, bUT, BUt, But.... the D3/5xxx and Rebels sell like hot cakes you say. True, they do. And they are money makers now, but that window is rapidly closing. Smartphones have been out over a decade now. Pretty soon all the new/young camera buyers will be people that grew up with camera phones and they would never consider buying a D3xxx camera to take pics of lil Timmy's soccer game. In a few years we'll see phones with small lens mounts that can attach small 200-400mm equivalent zooms that are 2-4" long and are only $200. Buh-Buy to all "cheap" cameras.
p.12 #18 · Hogan's article - The DSLR cash cow is alive for how long?
Forgot to add to my previous post:
The most important cost savings when manufacturing a DSLR v MILCs is the labor cost:
Less people needed to make less parts. More automation.
Less people needed to assemble less parts. More automation.
Less people needed to align/test/validate less parts. More automation.
Less people needed equals more QC. More automation.
So maybe right now some MILC parts cost more but those will rapidly fall in cost. Mirror box, pentaprism, AF sensors and shutters are all going away.
The only DSLRs that can support and "high cost" of camera manufacturing will very soon only be the mid/high to high/pro camera.
p.12 #20 · Hogan's article - The DSLR cash cow is alive for how long?
The mechanical shutter isn't going anywhere, Sony even added a new mechanical shutter to the A9 II - the camera with the most capable electronic shutter still needs the mechanical shutter, e.g., for flash photography and for situations where the e-shutter causes artifacts.
While the assembly of a mirrorless camera is simpler than that of a DSLR, some of the parts are really expensive to make, including the sensor where the read time is now a priority because AF is based on the sensor, and this is costly to develop. Sony have so far created one fast sensor that is cheap enough to put in a consumer still camera (the A9(II)'s 24MP sensor) and all of the A7 series cameras have e-shutters that take a good bit of time to read, so the AF isn't as fast and e-shutter produces more common rolling shutter and horizontal bands in artificial light. Nikon were asked about global shutter and they responded that it would be far too expensive. Also the lenses for mirrorless cameras need new AF motor technology and changes to the optical design to facilitate focusing with such motors, all of this costs a lot of money to develop. Sigma commented "In order to support contrast AF we need to implement 'wobbling': we need to move the focusing lens back and forth to detect the subject. In order to do that, we need to make the focus lens very light and small and this tends to degrade the performance in the short focal distance." Also the Nikon EVF for the Z6 and Z7 is finally good EVF which doesn't flicker annoyingly in artificial light, and is sharp to the edges, but it has a highly complex optical system between the display and ocular, adding to the number of parts needed. Mirrorless isn't cheap by any means. By comparison, the cost of assembly of consumer-level DSLRs seems to be a relatively trivial cost, judging from the prices of the products in the market. At the very high end (D5, 1DX II, etc.) the cost is higher, though. The A7R IV costs about 4200€ in my country, the D850 can be had for 2500€. I can get a 50/1.8 AF-S for 199€, whereas the Z 50/1.8 S cost is 529€, I am not seeing the fabled reduced cost in mirrorless products, it is clearly an alternate reality that some people imagine.
