Canon Leads in Sensor Tech

[jrista]

Another debate on sensors, and now Nikon is / or is not crushing Canon's sales because of it. The sensor is just a small part of the big "picture", so to speak.

If you haven't figured it out by Canon's actions with DSLR's in the last few years they seem to think Video is the future.

I doubt the sales lost or gained from Nikon are significant compared to the sales Canon lost to the iPhone and other phone manufactures. Canon should have gotten a clue from these companies. Where are the Canon phones built in to their cameras? The dedicated P&S is dead, add a smart phone to the back of it and people will come back. I would think seriously about buying a T5i phone, it would be smaller than bag phone I had in 1988. How about a new M model 1/8" wider with an iPhone on the back. I might pre-order.

Sensors, who's is better? Unless Canon sensors crash I don't care. I have my good glass, in a year or two this debate will be about a completely different set of specs on sensors and I will still have my good glass.

It's a good point about good glass. That's the reason I spent over ten grand on the EF 600mm lens. But there is a cost do buying good glass...the system has to last. If the onslaught of racing technological improvements in the CMOS Image Sensor sector continues, Canon's sensors will become irrelevant in a few years. Image quality took a jump with the D800...what happens when even your smartphone camera can achieve dynamic range like that, and every DSLR except Canon's does even better?

All anyone is saying is that Canon needs to step up their game, broaden their focus, and compete more effectively on multiple fronts. Dynamic range also doesn't just apply to stills photography, either. Canon is a new entrant to a fairly well established digital cinematography market. Again, sensors for the cinema segment are also improving. Red demonstrated a cine-sensor with over 20 stops of dynamic range! Sony has produced numerous smaller video sensors with incredible dynamic range (although nothing yet that quite compares with 20 stops). The sensor IQ debate doesn't actually end with stills, it is a video factor as well (and in many ways, DR is FAR more important for cinematography than it is for stills.)

Will Canon actually succeed in the video market? What'll happen when their cinema line remains stuck at 12 stops of DR while all of their competition continues accelerating to 20 stops and beyond? If the future is video, then Canon still needs to compete, and DR is still a factor. I hope they figure that out and do something about it before their (rather diverse) competition crushes them. I've invested too much.

 

[takesome1]

I do agree with some of what you've said about DSLR video. On a much earlier thread I suggested that video and stills, having converged for the past several years, may be at the point where they start to diverge again. I don't know enough about the technology, but I suspect it is unlikely that Canon can continue to improve their DSLRs in both video and stills without one starting to conflict with the other.

Sometimes I dream that the next 7D will not have a dual pixel sensor and instead Canon will use what they learned in developing that technology to make a sensor that performs better for stills photographers. I can dream can't I?

I don't think it is impossible for Canon to achieve that. I do, however, think they need to spread the R&D spending around a bit more. I can't really remember the last time there was a significant still photography innovation. The 61pt AF system might be the last thing I've heard that seriously improved still photography IQ in the Canon world...and that was quite a while ago now, couple years at least.

I've tried to use the video capabilities of my 7D a few times to record some of the interesting things I see out in nature, the stories that a few still photographs just can't tell. I've come to the conclusion that outside of your basic shaky-camera gig with quirky focus, you can't really use a DSLR for good quality video without investing in some of the tools that actually make it practical. A focus puller, a fluid-filled tripod head, a proper screen magnifier and shroud, etc. Once you get those things, and really want to start producing some higher quality cinematography, you start to realize you can't really do it all on your own, and you realize you need even more gear...maybe a dolly for smoother panning, and at least one other person to pull focus while you focus on everything else. Then you start thinking abut audio, the need for some external microphone jacks, cleaner video output, RAW video output, so on and so forth. Its just another rabbit hole.

You can do some basic things with DSLR video, and software like Adobe Premier helps (especially with its post-process image stabilization features...however then you really wish you had the full 4k 4096x3112 resolution so you have some extra pixels to support the cropping that comes along with that stabilization)...but anything more, and a simple DSLR just doesn't cut it, and it is an R&D funding black hole...

I find the viability of DSLR video in professional DSLR cameras to be limited without a lot of extra gear, and a growing number of additional features that would need to be added to stills cameras to make it really viable. Even for something as simple as filming the local fowl and fauna to make a short, but quality, video. So I totally agree...I think it is time the technologies diverge...at least at the professional level. I honestly couldn't really care what Canon does with their consumer grade products. ;P

If memory serves, when Canon released the 1D C they had released video of photographers doing frame grabs for stills.
This very well may be the way Canon see's still photography in the future. Instead of 12 fps you shoot a few seconds of video and cut what you need out. The problem I saw when watching the video, the IQ wasn't quit there yet.

I think diverge is a pipe dream, Canon's actions the last few years seem to always be pushing toward video and combining the two.

 

[takesome1]

Another debate on sensors, and now Nikon is / or is not crushing Canon's sales because of it. The sensor is just a small part of the big "picture", so to speak.

If you haven't figured it out by Canon's actions with DSLR's in the last few years they seem to think Video is the future.

I doubt the sales lost or gained from Nikon are significant compared to the sales Canon lost to the iPhone and other phone manufactures. Canon should have gotten a clue from these companies. Where are the Canon phones built in to their cameras? The dedicated P&S is dead, add a smart phone to the back of it and people will come back. I would think seriously about buying a T5i phone, it would be smaller than bag phone I had in 1988. How about a new M model 1/8" wider with an iPhone on the back. I might pre-order.

Sensors, who's is better? Unless Canon sensors crash I don't care. I have my good glass, in a year or two this debate will be about a completely different set of specs on sensors and I will still have my good glass.

It's a good point about good glass. That's the reason I spent over ten grand on the EF 600mm lens. But there is a cost do buying good glass...the system has to last. If the onslaught of racing technological improvements in the CMOS Image Sensor sector continues, Canon's sensors will become irrelevant in a few years. Image quality took a jump with the D800...what happens when even your smartphone camera can achieve dynamic range like that, and every DSLR except Canon's does even better?

All anyone is saying is that Canon needs to step up their game, broaden their focus, and compete more effectively on multiple fronts. Dynamic range also doesn't just apply to stills photography, either. Canon is a new entrant to a fairly well established digital cinematography market. Again, sensors for the cinema segment are also improving. Red demonstrated a cine-sensor with over 20 stops of dynamic range! Sony has produced numerous smaller video sensors with incredible dynamic range (although nothing yet that quite compares with 20 stops). The sensor IQ debate doesn't actually end with stills, it is a video factor as well (and in many ways, DR is FAR more important for cinematography than it is for stills.)

Will Canon actually succeed in the video market? What'll happen when their cinema line remains stuck at 12 stops of DR while all of their competition continues accelerating to 20 stops and beyond? If the future is video, then Canon still needs to compete, and DR is still a factor. I hope they figure that out and do something about it before their (rather diverse) competition crushes them. I've invested too much.

What will happen when my smartphone is as good as my 1D? I will be buying a smartphone adapter for my 500mm, and carrying a few less pounds on the airplane.

 

[Famateur]

I hear ya, Unfocused. My opinion (as expressed in another thread) is that Canon chose to concentrate on Live View AF for this generation of sensor. From all the YouTube videos I've seen of the 70D thus far, they pulled it off superbly.

I understand that, to many, Dual Pixel AF appears to be aimed at videographers, and while in practicality, it pretty much is right now, think of what this technology will do for the next mirrorless body!!! That's the first thing I thought of when they announced DPAF. Add a high quality EVF (for those who really want/need it), and you've got a mirrorless body that spanks the competition for auto-focus. I expect a DPAF EOS M will fly off the shelves.

One other thought: we know companies tend to release technology on a scheduled road map. While many are moaning that the DR/noise performance hasn't improved that much in the 70D compared to previous generations, how do we know that they didn't just hold it back in the release pipeline to release it with the 7D?

I know that's optimistic, and I'm perfectly content if they really did focus on just Live View AF, but maybe we should wait to see what the 7DII actually delivers. Just a thought...

Anyway, Canon demonstrated that when it sets out to solve something (like Live View AF), it can succeed. When Canon decides to "solve" the dynamic range and noise "problem", I have no doubt it will be equally successful. I'm excited for Canon. 8)

Of course, not everyone in this forum shares my optimism, particularly after having high DR/noise hopes for the last couple of years and having them dashed. They're like Buttercup, and I'm Wesley as we're fleeing to the Fire Swamp for safety:

Buttercup: We'll never survive!
Wesley: Nonsense! You're only saying that because no one ever has...

;D ;D ;D

 

[Loswr]

Of course, not everyone in this forum shares my optimism, particularly after having high DR/noise hopes for the last couple of years and having them dashed. They're like Buttercup, and I'm Wesley as we're fleeing to the Fire Swamp for safety:

Buttercup: We'll never survive!
Wesley: Nonsense! You're only saying that because no one ever has...

What are the dangers of the Fire Swamp? The High DR spurt, but there's a popping sound from the ADC before that, so those are easily avoided. Lightening pattern noise, but Aglet was clever enough to figure out what that looks like, so we can avoid that, too. What about the ROUS? Resolution of Unusual Suckiness? I don't think that exists.

 

[Famateur]

Of course, not everyone in this forum shares my optimism, particularly after having high DR/noise hopes for the last couple of years and having them dashed. They're like Buttercup, and I'm Wesley as we're fleeing to the Fire Swamp for safety:

Buttercup: We'll never survive!
Wesley: Nonsense! You're only saying that because no one ever has...

What are the dangers of the Fire Swamp? The High DR spurt, but there's a popping sound from the ADC before that, so those are easily avoided. Lightening pattern noise, but Aglet was clever enough to figure out what that looks like, so we can avoid that, too. What about the ROUS? Resolution of Unusual Suckiness? I don't think that exists.

LOL! Once again, your wit was faster than my reply to my own post (as evidenced by my other post being after this one). Resolution of Unusual Suckiness! Love it!

 

[Famateur]

How do we know that they didn't just hold it back in the release pipeline to release it with the 7D?

Just to expand on this a little more, it sounds like the 7DII will be released several months (or even a year) after the 70D. Am I the only one that thinks Canon has more up its sleeve than putting the 70D sensor in a bigger heavier body with more whistles and bells?

I know, I know -- that's essentially what they did with the 60D and 7D (only in reverse?). Maybe I'm getting back to Fire Swamp optimism... ;D

It just seems like that's a lot of time to wait for an already long-overdue successor to the 7D, and perhaps that time would make more sense to us if we knew that a new sensor tech (or ADC, or whatever) was being tested in conjunction with DPAF. We've all been thinking about the 70D sensor as it is and wondering what Canon might add to it for the 7DII. What if it was the other way around? What if they created the 7DII sensor and then opted to use a lesser "consumer" version with DPAF for the mid-level 70D?

"Rodents of Unusual Size? Personally, I don't think they exist." 8)

 

[jrista]

Of course, not everyone in this forum shares my optimism, particularly after having high DR/noise hopes for the last couple of years and having them dashed. They're like Buttercup, and I'm Wesley as we're fleeing to the Fire Swamp for safety:

Buttercup: We'll never survive!
Wesley: Nonsense! You're only saying that because no one ever has...

What are the dangers of the Fire Swamp? The High DR spurt, but there's a popping sound from the ADC before that, so those are easily avoided. Lightening pattern noise, but Aglet was clever enough to figure out what that looks like, so we can avoid that, too. What about the ROUS? Resolution of Unusual Suckiness? I don't think that exists.

Hah!! +10000

 

[LetTheRightLensIn]

I was too unfocused tonight to take the bait!

(plus everyone already hit most, enough, of the relevant points already)

 

[9VIII]

I would still be happy if they would just re-release the 1D4 at half the price.

I remember an interview where some Canon executive said that part of their long term strategy is to introduce larger sensors into lower price ranges.
Lowering the cost of large sensors actually seems like the most beneficial improvement they could make. Most people already use FF glass, now they just need the body to unlock its potential. Stick APS-H in the next Rebel and full frame in the 80D, now that's an improvement.
That also fits with introducing medium format for a new top tier.

 

[jhanken]

I remember an interview where some Canon executive said that part of their long term strategy is to introduce larger sensors into lower price ranges. Lowering the cost of large sensors actually seems like the most beneficial improvement they could make.

Thank you. The business reality is this:
- The collapse of the point-and-shoot market is an existential threat to Canon, Nikon, and Olympus, and perhaps Fuji and some others I am forgetting. The significance of this fact cannot be overstated when trying to understand each company's moves in the DSLR market. Other companies might survive a nuclear scenario in point-and-shoots, but with significant hits to their photo arms (e.g. Sony, Samsung, Panasonic),

- Fuji, Olympus, Panasonic, Sony and Samsung have all had strong technical responses in terms of new offerings in the MILC category or othe new categories Nikon's 1 system approach has been a bit iconoclastic, and Canon's M system was a least initially bungled, despite some promising hardware

- The vanilla middle seems to be a terrible place to be alone, i.e. APS-C sensor or smaller sensor, not interesting body, nothing crucially differentiating (e.g. the leaf shutter Fuji x100 family), based on critical reception. While the photo technology may be fantastic in this space, it is very difficult to differentiate. Cash cows are mooing but perhaps endangered.

- Canon, for sure, has a bit of an albatross with its current investment in sensor tech production capacity. That, and its confidence that it can wrestle competitive tech advances from current technologies, goes a log way to explaining the be split-receptor technology pioneered in the 70D. Probably no coincidence that this new tech also is biased toward video users in terms of attractiveness, that makes sense for any company trusting to shore up defenses against the IOS/Android phone camera onslaught.

- Given the need for companies to differentiate in some meaningful way, Canon and Nikon's relative position with respect to high-end DSLRs makes sense, given Nikon's reliance on more up-to-date chip fans, and canon's growing popularity in the increasingly competitive video hardware market.

- And finally this: Canon's most durable competitive advantage is in the development, production and sale of high-end lenses. Not that Tay aren't great at camera bodies, it's just that so are others, depending on what you prefer. A focus on full-frame users, even if at reduced ultimate margins, serves to further increase the size of a customer base with extremely high switching costs and generally high loyalty.

 

[AvTvM]

1. I clearly see Canon behind Sony/Nikon in sensor tech. And not just by a small margin.
Both Low ISO IQ [read noise] and even more so in DR which on average is 1.5 to 2 stops worse than comparable Nikon cameras, which in addition are priced roughly 1/3 lower [especially 5D 3 vs D800].

2. I do not believe, video to have a huge growth potential compared to stills. If this was the case, we would see a constant flow of excellent, dedicated video cameras with large sensors/shallow DOF without flipping mirrors in the light path and at prices comparable to DSLRs. But we are not.
The true reason is evident: post-processing of video material is more difficult and time-consuming by a magnitude of 100 to 1000 times compared to posting OOC jpgs to facebook. For every pro or enthusiast willing to edit video there are 1 million stills snapshooters not willing to spend the time and effort.

3. The new 70D dual-pixel sensor with on-sensor PD/CF-AF WOULD make a lot of sense - in a mirrorless camera! Definitely not in DSLRs. But up to now Canon has been dragging its feet over putting this technology to good use by building market-leading ultra-capable and ultra-compact mirrorless cameras [and lenses]. The were beat by Olympus, Fuji and others and they will soon be beat big time by Sony [upcoming FF-mirrorless].

4. Canon set the wrong priorities in supplying video-capable DSLRs which only a small, but very vocal minority of its customer base really needs. I do understand, why this small group of video-oriented customers is so happy: all of a sudden they could get video-DSLRs with high-quality, shallow-DOF for only USD 2500 [5D 2] instead of being forced to rent professional video-cameras costing USD 20.000+ Of course they were and are enthusiastic. But it is a very small minority.

5. Canon will suffer the consequences of this desicion. Due to "inertia" [large installed base + good brand recognition] it may still take a while, but it will come. For sure.

 

[Aglet]

Over the next few days we may see a what the competition's gonna be offering to tempt our wallets.
Oly's new EM1 looks like a deliciously compact little MFT that could be even sweeter than the EM-5.
Sony may be courting phone-camera users with real optics add-ons.
Pentax is rumored to have some new toys for us too.
Fuji's likely up to something interesting again.

All these are very innovative but likely still to remain niche players compared to the Canon Juggernaut.. who may be finally intro'ing improved M models. Almost seems like an anti-climactic finish if Canon doesn't have something meatier for a press release.

2013 is SUCH an interesting year for photo-tech.

 

[AlanF]

Over the next few days we may see a what the competition's gonna be offering to tempt our wallets.
Oly's new EM1 looks like a deliciously compact little MFT that could be even sweeter than the EM-5.
Sony may be courting phone-camera users with real optics add-ons.
Pentax is rumored to have some new toys for us too.
Fuji's likely up to something interesting again.

All these are very innovative but likely still to remain niche players compared to the Canon Juggernaut.. who may be finally intro'ing improved M models. Almost seems like an anti-climactic finish if Canon doesn't have something meatier for a press release.

2013 is SUCH an interesting year for photo-tech.

Be careful with "Oly", they are in big trouble - http://www.bbc.co.uk/news/business-23955003

 

[docsmith]

There are so many different directions to go with this thread....First, I think the dual pixel was a great advancement for the consumer/prosumer level 70D. But, I am not sure the 70D was the target for developing the dual pixel system, but rather the EOS-M series cameras. Canon will likely drop the dual pixel system into the next generation of M to improve AF and have a potentially very nice second offering into the mirrorless market.

I don't see video as the growth market that others seem to think. Consumer and professional level video camcorders have been around for decades. Affordable video cameras at least 20 years. That was a defined, mature market. The "growth" that Canon experienced was stealing away users from that market, both professional and consumer, so that they could have stills and HD video from one unit....and also use Canon's lens system.

As for stills...this isn't scientific, but everyone I know that is into image capture (video or stills) is primarily into stills. This goes from age 10 to age 70. Not only are they into stills, but almost no one cares for video. I am not really sure why, but it just seems people (that I know) prefer stills. For someone like Canon to ignore stills would be like (to stick to the car analogy from earlier) an automobile maker ignoring the family sedan. Sure, it may not be the "growth" but it is still a very significant market.

Finally, regarding Canon moving onto new sensor tech because their current sensor tech is 10 years old....I hope they do. I would love more DR (as long as they still have good contrast). I would love the best sensor tech possible. But often, companies stay with what works until they find themselves in a competitive disadvantage or a new technology comes along. I hope Nikon/Sony has put Canon at enough of a competitive disadvantage so that Canon improves their tech, but given the sales numbers, I am not sure that is the case. Going back to the automotive industry comparisons....anyone ever look under the hood of their car? At the core engine? Sure there are add-ons and turbochargers/etc..but most car makers core engine has been the same for decades...that is...until something new came along...like hybrids.

So, I am not holding my breath for new sensor tech from Canon. They do have to figure out how to get people to upgrade from the 5DIII to 5DIV and 1DX to 1DXII. But that is 2-3 years down the road. Sure the 7DII may (or may not) come out sometime in the next year...but Canon could easily drop a slightly improved sensor into it, give it the 1DX/5DIII AF and 10-12 fps and call it good. Granted, I am hoping for new sensor tech

 

[Eldar]

I may be one of the last Mohicans, since I have no interest in video, just stills. I have not once used the video funcionality of either the 5DIII or the 1DX. I firmly believe that photography represents an art form and a form of documenting our lives and history, that will survive, no matter how much video we get. I encourage you to check out Nick Brandt's photos from Africa (saw the exhibition last week). Video can never ever replace that.

An observation I have made over the last couple of years is that, after reading all the acclaim of the newer SLRs video functionality, those around me who buys cameras often talk a lot about video to begin with. But when I ask them to show what they have produced, they (almost) always show me their photographs.

One improvement I am looking forward to is the new live view AF. I use live view more and more, but up until now I have always used manual focus. I suppose that is a positive development for us still photographers, coming from Canon's focus on video.

 

[Loswr]

One improvement I am looking forward to is the new live view AF. I use live view more and more, but up until now I have always used manual focus. I suppose that is a positive development for us still photographers, coming from Canon's focus on video.

As I understand it, the advantage Dual Pixel AF brings to live view focusing is speed. It's not more accurate than contrast detect AF. In most circumstances where I used to live view, after carefully composing the shot, adjusting tilt and/or shift on the lens or positioning the flash heads for a macro shot, getting the polarization just right, etc., shaving a tiny bit of time off the focusing step is not really a significant benefit.

OTOH, for a camera like EOS M where live view focusing is the only option, PDAF would be a huge advantage.

 

[Don Haines]

Video is a must have feature in a DSLR for a lot of people. It is very hard to sell a camera without 1920x1080 video and people are starting to look for 4K video in cameras. For a company to compete, they have to offer video.

That said, how many of those people who MUST! have video use it? Myself, I tend to only use it for astrophotography and image stacking... My bet is that this is a must-have feature that is seldom used....

 

[RLPhoto]

Sony makes better sensors, but Canon makes better cameras.

 

[gbchriste]

I had a professor - a techie type - who coined an axiom that went something like this:

"All other factors being roughly equal, the most important technical factor in any system is cost."

Of course, we don't think of "cost" as a technical factor. But the point he was making as that the vast majority of rationale actors will not pay for an elevated level of technical performance if they don't need it.

One of his favorite cases in point was the Betamax vs. VHS video format. By any technical measure associated with image and audio quality, the Betamax was a far superior format, although it cost more. But for the average user, the ability to record 6 hours - or as many as 6 1-hour TV shows or 3 movies - on one cassette was more compelling that the difference in image and audio quality, and that extended record time could be had for less money.

Without getting in to a PC-vs-Mac flame war, I think there would be a very large number of people that consider the Mac a superior computing platform but won't switch because the cost difference between it and the PC overtakes any perceived benefit gained by adopting the Mac platform.

It's the same reason the vast majority of us drive a Chevy instead of a Mercedes. Our immediate need is for reliable transporation to get us back and forth from home to work. The Chevy meets that need, even though the Mercedes is clearly a far superior vehicle.

The bottom line is that "good enough" at a competitive cost will always win out over "far superior" at a higher price in a rational market.

In the present case, 5D3 is a more expensive consumer choice than the Nikon D800, but you have to look at it not from the consumer's vantage point, but from Canon's. For the present, the sensor technology they are using is "good enough" for them to continue to sell a lot of cameras and remain competitive.