Has Canon fallen hopelessly behind? (cross-posted)

[scooby70]

I'm completely sold on EVF/SLT for legacy SLR lens mounts and wouldn't consider another dSLR myself, but it seems that most photographers with lots of disposable income for purchasing aspirational photo gear are in love with optical viewfinders.

I actually prefer my EVF equipped camera for almost everything as it's WYSIWYG and I can have an in view histogram. Wonderful. There are only a few drawbacks that I can see...

There's a fraction of a second delay in the real world appearing in my EVF but it's not significant to me.

I find that the EVF allows focusing in lower light than an OVF but in almost total darkness the situation reverses as the EVF effectively blacks out whilst the OVF at least allows some outline to be seen.

In low light the EVF destroys night vision.

For most shooting though, I find my MFT set up is what I reach for rather than my 5D.

[hjulenissen]

Sony could release a FF 30FPS 36 MP camera for $1000 and the majority wouldn't consider purchasing it.

I might, but not because of the 30fps. I have no use for higher framerates.

It would have to have really good AF, ergonomy, image quality and the right line-up of lenses at the right price in order to make me switch brands.

On second though, I would be more likely to switch to a camera system that was more "open", i.e. 4/3 or something similar for 24x36mm.

I think this is, in large part, because serious photographers are typically over 40 or 50 and pretty well set in their ways.

That is probably part of the reason, but being heavily invested in lenses can make you as conservative as being heavily invested in knowledge.

If you use the tools as tools, and are primarily doing this for the images, then any change that might disrupt the flow of good images can be negative. It might pay off in the long run (due to revolutionary new features), but it might not.

-h

[Chris Pollock]

Chris, what adapter are you using? for the 14-24 on the Canon platform, is the one by Mark Welsh? or another.  How do you set aperture? 

I use the one from 16:9. You can read about their latest version here: http://www.16-9.net/nikon_g/

I got mine a few years ago, so it's an earlier version - probably version 2. They were having trouble producing them at the time, so it took months for them to fill my order.

I was surprised to see that on Nikon lenses the aperture is set via a mechanical linkage with the camera. The adaptor therefore doesn't need any electronics, just a little lever on the side which engages with the aperture mechanism on the back of lens. You just slide the lever to set the aperture. The are markings to indicate roughly where to slide it for F8 and F22. It's far from a precise way to set the aperture, but it does the job. Obviously it's manual focus only, and you don't get any EXIF information about the lens in your photos.

BTW, I agree with your findings on the MKIII, I purchased/returned one pretty quickly.

I'm going to hang onto mine. I've already sold my 5D Mark II, so it's too late to reconsider now. Besides, even with its mediocre sensor it's a worthwhile upgrade, which fixes most of my gripes about the 5D Mark II. You can't buy a better camera that you can attach a Canon lens to, with the possible exception of the much more expensive 1DX.

I think I will buy a D800 at some stage, once it's been out a while and the price comes down a bit. After 10 years of shooting with Canon, I'm curious to try the competition.

[Petrus]

The subject of these posts "Has Canon fallen hopelessly behind?" has started to annoy me a bit. I have shot with Canon pro DSLRs for 10 years now, because Nikon had fallen hopelessly behind in digital professional cameras. Now I am switching to D4 and D800e + a bagful of Nikon lenses. What happens in 2022? (except that I will be retired already, but with younger folks). I suppose you get my point. At least erase that "hopelessly" part...

[Ray]

Doesn't Ray own at least one Nikon dSLR?!

Indeed he does. For the past year I've been using almost exclusively a D700 and D7000 with two Nikon zooms, the 14-24/2.8 and 24-120/F4. BJL knows that I switched to Nikon some time ago, so I really don't know what he's trying to imply when he writes,

But I should give up: it is futile fighting "my imaginary camera is better than your real camera" hypotheticals: you are clearly in full-on brand-loyalty obfuscatory defense mode!

My arguments here have nothing to do with brand loyalty, and never have in the past on the subject of cameras. For me, cameras are merely tools. I buy them without the need to fondle them first to see if they fit in my hand. I consider myself flexible enough to be able to adjust to and accommodate the quirks of a different design and control layout, and I'm able to make rational purchasing decisions based upon information from reliable reviews, knowledge of basic specifications regarding price, features and weight, and a study of the results of tested performance from companies such as DXO.

[Les Sparks]

I think that the new Nikon D3200 presents a bigger challenge to Canon than the D800. Compare the two entry level cameras:
The D3200 has 24 MP and costs $699 at Amazon. The Canon Ti3 has 18 MP and costs $750 at Amazon. If you're looking to enter the DSLR game you're probably going to go with Nikon. Now over a year or so, you're going to have a reasonable investment in Nikon glass so when you're ready to upgrade to a new body, Canon is not likely to be a choice.

So, in my opinion, Nikon stands to gain a lot more new users over the next few years at the expense of Canon.

Les

[BJL]

Indeed he does. For the past year I've been using almost exclusively a D700 and D7000 with two Nikon zooms, the 14-24/2.8 and 24-120/F4.

Ray, that is actually a surprise: last I read, you used the Nikon 12-24, and thus one Nikon body, and Canon for the rest. When did you abandon your substantial collection of Canon gear (like the 100-400) for a relatively minimalistic two zoom lens kit?

[Ray]

Ray, that is actually a surprise: last I read, you used the Nikon 12-24, and thus one Nikon body, and Canon for the rest. When did you abandon your substantial collection of Canon gear (like the 100-400) for a relatively minimalistic two zoom lens kit?

BJL,
The Nikon D700 was the first Nikon camera I bought over a couple of years ago to use with my Nikkor 14-24. For a while I used this one Nikon camera and lens, plus my Canon D50 with 17-55/2.8 and occasionally with my 100-400.

However, a bit over a year ago, being amazed at the DR capabilities of the D7000, as shown on the DXOMark graphs, I started a thread entitled D7000 Dynamic Range in order to explore whether or not this extended DR was of real practical significance or just an engineering spec as some people had suggested.

As I recall you were a participant in that discussion and repeatedly made the point that the shot-noise floor would probably make such extreme claims of DR performance irrelevant, and that the DR of most cropped-format DSLRs was already limited by shot noise.

Not convinced this was the case, I took the plunge and bought a D7000 together with the new 24-120/F4 VR zoom. The first thing I did on receiving the camera was to compare its DR with some of my other cameras, including the D700 and 50D, and sure enough the DXO results were confirmed. I'm sure I posted the comparison images at the time; maybe on a different thread. I used Jonathan Wienke's Dynamic Range Test Chart, and I'm sure my methodology was sound.

I haven't got rid of my Canon gear. I'm very much lens-orientated. I bought the D700 primarily because of the availability of the Nikkor 14-24/2.8. The Canon 100-400 is a lens that suits me. It's reasonably light, flexible, and of reasonable quality, but like many people, I'd be very interested in an upgrade.

If Nikon were to develop its own 100-400/F5.6 VRIII which was demonstrated as being razor sharp at 400mm and F5.6 (which my Canon 100-400 isn't), I'd have little reason to retain my Canon gear. The D800 with 3 Nikkor lenses giving me a continuous focal length range from 14mm to 600mm, is all I need, 600mm being the equivalent FL with the D800 in DX mode, of course.

Time will tell whether or not Canon has the technology to deliver a camera equalling, and perhaps even exceeding the qualities of the D800. Perhaps the coming Photokina event this September will be too soon for a major announcement from Canon, but they must surely be pulling out all the stops.

[Chris Pollock]

I don't think it is feasible, at least for raw images. Technology is quite mature. Canon can reduce read noise at low ISO, but at high ISO your are running into limitations of physics. With JPEG the vendors can cheat a lot.

You're probably correct. I don't really expect noise-free images at 25,000 ISO any time soon - I used a little hyperbole for effect. My point is that I don't want to get rid of my Canon gear just because Nikon has the best sensor at the moment. Canon might come up with their own killer camera in a year or two.

[Alistair]

Am I the only one who feels Canon has taken it's eye off the ball by thinking all photographers want to really be cinematographers

Nemo

No, I think Canon took their eye off the ball in respect of their sensor technology. In respect of everything else, especially lenses, they are more than competitive.

For their DSLRs, Canon's strategy seems to centre on retention of in-house sensor Fab while Nikon has gone fab-less and use Sony's offerings which appear to be materially ahead of Canon's. Nikon then follow up with a very effective jab to the kidneys being highly aggressive pricing.   

Canon must respond and quickly because the Nikon offering is sufficiently compelling for a lot of people to swap systems, and that is quite a rare market advantage existing right now.

Either way it is a very fascinating case-study of both technology company strategy and human/consumer phsycology (have a quick look at the dp forums if you can stomach it - fascinating reactions).

[BJL]

I haven't got rid of my Canon gear. ...

Good; I am not losing my mind after all.

Time will tell whether or not Canon has the technology to deliver a camera equalling, and perhaps even exceeding the qualities of the D800. Perhaps the coming Photokina event this September will be too soon for a major announcement from Canon, but they must surely be pulling out all the stops.

Agreed all round: I am sure that Canon is pulling out the stops, and needs at some stage to deliver a significant step up in some aspects of its sensor technology, like its ability to handle scenes of very high subject brightness range (maximum dynamic range).  Also, given the established three to four year product cycle for 35mm format sensors and the fact that Canon (like Nikon) just replaced both of its previous generation of sensors in this format, I also agree that Photokina 2012 is way to soon. Nor is Canon going to turn around within months of finally shipping the 1D X and say that it was lying about that being _the_ new top of the line model.

[MatthewCromer]

Good; I am not losing my mind after all.Agreed all round: I am sure that Canon is pulling out the stops, and needs at some stage to deliver a significant step up in some aspects of its sensor technology, like its ability to handle scenes of very high subject brightness range (maximum dynamic range). 

DR is the key innovation IMO.

I took my Alpha 65 this evening to the local riverside to photograph the mountain laurel in bloom and captured a bunch of images with 10-11 stops of dynamic range in them.  After processing, they turned out pretty nice IMO.  I could not have done that with any of my previous cameras in a single frame (well, maybe the Shen Hao with Fuji NPS quickloads).  The shadows are a bit noisy -- the D800 could do a better job there (but a lot harder to nail the exposure exactly unless you use liveview).  Oh, and I was shooting handheld -- the auto horizon level was a delightful touch and the electronic first curtain shutter and complete lack of mirror slap as well as built-in stabilization all contributed to success.

BTW, to get all 11 stops to show up in the EVF you need to enable "DRO" as well as turn contrast down to -3 and use "settings effect -- on".  But it works!

I'm definitely looking forward to Sony's FF 36MP dSLR camera -- hope it comes out soon.

[Ray]

No, I think Canon took their eye off the ball in respect of their sensor technology. In respect of everything else, especially lenses, they are more than competitive.

For their DSLRs, Canon's strategy seems to centre on retention of in-house sensor Fab while Nikon has gone fab-less and use Sony's offerings which appear to be materially ahead of Canon's. Nikon then follow up with a very effective jab to the kidneys being highly aggressive pricing.   

Canon must respond and quickly because the Nikon offering is sufficiently compelling for a lot of people to swap systems, and that is quite a rare market advantage existing right now.

Either way it is a very fascinating case-study of both technology company strategy and human/consumer phsycology (have a quick look at the dp forums if you can stomach it - fascinating reactions).

I get the impression that Canon's R&D department is involved in all sorts of fascinating sensor development. I mentioned that I vaguely recalled many years ago reading a news item that Canon had succeeded in producing a 30mp sensor. Wayne Fox found a more recent news item reporting on a 200mp sensor developed by Canon, and just today whilst browsing the internet I came across the following report of an 8"x8" CMOS sensor that Canon has produced.

http://www.canon.com.au/About-Canon/News-Events/News-Press-Releases/Canon-develops-worlds-largest-CMOS

There are also details on the 'canonrumors' site of the very recent issuing of a patent to Canon for a back-illuminated sensor for APS-C and Full-Frame formats. (January this year)

http://www.canonrumors.com/forum/index.php?topic=2961.0

We should try to get things into perspective. Just a few years ago Nikon broke into the full-frame DSLR market with a relatively low resolution model (the D3) which had better high-ISO performance than any Canon model.

Well, it would have to have at least something better than the well-established competition, or there would be no point. How could you break into the market! But to get things into perspective, the D3 did not have lower noise or better DR at base ISO, only at higher-than-base ISOs. Canon still had the advantage of higher resolution with its 1Ds3 and 5D2.

Nikon later came out with the D3s which had even better high-ISO performance, but again not better performance at base ISO which was still about the same as the higher-resolving 5D2 and 1Ds3 models.

With the release of the D800, Nikon is now ahead in two major attributes; DR at low ISOs, and resolution, but the 5D3 is at least as good at ISO 1600 and above. In fact, at ISO 25,600 the 5D3 is very marginally ahead, according to DXOMark.

Now I have no doubt which camera I prefer. I happen to value both resolution and clean shadows. I already own a couple of good Nikon zoom lenses and my Epson printer is the 24" wide 7600.

However, I can understand that there are photographers for whom 22mp, a relatively fast frame rate, good high-ISO performance, a quiet shutter and accurate autofocussing, are sufficient.

Whilst its true that the D800 in DX mode using battery grip can also achieve the 6 fps of the 5D3, the 15mp DX image then loses all advantages but one, and that one advantage is better DR at low ISOs. In all other respects, the 5D3 image should be at least as good or better than the D800 in DX mode. SNR in the midtones is over one stop better across all ISOs. Resolution is also better of course, and DR at and above ISO 800 is also better by a degree which rises to a significant 1.5 EV at ISO 12,800.

If one needs to shoot fast action at high ISO, the 5D3 is superior to the D800. As always, the best tool for the job should apply.

[BJL]

I get the impression that Canon's R&D department is involved in all sorts of fascinating sensor development. ... Wayne Fox found a more recent news item reporting on a 200mp sensor developed by Canon, and just today whilst browsing the internet I came across the following report of an 8"x8" CMOS sensor that Canon has produced. ...

Unfortunately, none of those "extreme" sensors involve the sort of innovations that Canon needs in the DSLR market. They just adjust the size and number of photosites (the 120 MP [not 200] and the earlier 50 MP [not 30 MP] ones use compact camera photosites on APS-H sized sensors), without changing the fundamental design of the individual photosites or the subsequent signal transfer and processing. In short: no sign yet of doing the ADC in parallel on chip before the high speed transfer along the sensor's edge, which seems to be the next step that sensor design is taking, for the sake of reduced read noise combined with high frame rates.

Yet I am sure that Canon is also working on a next generation technology like an on-chip highly parallel ADC approach. By the way, there is a next step beyond column-parallel, which is ADC at each photosite, used in some exotic sensors for surveilance cameras. Those can use time-based measurement of when a highlight photosite gets full to greatly extend highlight handling. I mention this just as an example of what Canon (and others) might be working on.

[Petrus]

If one needs to shoot fast action at high ISO, the 5D3 is superior to the D800. As always, the best tool for the job should apply.

Those would be D4 and 1DMk4 and 1DX then...

[Ray]

Unfortunately, none of those "extreme" sensors involve the sort of innovations that Canon needs in the DSLR market.

Of course not. Secrecy is the name of the game. The high pixel count news items, whether 30 or 50 or 120, or 200, just indicate that Canon is continually developing new technology. Such announcements are essentially advertisements, but it's not difficult to imagine that in any R&D lab there will be all sorts of prototypes and possibilities that are being considered. Some will be shelved, and some will be given the nod from the board of directors for immediate development, and increased budget allocation will consequently be provided to bring such ideas to market.

[hjulenissen]

I get the impression that Canon's R&D department is involved in all sorts of fascinating sensor development. I mentioned that I vaguely recalled many years ago reading a news item that Canon had succeeded in producing a 30mp sensor. Wayne Fox found a more recent news item reporting on a 200mp sensor developed by Canon, and just today whilst browsing the internet I came across the following report of an 8"x8" CMOS sensor that Canon has produced.

http://www.canon.com.au/About-Canon/News-Events/News-Press-Releases/Canon-develops-worlds-largest-CMOS

There are also details on the 'canonrumors' site of the very recent issuing of a patent to Canon for a back-illuminated sensor for APS-C and Full-Frame formats. (January this year)

http://www.canonrumors.com/forum/index.php?topic=2961.0

I am sure that they are. Doing something for the heck of it, for defence/NASA, and doing something for a mass-market price-sensitive DSLR may be 3 different things.

We should try to get things into perspective. Just a few years ago Nikon broke into the full-frame DSLR market with a relatively low resolution model (the D3) which had better high-ISO performance than any Canon model.

I think that this figure is nice for perspective (dxo, brought to my attention by Guillermo)

[hjulenissen]

By the way, there is a next step beyond column-parallel, which is ADC at each photosite, used in some exotic sensors for surveilance cameras. Those can use time-based measurement of when a highlight photosite gets full to greatly extend highlight handling. I mention this just as an example of what Canon (and others) might be working on.

That would be interesting.

-h

[Chris Pollock]

Those would be D4 and 1DMk4 and 1DX then...

Cost is an issue for a lot of buyers.

[Ray]

I am sure that they are. Doing something for the heck of it, for defence/NASA, and doing something for a mass-market price-sensitive DSLR may be 3 different things.I think that this figure is nice for perspective (dxo, brought to my attention by Guillermo)

I thought we'd got beyond the DXO weighted and biased scores. Hasn't that been the cause of so much distrust and disbelief in DXO results?