New 7D mk ii, Samsung nx1, Sony a77ii

Discussion in 'Nikon' started by kylebybee, Oct 1, 2014.

  1. All these new bodies are exciting looking, all APS-C sensors, makes one wonder what Nikon is going to answer with and when and if. The thing that I like on all of these especially the Samsung is the touch screen focus ability, and ALL those focus points that they all have. I'm just marveling at the technology coming out.
     
  2. With my usual "I work for Samsung - I don't speak for them but might be subconsciously biased" disclaimer, I'm not all that convinced by touch screens for AF. They're generally not especially accurate, your finger is in the way of what you're trying to select, they're no use when using a finder, and prodding the camera is no way to get a stable shot. They rarely let you hold the camera steadily, either. I'll stick to my Nikon joystick, thanks. (I do have a GF2 with a touch screen, for reference; it's fine for menu access, it's just AF that doesn't convince me.)

    At least initially, Nikon seem to have answered these with the D750. Maybe they'll respond with a D7100 replacement at some point - it's obviously an old camera (though so was the D300s!) - or maybe they'll really assume they're getting nowhere with the high end of crop sensors against Canon and the mirrorless systems, and just go all in on full frame. I have a working theory that the Expeed 4 isn't vastly different from the Expeed 3 (see Wikipedia's observation that the 3 and 4 in the D4 and D4s have the same part number), and that - 1-series aside - all the current Expeed 4 bodies have roughly the same pixel rate. In the absence of Canon's trick of pairing up multiple processors, I suspect Nikon have technical limitations in trying to match either a 7D2 or a 1Dx for pixel pushing power. I'm sure they'll get a chip upgrade at some point (maybe for the D5), but they don't seem to be losing too much sleep going over the high frame rate market for now, 1-series aside. I have to expect that if Nikon had an easy way to match the A77, the original 7D or the 1Dx in frame rate, they'd have done so by now.

    Or I could be completely wrong, which is hardly unusual. If I'm not, I'll be interested to see whether the appearance of an Expeed 5 with a significant performance boost was enough to change the market positions that Nikon goes after.
     
  3. The touch screen AF on my OM-D EM-5 is really cool when taking pictures of a group of people or something. I love it.
    I love Nikon stuff, think their menus are the best, their ergonomics are top-shelf, but they lost me when they came out with such a weak (for me) mirrorless option.
    No D7300 or D400 or D9300 or whatever could have solved that, to be honest. I needed small and light and easy to carry around.
    Nikon DX mirrorless is what I wanted, but alas, it never happened. If it ever does, who knows, I may come back.
     
  4. Photographers that use their cameras to shoot wildlife, birds and sports don't have time to screw around with touch-screen AF. They need a fast, accurate, flexible AF that will manhandle a big super-telephoto lens when they start stacking teleconverters on their 500mm and longer lenses.
    If you don't shoot like that, then almost any APS-C sensor will do. If you need that, then the 7D MkII is the body to buy.
     
  5. Spearhead

    Spearhead Moderator

    touch screens for AF. They're generally not especially accurate, your finger is in the way of what you're trying to select, they're no use when using a finder, and prodding the camera is no way to get a stable shot.​

    I use a dSLR with touch screen AF. It's addicting how easy it is. It's easier to tap on a specific spot on the screen than it is to tap on a button in some phone apps. As long as one uses continuous focus, the camera stability is not an issue.
     
  6. I still think if the implementation of eye tracking (within the VF) could be perfected, it would be ideal....either with a 'follow/track' function and/or a 'look & lock' function.
    Nikon cannot have been so short-sighted as to not see the benefits of parallel processing, this is the 21st Century?
     
  7. I have some vaguely interesting ideas on the eye tracking front, but it's true that technology can do a bit better than the Eos 3 (though I'd still vaguely like one - for my couple of remaining Canon lenses - just for entertainment value). If we get much beyond the 51-point AF, I'd be on for an analogue joystick (Playstation-style) instead of the multi-way controller we've currently got - though the current one is good for selecting a single discrete point.

    Re. the touch screen, it probably takes me a short enough time to use the joystick with the 51 focus points that I wouldn't get the camera away from my eye and back that fast, let alone selecting on a touchscreen. If the devices is entirely touchscreen-controlled, I'd have my hands in the right place already, and be less bothered. I'm not blown away by this feature on a cell phone, but then I don't have high demands for AF on my cell phone anyway, especially when it comes to tracking. I suspect what's easiest depends on what you're comparing to and what you're controlling: selecting one of 51 points with a joystick is easy; selecting one of 41 focus points by trying to spin two dials (Eos 1v...) would be painful. Given a huge number of AF points for contrast-detect (or some on-chip phase detect) then I can see the joystick starts to be worse.
    Nikon cannot have been so short-sighted as to not see the benefits of parallel processing, this is the 21st Century?​
    I believe the phrase is "it's just a small matter of programming". This is easy to do if there's no interaction between pixels, and for reading data off the sensor there isn't, but for reconstruction (RGB conversion, JPEG output, video...) there is, at least to some extent. If the Expeed 4 really is just a respin of the 3, we might see Nikon do this in a future generation. Or - and this is depressingly often the reason that useful technology doesn't proliferate - Canon may own a patent on sensible ways to do this. Canon presumably built in a small amount of functionality to do this, allowing them to use the same chip across the product line and simply double up at the high end. There are some differences between what Nikon puts in a D4s and a D3300, but I could believe that the shipping numbers of the D4s, D810 etc. don't currently justify a custom ASIC just for them, and putting too much capability in would raise the price of the D3300 end of the range. So they may just have decided to take the hit.

    All of this is speculation, by the way. I know very little about the innards of Expeed; if I ever get time to do some hacking, I may learn more.
     
  8. The Expeed processor has multiple cores that run in parallel, and each code runs several instructions at the same time. How much more parallelism is really needed? I guess a dedicated processor for AF and other camera functions while the other processor works on the main image processing tasks even while the mirror is down and AF is operating could help achieve higher fps rates with high resolution.
    Nikon will no doubt respond to the 7D Mk II. However, to achieve high resolution, high fps rates, and AF that can work with high accuracy in the very short mirror down time at 8-10fps while delivering the expected 24MP resolution is not going to be easy to achieve. Without the 24MP I dont' think the camera would sell. Customers in the $1-$3k price classes of DSLRs have shown that they only care about other features as long as the image quality is better than or at least as good as the best. This includes resolution of course, and especially in the case of the high end DX category where the purpose of the camera for many people is to put as many pixels as possible on the subject, in the absence of, or lack of practicality of a longer lens. Canon took five years to work on the 7D Mk II. I think Nikon will respond in 2015 with a dedicated head to head competitor, as well as a D7100 follow up with lifted buffer restriction. I think a major AF system overhaul (increased number of focus points?) is not due until the D5 though, if we don't consider the Advanced I and II major overhauls, not to mention the smaller differences from camera to camera.
    Touch screen AF seems to be suited for very small cameras with few dedicated controls; with a wide angle or normal lens, I think it can work ok, indoors or in otherwise subdued light. In bright light, or when hand-holding a 200mm or 300mm lens, I would prefer a touch pad to move the AF point around and an OVF that gives a high dynamic range, real-time image even in bright light. A long lens is much easier to hold steady without fatigue when the camera is pressed against the photographer's face than held at arm's length.
    With regards to AF systems, I think the D7100's 51-point system works extremely well and is enjoyable to use. On FX DSLR cameras the focus point spread is narrower, which gives some margin of safety around the main subject but then the images often have to be reframed in post-processing, which is an additional step that is not needed as often on D7100.
     
  9. Ilkka Nissila said:
    Without the 24MP I dont' think the camera would sell.​
    When did 24MP become the "expected" resolution for an APS-C sensor? The 7D MkII already has resolution equivalent to 50MP in a full-frame camera. It's got the latest dual processors, new mirror technology and "reasonable" file sizes to get 10-fps. The 1D X has much courser pixels (fewer pixels) and delivers 12-fps. The answer to more responsive bodies isn't more pixels, at least not without the processing horsepower to keep up.
    Nikon never fully responded to the 7D, so why are they now going to suddenly respond to the 7D MkII?
    Since the 1D X, Canon has practiced true trickle-down benefit with its AF system, first taking it to the 5D MkIII and now the 7D MkII. Nikon needs to take its D4 AF technology, upgrade it and then put it in an $1800 body. They also need to update their super-telephoto lenses to reduce weight, improve IQ and improve stabilization. Canon seems to be catering to their pro sports photographers, but also bring the related benefits to serious sports and wildlife amateurs. Nikon seems focused on making pixel-counters happy, at the expense of speed.
     
  10. ShunCheung

    ShunCheung Administrator

    When did 24MP become the "expected" resolution for an APS-C sensor?​
    For Nikon, it was the D3200, introduced 2.5 years ago in April, 2012: http://www.photo.net/nikon-camera-forum/00aHt7
    Since the D3200, Nikon has not introduced any DX-format DSLR with fewer than 24MP, and for that matter, with more than 24MP. In fact, other than the D4, D4S, and Df, which are special-purpose DSLRs, 24MP is the minimum pixel count for all Nikon DSLRs introduced since the beginning of 2012.
    I think 24MP is an overkill for a consumer DSLR such as the D3200 and D3300, but for whatever reason, Nikon seems to believe that pixel count sells cameras so that 24M is now their bottom (except for specialty DSLRs).
    Nikon introduced the Multi-CAM 3500 AF module in 2007 and immediately put that into the D3 and D300, introduced simultaneously on August 23, 2007. Today, the top-of-the-line D4S as well as the $1200 D7100 (now with a discount) use that AF module design, with various minor improvements over the years.
    Needless to say, now 7 years later, that Multi-CAM 3500 design is dated, but I am afraid that we need to wait until Nikon introduces the D5 before we'll see major improvements with more cross-type AF points, which Canon has employed since 2012. I have no insider information, but I have little doubt that we'll see a D5 well ahead of the 2016 Rio Summer Olympics (August 2016).
     
  11. When did 24MP become the "expected" resolution for an APS-C sensor?​
    The short answer is when the camera companies discovered that the higher that number, the greater the interest from the public, regardless of whether it's actually better for them. All these companies are in business to make money, not design cameras.
    I wish "low light performance" had been the chief measuring stick. If so, we might have 6 and 12MP cameras that can take a photo in absolute darkness with no noise.
     
  12. So, now I understand why Nikon ran off the tracks with the sports/wildlife crowd.
     
  13. ShunCheung

    ShunCheung Administrator

    So, now I understand why Nikon ran off the tracks with the sports/wildlife crowd.​

    Wildlife, yes. However, while the D750 is not specifically designed for sports photography, it is going to be a very good alternative for people who cannot afford the D4/D4S. Its low light capability is a huge advantage for night and indoor sports. See this current discussion on DPReview: http://www.dpreview.com/forums/post/54483840
     
  14. The D750 is a full-frame camera, so 24MP is no big deal and it's not the category of camera we're talking about. It would be more comparable to Canon's 5D MkIII.
    The OP was talking about APS-C bodies.
     
  15. David, the 7D Mk II has a 1.6x crop so its 20MP is not too far from the pixel pitch of the 1.5x 24MP D7100. I believe that a large number of high end DX users specifically want it to be DX so that they can get the maximum subject detail using the longest lens that they have, and they can further often crop into the image. I do not think a lower resolution such as 16MP would be met with enthusiasm among people who are used to D7100 image quality but want a larger buffer and higher fps. They want to keep the image quality of the less expensive body, so that they can get the maximum (feather! ;-)) detail possible from their lenses. It is a little different for the case of the D4s which is primarily targeted for journalists who have the long lenses they need, and the subjects aren't especially small (they're typically human sized and often reside in dark places ;-)) and the images of whom mostly get printed on low quality media.
    Nikon never fully responded to the 7D
    Canon 7D has a 19-point AF point coverage vs the D300(s)/D7100 51 points (basically Nikon's best at the time) so in a way only the 7D Mk II fully responded and improved upon the original D300's AF system from 2007. Also dxo's overall score for the D7100, D300, and 7D sensors are 83, 70, and 66, so it could be said that Nikon was ahead the whole time in both image quality (signal to noise) and AF; in the D7100 Nikon also moved ahead in resolution but fps rate and buffer size of the D7100 are lower than 7D/D300s, which is the root of all this hubbab and outrage. In FX it was shown by the 5D Mk II vs. D700 that people valued resolution above all else, since the D700 was better in virtually every other aspect but the Canon nevertheless sold better. Nikon may have simply assumed that the same is true in DX, hence the D7100 instead of a 8fps 16MP model which they no doubt could have done in 2012 if it were for technology only, but they were hit by a series of natural disasters and a nuclear accident in the previous year, so their resources were limited and had to make choices.
    Since the 1D X, Canon has practiced true trickle-down benefit with its AF system, first taking it to the 5D MkIII and now the 7D MkII.
    Both manufacturers are introducing their best AF systems in the mid grade models (D750, 7DMk II) now, so they're not "trickled down" from 1D X / D4s. D4 and 1D X are old tech now.
    They also need to update their super-telephoto lenses to reduce weight, improve IQ and improve stabilization.
    Nikon are doing just that, with the 800/5.6 and 400/2.8 which incorporate fluorite elements for reduced weight and improved balance (their weight is now comparable to their Canon equivalents). Nikon have indicated that they are in the process of revising the whole lineup of fast supertele primes, though I'm only aware that the 400/2.8 800/5.6 patent also includes the 600/4 fluorite, which hasn't been brought out yet. Unfortunately both manufacturers' fast supertele prices are skyrocketing as they are made lighter. I am intentionally holding out and not buying the current 300/2.8 because I want to see the fluorite version and also hoping the tripod mount will be better. But if it's 8000€ then forget it, there's no way I would pay so much extra to reduce half a kg of weight. To me the rapid update cycles of the expensive superteles is a worrying phenomenon, as it means Nikon doesn't have to stock spare parts as long for old lenses either. 10 years after discontinuation and that's it. I would expect at least 20 year service life if I buy an expensive lens like that. So my enthuasiasm is dampened by servicability concerns.
    Nikon seems focused on making pixel-counters happy, at the expense of speed.
    Past history of camera sales suggests that customers in the $1000-$3000 price classes overwhelmingly prefer high image quality (including but not limited to high resolution) over high fps rate, which is what Nikon's 2012 generation responded to. High fps, if it comes with a compromise in image quality, is of interest to a small group of photographers only. Only if it is combined with the highest image quality will it be met by widespread enthusiasm.
    I'm not particularly counting pixels, but I do make prints, and sometimes need to frame a bit loose so that I can make sure I'm keeping the action in the frame. I also find that it is useful in e.g. figure skating photography to have enough pixels so that I can make a tight composition of a fast spinning athlete and yet when they extend their arms and legs, and slow down, I still include them without cutting their limbs. High resolution capture works well for this kind of situations especially with a prime that helps freeze the movement with its wide aperture. I find that when I'm shooting concerts outdoors, the best combination I have used is the D7100 with the 200/2 II, as it allows me to keep fast shutter speed and maintain excellent A2 print size image quality while blurring extraneous stage clutter. A 300/2.8 would let me use a smaller pixel density camera but the desire for longer and faster never ends if we go down that road. I used the 16MP D7000 before and find the D7100 a big improvement, mostly due to its better AF system but also the sensor is amazing when combined with a good lens, both low ISO as well as intermediate to (for DX) high ISO.
    If you're suggesting that the D300s successor should be lower pixel count, e.g. if it is 16MP then it would be no better image quality than cropping from the D810, so that would mean I have no interest in it. I think that boat (16MP DX) sailed a long time ago and today most people after a high end DX camera expect amazing detail instead of just less than one half of an FX camera's image area. In the past I didn't believe that 24MP DX made much sense, but this was mainly because lenses of that time, and AF systems back then could not deal with the challenge, or so I believed, but today both several lenses that I have and the best AF systems give good results on 24MP DX. Now the 24MP means a deep buffer will cost more money and processor has to be fast to be possible to work at increasing fps rates, but that's just the kind of tradeoff that need to be made to get the best image quality and performance in a given point in the technology development. The question is where to position the product in the performance vs. price curve, and how that will affect sales.
    I still make prints, in fact a lot of them and my whole editing process is based on the prints as I can then compare different images and see how they fit together to tell the story of e.g. a concert, or a wedding, or a sports event. I make adjustments to individual images based on how they feel together with other images. I only crop when it is the only way I can get the image, but I do use cropping when an important image can be salvaged that way. I have discovered that there are a new generation of shooters who think the purpose of high resolution cameras is so that they zoom into a small part the picture and ooh and aah at the detail. I am rolling my eyes when I see this as I feel as much of the image should be used as possible, to get good tonality and colours. But anyway, either which way, high resolution is appreciated and used by many today. Even if the seasoned professional understands that high resolution is not the be all and end all of the image, and that sometimes a fast response time is more important, I believe there aren't enough seasoned professionals who think like that to justify the production of a camera at a price point that requires a million units to be sold to make it profitable. I believe a 16MP DX at $1800 would be met by an extremely cold reception if made today. Thus, Nikon has to strive to keep the current 24MP DX in the new models and yet improve upon speed, to keep as many potential buyers as possible. The reason the D4s is 5500€ is not because it's expensive to make one, but because it is a special interest product made for a small group of people who need the specific combination of features. A 16MP DX 8-10fps camera would be the same, not of enough interest to the wider population to sell well (since the buyers would be divided between the higher resolution cameras that are less expensive, and the high speed camera, and usually in such a case they always buy the camera with the better image quality and lower price), so the price would have to be much higher than desired. This, is why I believe it doesn't exist yet; Nikon are unsure of the market's reaction to it if it doesn't meet the quality of lower priced models. Canon set the ball rolling with their 20MP 10fps model; usually Nikon are happy to make something that targets a Canon model either by going a little lower or a little higher in features and price. They will not be standing idle.
     
  16. ShunCheung

    ShunCheung Administrator

    David, my point is that the subjects in sports photography are people, who are relatively large. Therefore, an FX body will do fine and that is why I disagree with your point that Nikon is not delivering for non-pro sports photographers who cannot afford a D4/D4S. In fact, there are still plenty of used D3, D3S, and D4 to choose from.
    For wildlife photographers like me, with Canon introducing the 7D Mark II, Nikon is definitely behind. For smaller wildlife, including a lot of birds, DX clearly has an advantage. And by far the majority of wildlife photography is with sun light. Low-light capability is not as critical. I am relatively happy with the D7100, but at least based on the specs, it is no 7D Mark II.
    Meanwhile, as soon as people start comparing the D750 against the 5D Mark III, Canon is going to have a huge problem as their entry is 2.5 years older and $1000 more expensive.
     
  17. I don't have time to respond to all of Ilkka's Nikon-praise, but here are a couple of responses:
    David, the 7D Mk II has a 1.6x crop so its 20MP is not too far from the pixel pitch of the 1.5x 24MP D7100​
    I was responding to Shun's reference to the D750.
    Both manufacturers are introducing their best AF systems in the mid grade models (D750, 7DMk II) now, so they're not "trickled down" from 1D X / D4s. D4 and 1D X are old tech now.​
    The 1D X AF system predated the 5D MkIII and the 7D MkII, which use the same AF system, with only very minor differences. Thanks to higher battery voltage, the 1D X still handily bests the newer bodies at manhandling super-telephoto lenses combined with 2.0x TCs.
    Regarding lenses, I'm a wildlife guy, so the 500mm and 600mm lenses are my primary interest. Maybe Nikon will catch up in the next year or two.
    About MP, I print large landscapes, up at 72" on the long side, and my old 5D MkII gave great resolution at that size. That sensor is six-years old, but it's only retired from my lineup, because it couldn't keep up in the field, with slow AF and slow fps. If all I were doing was shooting landscapes, portraits and travel, it might still be active in my bag.
     
  18. Shun Cheung
    David, my point is that the subjects in sports photography are people, who are relatively large. Therefore, an FX body will do fine and that is why I disagree with your point that Nikon is not delivering for non-pro sports photographers who cannot afford a D4/D4S. In fact, there are still plenty of used D3, D3S, and D4 to choose from.
    For wildlife photographers like me, with Canon introducing the 7D Mark II, Nikon is definitely behind. For smaller wildlife, including a lot of birds, DX clearly has an advantage. And by far the majority of wildlife photography is with sun light. Low-light capability is not as critical. I am relatively happy with the D7100, but at least based on the specs, it is no 7D Mark II.
    Meanwhile, as soon as people start comparing the D750 against the 5D Mark III, Canon is going to have a huge problem as their entry is 2.5 years older and $1000 more expensive.​
    Okay Shun, I agree that for athletics, a full-frame is fine, usually, and in fact, for indoor photography high-ISO performance may be more critical, which leads you toward a body like the Canon 1D X instead of a crop sensor.
    My main shooting is also wildlife, but I find myself often shooting in not so great light, hand holding 500 to 1000mm and needing ISOs in the 800 to 1600 range to get enough shutter speed. This was my main problem with the old 7D is that it got noisy fast above ISO 800 and it could be ugly posterized noise. My fingers are crossed that the 7D MkII will improve on that substantially, so that I can use ISO 1600 with abandon.
    I agree about the cost difference of the 5D MkIII, but I don't think that Nikon has matched its AF system. It's only going to matter for things like birds-in-flight and some fast sports, but it still betters the Nikons that I've shot, but I haven't tried the D750 yet. I agree that new users might be inclined toward the D750, but then they might also consider the 6D.
    Us wildlife and bird guys, along with sports guys, have totally different priorities from the landscape, portrait and travel guys and gals. The uninformed, buying their first DSLR is very likely influenced by MP count. Once in a system, I don't believe in swapping teams, unless it's critical to making a living.
     
  19. (I turn my back for five minutes...)
    The Expeed processor has multiple cores that run in parallel, and each code runs several instructions at the same time. How much more parallelism is really needed?​
    The point was less about internal parallelism and more that Canon seems to be able to use SMP techniques to put a configurable amount of processing power in the system. They don't have to make a Digic that can go as fast as the 1Dx - they can make one that's fast enough for the 70D, but put two of them in the 1Dx and get better performance. Nikon don't seem to use the same approach with the Expeed, as far as I know. Even with whatever internal parallelism is exposed, there are additional issues when multiple ICs are required - such as synchronisation and cache snooping across the devices, avoiding memory conflicts, etc.
    When did 24MP become the "expected" resolution for an APS-C sensor?​
    In a true professional body (and remember that the D300 was always a prosumer body, as was the D700), I'm sure 24MP is irrelevant. Someone buying a camera solely for maximum frame rate, low noise performance and autofocus - a sports pro who'll never print more than a page size, can control their shooting position and who has access to big telephoto lenses - can do fine with much less. But the consumer Nikon cameras are all 24MP, excepting the Df (which is odd). There are still people who buy the D810 because it's the "best camera" based on the megapixel count. Honestly, it doesn't hurt to have it. Any camera priced in the range that might appeal to someone just wanting "the best Nikon DX body" and not interested in the technical details will have to deal with those who go through the spec lists looking for comparisons. How many would buy a "D400" at 16MP if it had a premium over a D7100? Not as many as if it was just the "best DX camera you could get, at everything".

    I'm not saying there isn't a valid market for a 16MP DX camera designed for frame rate. But it does make the marketing message more confusing. Nikon have enough of a problem with that already, especially in the FX space (hence the idea of "upgrading" from a D800 to a D750 and the ongoing complaints from people who want something that's universally better than a D700 but costs less than a D4), but I would be surprised if the - cheaper and likely populated by less experienced photographers, on average - DX market wasn't more sensitive to this.
    The 7D MkII already has resolution equivalent to 50MP in a full-frame camera.​
    Only if my D800 becomes a 72MP camera the moment I attach a TC14 to it. Except in very specific situations, this isn't a useful way of thinking about things. Though if anyone wants to swap their D810 for my (calculator rummage) 75MP 1 V1, I won't argue.
    The 1D X AF system predated the 5D MkIII and the 7D MkII, which use the same AF system, with only very minor differences.​
    Huh? I thought - and I could be wrong - that the 1Dx and 5D3 used essentially the same AF module, but the 1Dx supplemented it with colour tracking from the matrix meter (as do Nikon). 61 AF points, 41 cross type. The 7D2 has 65 AF points, all of which are cross-type, the centre one of which is "special" in terms of running at EV -3 and high precision. Nikon's multicam 3500 is 51 points, of which 15 are cross-type, and has been so since the D3/D300 - though more recent tweaks have given additional sensitivity and software improvements. I'm sure the actual hardware has had changes, but "very minor differences" applies more to that than what Canon are doing. But I agree that it would be weird from a marketing perspective for Nikon to ship a camera with better autofocus than the D4s before the "D5" is ready. Likewise, a DX camera that has higher resolution than the D4s and beats it on frame rate would raise some interesting marketing questions - and possibly result in some 300 f/2.8s being swapped for 200 f/2s.
     
  20. Incidentally...
    When did 24MP become the "expected" resolution for an APS-C sensor?​
    On the Canon forum, we get three posts in before someone complains that the 7D2 doesn't have 24MP (or even 22). Everyone wants everyone else's features. The NX-1 is interesting in that it offers both more resolution and a very high frame rate (though it doesn't have to flap a mirror, which helps), but I've yet to see a review of the image quality (and it's still missing more than a few lenses).
     
  21. Spearhead

    Spearhead Moderator

    Everyone wants everyone else's features.​

    Anyone who thinks 24MP instead of 20 is a big deal doesn't understand digital photography.
     
  22. Only if my D800 becomes a 72MP camera the moment I attach a TC14 to it. Except in very specific situations, this isn't a useful way of thinking about things. Though if anyone wants to swap their D810 for my (calculator rummage) 75MP 1 V1, I won't argue.​
    In focal-length limited situations, it is a useful way to think about the sensor. If the pixel-pitch (pixel-density) is exactly the same, then all that changes is the angle of view from the same lens. However, when pixel-density increases, then the effective reach of the lens increases, because more pixels are placed on that same area. Bird and wildlife photographers make use of this often, because they often shoot in focal-length limited situation, taking a shot with a 1,000mm and then cropping it further. Of course, there's often a tradeoff in noise when you make pixels smaller.
    Huh? I thought - and I could be wrong - that the 1Dx and 5D3 used essentially the same AF module, but the 1Dx supplemented it with colour tracking from the matrix meter (as do Nikon). 61 AF points, 41 cross type. The 7D2 has 65 AF points, all of which are cross-type, the centre one of which is "special" in terms of running at EV -3 and high precision.​
    You're right. I was commenting from the point of view of someone that hardly ever goes beyond single-point expanded. With the 1DX and the 5D3 I'd hoped that the glorious day had arrived that all I had to do is get 1 of 65 AF point on a bird flying by and the eye would be in focus. No such luck. However, you can bet on it that one of the first tests that I'll do when my 7D MkII arrives is try to get a bird's head in focus with 65-AF points.
    In most other situations, I can see where the 7D MkII's AF system would be considered more advanced.
    About marketing decisions about sensor sizes that serve niche sectors of photography, I think Canon has way more flexibility because of the multiple product line, including lots of non-photography related income. They can afford loss-leaders much more easily than Nikon. If you look only at unit-sales, then MP are going to be a bigger criteria, even if some niche doesn't want it or it actually damages functionality for a particular small niche.
     
  23. Too bad Samsung doesn't make the NX1 with an over-sized lens mount so users of other camera systems could use their other lenses (via an adapter with electronic contacts). I believe that Samsung has the electronics knowledge to do this. Of course, there might be some legal issues but at least Nikon users would have a cropped frame, 15 FPS body to put their expensive Nikon telephotos on. If Sigma can offer to change lens mounts on its latest lenses, I'm sure Samsung could come up with something.
     
  24. In focal-length limited situations, it is a useful way to think about the sensor. If the pixel-pitch (pixel-density) is exactly the same, then all that changes is the angle of view from the same lens. However, when pixel-density increases, then the effective reach of the lens increases, because more pixels are placed on that same area. Bird and wildlife photographers make use of this often, because they often shoot in focal-length limited situation, taking a shot with a 1,000mm and then cropping it further. Of course, there's often a tradeoff in noise when you make pixels smaller.​
    David: Yes, but exactly the same is true of teleconverters - you get more reach in return for reduced low-light performance and increased depth of field. I don't claim that DX doesn't make sense for those who always want long lenses - one can argue the merits of larger AF area coverage (vs density of focus points), we would be talking about a perfect 1.5x teleconverter, etc. (and, of course, the DX camera is smaller, cheaper and lighter). Equally, one could talk about the merits of a larger finder and the potentially larger dynamic range from larger sensor sites. I just don't think the phrase "equivalent in a full frame camera" is that informative unless you really want to consider a full-frame sensor as being equivalent to its DX crop, or a 5x4 film as equivalent to an APS roll of the same emulsion.
    This doesn't, of course, mean that a D4s is entirely a substitute for a "D400", at any price. But while I'm sure Nikon could make and sell an 11fps, 16MP "mini-D4", the D4s wouldn't be "worse" most of the time. And if we're comparing pixel density, the suggestion on this thread that Nikon could drop below 24MP doesn't help. Though I note that Canon's smaller crop sensor area does make the 7D2 roughly equivalent to a 24MP Nikon in terms of reach.
    Anyone who thinks 24MP instead of 20 is a big deal doesn't understand digital photography.​
    Absolutely. Though one might make the same argument about the frame rate advantages of Canon (albeit proportionately larger in the case of the 7D2). Still, if I was choosing one of two bodies that were different only by 10% in some area, I'd still err towards the slightly better one.
    Too bad Samsung doesn't make the NX1 with an over-sized lens mount so users of other camera systems could use their other lenses (via an adapter with electronic contacts).​
    Well, you can absolutely adapt (according to ebay) F mount lenses to NX mount - there's plenty of room. Uniquely, I think, among mirrorless digital mounts, there isn't room to adapt an M-mount lens, which is one reason I've always ignored NX (and yes, if I'd joined Samsung before the launch of the NX series, or if I had any input to the camera team, I would have pointed this out before launch - not that anyone listens to me!)
    As for Samsung producing an electronic mount adaptor for another system, I'm sure they'd rather take your money for new lenses (and yes, I know that the lenses often don't exist in NX mount, but you need to trade the benefits of getting access to a 400 f/2.8 against the number of lenses you don't sell of the types that you do make). If they were going to make an intelligent adaptor, EF would make more sense, because it's fully electronic and doesn't require a load of flapping and turning mechanicals (the same bits that make a manual adaptor easier for the F mount). Both Canon and Nikon keep their mount protocols private, and occasionally seem to break things deliberately (as Sigma find out). I can't imagine them wanting to license the use to Samsung, who would be competing with them. The Lens Rentals blog has had a few choice things to say about adaptor quality, which won't help.
    So technically possible? Yes. Sales sense? Probably not. Legal minefield? Indubitably.
    Perhaps we should at least confirm the image quality of the NX1 before we go further? If it's amazing, I'll be drooling with everyone else. Previous NX bodies have been slightly behind the state of the art in sensor quality, though the NX1's certainly has some innovations. And let's see if the AF performance keeps up with the DSLRs too.
    (Usual disclaimers for posts mentioning an NX camera: I work for Samsung, but I have no internal knowledge about these devices or the decisions behind its specifications, the above is all speculation, I've never seen an NX1, and nothing I say represents Samsung's official position. Sorry, I have to be careful about this kind of thing.)
     
  25. Andrew, since Nikon is primarily am optics company and Samsung is primarily an electronics company, I don't see why Nikon would complain since a Nikon mount NX1 body could sell a lot of Nikon lenses. Nikon doesn't feel there is consumer demand for a cropped frame, high speed, sports body so they should be OK with this. I suspect that Samsung already gets its glass and polishing done in Sendai so they only assemble in Korea anyway. My understanding is that due to the recent drop in camera/lens sales across several types of digital cameras (due to smart phones), there is a tremendous amount of unused optical glass and polishing capacity in Japan.
     
  26. David: Yes, but exactly the same is true of teleconverters - you get more reach in return for reduced low-light performance and increased depth of field. I don't claim that DX doesn't make sense for those who always want long lenses - one can argue the merits of larger AF area coverage (vs density of focus points), we would be talking about a perfect 1.5x teleconverter, etc. (and, of course, the DX camera is smaller, cheaper and lighter). Equally, one could talk about the merits of a larger finder and the potentially larger dynamic range from larger sensor sites. I just don't think the phrase "equivalent in a full frame camera" is that informative unless you really want to consider a full-frame sensor as being equivalent to its DX crop, or a 5x4 film as equivalent to an APS roll of the same emulsion.​
    Wonderful discussion Andrew, thanks for engaging.
    Yes, as a user of 1.4x and 2.0 time teleconverters, you do get more reach and reduced low light performance, not unlike with a crop-sensor body. It's a trade off or compromise, for sure. In good light, there's hardly any difference, EXCEPT the bodies ability to AF at f/8. For example, with a 2.0x TC-III on the Canon 5D MkIII, the AF lag can be up to one-second for ANY mistake in AF technique by the user. With the 1.4x TC-III on the 5D MkIII, the lag is only around .1-sec. That's a huge difference in practice. d
    If you're able to use the 1.6 crop of the 7D MkII, with a 1.4x TC and get an effective focal length that exceeds a 5D MkIII with a 2.0x TC-IIIk then you'll get much fast initial AF capture, assuming that the 7D MkII is at least as good as the 5D MkIII.
    With 1.5 and 1.4x teleconverters, there may not be much practical difference, but when you start using the 2.0x converters and get the native aperture down to f/8, then there's a big difference, at least on the Canon cameras below the 1D X.
     
  27. Thanks David (and you!)
    I agree that there's an autofocus difference. My take on that - which is not entirely qualified - is that the same AF module behaves like the same pixel density. In the FX bodies, the AF module covers much less of the frame - bad for positioning, but good for differentiating small parts of the image - compared with DX. Stick a teleconverter on and the AF module gets less light (and light from a narrower angle). If the FX AF module had the same coverage relative to the sensor size, the sensor sites would be larger - I suspect making up for the effective sensitivity loss caused by the teleconverter. Of course, the optics of the mirror box mean that the FX AF sensors can't cover such a large area as in DX. By symmetry, if the DX AF module increased in density to match that of the FX bodies, I'd expect it to lose low-light performance. But I may be vastly over-simplifying.
    I certainly agree that AF performance is important, though (it's one factor in my desire to upgrade from a D800e to a D810). Nikon, too, can't handle f/8 with most of the AF points. I guess we'll see how AF modules - and particularly on-sensor phase detect - evolve in the future.
    Michael: I agree about the history, but I don't know how much of Nikon's income comes from cameras compared with lenses. If Nikon really felt that way, I'm sure they'd be making more lenses for other mounts (which hasn't seemed appealing since the Canon rangefinder days). A mount deal that applied only to the NX1 and didn't allow competition with consumer DX DSLRs or the FX range seems a bit convoluted, though. I've no idea where Samsung's glass is ground, but given that Nikon do so much in China and given the bad history between Korea and Japan, I'd be a little surprised if that was where they chose to source. But I've no internal knowledge, and successful companies make practical decisions, so you may be right.
     
  28. Andrew, with Canon and its big super-telephotos when combined with teleconverters, it's not a crop vs. full-frame difference, but a battery voltage difference. The 1D X and the 5D MkIII have essentially the same AF systems, with only a difference in the x-type AF cells. They're both full-frame, but the 1D X has around 50% more battery voltage. With a 500mm plus 2.0x TC, the 1D X locks in after only .1-sec., where the 5D3 can take up to a full second in the same situation.

    The 7D MkII is equipped with a new battery, that's got about 10% more voltage and is backwards compatible, so I'm anxious to see if it narrowed the gap. Even if it hasn't, then I can use a 1.4x TC instead of a 2.0x to get similar effective focal length.
    At shorter focal lengths, such as 300mm and 70-200mm, you can notice the differences, but they're not near as extreme as with the 500 and 600mm lenses.
     
  29. Belatedly (sorry, away), thanks David, and interesting. Nikon also have higher voltages on their bigger bodies, and some lenses can make use of it. I'm not sure how relevant it is to the big superteles, though - practically, they're all using relatively lightweight internal focus groups anyway. I don't know that a TC should make a difference in the speed of an AF group, beyond requiring more sensitivity and accuracy. Ironically, my TC-16A (Nikon's autofocus 1.6x teleconverter for manual lenses) is actually very fast, because it doesn't have much to move. It can't really cover much of the focus range of a 500mm, though - you still have to get focus roughly right manually.
     
  30. Thanks Andrew, but I wasn't making that stuff up about the 5D MkIII taking a full second to acquire focus while the 1D X will do it in a tenth of a second. For my side-by-side comparison, I was using the latest EF 500mm f/4L IS-II and the EF 2.0x TC-III on relatively new 5D MkIII and 1D X. The difference was HUGE.
    Cash concerns kept me from buying a 1D X (I used a loaner from Canon for the comparison) so I've continued to struggle with my 2.0x converter on my 5D3. The slow down with the 1.4x TC is not near so bad, but there is a little slow down. With smaller lenses, such as my EF 70-200mm f/4L IS, there only a small slow down, that's easy to manage. The super-telephotos are where it becomes a big problem, particularly when trying to track birds in flight.
    Maybe it's a Canon-only issue. What happens when you put a 2.0x TC on your big Nikon?
     
  31. I believe you, David. I'm just not sure why it should be happening! I'm curious whether it's definitely the batteries, or if it's something in the AF system. I don't suppose the battery grip on the 5D offers a higher voltage? (The D700 - and probably D800 - does let you increase the voltage with the right batteries, so you can see whether some of the limitation on those "consumer" cameras is power based rather than AF based.)

    I'm assuming that question was to the group - I don't have a 2xTC, and my only "big Nikon" is an F5... but many reviews have suggested that the D810's AF seems to keep the D4s's quite honest. I don't know how many lenses have been tried.
     
  32. Thanks Andrew.
    My experience is consistent with other Canon users on birdphotographers.net. The battery grip does not offer extra voltage. It'll be interesting to see the 7D MkII, which has slightly higher voltage than the 5D MkIII and the battery is backwards-compatible.
    As for the AF of the D810, I don't know, but I've stood by Nikon users with the D800, shooting eagles in flight, and they were having trouble with both acquisition, using a bare 600mm, and buffer capacity. My friend with a D4 was having much better luck and keeping up with me easily, with his bare 600mm. The D810 must be a big step forward from the D800 in regards to AF and buffer capacity. (The files are fantastic, IF you get the subjects in focus and the camera lets you take the shot).
     
  33. With a 500mm lens, full frame camera, at f/4, with subject distance 10m, the depth of field is about 9cm. If a 2X TC is attached, the lens becomes a 1000mm f/8, and the depth of field at the same distance is reduced to about 4cm. Thus the lens needs to be focused more precisely to obtain satisfactory results consistently with the 2X TC attached. This increased precision can be achieved by slowing down the focus group movement so that the camera has time to process the AF sensor data and respond to it adequately. Another approach is if the processing power can be increased, run the AF motor at full speed and focus the data from the AF sensor more quickly, leading to the combination of high speed and high precision. I guess the dual processor architecture of the 1D X is responsible for this feat.
    With Nikons, I felt the TC-20 E III yielded high quality images with the 200/2 stopped down 1-2 stops, but autofocus of this combination (with the D800 body) was not precise enough for use for sailing and aviation photography. With the TC-14 E II the autofocus was very precise, at all distances, however. I experienced similar problem in my brief testing of the 200-400II, i.e. at close distances good results were obtained using the 2X TC, but a greater degree of variability in focus was obtained at longer (30m+) distances. With the 200/2 I know it's not about the optical quality of the combination since I got very good results when using the lens with the 2X TC for landscape photography, with live view focusing. My conclusion was that the 2X TC isn't very useful to me. Currently I use the TC-14 E III with D810 and it works well but I cannot test the 2X with the new camera since I no longer own that TC.
     
  34. David: Yes, that's one reason I want to upgrade to a D810. :)

    Ilkka: Yes, I can see why AF would be slower with the teleconverter, just not why body voltage should be an issue. (If you need to run the AF more slowly over a teleconverter, if anything a higher voltage should help more in the non-teleconverted case.) I completely believe that the speed of the AF electronics may be involved. So I'm in no way disputing what David is saying, I'm just trying to understand the mechanism. And I'll take your advice about a 2xTC on the 200 f/2 as a way to persuade my wife that I really need a 400 f/2.8. :)
     
  35. Just for the record (I know this is a Nikon forum, but a Canon is in the thread title), the Canon 1D X, the 5D MkIII and, hopefully, the 7D MKII, all have excellent AF accuracy with the EF 2.0x TC-III on Canon's 500mm and 600mm super-telephoto lenses, either Series I or II. I'm addicted to mine and only curse it when trying to keep AF on birds in flight while the AF hunts around for a subject.
    [​IMG]
    Canon 5D MkIII, EF 500mm f/4L IS, EF 2.0x TC-III, hand held.
     

Share This Page