Has Pentax trumped Nikon crop sensor cameras at high ISO?

<blockquote>

<p>Jeff may be confusing some of Laurentiu's comments with my own.</p>

</blockquote>

<p>That must be it. I never owned a FF digital camera - I started with Pentax APS-C and then moved to MFT. I only stated that I am interested in purchasing a FF MILC rather than an APS-C one (when the Fuji cameras were mentioned). And that is simply because I have many more FF lenses than APS-C ones. If I wouldn't have those FF lenses, I wouldn't have any interest in FF either.<br>

<br>

To see why I am perfectly happy with MFT, just take another look at the comparison from the first post:<br>

<a href="http://www.dpreview.com/reviews/nikon-d7100/22">http://www.dpreview.com/reviews/nikon-d7100/22</a></p>

<p>But this time change the cameras used in the ISO 6400 comparison:<br>

D5200->D800<br />A77->E-M5<br />K-5IIs->K-5II<br>

I changed the K-5IIs just to eliminate the advantage of removing the AA filter. Looks like Pentax is actually using a pretty strong AA filter in their K-5II because I find the details better captured by the other cameras.</p>

<p>Now, with those comparison changes I don't see that much difference between the three formats and I am pixel peeping. And I am doing that at ISO 6400, a setting I use maybe once a year when I want to have fun. I rarely need more than ISO 800. Sensor technology is just so good these days that I don't look forward to the next generation of sensors anymore.</p>

<blockquote>

<p>That's a <em>pixel density</em> advantage, Laurentiu - not an inherent FF advantage.</p>

</blockquote>

<p>I never said it was a FF advantage. It is just that until recently FF sensors did not offer the same pixel density as smaller ones, so companies could pitch APS-C/MFT sensors for bird photography and claim that their reason of existing is to give you more "reach". But once you offer similar pixel pitch in different formats, what advantage are you left with? And I think Sony now has the technology to produce large sensors with the same pixel pitch as smaller ones - give them a few more years and that technology will become more and more affordable.</p>

<p>At least MFT offers an alternative system with smaller lenses and some smaller bodies. But APS-C has never got to that point - Fuji may take it there, but otherwise, the gain for APS-C has mainly come in the form of slightly smaller (and less expensive) camera bodies - and even there I assume the problem is that Canon and Nikon likes to make their cameras big rather than that they cannot be made smaller. Pentax has been dedicated to APS-C for years now but for reasons related to their economical state and to their clarity of vision, they failed to produce a consistent APS-C lens lineup. Their cameras are weather sealed, but their nice Limited prime lenses are not and they are slow to boot (to make them more compact). The APS-C zooms are OK, not amazing, and rather expensive. The result is an incomplete system that can be very appealing in some areas and very puzzling in others. And with the move from DSLRs to MILCs, I am not sure what will happen to current Pentax technology - I don't see much future in their K mount.</p>

<p>APS-C never took it off as a standalone format - lenses like Sigma's 18-35/1.8 should have been developed years ago. Instead of that, APS-C was crippled to not compete against FF. And now the market is changing drastically and it looks like it's shrinking faster than I would have imagined. That compact cameras are disappearing should not surprise anyone, but when <a href="http://www.dpreview.com/news/2013/08/01/camera-shipments-2013-CIPA">DSLR and MILC shipments <strong>each</strong> decrease by 18%</a>, that is a sign of big changes. So I am wondering: with a shrinking market and entry level users moving away from dedicated devices and new technology developments, will the APS-C/FF tiering that Canon/Nikon/Sony practiced still continue to make sense? And if not, where will APS-C fall? I don't have an answer, but I think the question is worth asking. And the same question is worth asking about MFT, of course, but there I feel that they carved a better niche by actually achieving compactness and now with the video offerings (Panasonic and Blackmagic).</p>

<blockquote>

<p>More pixels <em>does not </em>mean more noise, Lannie - far from it.</p>

</blockquote>

<p>This is one of the interesting observations made by dxomark some time ago: <a href="http://www.dxomark.com/index.php/Publications/DxOMark-Insights/More-pixels-offset-noise!">Contrary to conventional wisdom, higher resolution actually compensates for noise</a>.</p>

<blockquote>

<p>Here is where I and Laurentiu begin to part company. I think that crop sensor cameras have a great future. I have always insisted upon full-frame for low light applications, but I am beginning to see that crop sensor cameras just might be the best all-around format, if one must have only one. </p>

</blockquote>

<p>I am not sure if we're disagreeing. I do not expect the future to be FF. Where we might disagree is on which crop format will be more popular over the next decade.<br>

<br>

APS-C has been around, but it is closer to MFT (0.7Ev) in performance than to FF (1.3Ev). The main advantage it has is that it's been used by several companies so far. The main disadvantage is that each company offers its own APS-C system and even the sensor size is slightly different. Also, APS-C cameras were always designed to offer the upgrade option to FF, rather than being designed from scratch around the smaller sensor size, to take advantage of it. The only system I know that was designed from scratch around an APS-C sensor was Fuji's. Sony E-mount was meant to host a FF sensor from the beginning - it already did so in a video camera.<br>

<br>

On the other hand, MFT is relatively new - that and the lack of upgrade path to FF are its main disadvantages. But, it also has some important advantages: it is not just a sensor size, it is a standard mount around it that was built with no consideration for other sensor sizes. It is supported by two companies making cameras, more if you count video cameras and there were rumors of some other photo cameras planning to share the mount. Only the Pentax K mount got shared like this. And then the system has great wide support from lens manufacturers as well.<br>

<br>

Jeff had shared an article earlier that was putting much weight on Zeiss supporting NEX/X systems vs MFT. But that was just a common sense decision on Zeiss's part. They already have a partnership with Sony and Sony's system is lacking good lenses. Zeiss can easily share the optical designs between the two, so why not sell those designs for Fuji as well (even though they overlap with Fuji's own offerings rather than filling gaps). On the other hand, producing lenses for MFT does not make much commercial sense to Zeiss - they cannot reuse the work they do for NEX and they would face stiffer competition from Olympus/Leica/Voigtlander. Simple choice for them, really. Start with NEX, get some profits, then reevaluate the situation later if necessary.</p>

<p>And now look at one more thing about Sony. They have the A mount cameras, the APS-C E-mount cameras and the rumored new FF E-mount camera. How exactly are they going to develop all these if they do come with a FF E-mount camera? I'll bet they'll want the support of Zeiss on FF more than they want it for the APS-C cameras. They'll probably scratch the A mount ones like Olympus is finally doing with FT and APS-C E-mount will get back to being offered as a low price alternative to FF bodies with few lenses specially designed for it. There won't be any advantage in choosing APS-C over FF other than price and the upgrade path.</p>

<blockquote>

<p>It's not a pixel density advantage, Keith. It's a format advantage.</p>

 

</blockquote>

<p>It's not, Q.G - at best it's a format <em>and </em>pixel sensity advantage, but it is simply not the case (as Laurentiu seems to be suggesting, given that the whole thrust of his posts this far have been in favour of FF over crop) that in and of itself, cropping a full frame image will inherently provide all of the advantages that we APS-C users recognise as benefits of that sensor format.</p>

<p>You <em>will not </em>get that benefit unless the sensor <em>also</em> has a heap of pixels to crop from - and in that regard, being FF is no more relevant than being MF or any other larger format. The sensor can be as big as the side of a house, but without a high pixel count <em>too</em>, cropping into images from it isn't going to give you the pixels per duck that crop sensors provide...</p>

<p><em>Ergo, </em>it's not a format advantage.</p>

What i read is that Laurentiu is saying that when you crop a FF to APS-C, you get the same resolution, as in <i>"You can crop a D800 image and still maintain as much resolution as you would get from an APS-C camera."</i><br>And you saying that that is not an format advantage.<br><br>It indeed <i>is</i> a format advantage. The D800 has slightly higher pixel density than the Canon 7D. But even if we ignore that, assume that pixel density is the same, the simple truth is that it's nothing but the larger format that provides an advantage. It being your <i>"more pixels per duck"</i>.<br>Your response, <i>"That's a pixel density advantage, Laurentiu - not an inherent FF advantage."</i> is incorrect.<br><br>But true, if we no longer ignore that the FF D800 also (!) has a slightly higher pixel density, what Laurentiu said has to be adjusted: you do not get the same resolution when you crop a D800 image to APS-C size, you get even more resolution. <i>More "more pixels per duck"</i> The result of both (!) a format and a pixel density advantage.<br><br>Now you obviously like many other aspects of APS-C over those of FF cameras. That, of course, is up to you. But what you said there just wasn't right.<br><br>Now of course you will not get that benefit if the FF cameras has less pixels than the APS-C camera. Duh... Just like you do not get the benefits being rich entails when you're not, when you are poor... Etcetera. So what are you trying to tell us here? Sure being poor doesn't bring you the benefits that being rich entails. Assuming (!) that a camera, a FF one (not that that matters) has fewer pixels when cropped down to APS-C format than a native PS-C format camera, you're absolutely right. But why assume?<br>Forgetting about comparisons set up to show that a big mouse can appear bigger than a small elephant, and furthermore considering that the FF in question is a D800, the APS-C a 7D, what you say just isn't correct.

<blockquote>

<p>The sensor can be as big as the side of a house, but without a high pixel count<em>too</em>, cropping into images from it isn't going to give you the pixels per duck that crop sensors provide...</p>

</blockquote>

<p>Right, but this is exactly what is becoming possible now. There are technical challenges to making large sensors. Few companies even make MF sensors. There was an interview with dxomark that I cannot find anymore, but it had an interesting response where the dxomark guy said that if they could only get the same efficiency in large sensors that they get from tiny P&S sensors, the IQ would be oh, so amazing. There was also this perception that high pixel density equates more noise and a lot of people still think about it that way. But technology changes pretty fast and I think in a few years we will see more cameras like the D800. </p>

<blockquote>

<p>you do not get the same resolution when you crop a D800 image to APS-C size, you get even more resolution</p>

</blockquote>

<p>It depends on what APS-C sensor you use as reference.<br>

<br>

The ratio of the surface of FF to Nikon/Pentax/Sony/Samsung/Fuji APS-C is 860/370 = 2.32 (for Canon APS-C it is 860/329 = 2.61). Reference - <a href="http://en.wikipedia.org/wiki/Image_sensor_format#Table_of_sensor_formats_and_sizes">wikipedia sensor dimensions table</a>.<br>

<br>

That means that the D800 would give you the same resolution (when cropped) as a 36.3/2.32 = 15.6MP Nikon/Pentax/Sony/Samsung/Fuji APS-C camera (or a 36.3/.61 = 13.9MP Canon APS-C camera).<br>

<br>

One could argue that the 24MP Sony APS-C sensors or 18MP Canon APS-C sensors still offer some advantage, but to me that is becoming an academic argument because I was perfectly happy with 15MP on my APS-C cameras. And now the question becomes: should I get an APS-C camera or should I pay a bit more and get a camera that works in both cropped and non-cropped modes? To me (and I realize that is just me) the latter option makes more sense - even financially. And if you are not like me and you really care about the "reach" argument, then why not go for MFT from the beginning? A 15MP MFT sensor gives as much "reach" as a 24.6MP APS-C sensor (or 21.93MP Canon APS-C). To me this shows that people choosing APS-C over MFT do not do it for the reach argument alone. I really can't tell why anyone would want to invest in APS-C when it can be fairly well emulated by a cropped mode of a FF camera or challenged in "reach" by a smaller MFT system. The only practical explanation I have is the availability of the upgrade path to FF - you start with APS-C because it is more affordable and eventually replace it with FF.</p>

It does indeed depend on what camera you compare to what other camera. That was a large part of the point i was trying to make.<br>Comparing the D800 to a Canon 7D, what we get is higher pixel density in the D800. So no matter how many more pixels the D800 has outside an APS-C size crop, it has more pixels in that APS-C crop than the Canon. More pixels per duck.<br>That is the pixel density advantage. The format advantage is that the D800 offers quite a bit more pixels outside the APS-C crop. If you use them all (and why wouldn't anyone?), even yet more "more pixels per duck".<br><br>And then there is physics. Not all pixels are created equal. Pixel count is not the begin all end all of image quality. Big pixeled, lower pixel count cameras may do better than high pixel density high pixel count cameras. Or not, depending on how well the things are made, and (!) what the processor inside the camera does to it (don't even imagine there is something as raw "raw". Everything (!) that comes out of a digital camera is precooked. The difference between CCD and CMOS, for instance, is one in how the charge is converted into an anolog signal, and that in turn into a digital signal: "cooking styles").<br><br>Which camera (system) is the correct for whom is something i doubt we can decide in a discussion.<br>It would be even harder if that discussion misrepresents the things discussed.<br><br>

<blockquote>

<p>I really can't tell why anyone would want to invest in APS-C when it can be fairly well emulated by a cropped mode of a FF camera or challenged in "reach" by a smaller MFT system. The only practical explanation I have is the availability of the upgrade path to FF - you start with APS-C because it is more affordable and eventually replace it with FF.</p>

</blockquote>

<p>The price of "full frame" cameras was placed out of proportion to those of "crop format" cameras because the former were rarer, the market was less competitive and they were placed as premium products. The price difference between the D300 & D700 was hardly accounted for by the difference a bigger sensor made to the camera itself and they are very similarly built. There was simply so much more competition for customers in the smaller sensor market and smaller profit margins. These small margins antagonise the big two, hence no new "pro" crop-only cameras.<br /> Pentax does have an opportunity there, but even though they put out a credible crop body they just don't seem to have the size or decisive direction to put out really convincing lenses and pull the market away from the big two. The K5II has advantages over the D7100, but not so much that people would switch to it with their lens choice in mind.</p>

<blockquote>

<p>[Pentax] just don't seem to have the size or decisive direction to put out really convincing lenses</p>

 

</blockquote>

<p>This is absolutely true. I simply can't find Pentax lenses that are getting rave reviews. I once took the kit lens off my Canon Rebel T2i and replaced it with the EF 24-70mm f/2.8L--and then looked at the screen in awe. Put the best lenses on the better Pentax bodies, and there could be some converts, lots of converts.</p>

<p>--Lannie</p>

<blockquote>

<p>Comparing the D800 to a Canon 7D, what we get is higher pixel density in the D800. </p>

</blockquote>

<p>That is not true - have a look at my calculations. The 7D has 18MP and the D800 can emulate a ~14MP 1.6crop APS-C camera, so it has *less* pixel density than the 7D. Or did I get some numbers wrong?</p>

<blockquote>

<p>The price of "full frame" cameras was placed out of proportion to those of "crop format" cameras because the former were rarer, the market was less competitive and they were placed as premium products.</p>

</blockquote>

<p>Yep.</p>

<blockquote>

<p>The price difference between the D300 & D700 was hardly accounted for by the difference a bigger sensor made to the camera itself and they are very similarly built.</p>

</blockquote>

<p>I think someone mentioned that sensor alone accounts for around $200 or so (unreliable memory here). And you would need a more powerful processing system to manage 40MP vs 20MP images. But all in all, the difference should be probably less than $500 for equally featured cameras. But FF cameras always had a lot more features than APS-C ones, so price difference tended to be above $1500.</p>

<blockquote>

<p>Pentax does have an opportunity there, but even though they put out a credible crop body they just don't seem to have the size or decisive direction to put out really convincing lenses and pull the market away from the big two.</p>

</blockquote>

<p>I had hope that with Ricoh having deeper pockets than Hoya and more interest in imaging (Hoya acquired Pentax to get their medical division), they would turn around the company faster, but so far they have only issued incremental product updates - several K camera bodies and a couple of Q cameras, but no new lenses at all. Even the development of their GXR is stalled. I'd like to think they are working on some big thing in secret, but the interviews with the Pentax execs never gave me a good vibe - I felt more like they had no idea what is going on. Their current approach is too conservative to get them anywhere. The GR is the only nice new product they released since they acquired Pentax.</p>

<blockquote>

<p>I simply can't find Pentax lenses that are getting rave reviews. </p>

</blockquote>

<p>Oh, there are some - all three FA Limiteds have been highly regarded since they were introduced: FA 31/1.8, FA 43/1.9, and FA 77/1.8. The FA 43/1.8 was even produced for M mount. <a href="http://www.luminous-landscape.com/columns/sm-02-05-02.shtml">Popular Photography called the FA 31 one of the best AF lenses they tested</a>. That rave review link mentions some other good Pentax lenses too. I had the FA 31 and you could press the shutter all day just to see what comes out.</p>

<p>Some of the newer budget APS-C lenses seem to be quite good too. They introduced a 35/2.4 (2010) and a 50/1.8 (2012). These filled a hole by providing some affordable primes at prices around $200 (even less now with some discounts). But the pace of their introduction is too slow.</p>

<p>My 35 mm f/2.8 macro also tested extremely well and is an extraordinary lens, one of the 2 or 3 best that I own. I think that previous comments about Pentax not pursuing the market effectively are absolutely correct, and I hope that they can make a new start under Ricoh. But I also think that some photographers don't get into Pentax because of a perceived lack of a lens that they probably don't need or wouldn't buy anyway. Since I am a working pro, when I buy a camera or lens it has to help me make money. I don't do sports or wildlife, so the lack of 300,400,600 f/2.8 lenses and a body that shoots 10 frames per second doesn't bother me. For other people, these would be features that they absolutely have to have to do their jobs or to enjoy their hobby, but I think that some amateurs who could be perfectly happy with Pentax have passed up a good system because of the lack of a lens or two they would never buy.</p>

<i>"That is not true - have a look at my calculations. The 7D has 18MP and the D800 can emulate a ~14MP 1.6crop APS-C camera, so it has *less* pixel density than the 7D. Or did I get some numbers wrong?"</i><br><br>You're right. The D800 gets 7360 pixels out of a 35.9 mm wide area. That's 205 pixels per mm. The 7D gets 5184 pixels out of 22.3 mm, which is 232 pixels per mm. About 1.1x more per mm.<br>That still doesn't tally with 18 vs 14 MP. The D800 crops to 24x16 mm, not 22.3x14.9. And it then produces 15 MP.<br><br>But does it matter?<br>Comparing this FF to an APS-C shows that the difference in quality at cropped format is negligible.<br>But who on earth would want to crop a FF to DX and judge the FF at what it can then deliver? Only if you don't have FF lenses you would resort to throwing away the main reason why you would buy a FF. What you would be judging then is the ability to use a camera the way it was meant to. Not an advantage of either one over the other.<br>The FF, Keith, delivers "more pixels per duck" unless you decide to limit it to equal or less pixels per duck. That's then something a photographer does. Not the camera. It's then also a personal trait of the photographer you would be judging. Not the quality of a tool that person has in his hands.<br><br>You pay more for FF. Indeed - as you mention, Laurentiu - because you get more. If you need it and can pay for it, do. Else don't. But what's there to discuss?

<blockquote>

<p>What i read is that Laurentiu is saying that when you crop a FF to APS-C, you get the same resolution, as in "You can crop a D800 image and still maintain as much resolution as you would get from an APS-C camera."</p>

</blockquote>

<p>I agree - that's what he's saying.</p>

<blockquote>

<p><br />And you saying that that is not an format advantage.</p>

</blockquote>

<p>I am.</p>

<blockquote>

<p>It indeed is a format advantage.</p>

</blockquote>

<p>It <em>is not. </em><br /> <br /> I'll say it again - you can have as big a sensor as you like, but <em>unless it is also a densely-packed sensor</em>, cropping into it will not give you "an APS-C image" in all but name.</p>

<p>A little thought experiment: imagine a FF sensor with <strong>one</strong> pixel, the entire size of the sensor.</p>

<p>How do you think that's going to look in resolution terms, cropped into, compared to an image from an APS-C sensor?</p>

<p>Not much detail or "reach" coming from that crop. And yet it's still a crop from a FF sensor, isn't it?</p>

<p>No <em>intrinsic</em> format advantage there, then... It's <em>pixel density alone</em> that lends the bigger sensor the ability to reproduce the reach/resolution advantages of crop sensors - not enough pixels, no advantage, even though it's still FF.</p>

<p>The larger format does not, in itself, provide the advantage. The larger sensor is just a place to put the pixels, but it's the pixels that provide the benefits.</p>

<p>Here's a clumsy but essentially correct analogy; a bigger car <em>might</em> be faster than a smaller car, but it's not the larger size of the vehicle that makes it faster, it's the engine under the bonnet (hood); well, for "car" read "sensor", and for "engine size" read "pixel count".</p>

<p><em>The size of the container isn't the crucial factor.</em></p>

<blockquote>

<p>so the lack of 300,400,600 f/2.8 lenses and a body that shoots 10 frames per second doesn't bother me</p>

</blockquote>

<p>That is not bothering a lot of Pentax users either, but that lack (together with the lack of professional support) also doesn't attract the pros that could help improve the brand visibility (a strategy that worked for Nikon and Canon but that Pentax never seems to have considered). The main ask from users (at least of pentaxforums users) these days is for a FF camera given those great FF FA Limiteds. It ended up being a joke on pentaxforums that every thread must make some mention to FF at some point or another.</p>

<p>The most puzzling omission is a 24mm that would offer the classical 35mm perspective. There is a 21/3.2, but that is closer to a 28mm and is rather slow too. There are also no modern extension rings (with SDM contacts - for their relatively new in-lens focusing motor) nor a 1.4x teleconverter. You can fill some gaps through Sigma, which ironically makes more lenses for Pentax K mount than Pentax does. But third party support is also weaker. Many interesting Sigma or Tamron lenses are not issued for Pentax (Sigma 150 macro, Sigma/Tamron 180 macro, the newer Tamron 70-300 VC). And as I mentioned earlier, Cosina and Zeiss decided to abandon the mount entirely some years ago.</p>

<p>The most attractive part of the lens lineup may be the set of DA* weather resistant lenses that go well with the high end bodies: 16-50/2.8, 50-135/2.8, 55/1.4, 100/2.8 macro (a D FA, not a DA*, but same properties), 200/2.8, 300/4 - these offer an attractive value that probably cannot be matched easily by Canon/Nikon. The Pentax weather sealing is pretty good too - I collected links demonstrating that <a href="http://laurphoto.blogspot.com/2010/11/pentax-weather-sealing.html">here</a>. But these lenses got up to a slow start because of rumors of SDM failure with early models (all DA* use SDM). The problem seems to have been fixed, but with Pentax never acknowledging it and their fixing of it, the doubts around SDM continued and still linger.</p>

<p>And Pentax also wasted effort on some recent products: K-01 was pointless. Q was interesting but releasing it with a 1/2.3" sensor when compacts started to be issued with 1/1.7" ones did not help. Now they finally came up with a 1/1.7" model just when Nokia is releasing the 1020 with a 1/1.5" sensor - they seem to be a couple of years behind the times. I also wonder about the point of developing the 560/5.6 lens - a 300/2.8 and a couple of teleconverters would probably have had more market appeal.</p>

<p>Their best success was the work they did with the 645D and the K-7, which shared a lot of technology among them. And then putting a Sony sensor in the K-7 resulted in the K-5. Some people were speculating that the experience with the noisier Samsung sensors is what enabled Pentax to get a bit more out of the Sony sensor than anyone else.</p>

<p>What I like most about Pentax is that their cameras are well featured at their price point, controls are very well placed and very easy to pick up, and their shape is very comfortable - the K-7 body felt great in my hand with its controls falling right where they should so I don't have to stretch my fingers or feel around for buttons. Their Limited lenses are also built very nicely, all metal - they had some quirks that favored style vs practicality (like screw-on caps), but they are otherwise beautiful products, even leaving aside their optical performance. When they'll make a MILC, I'll have a close look at it.</p>

<blockquote>

<p>But does it matter?</p>

</blockquote>

<p>It doesn't. That is what I mentioned too - once the crop mode gets you around 15MP, it's good enough to use shorter, lighter FF lenses to get decent results. Especially for people like me that could only make money out of photography by selling their equipment :)</p>

<blockquote>

<p>It <em>is not. </em></p>

</blockquote>

<p>It is a theoretical advantage though, because theoretically there is nothing preventing you from having the same pixel density. It is just practical economical aspects that make a difference - and they are derived from technological challenges, not from technological barriers. And as it happens with technological progress, there comes a point where what you get for your buck is good enough and getting more won't make you a lot better. And betting against technological progress cannot be a meaningful strategy for promoting a sensor size. Sensors will get better and what will matter is what lenses you have available. How many APS-C lenses do you know that you would want to buy a system for just to get to use them? Even when I was using Pentax (which has some interesting APS-C primes), all my best lenses were FF. The best APS-C lens I got - the DA 15 Limited - it is good, but I wouldn't miss it after I sell it.</p>

<i>"I'll say it again - you can have as big a sensor as you like, but unless it is also a densely-packed sensor, cropping into it will not give you "an APS-C image" in all but name."</i><br><br>And i'll then say again, Keith, that you are assuming (why?) that the APS-C sensor wil have a higher pixel density.<br>Don't. If the APS-C thingy can be <i>"a densely packed sensor"</i> there is absolutely no reason why the FF would not also be. And look at the thingies: just about no difference.<br>What the FF brings is the full format advantage. You do get those <i>"more pixels per duck"</i>, Keith. Get yourself a D800 (and a decent set of lenses) and you will get <i>"more pixels per duck"</i> than your 7D can deliver. Crop the D800 to DX format, and you're still in the same ball park.<br>Format advantage. Nothing else. Format advantage.<br><br>But if you insist i will again agree that if we remove all the advantages one thing has over an other thing, that one thing with all advantages removed will hold no advantage over the other thing.<br>And if we do that, what is the point of pointing out that the (in our fantasy) no longer existing advantage is of one sort and not the other?<br><br>In short, if you want more pixels per duck, having a choice between the two cameras mentioned, you need the FF. There are no two ways about it.

<p>Accutance, Q.G., it's about accutance. A high rez FF will trump an APS-C especially with regard to landscapes which have distant detail. Getting closer or using higher zoom lens with an APS-C sensor won't cut it. </p>

<p>My examining full rez sample Raws from the Nikon D800 over at Imaging Resource convinced me. There is a trade off with lens to sensor size depending on how fine a pixel mesh covers that much larger FF area. Same framed landscape captured and compared between the two sized sensors regardless of higher resolution on the smaller sensor is going to render more finer distant detail on a higher resolution FF. I've seen it. </p>

<p>You all get bogged down in the simplest technical issues. 12MP does as good or better than slow slide film and I so rarely have the need to print anything out anyway. When I am cropping and resizing for the typical output of a screen what gives me problems is image noise, not megapixels. If you lot ever print pictures 40 inches wide like these cameras deserve then give me an invite to your exhibition ;)<br /> Knowing that the K5 has the same sensor as the D7100 and has slightly lower raw noise in return for already applying noise reduction to the data, I am curious to know if we actually gain anything by this exchange or if it is just a shallow cheat to impress people in image comparisons.</p>

<blockquote>

<p>Knowing that the K5 has the same sensor as the D7100. . .</p>

</blockquote>

<p>Actually, the sensor for the D7100 has 24 megapixels, and so I presume that you are referring to the D7000 sensor.</p>

<p>In any case, I certainly do agree with your central point. My D3s only has 12 megapixels, which is one reason that its images are so clean. </p>

<p>--Lannie</p>

<p>I use a third-party lens as my main, so I could bypass the lethargy of Pentax in that area, if it weren't for the very small market in used lenses there. The price:performance of Pentax evaporates if you take depreciation on lenses. I could accept a K5II over a D7000 with a 3rd party lens because neither has optimal AF, but when I want another lens I have a problem of choice and cost...<br>

So it seems that the fine implementaton of the sensors in the Pentax line counts for nothing in the end.</p>

<blockquote>

<p>I am curious to know if we actually gain anything by this exchange or if it is just a shallow cheat to impress people in image comparisons.</p>

</blockquote>

<p>Some call it a shallow cheat to impress folks, others call it providing information folks can use to make up their own minds which camera to go with. It's clear you have made up yours.</p>

<p>I'm just glad I could shallowly cheat to impress you in that direction. Now go take some pictures.</p>

<p>But before you do just keep this in mind. If noise is your deciding factor, cramming more pixels into the same size area sensor is most assuredly going to produce more noise unless the manufacturer employs processing tricks before or after the A/D converter to clean the data up. How ever it's done just check to see how much clean detail is left afterward.</p>

<blockquote>

<p>So it seems that the fine implementation of the sensors in the Pentax line counts for nothing in the end.</p>

 

</blockquote>

<p>There are some high-end Pentax lenses that can certainly take full advantage of the strengths of the sensor(s).</p>

<p>If one is going to stay with Pentax, one does not think so much about getting rid of one's lenses, but about using them.</p>

<p>I will likely stay entirely with Nikon from here on out because of the expense of changing systems or addiing additional cameras or lense of other brands, but I do think that Pentax is still a major player. People just starting out might want to give this fine old brand a careful look.</p>

<p>--Lannie</p>

 

<blockquote>

 

</blockquote>

<blockquote>

<p>If noise is your deciding factor, cramming more pixels into the same size area sensor is most assuredly going to produce more noise unless the manufacturer employs processing tricks before or after the A/D converter to clean the data up.</p>

</blockquote>

<p>No, that is not true. Recycling a link I posted earlier in the thread:</p>

<blockquote>

<p><a href="http://www.dxomark.com/index.php/Publications/DxOMark-Insights/More-pixels-offset-noise!" rel="nofollow" target="_blank">Contrary to conventional wisdom, higher resolution actually compensates for noise</a>.</p>

</blockquote>

<p>The D7100 has higher resolution than the D7000 yet it performs a tiny bit better, rather than a tiny bit worse because of the "cramming of the pixels". See <a href="http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/(appareil1)/865%7C0/(brand)/Nikon/(appareil2)/680%7C0/(brand2)/Nikon">D7100 vs D7000</a> comparison on dxomark.</p>

Contrary to clever ways of representing noise reduction as "Contrary to conventional wisdom", higher resolution does not (!) compensate for noise.<br>You, in fact, get more noise, even when the signal to noise ratio per sensor element is the same.<br>More pixels = more pixel signals with noise.<br><br>Yes, you can reduce the resolution by downsampling, and the noise (a pixel level thing) will be 'averaged', i.e. blurred. Lost, just as fine detail will be lost.<br>But it is a pretty big ask to consider that as "higher resolution actually compensat[ing] for noise". Another misrepresentation of things that discussions like this do not need.

<blockquote>

<p>This Insight uses specific DSLRs to demonstrate the technique for objectively comparing noise for cameras with different levels of resolution.</p>

</blockquote>

<p>The first freakin' line out of that linked article! GEEZ!</p>

<p>I'm sorry, Laurentiu, I can't get past the logic of using specific DSLRs to demonstrate a technique to determine sensor noise levels whose results can be applied across a wide range of non-specific cameras with higher resolution to sensor size as useful and verifiable assessment on sensor behavior in general as proving what I indicated above was not true. That link makes no sense and no where even close to using an objective process.</p>

<p>Unless you're an electronics engineer with specific training in discerning the various nuances of sensor design and voltage signal purity, I'm going to stand by what I said as being true.</p>

<p>No one can prove beyond a shadow of a doubt that all or very few camera manufacturers are applying trickery to improve the signal by only assessing A/D Raw data dumps which DxO solely relies on. Unless DxO can hook up electronic analyzer hardware to the camera's sensor feed before the voltage readings go through the A/D converter all this is just conjecture and measurebating off of software driven interpretation of the data which can vary according to the software used.</p>

<p>It proves nothing and provides no useful data to help a photographer make better images. There's detail and there's noise, the software on the user end is the only deciding factor on whether it's useful for the photographer. Everything else is proprietary according to the manufacturer and hidden from the end user, the third party software interpreter including DxO's even if their findings favor Pentax over another brand. They can't prove they can separate the hardware from the software as the main influence over the final results.</p>

<p>DxO impresses me on how they can measure things, but don't impress me on how their measured data helps me make better images.</p>

<blockquote>

<p>You, in fact, get more noise, even when the signal to noise ratio per sensor element is the same.</p>

</blockquote>

<p>Well, apparently you do not get more noise. You can check the SNR graphs. If you click on the Screen button in the upper left of the graph, you will see how the comparison at pixel level looks like and the low resolution sensor does indeed look less noisy in that case, but once images are brought to the same resolution (click Print button to get back to the default graphs), the higher resolution sensor actually gets better SNR results.<br /> <br /> Even if you do not trust dxomark, you can take shots from the D7000 and D7100 and scale them down to the same resolution, then compare them. You can even start with the OP link and throw the D7100 in that comparison.<br /> <a href="http://www.dpreview.com/reviews/nikon-d7100/22">http://www.dpreview.com/reviews/nikon-d7100/22</a><br /> <br /> Unfortunately, the focus is slightly different in those images, so that kind of defeats the purpose. dxomark can do these tests better than anyone because they have the equipment for it - they can control the lighting in the test room very accurately - something that you and me cannot do at home.<br /> <a href="http://www.dxomark.com/index.php/About/In-depth-measurements/DxOMark-testing-protocols">DxOMark image quality testing protocols for lenses and sensors</a></p>

<p> </p>

<blockquote>

<p>The first freakin' line out of that linked article! GEEZ!</p>

</blockquote>

<p>Well, get to the second page and it won't get any better. But they seem to have worked with some mathematical model to come to the conclusion that 4 smaller photosites have better SNR than a larger one that covers their surface. I've done a similar exercise for a monochrome sensor and the math was beyond me although someone assured me the noise would be identical. With a Bayer sensor it becomes even trickier because you replace a red, green, or blue photosite with 2 greens, one red, one blue. You lose half the ability to capture green but you gain more ability to capture red and blue - perhaps that works out for the better - I couldn't work the math out.</p>

<blockquote>

<p><strong>Unless you're an electronics engineer with specific training in discerning the various nuances of sensor design and voltage signal purity</strong>, I'm going to stand by what I said as being true.</p>

</blockquote>

<p>Well, <a href="http://en.wikipedia.org/wiki/DxO_Labs">DxO Labs</a> engineers are as close as you can get to your requirements. So unless you happen to be an even more experienced engineer than they are, what makes you think that your opinion is worth anything?</p>

<blockquote>

<p>They can't prove they can separate the hardware from the software as the main influence over the final results.</p>

</blockquote>

<p>Actually, they can and that is how they detected Pentax's "cooking":<br /> <a href="http://www.dxomark.com/index.php/Publications/DxOMark-Insights/Half-cooked-RAW">Can RAW images be already somewhat cooked?</a><br /> <a href="http://www.dxomark.com/index.php/Publications/DxOMark-Reviews/DxOMark-review-for-Pentax-cameras">DxOMark review for Pentax cameras</a></p>

<blockquote>

<p>DxO impresses me on how they can measure things, but don't impress me on how their measured data helps me make better images.</p>

</blockquote>

<p>Wow, that is a big burden to put on some engineers - now they're supposed to make your images better before you can give them any credit for knowing their stuff. dxomark isn't about people making better images, they are about providing an objective comparison of sensors and lenses.</p>