Rumors about Canon Matching Pentax K-5 Sensor?

[JDMvW]

<p>The specific rumor I have heard is that Canon is going to buy Pentax, and then using the new Pentax sensor the new joint company will produce a digital version of the first Prakiflex camera from KW.</p>

<p>Remember, you heard it here first.</p>

<p>I'd also like to hear more about why so many DxO comparative data just seem screwy (translation: disagree with my own deeply held prejudices).</p>

[michael_kuhne]

<p>They are both very fine cameras. And one's already-acquired lenses have a huge bearing on decisions for a body. New sensors and technology emerge on a regular basis. This year this body is the sensation, next year another body is- but the better grade lenses that especially suit one's needs are ongoing.</p>

<p>I do not yet have the K-5, but still get fine results with my K20D and K200D, and do not feel deprived. Nor would I feel so if I owned a 7D with good lenses. I will probably eventually buy a K-5, but I'm in no rush.</p>

[jim_ducey]

<p>You can turn on the noise reduction feature in the custom settings on the 7D, it is off by default if you are going to use very high iso<br>

Jim</p>

[charles_wood]

<p>I would suggest to all the pixel peepers arguing the merits of the 7D, K-5, D7000, etc., that you go to DPREVIEW, select the K-5 review, go to the RAW COMPARISON screen, and then look at various parts of the scene in detail. From ISO 200 to upwards, the 7D is visibly the noisiest camera of the group. The K-5 actually comes closest to a 5Dmk2, when you bring it into the comparison. The 5Dmk2 renders the best fine detail, of course and comparing fine detail puts you at the mercy of lens selection. But in terms of freedom from noise the K-5 is a very strong contender.</p>

[zack_zoll]

<p>Daniel, you've said you were asking "people" to try to find the difference. This is a very vague polling group.</p>

<p>Let's try this from the other direction: fashion magazines are generally 9x12 sized, and it's generally accepted that 300 dpi is the most resolution that our eyes can see. Some four-colour or chemical printers use 600 dpi as a default, but RGB usually hits around 300 dpi. So a 9x12 at 300 dpi is about 9.7 megapixels.</p>

<p>You are entirely correct when you say that in terms of resolution and IQ, your average Vanity Fair reader (and I include myself there) won't be able to see the difference in sensors between an H4D and a 30D, or even a Rebel XTi. But that wasn't the point I was making. My point is that if you hand, for instance, Prada's art director a disc of images that you shot with your Rebel after they paid you X thousand dollars and you said, "No one can tell the difference," they probably won't rehire you. They may not even want to pay you for the job you already did. Guys like Terry Richardson. who shoot with a point-and-shoot 35mm camera, can do that because they already have a great reputation that affords them much more creative freedom.</p>

<p>I'm not saying you need expensive gear to get good jobs. What I am saying is that if you have demanding clients, and you and another guy with the exact same skill level and connections - and much better gear - are going after the same job, the other guy is going to get it. The final result might be the same to your average reader, but the people who are actually employing you will be viewing the images at 100% on a 27" Apple screen or somesuch.</p>

[laur1]

<blockquote>

<p>If you operate at a certain level, you may find yourself (and your clients) seeing differences in equipment that 99.9% of the public doesn't even know exist.</p>

</blockquote>

<p>I think this is the idea. People bash dxomark because they can't translate their scores into their experience. I did the same some time ago, when after looking at the scores of my cameras I found out that my favorite had the lowest score and the camera I upgraded from was rated actually higher. But as I continued to use the cameras, I started to get a feeling that those ratings were onto something. Then I looked closer at their methodology and it started making more sense.</p>

<p>The differences measured by dxomark are not of the kind you can easily evaluate by looking at test shots, except maybe for the high ISO performance. No one can look at test shots and say - "oh, by the looks of it this camera has 22 bits color depth and 13 EV dynamic range, and this other one looks to have just 20 bits color depth, but has 14 EV dynamic range". You just don't get to see those differences unless you happen to have both cameras available for shooting a challenging scene and you get to compare the results heads on - but who gets to do that? The imaging resource test shots aren't even properly focused (in the sense of always being focused on the same spot), so it's very hard to compare even basic stuff like sharpness.</p>

<p>Here's one excerpt from dxomark on high ISO score (<a href="http://front1.dxomark.com/index.php/en/Learn-more/DxOMark-scores/Sensor-scores">source link</a>):</p>

<blockquote>

<p>Low-Light ISO is then the highest ISO setting for the camera such that the SNR reaches this 30dB value while keeping a good dynamic range of 9 EVs and a color depth of 18bits.</p>

</blockquote>

<p>9EV and 18bits is the quality bar they use in this score. Even the Olympus EP2 passes this bar up to ISO 505. Most people are happy shooting images at ISO values where they no longer get this IQ - yet they are happy with it. What this tells me is that people can't really objectively appreciate IQ. Moreover, they are happy with lower IQ as long as other elements come together fine in the image. Nothing wrong with that either.</p>

<blockquote>

<p>Just to make the point that sensor IQ at this stage of the game is a neck and neck race with differences so minor as to be hardly relevant.</p>

</blockquote>

<p>The differences are not minor, but they're not easy to notice anymore.</p>

<p>So where does this leaves us? dxomark scores are still the best way we have of comparing sensor performance - I am happy they do this work. But one has to know how to read them. Also, just because a camera sensor is scored higher, doesn't mean it is superior in *all* aspects - it may still be inferior in some area while being superior in others, not to mention that attributes like resolution don't even matter for the score. Finally, most of our shots don't need the most performant sensor - the technology is already developed well enough that it doesn't provide a bottleneck for our creative output.</p>

<p>As for Canon coming up with a more performant sensor, it is very unlikely unless users ask for it. If the perception in Canon market is that the 7D is the best sensor, that's all that's going to be available until the perception changes. There is no offer if the demand is lacking.</p>

 

<blockquote>

<p>But after reading it, I'm going with the 7D</p>

</blockquote>

<p>And so you should. It makes no sense for you to change brands when you have an existing investment in lenses (and good ones too). The sensor is just one piece of the equation. I don't intend to upgrade any of my cameras until they break. And I don't have the K-5 either.</p>

 

[jim_ducey]

<p>For my own use I don't do much shooting in the dark with ISO above 1600 without supplemental light. The high frame rate is important for sports, air shows and wildlife shooting.Perhaps so do shoot with extremely high ISO but I am not one of them. The medium format cameras I have and have had don't go to extreme high ISO. One reason I like the Canon 7D is quick local repair. For me the Canon is a better tool.<br>

Jim</p>

<p> </p>