Has Pentax trumped Nikon crop sensor cameras at high ISO?

<p>I now shoot Nikon after shooting Canon since 1982--nothing against Canon; it simply did not have the D800E, and I love high resolution for much of my shooting.</p>

<p>Even so, landscapes are not all I shoot. One of my favorites things to shoot is <a href="/photodb/folder?folder_id=1032268"><strong>urban settings at night</strong></a>. I do now have a second-hand D3s, which is of course magnificent in low light, but both the D3s and the D800E are a lot to risk in some of the neighborhoods I shoot in. </p>

<p>This led me to look at some less expensive alternatives. I was looking at some of the smaller (i.e., crop sensor) Nikons when I happened to notice the results from the Pentax K-5 IIs. <br>

<strong> </strong><br>

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

<strong> </strong><br>

<br />I mention all of this solely in order to give some background. (That is, I am<em> not</em> asking for advice on how to spend my money or plan my life. If you respond, please address camera performance only.)</p>

<p>I turned the ISO up to 6400 on the DPReview page above to see how the Pentax and Nikon crop sensor cameras might do in low light, high ISO situations. What I found was that the Pentax K-5 IIs really seems to be quite clean at ISO 6400. Yes, I am aware that it has "only" 16 megapixels, but low megapixels are one of the prices we often pay for getting clean shots in low light with high ISO.</p>

<p>I would just like to see how others would respond to these findings. I'm not looking to jump systems again, just looking for a specialty camera for low-light situations. That the Pentax is a crop sensor camera is icing on the cake. I also am not planning to buy more than one or two Pentax lenses--possibly no more than a kit zoom.</p>

<p>I would go so far as to say that (based strictly on the DPReview tests) the Pentax K-5 IIs just might be the cleanest crop sensor camera that I have ever seen at high (6400) ISO.</p>

<p>Does anyone have an opinion on these matters?</p>

<p>PLEASE NOTE: This post is not a request for advice as to how best to spend my money. It is about cameras.</p>

<p>--Lannie</p>

<p>I went to the DXO site and the two camera's have similar specs. No actual experience with either camera. Thinking of it I cannot remember the last time I have even seen a Pentax camera.</p>

<p>Nor I, but it appears that Pentax has been doing some interesting things since most of us last looked.</p>

<p>--Lannie</p>

<p>My research has showed the same thing, Pentax is now the king of low light for APS-C. And speaking of minimizing your financial exposure in risky neighbors, the sensors in their less expensive models are just as good.</p>

<p>You might also look for cheap lenses. Though for wide angles you need the newer ones designed for APS-C sensors, you can easily find the older M42 and K mount lenses in longer focal lengths with wide apertures for very little money. I own three, a 50 f1.4 M42 that cost $100, a Sigma 135 f1.8 in K mount I picked up for $75, and a Pentax 135 f2.5 K also for $75 I use when I don't want to carry around the very heavy f1.8. Just a thought </p>

<p>I have never shot Nikon or Canon digital cameras, but I have shot a lot of theater dress rehearsals for many years, and have been impressed with the performance of my Pentax K 5 at high ISO--typically 1600. I also remember when the Pentax K 10 and Nikon D 200 came out (both using the same Sony sensor) that many test reports rated the Pentax as having lower noise at equivalent ISOs. Is it better than Nikon/Canon equivalents? I don't know since I haven't used them.</p>

<p>Thanks, guys. Again, if one simply goes<a href="http://www.dpreview.com/reviews/nikon-d7100/22"> <em><strong>to this link</strong></em></a> and plugs in ISO 6400 for all of the cameras, the cleanness of that image on the Pentax is simply incredible.</p>

<p>Also (again), I simply have not seen ANY crop sensor camera produce such clean shots at high ISO. Yes, the D3s is somewhat cleaner, but that is full-frame and many, many dollars more. Even there, the difference is not massive.</p>

<p>Since I own no Pentax cameras (and never have), I hope that it is clear that I have no axe to grind in any of this. I am simply blown away by what Pentax is accomplishing these days.</p>

<p>--Lannie</p>

<p>Don't know if you've seen this PentaxForums Dynamic Range test between the Canon 7D vs Pentax K-5 so I'll post the link... http://www.pentaxforums.com/reviews/canon-7d-vs-pentax-k-5-review/image-quality.html#drtest</p>

<p>View Full Size 100% zoom on only the -5EV Developed Raw shots that are normalized.</p>

<p>This is not an indication of low light performance because full spectrum light that's under exposed like this is going to show much less noise compared to very little to no full spectrum light in night scenes. It is the quality of light that determines the amount of noise in combination with sensor performance and exposure.</p>

<p>Just thought this may help you decide on what to get. I've been wanting to upgrade to a newer DSLR from my 6MP Pentax K100D that I've been happy with since 2006 but this K-5 DR test just blew me away enough to make me commit to upgrading and I have a slew of lenses that'll work just fine with it.</p>

<p>Hi ISO Pentax files have NR baked in to the Raws, Lannie - nothing you couldn't do (and probably better) in PP with a Nikon.</p>

<p>Maybe so, Keith, but they don't look smudged. There is still a lot of detail. I'm not trying to compare this camera to the D3s, which has other strengths as well. I'll give up the D3s when they pry my cold dead fingers, etc. . . .</p>

<p>Thanks to Tim and others. Tim, those are awesome shots. The idea of getting these results with a crop sensor camera at a modest price is the appeal for me. I also don't feel like laying out a lot of money for lenses or otherwise letting things get too complicated. Just a specialty camera with one primary application--although surely a good all around camera as well.</p>

<p>In fact, if the Pentax were the only camera I had, I would still be happy. I've gotta have low light, high ISO to be happy, though.</p>

<p>--Lannie</p>

<p>Lannie, wow, that's indeed impressive performance from a crop, nipping at the heels of a few cameras that have every expectation of being much better!</p>

<p>I was a Pentax shooter long ago, and today, just as then, Pentax seems to be the company that puts out a solid, economical product that gets no respect. I would probably still be a Pentax shooter today if only Pentax had a dSLR at the time I was moving that direction. I don't regret Canon's superb lens offerings, though, and I'm impressed with the image quality from the new 6D.</p>

 

<blockquote>

<p>Has Pentax trumped Nikon crop sensor cameras at high ISO?</p>

</blockquote>

<p>No.</p>

<p><a href="http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/(appareil1)/865%7C0/(brand)/Nikon/(appareil2)/830%7C0/(brand2)/Pentax">If you compare the K-5IIs with the Nikon D7100</a>, you'll see that they are practically offering the same performance. The D7100 actually scores a bit higher on dxomark, but I doubt you would be able to notice that small of a difference.</p>

<p>I attribute this performance to Sony's technology. I also don't expect to see much improvement going forward. <a href="http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/(appareil1)/865%7C0/(brand)/Nikon/(appareil2)/830%7C0/(brand2)/Pentax/(appareil3)/676%7C0/(brand3)/Pentax">If you look at the K-5 as well</a>, you will notice that performance has not improved since it was introduced in 2010.</p>

<blockquote>

<p>Again, if one simply goes<a href="http://www.dpreview.com/reviews/nikon-d7100/22" rel="nofollow" target="_blank"> <em><strong>to this link</strong></em></a> and plugs in ISO 6400 for all of the cameras, the cleanness of that image on the Pentax is simply incredible.</p>

</blockquote>

<p>Oh, about that:</p>

<ol>

<li>You are comparing 100% crops of cameras with different resolution. The Nikons will look noisier. But if you pick an image size and scale down the Nikon images, you'll get cleaner results. Add the NR to compensate for the Pentax NR that Keith mentioned and you will get pretty much identical results. (this is why dxomark compares at a fixed resolution and estimates the NR benefits for Pentax)</li>

<li>Also, the K-5IIs lacks an AA filter, so it will capture more detail at pixel level than other 16MP cameras. To see the impact of the lack of the AA filter, pull the K-5II into the comparison - not as clean anymore.</li>

</ol>

<p>The fact that even at 100% the differences are so small should tell you that there is not really a big gap of performance between these cameras. The A77 looks worse because it has to deal with 1/3Ev loss but even in its case this difference is only obvious at this magnification.</p>

<p>I shot Pentax MF during my film days but went digital with Canon for the same reason as many other Pentax shooters. Pentax simply waited too long to go digital. But I grew tired of the shadow and 'read' noise in Canon sensors, including my 5D2. At that point I went back to Pentax with the 645D and bought a K5 as a secondary backup and never looked back. Both Nikon and Pentax have done a superb job of implementing the SONY sensors. That said, for the bucks invested, the K5 and current variants, give you weather and dust sealing, internal image stabilization, optional GPS accessory (also useful for tracking night sky movement up to five minutes), extremely quiet shutter for street or shooting in situations requiring low noise, 100% viewfinder. All in all, a superb camera for the bucks, with a couple of features not available in other cameras at any price. If you want another surprise in terms of overall performance up to moderate ISO levels, check out the comparison shots of the Ricoh (Pentax parent company) GR. It uses the same sensor as the K5 with superb optics in a small P&S. It might be what you're really looking for. Or, if you can still find one, the K01 on closeout for about $350-400 with a kit lens...essentially a stripped K5 in a mirrorless design but with the same sensor and high ISO performance.</p>

<blockquote>

<p>Both Nikon and Pentax have done a superb job of implementing the SONY sensors. </p>

</blockquote>

<p>Did anyone do a poor job using Sony sensors?</p>

<p><em>Did anyone do a poor job using Sony sensors?</em><br>

<br />Yes, relatively speaking compared to Nikon and Pentax, SONY implementation of the 16 meg sensor wa<em>s </em>marginally not as good in some performance areas as the K5 or the 7000...if you believe DXO Mark. Pentax scored an overall 82, SONY and Nikon scored 80.<em><br /></em></p>

<p>Laurentiu, since all things mentioned being the same as you've indicated with your DxO mark links, are the prices the same?</p>

<p>Has DxO ever created real world sample images to support and correlate their lab findings AND also show how these image's previews behave with regard to a combination of color distortions, noise suppression, WB and other basic editing functions that point to the value and efficacy of their lab measurements? Or maybe they just like to create an art out of measuring. I get that.</p>

<p>I guess it doesn't really matter anyway since all these cameras perform the same.</p>

<blockquote>

<p>I guess it doesn't really matter anyway since all these cameras perform the same. --Tim</p>

</blockquote>

<p>Maybe I should have asked,</p>

<h1>Has Pentax achieved parity with Nikon crop sensor cameras at high ISO?</h1>

<p>If parity has been achieved--at a lower price--then that is surely worthy of note. Given that I already have Nikon lenses, I won't likely buy the Pentax. If I were starting from scratch, however, I would surely be looking long and hard at the Pentax.</p>

<p>Also worthy of consideration is how far Pentax has come (in a short perio of time) in the development and marketing of DSLRs. If the data on DXOMark and the pictures on DPReview.com are to be believed, Pentax has come a long way in a hurry. Combine that with a loyal customer base and the fact that a lot of people really don't want the expense and weight of full-frame gear, Pentax seems to have the prospect of a pretty bright future--if DSLRs have a bright future.</p>

<p>--Lannie</p>

<blockquote>

<p>Yes, relatively speaking compared to Nikon and Pentax, SONY implementation of the 16 meg sensor wa<em>s </em>marginally not as good in some performance areas as the K5 or the 7000...if you believe DXO Mark. Pentax scored an overall 82, SONY and Nikon scored 80.</p>

</blockquote>

<p>Sony doesn't make SLRs anymore and it does look like a sensor that is optimized to always record looses some DR in MILCs and even more in SLTs due to their need of using higher gain to compensate for the 0.3Ev loss of light. <a href="http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/(appareil1)/865%7C0/(brand)/Nikon/(appareil2)/783%7C0/(brand2)/Pentax/(appareil3)/676%7C0/(brand3)/Pentax">The Pentax K-01 also scores lower than the K-5 despite sharing the same Sony sensor technology</a>. Do you think that is because Pentax "implemented" the sensor badly in the K-01? Whatever tweaks Pentax and Nikon made to their use of Sony sensors only produced minimal differences that are barely noticeable.</p>

<blockquote>

<p>Has DxO ever created real world sample images to support and correlate their lab findings<br /></p>

</blockquote>

<p>I was thinking the same a few years back. That was before I spent the time to understand what the dxomark numbers mean. Now I can say they correlate with what I see on the cameras I used - the problem is that people don't understand them and misinterpret what they mean - they fight over differences of less than 0.3Ev despite the fact that dxomark mentions that such differences are hard to notice in practice. If you have some time, read <a href="http://laurphoto.blogspot.com/2012/04/quick-guide-for-understanding-dxomark.html">this article</a> where I tried to explain the basics of what they measure.<br>

<br>

For an anecdote - Pentax users kept dissing dxomark for rating Pentax cameras low. Most users on PentaxForums were stating their indifference for dxomark scores. Then the K-5 came out and leaped to the top of the APS-C cameras and suddenly dxomark made sense to everyone - they were all quoting dxomark.<br>

<br>

For the cameras discussed here dxomark shows insignificant differences although some will pick on the 80 vs 82 difference which is meaningless in practice (and in theory as well, because the Nikon D7100 scores 83). Nikon APS-C cameras (5200 and 7100) actually rank now above Pentax ones. But that rank is rather meaningless - it is just an ordering measure - what one has to look at are the graphs that dxomark produces. Even then you have to pay attention - the only score I tend to look at is the High ISO score. For the rest of the evaluation I just look at graphs. The DR and color depth scores can be misleading - I cover this in my article.<br>

<br>

BTW, I still use Pentax, although not as much as I used to. I am now mainly using MFT. I never used Nikon, not even a P&S.</p>

<i>"You are comparing 100% crops of cameras with different resolution. The Nikons will look noisier. But if you pick an image size and scale down the Nikon images, you'll get cleaner results."</i><br><br>Since noise is a pixel-level phenomenon, the only difference higher resolution makes in the comparison of those images is that the image content (the bit of the label) appears larger in the higher resolution camera sample. Ignoring noise reduction measures that may have been applied by some, not by other cameras, cameras will only look noisier in comparisons of 100% crops if they indeed are.<br><br>But yes: if you would compare same field of view images, printed the same size, or displayed on the same size monitor, the higher resolution camera will not show as much noise as lower resolution cameras, assuming that they both have the same level of noise. That's the age old format advantage showing itself.

<blockquote>

<p>BTW, I still use Pentax, although not as much as I used to. I am now mainly using MFT. I never used Nikon, not even a P&S.</p>

</blockquote>

<p>Laurentiu, if you are using micro-four-thirds, then you clearly are interested in smaller and lighter cameras. It seems that, apart from the Pentax 645D, Pentax is not moving in the direction of medium format or full-frame. (If I am wrong on that, please correct me.) Pentax seems clearly to be trying to optimize offerings in the 1.5x crop sensor format. Since I have never even held a Pentax in my hands (except for a few lenses that I used with Canon adapters when I shot Canon), I am not really in a position to evaluate Pentax based on personal experience.</p>

<p>Q.G., thanks for directing us back to the issue of making comparisons on scaled down images--but as for comparing images that are scaled down, Laurentiu, I am not sure that I can make a valid inference based on the tiny samples shown on the DPReview page that I cited. I also cannot tell to what extent <em><a href="http://www.dpreview.com/reviews/nikon-d7100/22"><strong>the "clean" images at ISO 6400 for the Pentax K-5 IIs</strong></a></em> are the result of noise reduction. I think that we need more data (hands on, visual data from real shots) in order to evaluate further.</p>

<p>In any case, it does seem that Pentax has made substantial strides since the *ist--not only the 645D, which appears to give excellent "MF" results (the scare quotes serving only to remind that the sensor is not really 60 x 45mm, only that it takes lenses from that film format). Those strides, that is, seem to be primarily in the 1.5x crop market--which I consider to be an interesting development, if only because MFT images tend to break down at ISO 6400, if not before--the format is a bit too small to interest me, given the type of shooting (low light, high ISO) that I like to do.</p>

<p>I will be interested in seeing what Pentax comes up with next, but with my current gear I will likely only be watching from the sidelines--unless the day comes when I have to move to a smaller format, whether through infirmity or for financial reasons. Right now, that is, I shoot full-frame in most cases. If 1.5x crop comes to match where full-frame is at this moment where ISO is concerned, then that would be one less reason to lug around heavier, more expensive full-frame gear.</p>

<p>--Lannie</p>

<blockquote>

<p>I am not sure that I can make a valid inference based on the tiny samples shown on the DPReview page</p>

</blockquote>

<p>It took me a little while to realize I could pan around over the entire image. Click the big image to make a jump. Drag over any of the images to pan around.</p>

<p>I find it especially telling to read tiny text (e.g. the white on red to the right of the Greek figure and beneath the US dime). If you crank the ISO to the point that the text is borderline readable, then you can make very good "go/no-go" comparisons that take into account both noise and resolution. Assuming the overall angle of view of the sample image is the same (which it is, regardless of format or resolution), the more readable text indicates a higher quality of image. </p>

<p>Thanks, Sarah. I see what you are talking about. I had not tried that before.</p>

<p>I just downloaded the images and <strong>put the Nikon image through noise reduction</strong> until it looked as clean as the Pentax image. It is almost a wash, with the Nikon still having a slight resolution advantage in spite of the NR. </p>

<p>(Keith Reeder's comment above comes to mind at this point.)</p>

<p>Even so, it looks very close. The Nikon clearly can be used at ISO 6400 with a bit of noise reduction--and even without that for certain size displays Given that I have Nikon lenses, I would go for the Nikon. Otherwise, it appears too close to make any difference to me. Below 6400, however, the advantage goes clearly to Nikon, given its additional resolution.</p>

<p>That is my best call on the matter. I am still impressed with the Pentax, however, and I will be watching to see what that company comes out with next.</p>

<p>--Lannie</p>

<p>Why did I do this little exercise? I think that it is because I suspect that at some point crop sensor cameras will give such clean images in low light/high ISO shooting that most persons won't see the need to shoot full frame--and I might be one of them.</p>

<p>--Lannie</p>

<blockquote>

<p>Also worthy of consideration is how far Pentax has come (in a short perio of time) in the development and marketing of DSLRs. If the data on DXOMark and the pictures on DPReview.com are to be believed, Pentax has come a long way in a hurry.</p>

</blockquote>

<p>I wouldn't say so - they made incremental progress. Their partnership with Samsung got dissolved and they ended up using Sony sensors, which gave them this performance. The cameras in themselves didn't really improve in dramatic steps, but the performance boost from Sony technology was easily visible in images. The K-5 is just a K-7 with a Sony sensor. If they would have sticked to the K-7 sensor, you wouldn't have given them a second look despite all the nice features that camera body includes. IMO the K-7 was a historical camera model for Pentax.<br /> <br /> One thing Pentax lags in is continuous AF - I don't know how much better it got, but I doubt it can match Canon/Nikon implementations. Number of AF points has also not increased a lot. This didn't matter for me, because I am a MF guy, but for the regular user, it would make a big difference. There are basically other features that are necessary for obtaining nice images than a good sensor and once you look at those, Pentax stops looking as appealing. Their lens lineup also has holes and accessories are lacking (no modern extension tubes or teleconverters).</p>

<blockquote>

<p>Combine that with a loyal customer base and the fact that a lot of people really don't want the expense and weight of full-frame gear, Pentax seems to have the prospect of a pretty bright future--if DSLRs have a bright future.</p>

</blockquote>

<p>Not sure how well they are doing. Many people I knew at PF had moved away from Pentax before I did. And whatever future DSLRs have, it still is safer to go with Canon or even Nikon. I have the feeling that Canon will transition better out of the DSLR era than either Nikon or Pentax.</p>

<blockquote>

<p>Laurentiu, if you are using micro-four-thirds, then you clearly are interested in smaller and lighter cameras.</p>

</blockquote>

<p>No. I am interested in lenses. I like Voigtlander MF lenses. When Cosina and Zeiss dropped support for Pentax K mount so they could make lenses for other mirrorless mounts, I started looking at what mounts they decided to support. Cosina went for MFT.</p>

<blockquote>

<p>I am not really in a position to evaluate Pentax based on personal experience.</p>

</blockquote>

<p>I bought into Pentax in 2007 with the K10D, which I still have and use on occasion. I think they did well in the APS-C DSLR segment, especially considering that over the past years they were bought out by Hoya for their medical division and then the camera division was picked up by Ricoh for a mere $125mil. The problem is that the market is changing - I think DSLRs will go the way of rangefinders and I am also wondering about the future of the APS-C format, which has only been fully supported by Fuji. With FF becoming close enough in price to high end APS-C and with MFT providing a really nice system overall, I don't know how exactly APS-C will survive. If I would start from scratch today, I'd just go MFT. But I have a bunch of FF lenses from film era, so I am interested in a FF MILC - perhaps the rumored NEX that is to be announced soon.<br /> <br /> It will be interesting to see how companies adjust to the changing market during the rest of this decade.</p>

<blockquote>

<p>Has Pentax achieved parity with Nikon crop sensor cameras at high ISO?</p>

</blockquote>

<p>The answer to this would be yes - because they use pretty much the same sensor technology. The only "improvement" they added is the low-level RAW NR that you cannot disable and that makes their results above 3200 look cleaner. BTW, this processing was first pointed out by dxomark <a href="http://www.dxomark.com/index.php/Publications/DxOMark-Reviews/DxOMark-review-for-Pentax-cameras">here</a>.</p>