Here are Popular Photography's noise data, up to ISO 3200, on the K-5 and the K30, which have the same sensor:ISO K-5 K30100 1.3 1.0 200 1.5 1.3 400 1.9 1.6 800 2.2 2.5 1600 2.7 3.9 3200 3.4 4.4(Incidentally, I formatted this table nicely, but I see in Message Preview that Dpreview ruined my nice formatting.)So Pentax improved noise slightly up to ISO 400. But above that, noise is worse on the K30. The difference is particularly noticeable at ISO 1600 and up.Any thoughts on why Pentax did this?Joe

Any thoughts on why Pentax did this?Same sensor an' there ain't no free lunch. The equation's gotta balance 'cause there's only so much you can do with software tweaks.At least now you've got a choice.H2To para-phrase the RNC -- are your noise numbers better off now than they were four years ago?

Very confusing...Looking at the DPreview comparison gizmo it looks like the K30 outperforms nearly every camera in image quality even at high ISO.It made me quite tempted to change brand

EvanDavies1wrote:Very confusing...Looking at the DPreview comparison gizmo it looks like the K30 outperforms nearly every camera in image quality even at high ISO.Confusing indeed. The Dpreview results seem to show the K30 outperforming the K-5.Joe

Joseph Tainterwrote:Here are Popular Photography's noise data, up to ISO 3200, on the K-5 and the K30, which have the same sensor:ISO K-5 K30100 1.3 1.0 200 1.5 1.3 400 1.9 1.6 800 2.2 2.5 1600 2.7 3.9 3200 3.4 4.4(Incidentally, I formatted this table nicely, but I see in Message Preview that Dpreview ruined my nice formatting.)So Pentax improved noise slightly up to ISO 400. But above that, noise is worse on the K30. The difference is particularly noticeable at ISO 1600 and up.K-30 has less noise reduction at ISO1600 and above, the pictures are sharper.As for ISO100-800 - no any serious difference. K-30 has another sensor, not K-5's.

oglwrote:K-30 has less noise reduction at ISO1600 and above, the pictures are sharper.As for ISO100-800 - no any serious difference. K-30 has another sensor, not K-5's.Both of these statements are most likely true. GordonBGood has documented that the output pattern of the K30/K-01 sensor is different than the K5; even the pixel area covered differs in both horizontal and vertical measures. The processing architecture differs (most obviously the shift from 14 bit to 12 bit - not that this has any real-world impact). GordonBGood is also probably correct that Pentax has shifted emphasis toward improved Live View and video capability which can very slightly degrade still imaging capability.Pentax took a hit for greater high ISO auto noise reduction evident in some bodies (K20D, K7, K5) and likely backed it off somewhat in the K-01/K30 sensor/processor (they appear to be identical). That's a good thing for those of us who prefer to do it ourselves. Keep in mind that the big boys with the huge marketing budgets do the same thing with smoothing (Canon is especially aggressive about it) - but are never called out.

JNRwrote:That's a good thing for those of us who prefer to do it ourselves.I am no expert but from what I remember and understood of what Gordon said is that the noise reduction done on chip cannot be in anyway replicated as effectively post read off. -- Chris.A weather sealed ultra wide, is that too much to ask?http://www.pentaxphotogallery.com/chrisideGMT +9.5Pentax SLR talk FAQ http://forums.dpreview.com/forums/read.asp?forum=1036&message=23161072

I didn't see that comment from Gordon. I suppose it is just a matter of how effectively it is addressed on the chip. The complaint against the K20D was that the smoothing lost detail above 1600. My Canon 5D clearly was doing something at the same ISOs - and the loss of detail was very evident.I'm not saying that the K20D imaging overall was better. At ISO 100-200, the Canon FF was wonderful - but it didn't hold up well in the higher ISOs. In this regard, I have the impression that DXoMark is not at all consistent in its ratings.Overall, it is my impression that it depends on the sensor, and how noise reduction is implemented. Newer designs apparently are more successful. -- JNR http://www.jamesrobins.com

JNRwrote:I didn't see that comment from Gordon. I suppose it is just a matter of how effectively it is addressed on the chip. The complaint against the K20D was that the smoothing lost detail above 1600. My Canon 5D clearly was doing something at the same ISOs - and the loss of detail was very evident.Overall, it is my impression that it depends on the sensor, and how noise reduction is implemented. Newer designs apparently are more successful.James, you are correct that the compulsory raw Noise Reduction (NR) algorithm as applied by Pentax bodies has been improved considerably from that used by the K20D/K200D/K-m/K2000D bodies, as those had quite a coarse "smudging" pattern and was applied as low as ISO 800 due to the noisier CCD sensors used and at ISO 1600 for the K20D due to the noisier Samsung CMOS sensor used for that model.I'm not sure I ever made the comment that the Pentax type of compulsory raw NR as applied from ISO 3200 upwards couldonlybe applied during data acquisition; however, I have stated that it is applied to individual raw photositesbefore demosiacingmeaning that the amount of smudging as to radius is very small and thus it does the least damage as compared to the smudging (or artefacts) contributed by Bayer pattern demosiacing anyway. If my guesses about how this NR is implemented are correct, the NR could be applied to actual raw data in post processing before any other processing would take place.As to the OP's comments on Popular Photographies respective noise tests for the K-30 versus the K-5, it appears that these are just a measure of standard deviation of JPEG levels from a standard grey patch for Out Of Camera (OOC) JPEG's for default settings and as such are only indicative of individual default JPEG development settings for the respective cameras; it seems that the respective high/low ISO balances are different, but then that could be just due to different sharpening algorithms applied to OOC JPEG's. This isn't really all that useful as one isn'tforcedto use the Bright default setting for OOC JPEG's and most of us don't. With different OOC JPEG development settings, the K-5 (and the K-30) could well be listed to have "moderately low" noise all the way up to ISO 3200 sensitivity according to Popular Photography's noise standards.As has been stated, one would expect that the K-30 would provide image quality comparable to the K-01 as the sensors appear to be identical with one very important caveat: it likely will have a slightly better raw capture low ISO sensitivity Dynamic Range (DR) when used in non Live View (LV) mode due to less heating of the sensor as compared to the K-01 which only has LV mode and no other. From my quick measurements, raw image quality is not any better than that of the K-5 although both are excellent and one wouldn't choose one over the other only based on raw image quality.As to having only 12-bit raw depth rather than the K-5's 14-bit, although that can be demonstrated to have a slight disadvantage for ISO's below 400, one would be hard pressed to find practical images where this could be shown to be a limit; of course the advantage of the smaller bit depth is that raw file sizes will be six Megabytes (25%) or more smaller at low ISO's and ten Megabytes or more (two thirds the size) or more savings for higher ISO's.Hopefully, the K-5's replacement may have an option of 12-bit or 14-bit raw depth so we can make this choice ourselves. From the trends of the most recent Pentax bodies, I doubt we will see any real changes in raw NR applied but likely will see further tweaks to development of JPEG images.Regards, GordonBGood

JNR, FWIW the k-30 RAW samples on dpreview look less noisy at high iso than k-5's. -- Ask not what your camera can do for you, but what you can do for your camera http://www.fluidr.com/photos/gs-photos

GordonBGoodwrote:James, you are correct that the compulsory raw Noise Reduction (NR) algorithm as applied by Pentax bodies has been improved considerably from that used by the K20D/K200D/K-m/K2000D bodies, as those had quite a coarse "smudging" pattern and was applied as low as ISO 800 due to the noisier CCD sensors used and at ISO 1600 for the K20D due to the noisier Samsung CMOS sensor used for that model.I'm not sure I ever made the comment that the Pentax type of compulsory raw NR as applied from ISO 3200 upwards couldonlybe applied during data acquisition; however, I have stated that it is applied to individual raw photositesbefore demosiacingmeaning that the amount of smudging as to radius is very small and thus it does the least damage as compared to the smudging (or artefacts) contributed by Bayer pattern demosiacing anyway. If my guesses about how this NR is implemented are correct, the NR could be applied to actual raw data in post processing before any other processing would take place.Honestly often what you say goes well over my head so I guess I misinterpreted what you said and as a result when paraphrasing changed the meaning. -- Chris.A weather sealed ultra wide, is that too much to ask?http://www.pentaxphotogallery.com/chrisideGMT +9.5Pentax SLR talk FAQ http://forums.dpreview.com/forums/read.asp?forum=1036&message=23161072

Thanks for chiming in! Always worth reading.

gaddigadwrote:JNR, FWIW the k-30 RAW samples on dpreview look less noisy at high iso than k-5's. -- Ask not what your camera can do for you, but what you can do for your camera http://www.fluidr.com/photos/gs-photosThey look a little softer to me.

A similar comparison might be made between the K-01 and the K-5. In terms of JPG, the noise measurements favor the K-01 in a significant way. When you look at raw, that gap closes a bit. This speaks to more aggressive NR in the JPG engine rather than a vastly better sensor. But if you are a JPG shooter, its a big deal.It will be interesting to see how much of the K-30's advantage is in the JPG processing vs the sensor, and how that translates into better picture taking / versatility.

chris gunnwrote:gaddigadwrote:JNR, FWIW the k-30 RAW samples on dpreview look less noisy at high iso than k-5's. -- Ask not what your camera can do for you, but what you can do for your camera http://www.fluidr.com/photos/gs-photosThey look a little softer to me.Yes, just on the brink of being soft but not really crossing it in my opinion, as it seems to me d7000's noise at comparable isos for instance looks like it has more texture but does not really add any real, discernible detail to the image. So Pentax is doing a great NR job here saving some dear PP time & dollars for those not inclined to update their NR packages.

Canon girl here...but am very intrigued by the K-30. Unfortunately, I can't find anywhere here that sells Pentax DSLR cameras so I can't go try one out and see how it feels. I don't want to spend $1000.00 only to find out I don't like it and then I have to pay to send it back somewhere. So I guess I'll still with my Canon.

chris gunnwrote:gaddigadwrote:JNR, FWIW the k-30 RAW samples on dpreview look less noisy at high iso than k-5's.They look a little softer to me.From DxOMarks' raw evaluation of the K-01, it looks like the new Pentax hardware "imaging engine" is applying slightly more compulsory high ISO raw Noise Reduction (NR) to the raw files, which is likely what makes the K-30 raw conversions both look a little less noisy and a little softer than those of the K-5 with the same conversion settings.Regards, GordonBGood

eawhitcombwrote:Canon girl here...but am very intrigued by the K-30. Unfortunately, I can't find anywhere here that sells Pentax DSLR cameras so I can't go try one out and see how it feels. I don't want to spend $1000.00 only to find out I don't like it and then I have to pay to send it back somewhere. So I guess I'll still with my Canon.Hi there. Pentax bodies generally feel and handle better than the competition and judging by the body pictures k-30 seems to be very ergonomic. FWIW I bought my own dslr (k100d) after getting an impression only from web reviewers (professional & user reports) and it turned out extremely well. -- Ask not what your camera can do for you, but what you can do for your camera http://www.fluidr.com/photos/gs-photos

Thanks gaddigad! I haven't ruled it out --- I keep a web page opened to the camera/lens kit at B&H just in case. I have until 10/31/2012 before the deal on B&H disappears.