Web Analytics RiceHigh's Pentax Blog: Noise/Details Shootout and Quick Comparisons: K-7 Vs K20D Vs 50D

Friday, July 10, 2009

Noise/Details Shootout and Quick Comparisons: K-7 Vs K20D Vs 50D

The Japanese Digital Cameras Information Blog has posted a shootout with detailed head-to-head comparisons for the three cameras. Crops of original image samples taken by the Imaging Resource and posted in their reviews are pasted together one by one for easy and quick inspection and review:-

http://informationondigitalcamera.blogspot.com/2009/07/pentax-k-7-vs-pentax-k20d-vs-canon-eos.html

It should be noted that Noise Reduction has been turned off for each camera and high ISO images from 1600 to 6400, which are direct Jpegs from the three cameras, are compared.

It can be seen from the many comparisons that the K-7 is most noisy at all ISO speeds but details could be less. The 50D images are the least noisy and more details are retained. Not toally surprisingly, for what people have posted so far, the K20D actually performs better than K-7 for both noise control and details retained, in general. For a few crops, the K-7 and K20D both retain some more details than the 50D but not so for most cases.

Well, maybe you can just go view and measurebate yourself and compare endlessly before you would know the differences! ;-) I think that Blog has done a great job in making those easy and quick comparisons, which are just very useful afterall. Well Done, Guy(s)! (Thumb Up)

25 comments:

  1. Anonymous10/7/09 14:10

    http://3.bp.blogspot.com/_eQLWnTU0IRM/SlSzFz95GzI/AAAAAAAACgU/-kr_ciBYLVs/s1600-h/1600d.jpg

    where is more details of 50D?

    different objects from different cameras are different

    ReplyDelete
  2. Anonymous10/7/09 14:12

    who care about in-camera JPEG at ISO3200/6400?

    ReplyDelete
  3. > where is more details of 50D?

    View the red object, the 50D has more details. For the pink background, the K20D wins. Noise wise, the K-7 is most noisy.

    > who care about in-camera JPEG at ISO3200/6400?

    You should ask: who pretends not to care? Ans: Fanboys who do not want to see or just hear anything that is "negative" and unfavourable to "their" brand!

    ReplyDelete
  4. Anonymous10/7/09 15:18

    Comparison article at the same blog to compare k-7, Nikon D90 and Canon 50D.

    http://informationondigitalcamera.blogspot.com/2009/06/pentax-k-7-vs-nikon-d90-vs-canon-50d.html

    ReplyDelete
  5. Anonymous10/7/09 15:33

    Might be from "obscures chinese sites", this is quiet a bundle of bad marks for the k-7.
    While we europeans are still waiting for thorough testing from our "unquestionable" print medias, I'm sure manny of us wonder :
    Is there really something wrong with this new model ?
    Rice, 2 questions :
    1 - Are you on a mission for Canon ?
    2 - If not, What is your innermost opinion ? Can you truly believe Pentax botched the K-7 so much ?
    My Flickr

    ReplyDelete
  6. The test images are from the Imaging Resource: original Jpegs, NR off. Have you ever read? Unable to face the truth??

    ReplyDelete
  7. Anonymous10/7/09 15:54

    You'r bashing your readership.
    At least that looks like an answer to question 1 :-)
    Thanks you very much.

    ReplyDelete
  8. Hi Rice,

    It might surprise you, but actually I like the k-7 results better:
    -Pentax seems to have taken a "film grain like" approach to control the noise
    -The very good thing is that there is almost no color noise in grey/black areas (no green or red spots in these areas). I find color noise very annoying in pictures from the k20d and the 50d (and my k10d)
    -If you want to have the same results as with the k20d and the 50d, you probably just have to apply to smoothness. The details are much more present in the k7 images

    So I think the k7 will give much nicer prints than the k20d and the 50d, with a delicate taste of film grain. Much better imho than classical noise patterns from other dslr....

    Best,
    Paul

    ReplyDelete
  9. > Leodileo said...
    > You'r bashing your readership.
    > At least that looks like an answer to question 1 :-)

    Well, at least I am answering your bashing questions!

    As for question 1, that's clearly an uncivilised personal attack and libel. But the truth is obvious. I paid both Canon and Pentax. I am the customer of both companies. I paid Pentax for long, much more than what I bought Canon's products.

    ReplyDelete
  10. I see that noise is different in all three cameras at ISO 6400, but I can't see that the 50D produces less noise generally. The K-7 seems to be the best in terms of color noise at the expense of luminance noise (best seen here).

    I find the noise level in all three cameras is acceptable. That's what counts in the end.

    ReplyDelete
  11. The image linked by schmuki clearly speaks for itself: k-7 result is better (more details, no color noise, grain).

    http://2.bp.blogspot.com/_eQLWnTU0IRM/SlSztjIFziI/AAAAAAAAChk/fAW7vIdNwss/s1600-h/6400h.jpg

    I am beginning to wonder if I should not upgrade from the k10d, as I finally see a clear upgrade in term of noise control.

    However I would like to check some high iso raw files.

    Paul

    ReplyDelete
  12. Anonymous10/7/09 17:02

    Look at brush in the first picture, ISO 1600!

    With the K-7 the fine hair details from the brush is reproduced best!

    And who cares about 3200/6400?

    ReplyDelete
  13. Who care about in-camera JPEGs at all.
    I use Raw Therapee for all my pictures and in comparsion with original JPEGs from my Samsung GX20 (even Silkypix RAW developer) it´s another league. I don´t have K20D, but from what I see on internet, JPEGs are not so much different from GX20.
    Undoubtedly , Canon has best denoising algorithm. Before I found Raw Therapee, I even used Canon DPP for denoising all my JPEGs.
    Comparing JPEGs from camera means comparing in-camera raw developers. When you use one raw developer, the differences are much lesser.
    I´m not Pentax nor Samsung fanboy, but I am very satisfied with IQ which i can get from RAWs.
    K7 is letdown for me. Sooo much useless features (video, HDR JPEGs - why not HDR RAW?), but nothing important (better dynamic range and lower noise sensor). This does not mean it´s bad camera, just I don´t think it´s worth upgrading.
    And the Canon thing... My friend has got 500D (its same price range as K20D here). In comparsion to my GX20 it´s child toy, with lesser features, worse IQ (even with higher MPx) and without in-camera IS (very important feature to me, I have 20+ M42 stabilised lenses with my GX20 :-). The only plus on Canon side is HD video.
    After K7 release, my GX20 still takes great pictures for me, thus I will wait for next offering from Pentax or Samsung...
    PS: sorry for my bad english.

    ReplyDelete
  14. Anonymous10/7/09 18:28

    Didn't mean to be any offensive, Rice.
    I tend to be permeable to all this comments about your integrity, but this has nothing to do in that comment thread.

    ReplyDelete
  15. Anonymous10/7/09 20:35

    I would like to point out what this camera is capable of in the hands of a talented photographer (or in this case, cinematographer.)

    http://vimeo.com/5520766

    Pixel-peeping is no substitute for having a good eye.

    ReplyDelete
  16. Anonymous10/7/09 23:59

    Well, what this test shows is, that there is hardly any noticable difference in noise performance. Come on, the small difference you imagine to see might be a result of the optics or the exact alignment of the camera towards the subject and/or focus differences. On average all three cameras show pretty much the same noise level.

    ReplyDelete
  17. My pardon? The noise levels are the "same"???

    ReplyDelete
  18. Anonymous11/7/09 00:41

    No they are not but I consider the difference to be not relevant under field conditions. At ISO 6400 they are all above the edge.

    ReplyDelete
  19. Anonymous11/7/09 04:21

    DPREVIEW review of GH-1 is up

    http://www.dpreview.com/reviews/panasonicdmcgh1/

    Resolution:

    http://www.dpreview.com/reviews/panasonicdmcgh1/page28.asp

    Video Mode:

    http://www.dpreview.com/reviews/panasonicdmcgh1/page19.asp

    "The GH1/14-140mm combination's silent and quick contrast detect AF, coupled with ergonomics that have been designed with video in mind, certainly take it beyond the current crop of video-enabled DSLRs which suffer from clunky video controls and largely unusable AF systems. Like those DSLRs though it can shoot video with the shallow depth-of-field that you would only get from significantly more expensive large-sensor camcorders, and for this reason alone it will certainly find its way into the equipment bags of more than one professional
    videographer - if not as the main, at least as an additional tool."

    "As we've said before we don't review dedicated high-end video cameras, and therefore can't really comment on how well the Panasonic GH1 video quality
    compares to this type of equipment, but we can say that compared to the video-enabled DSLRs we have reviewed so far the GH1 is up there with the best."

    Conclusion:

    http://www.dpreview.com/reviews/panasonicdmcgh1/page30.asp

    Detail (D-SLR) Rating (out of 10)
    Build quality 8.5
    Ergonomics & handling 8.5
    Features 9.5
    Image quality 8.5
    Performance (speed) 9.0
    Value 8.0

    Highly Recommended

    My analysis - for someone wanting real video capability, all of the DSLR's including K-7 are well behind AT THE MOMENT. I doubt that a majority of the PK legacy set care about video, which leaves the K-7 at $1300 for the body only.
    That creates an opening for very competitive pricing on the Samsung GX-30, whenever it gets here)

    ReplyDelete
  20. Anonymous11/7/09 06:14

    Below are some observations I made of an Imaging Resources review of the Canon 1200. The image quality on IR's test appeared worse than the IQ of the earlier 8mp 1100. I sent my comments to IR and they acknowledge they might have a "bad copy" of the 1200 and will retest with another. Not sure if they did or not.

    The copy of the 1200 I purchased did not have the same soft focus problem as their test 1200.

    This is to say that testing can have unaccounted for variables that can lead to misleading results about the performance of an individual camera.

    I don't think we will have adequate knowledge of the IQ of the K7 until we have full blown tests from 3 or 4 reputable reviewers using multiple copies of the camera.

    Here are several forum comments on IR concerning the Canon 1200 reivew:


    ********** - 01:48am May 28, 2009 EST

    I compared several images shot with the new 10 mp Canon Powershot SD 1200 with 5 other cameras in the 8 to 12 mp range and $150 to $240 range. These included a Kodak, a Pentax, a couple of Nikons, and the prior 8 mp model Canon SD 1100. Every one of these OTHER cameras had MARKEDLY superior (cleaner, sharper, less fuzzy) photos than the new SD 1200. I gave up comparing - it appeared that virtually every camera over $150 in price exhibited cleaner photos than the SD 1200.

    There are four possible explanations for this:

    1) Canon screwed up with image quality on this camera
    2) A bad unit was used for the tests
    3) The images are misclassfied on the IR web site
    4) I misread the results (I doubt this is the case - I double and triple checked the cameras and results at 80, 100, and 200 ISO)

    Does anyone have an opinion about my findings?





    < Previous Discussion | Next Discussion >



    --------------------------------------------------------------------------------


    ********* - 02:09pm May 28, 2009 EST (#1 of 4)
    Image Resources (IR) confirmed the very (unacceptably, in my opinion) soft images on the left half of the photos taken with the 10 mp Canon SD 1200 IS Elph camera for their tests. The images of the camera it replaced, the 8 mp SD 1100 had better images overall than what is indicated by the new SD 1200 test photos.

    The facts to be determined are:

    Did IR have a defective copy of the camera shooting the test photos?
    or
    Is this what is to be expected of ALL SD 1200s?

    IR said they will shoot another series of test shots with another copy of the SD1200.

    It is apparently not uncommon for just about any make digital camera, mostly the smaller sized pocket versions, to have an alignment problem concerning either the sensor or the lens that causes uneven focus across the image.



    maxx - 08:09pm May 28, 2009 EST (#2 of 4)
    Yes, I think that's very common. I exchanged my Canon A720IS twice before getting one I was happy with, but I suspect most buyers wouldn't have. It's actually good to know sample cameras provided for testing aren't being cherry-picked by the manufacturers.

    It'll be interesting to see how your copy performs.



    ********** - 01:41pm May 30, 2009 EST (#3 of 4)
    Here is an update on the performance of the Canon SD 1200 IS I just received.

    The image had uniform focus across the photo. It's sharpness was a bit better than that of my Pentax K100D 6mp, understandably due to the 4mp difference. The depth of color was a bit better on the Pentax. But I think if I tweak the exposure down on the Canon 1/3 to 2/3 of a stop, the color depth would improve.

    I am very happy with the performance of the SD 1200.

    ReplyDelete
  21. Anonymous11/7/09 06:36

    I just finished doing a couple of quick checks on IR's website, using their "Comparomter" to compare the K7 with the K200D (10mp).

    I noted the following inconsistencies in the set ups for the photo comparisons in their salt and pepper grinder images:

    1) The angle from which the photo was taken is greater with the K7 - it looked "down" on the objects; the K200D look more straight ahead.

    2) Reflections were different. Either the lighting was different or the angle of view caused a shift of reflection on reflective objects.

    Whether caused by these and other differences in testing set-up or procedures or caused by differences in lens or camera design, at ISO 200 (the only comparisons I made) the K200D appeared sharper.

    ReplyDelete
  22. Anonymous11/7/09 06:44

    This is why I leave it to the reviewers - they will often kick the lens or camera back for a substitute, rather than do a review that is poor.

    DPREVIEW did this with K20 because of the hot pixel issue, and it delayed the full review. The whiners complained about the delay, but DPREVIEW people got on the forums and explained what happened.

    Some of the Pentax Groupies act as though they're entitled to special treatment - but it was Pentax' problem, not DPREVIEW.

    MY qualms on K-7 are not about the IQ, but about the capabilities - and what I see as needing Samsung to preserve the PK mount. Not every Canon nor Nikon meets my every criteria, but there's a broad model range, and they will most certainly be in business years down the road, as will their lens mounts.

    I see no reason to settle for less than OIS lens capability, period. I see no reason, for the low-light shooting when I NEED it, to settle for some ancient AF setup and s/s. Your "good enough" may not be someone else's "barely adequate".

    I see no reason to be forced to buy an external mic for video+sound, when the GH1 has a port for an optional mic, but does not require it to operate.

    I see no reason not to be able to mount a modern RING-USM lens.

    In short, I live in 2009, not 1999.

    ReplyDelete
  23. Anonymous11/7/09 06:48

    I came across this awhile ago - it's a slide-show presentation. It's 100 slides long, and the keyboard right-arrow will step them along using the viewer. The first 45 or so are about Pentax' product history, so those who enjoy that will get a taste. Then comes the digital period, and he identifies how they got to where they are through shared technologies, including some management quotes.

    Around slide 86, he gets to the
    core argument.

    There is a full-screen option, and ESC would drop you out at anytime to the prior state. I would recommend using right-arrow to step through it, as the auto-player-mode on auto is a bit slow - somehow reminiscent of the AF system to me, oddly - so I manually stepped through it

    http://www.slideshare.net/Christiansandstrom/disruptive-innovation-and-pentax

    ReplyDelete
  24. Anonymous11/7/09 12:29

    Hey, guys...
    K-7 is not better than K20D in terms of noise. Ths structure of noise is different, because the sensor of K-7 is not the same as K20D's. Maybe weaker AA filter.
    But, don't wait for miracles. It's 14.6 MP APS-C camera. Not FF.

    ReplyDelete
  25. RH, thanks for the link. Unfortunately, the in-camera JPG engines were used and even with NR off, there will always remain some NR being applied.

    Also, from the date of test, the images are from a pre-production K-7 which I had the opportunity to test against a K20D as well. I made a comparison based on RAWs and found no significant differences between K-7 and K20D. You may look up the sample images on my blog.

    ReplyDelete