Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
08-25-2010, 08:23 AM   #46
Pentaxian
audiobomber's Avatar

Join Date: May 2008
Location: Sudbury, Ontario
Photos: Albums
Posts: 6,723
QuoteOriginally posted by rparmar Quote
Yes, the K-x has two more stops of ISO, but one could always underexpose the K20D two stops to achieve the same result.
The K-x has one extra stop, ISO 12800. Both have ISO 6400.

08-25-2010, 09:34 AM   #47
Pentaxian
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,686
QuoteOriginally posted by audiobomber Quote
I don't have time to look at the metadata, but AFAIK the Comparator uses the factory defaults for these tests. For the K20D that means NR OFF. For the K-x, it means NR MEDIUM, starting at ISO 800. So the photos you posted are not directly comparable.
True. I had long ago downloaded the RAW files they also make available for these shots, so I've had the opportunity to compare them more directly. I'd say the K-x still beats the K20D at ISO 1600 and ISO 3200 (I actually didn't try 6400), but perhaps by a smaller margin than the Comparometer JPEG's suggest.

In my experience, the color of light and the nature of the scene itself do have a *huge* influence on the appearance of noise, so having more comparisons is always good to provide a more balanced view of the how the cameras might perform in different situations. One camera might do better than another in relatively well balanced light, while another might do better in warmer light (or at least, the gap might be narrowed).
08-25-2010, 09:41 AM   #48
Veteran Member
rparmar's Avatar

Join Date: Jan 2008
Photos: Gallery
Posts: 8,783
Original Poster
QuoteOriginally posted by audiobomber Quote
The K-x has one extra stop, ISO 12800. Both have ISO 6400.
As stated, I was using the extended mode to test the full range the K-x makes available.
08-25-2010, 09:53 AM   #49
Pentaxian
Marc Sabatella's Avatar

Join Date: Sep 2006
Location: Denver, CO
Photos: Gallery
Posts: 10,686
I think the question here is, why didn't you do the same (use extended mode to test the full range) for the K20D as well? Might have been interesting to have seen the K20D at ISO 6400 as well. But heck, I know doing these tests is a lot of work, and tough to get "right" to anyone else's satisfaction.

BTW, regarding the color differences, I'm curious how much of that might have been just a difference in the selected WB. Also regarding the apparent shadow color casts in the full shots you posted above, whether that could have had anything to do with the matter of trying to WB for mixed lighting (flash versus whatever ambient was). I find those sorts of variables very hard to control when i do tests, so usually I punt on it and just go for quick and dirty.

08-25-2010, 10:12 AM   #50
Veteran Member




Join Date: Sep 2007
Location: USA
Posts: 1,812
Like I said any test will have some conditions that do not reflect how one shoots and process - default settings used in established tests may seem on a level playing field but as pointed out camera's default settings may well be different - which leads us to using RAW - but that also opens a whole Pandora's box in post processing variables - so it is down to the individual to do one's own testing - which I have said is the most valid way for that person - but unfortunately may not be applicable for anyone else......

For example I would always try to get the "best" image I could by whatever means I have (and have skills to do) which for one camera may well be different - if I used the "best" processing from one camera and applied exactly the same to another camera's image - that processing may not be optimal - so would that be a level playing field? Not for me - I prefer to get the best/optimal image I possibly can (with the caveat that I have limitations in my skills and tools available) - but my method obviously may not be "level" for anyone else - but it would be to me, as those were the "best" I could do - until I am shown otherwise or learn more - for me that seems "fair" - but as I said YMMV.

Also I have found from a number of experiments on K-x paired RAW+JPG - that is is very hard to beat the K-x JPG rendition - I do realize and acknowledge fully that RAW may be more flexible for difficult conditions - but for my typical photography and some extreme lighting conditions the JPG result seems to be better (for me) than the converted RAW from LightRoom 3 Beta, ACR (Adobe Camera RAW) 5.6 and even Pentax DCU (Digital Camera Utility - based on SilkyPix) 4.11 which seemed best for RAW for me.

Hopefully that is clear enough -

dpReview have their standardized noise tests for the K20D they only did the default JPG but for the K-x they did both standardized JPG and RAW:
dpReview's words on the K-x:
" RAW noise

Finally let's take a look a the K-x's RAW output next to the competition. Removing any in-camera noise reduction and processing the images using Adobe Camera Raw (V5.6 Beta in this case, all NR set to 0) gives us the nearest thing to a 'level playing field' for assessing the relative noise levels of the four cameras' sensors.

With noise reduction turned off we get a more accurate idea of how noisy these sensors are and the image looks slightly different to what we've seen above in the JPEG section of this page.
"

ISO6400 -


ISO3200 -
08-25-2010, 10:31 AM   #51
Pentaxian
audiobomber's Avatar

Join Date: May 2008
Location: Sudbury, Ontario
Photos: Albums
Posts: 6,723
QuoteOriginally posted by rparmar Quote
As stated, I was using the extended mode to test the full range the K-x makes available.
Sorry to belabour, but I don't see it. You said "The fact that the entry-level K-x achieves the feat of matching the previous top-of-the-line body, plus gives two more stops if one needs them..."

Normal range with the K20D is 100-3200. Normal range with the K-x is 200-6400.

Extended range on K20D is 100-6400. Extended range on K-x is 100-12800.

Either way, there's one stop difference in high ISO, not two.
08-25-2010, 11:07 AM   #52
Veteran Member
mysticcowboy's Avatar

Join Date: Apr 2009
Location: port townsend, wa
Photos: Albums
Posts: 968
rparmar I just want to thank you for all the time you put into this test and the willingness to retest.
08-25-2010, 11:10 AM   #53
Veteran Member
rparmar's Avatar

Join Date: Jan 2008
Photos: Gallery
Posts: 8,783
Original Poster
QuoteOriginally posted by Marc Sabatella Quote
I think the question here is, why didn't you do the same (use extended mode to test the full range) for the K20D as well?
Oh dear, all this time and I didn't know you could do this with the K20D.

QuoteOriginally posted by audiobomber Quote
Either way, there's one stop difference in high ISO, not two.
As above.

08-25-2010, 11:13 AM   #54
Veteran Member
rparmar's Avatar

Join Date: Jan 2008
Photos: Gallery
Posts: 8,783
Original Poster
QuoteOriginally posted by Marc Sabatella Quote
BTW, regarding the color differences, I'm curious how much of that might have been just a difference in the selected WB.
I kept both on Auto WB and colour mode (or whatever it is called) on Portrait, same way I shoot every single shot in my life. Every (serious) shot I take is also RAW. Thus how a shot looks has more to do with PP than in-camera settings.

QuoteOriginally posted by UnknownVT Quote
Like I said any test will have some conditions that do not reflect how one shoots and process - default settings used in established tests may seem on a level playing field but as pointed out camera's default settings may well be different
Yep, establishing a baseline is a problem. I prefer to test with cameras set up as I normally use them, which I hope is optimised for the best possible shot for me. Defaults are meaningless.

For the tests I turned any enhancement off and kept all sliders at baseline in ACR, which is certainly not what I would normally do.
08-25-2010, 11:50 AM   #55
Veteran Member




Join Date: Sep 2007
Location: USA
Posts: 1,812
QuoteOriginally posted by rparmar Quote
Yep, establishing a baseline is a problem. I prefer to test with cameras set up as I normally use them, which I hope is optimised for the best possible shot for me. Defaults are meaningless.

For the tests I turned any enhancement off and kept all sliders at baseline in ACR, which is certainly not what I would normally do.
Which as I said is fine for you - and no disrespect meant - there will always be others who may think those are are still not "level".

Default settings are very meaningful for a lot of people (perhaps not for us here) as I would say the majority do buy and just use the dSLRs without changing many settings. Also another way of thinking - is that default settings are what the manufacturers think is the best image they can offer (realizing their cameras are going to be thoroughly tested with those images) - of course we are unlikely to accept that - and will think we can do better .

The other thing is using ACR at baseline settings - may be open to question as ACR's default (sorry about this "nasty" word) is to offer "As Shot" - which as far as I know means to apply the camera settings to the image "As Shot" - implying it tries to get the image as close to the "As Shot" paired JPG (if taken) as possible.

This is certainly the case for Pentax DCU for example, when any RAW image is opened the "default" processing offered is "Camera Setting" - once again it tries to give the closest rendition to the paired JPG image......

Another point is about ACR - which I do acknowledge as the de-facto industry standard and I use it - but this is what Imaging-Resource say about RAW conversion - taken from their review of the K-x on the HighISO RAW conversions but I'm pretty the same words are used elsewhere (eg: it's the same words in the Canon T2i review)

" We've recently started looking at RAW files converted with dcraw, an excellent freeware raw converter. dcraw usually offers timely support for the latest cameras, but more importantly, it does not apply any noise-reduction, sharpening or other corrections such as geometric distortion correction to the output files. (We found that Adobe Camera Raw still applies some limited noise-reduction when its NR settings are set to zero, and it also applies other corrections depending on the make and model of the camera). There will always be differences between RAW converters, in terms of the sort of demosaicing algorithms they use (the processes by which they convert the separate Red, Green, and Blue data sets to an array of full-color RGB pixels), but dcraw seems to use a fairly generic algorithm that delivers good sharpness with relatively few artifacts, and can be counted on to not apply any noise reduction if you don't want it to. "

Like I have said - there will always be some objections and you just cannot please everyone - however I do applaud your attempts to test for yourself and even responding to the myriad of objections/suggestions.

However when it comes down to it I make my purchase decisions based on those review images.
08-25-2010, 12:15 PM   #56
Pentaxian
audiobomber's Avatar

Join Date: May 2008
Location: Sudbury, Ontario
Photos: Albums
Posts: 6,723
QuoteOriginally posted by UnknownVT Quote
However when it comes down to it I make my purchase decisions based on those review images.
Yeah, you have to look at lots of reviews and samples. What better method is there? But even people who look at lots of reviews and samples don't necessarily know what to look for. Noise is easy to spot, detail comparisons are trickier, settings and processing can make a night and day difference and can mislead.

AFAIC, the default NR settings on the K-x are unusable. At high ISO they produce the smoothed plasticky look that I detest. It's hard to see that from review images. I didn't notice it until after I'd used the camera for a while. I find the little DPR stamp photos particularly useless for judging real world IQ.
08-25-2010, 12:30 PM   #57
Site Supporter
Site Supporter
GeneV's Avatar

Join Date: Oct 2008
Location: Albuquerque NM
Photos: Albums
Posts: 9,775
Interesting heads up about dcraw. I've always suspected that some NR was happening in ACR.
08-25-2010, 01:10 PM   #58
Moderator
Site Supporter
Blue's Avatar

Join Date: Jun 2008
Location: Florida Hill Country
Photos: Gallery | Albums
Posts: 17,213
QuoteOriginally posted by rparmar Quote
Oh dear, all this time and I didn't know you could do this with the K20D.



As above.
They hid the access to the extended mode on the K20d. Its a lot like the menu setting to allowing aperture ring control.
08-25-2010, 01:33 PM   #59
Veteran Member




Join Date: Sep 2007
Location: USA
Posts: 1,812
QuoteOriginally posted by audiobomber Quote
Yeah, you have to look at lots of reviews and samples. What better method is there? But even people who look at lots of reviews and samples don't necessarily know what to look for. Noise is easy to spot, detail comparisons are trickier, settings and processing can make a night and day difference and can mislead.

AFAIC, the default NR settings on the K-x are unusable. At high ISO they produce the smoothed plasticky look that I detest. It's hard to see that from review images. I didn't notice it until after I'd used the camera for a while. I find the little DPR stamp photos particularly useless for judging real world IQ.
Like I have tried to make clear throughout all my posts in this thread - this is really a case of "to each their own".

However I like the images I used - as they are the ones I look to to make a decision ... so I guess I base my decisions on some "particularly useless" images ..... (oh well, I hate having to sit in the remedial class - again? )

Here are the RAW converted images (w dcRAW) from Imaging-Resource.com respective reviews of the K20D and K-x

ISO6400:




ISO3200:





QuoteOriginally posted by GeneV Quote
Interesting heads up about dcraw. I've always suspected that some NR was happening in ACR.
I haven't quite figure out how to use dcRAW yet.

It is I think (in Windows) a DOS command line program.

I have not figured out how to apply the parameters and see the resultant change.

I have managed to drag and drop a K-x DNG/RAW file into dcRAW.exe and the result is in default .ppm format - and looks like no parameters are applied - since that DNG was shot under household lighting - and the result looks like tungsten lighting shot with Daylight WB - unlike the paired JPG or any of the conversions from ACR 5.6 or Pentax DCU 4.11.

Last edited by UnknownVT; 08-25-2010 at 01:39 PM.
08-25-2010, 02:33 PM   #60
Veteran Member
rparmar's Avatar

Join Date: Jan 2008
Photos: Gallery
Posts: 8,783
Original Poster
QuoteOriginally posted by UnknownVT Quote
Which as I said is fine for you - and no disrespect meant - there will always be others who may think those are are still not "level".

Default settings are very meaningful for a lot of people (perhaps not for us here) as I would say the majority do buy and just use the dSLRs without changing many settings.
No disrespect taken at all. On person's level playing field is another person's quagmire.

Defaults may in fact be optimised for "the majority" who will never change a setting, but then they are not the ones who will be nit-picking between 100% crops on a forum like this! In other words, I would not expect any serious reviewer to stick with the defaults if they can take some time to find "better" settings (using whatever measure of "better" suits them).

However, I leave the minutiae of RAW converter comparison to others. Frankly, I can't see it being worth my time. In the same way that there may be small differences between ACR and something else, there will be between Noise Ninja and something else, or even -- as I might be concluding -- between the K-x and something else. But these differences are not great enough to have a decisive impact on a picture, above and beyond the usual factors: subject, light, composition, optical parameters, etc.

I hope it is enough to provide some further test images to supplement the many already out there already. I leave the details to the experts!
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
camera, dslr, iso, k-x, k20d, ninja, noise, photography, sensitivity, settings, shot, shots
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Interesting test - k7 versus k10. ISO800 Dubious Drewski Pentax DSLR Discussion 22 09-02-2009 09:10 PM
k20d High ISO test shots 200-6400 PentaxPoke Pentax DSLR Discussion 44 04-27-2009 08:02 PM
k10d vs k20d ISO 1600 test PentaxPoke Pentax DSLR Discussion 9 05-11-2008 02:36 PM
K20D vs A700 high ISO test (not 56K friendly) codiac2600 Pentax News and Rumors 19 04-21-2008 09:36 AM
Unscientific test of firmware versus autofocus Rick Pentax DSLR Discussion 17 05-13-2007 11:41 PM



All times are GMT -7. The time now is 08:06 PM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top