Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
11-30-2009, 05:28 PM   #1
Forum Member




Join Date: Jan 2009
Location: Berlin, Germany
Posts: 63
Kx / K20D comparison on high ISO

As I just evaluate Kx as a companion to my beloved and often critisized K20D, I thought I might share a few results with you. As I was mostly interested in high ISO performance (on of the lesser strenghths of this model) and didn't like the results I saw from Kx initially due to its obvious NR influence, I directly deactivated it and shot RAW to see the images as straight from the cam as possible and viewed them in PS with prerelease Camera Raw 5.6 or whatever the verion is, that includes support for Kx. As I saw no relevant difference from the JPGs, the cam produced, I switched to JPGs with natural setting on both, K20D and Kx. I yet have to admit, that I left a color correction on in K20D where I tweaked color tone +1 to make red appear warmer and more natural in my eyes. But evaluating color wasn't my target anyway. To have both cams shoot at 6400, I set f-stop to 11 using an FA50 1.4 on both. Both were in Av-mode, although I had to give K20D an extra push of +2EV to achieve about the level of brightness of Kx. So actually K20D tends to underexpose under dim conditions heavily, what had been fixed in Kx. For what IMO is about the same level of high light rendering, K20D took 1/15th sec., Kx only 1/25th (yes, I ommitted using a tripod, hence the scene isn't exactly the same, which also might influence lighting). Kx rendered shades much better than K20D giving a wider DR, can see color where K20D just produces grayish grain. More convenient color rendition however seems to be easier with K20D, as it represents my view on the scene much better even with AWB whereas I chose warm white fluorescent tube on Kx.
I like K20D as it fits my hands perfectly. Furthermore, I very much like the fine grainy film like look all images from that model have even al lower ISO; wouldn't call that noise. For certain occasions, however, I would prefer Kx over K20D, because it is small and better in liw light. Initially I thought I could get along w/o it, now I feel I have to keep it.
Fotos attached named accordingly, complete diminished complete shot and crop.

Attached Images
 
View Picture EXIF
PENTAX K20D  Photo   
View Picture EXIF
PENTAX K-x  Photo 
11-30-2009, 06:56 PM   #2
Veteran Member




Join Date: Mar 2007
Location: Toronto (for now)
Posts: 1,748
One day, someone, somewhere will get the idea that you need to MIMIC the settings to do ao valid test. You can't shoot JPG and you can't damn well post push by 2ev and wonder why the image is noiser than the other.

Man.

Manually set the bloody exposure. Please.
11-30-2009, 07:09 PM   #3
Veteran Member




Join Date: Jun 2009
Posts: 11,913
QuoteOriginally posted by Alfisti Quote
One day, someone, somewhere will get the idea that you need to MIMIC the settings to do ao valid test. You can't shoot JPG and you can't damn well post push by 2ev and wonder why the image is noiser than the other.
Man.
Manually set the bloody exposure. Please.
+1 re your comments on the testing procedure.

PS - use of 'bloody' = are you an Aussie?
11-30-2009, 07:19 PM   #4
Veteran Member




Join Date: Mar 2007
Location: Toronto (for now)
Posts: 1,748
yeah another JAFA.

11-30-2009, 07:34 PM   #5
Ash
Loyal Site Supporter
Loyal Site Supporter
Ash's Avatar

Join Date: Mar 2007
Location: Toowoomba, Queensland
Photos: Gallery | Albums
Posts: 23,920
So the first two are from the K20D and the last two from the K-x?

QuoteQuote:
Both were in Av-mode, although I had to give K20D an extra push of +2EV to achieve about the level of brightness of Kx
This is way too crude a measure of "brightness" as you call it. Every setting - and that includes, WB and NR, *must* be the same between cameras, and *must not* be on any auto settings for any credible comparisons between sensors in performance.

Try it again if you have the time.
11-30-2009, 08:14 PM   #6
Junior Member




Join Date: Aug 2009
Location: Melbourne
Posts: 25
I don't have experience with this but would I be close suggesting that use of a manually set aperture (mechanically on the lens) would also eliminate any possibility the camera could purposely or otherwise set the aperture differently between cameras?

I expect the shutter time is the most acurate between cameras, so using the same shutter speed I'd then adjust the camera ISO to achieve the same exposure for comparison.

From what I've read, the ISO calibration (unlike film) is the least likely to be consistent and is apparently often used by the manufacturer to "adjust" exposure for their own reasons. Setting the same ISO on each camera seems to be starting from the wrong point.

Once these three exposure variables are made the same, only then could the results be evaluated.

The fact that you might arrive at a different value for ISO is of little interest when actually taking photos because you are only interested in adjusting the variables to capture the image. So whether one camera's 6400 ISO is another's 3200 is not as important as is whether one camera can capture the image with more or less noise (or DR or whatever you want to compare) than another, in identical conditions - i.e., those that exist when you take the picture.

Last edited by _phil_; 11-30-2009 at 08:25 PM.
12-01-2009, 01:44 AM   #7
Forum Member




Join Date: Jan 2009
Location: Berlin, Germany
Posts: 63
Original Poster
QuoteOriginally posted by Alfisti Quote
One day, someone, somewhere will get the idea that you need to MIMIC the settings to do ao valid test. You can't shoot JPG and you can't damn well post push by 2ev and wonder why the image is noiser than the other.

Man.

Manually set the bloody exposure. Please.
seems you got me wrong (prop. used wrong term; am not native speaker, as you may assume). I set +2 exposure correction ON THE CAMERA itself, not by PP. K20D automatic exposure was way too dim to compare. See pic enclosed, which is w/o exposure correction.
May redo some of the stuff when I have time.

Attached Images
View Picture EXIF
PENTAX K20D  Photo 
12-01-2009, 04:43 AM   #8
Veteran Member




Join Date: Jul 2008
Location: Sweden, Umea
Posts: 876
Dont shoot stright into a bright lamp
12-01-2009, 05:46 AM   #9
Veteran Member




Join Date: Jun 2009
Location: Bridgetown West Australia
Photos: Gallery | Albums
Posts: 903
I took some photos today with my K20 accidently set to ISO 800, when I went to PP them I couldnt beleive how noisy they are, almost unusable, you certainly couldn't print them any larger than about 8x10. I am guessing that as the shutter speed was fast (this was in quite good light) that the camera didn't perform a DFS. Even so I wasn't very impressed.
12-01-2009, 05:57 AM   #10
Loyal Site Supporter
Loyal Site Supporter




Join Date: Mar 2009
Location: Gladys, Virginia
Photos: Gallery
Posts: 27,602
QuoteOriginally posted by ozlizard Quote
I took some photos today with my K20 accidently set to ISO 800, when I went to PP them I couldnt beleive how noisy they are, almost unusable, you certainly couldn't print them any larger than about 8x10. I am guessing that as the shutter speed was fast (this was in quite good light) that the camera didn't perform a DFS. Even so I wasn't very impressed.
This is hard for me to believe. I shoot to ISO 1600 routinely and often run a little bit of noise reduction on them, but I wouldn't need to. The only way that I see noise is if I have significantly underexposed and am trying to bring back the photo or, if I have the enhanced dynamic range function on (I don't use it in general -- adds a lot of noise to the shadows). Dark frame subtraction is not so much about noise as it is about hot pixels from long exposures (greater than 5 or 30 seconds depending on your firmware).
12-01-2009, 06:16 AM   #11
Veteran Member




Join Date: Mar 2007
Location: Toronto (for now)
Posts: 1,748
QuoteOriginally posted by ozlizard Quote
I took some photos today with my K20 accidently set to ISO 800, when I went to PP them I couldnt beleive how noisy they are, almost unusable, you certainly couldn't print them any larger than about 8x10. I am guessing that as the shutter speed was fast (this was in quite good light) that the camera didn't perform a DFS. Even so I wasn't very impressed.

ISO800 ......


12-01-2009, 08:06 AM   #12
Veteran Member
er1kksen's Avatar

Join Date: Oct 2008
Location: Forestville, NY
Photos: Albums
Posts: 1,801
QuoteOriginally posted by Rondec Quote
This is hard for me to believe. I shoot to ISO 1600 routinely and often run a little bit of noise reduction on them, but I wouldn't need to. The only way that I see noise is if I have significantly underexposed and am trying to bring back the photo or, if I have the enhanced dynamic range function on (I don't use it in general -- adds a lot of noise to the shadows). Dark frame subtraction is not so much about noise as it is about hot pixels from long exposures (greater than 5 or 30 seconds depending on your firmware).
It all comes down to noise tolerance and how much you push your exposures around in post. For sports shots on my K20D I would use 1250-1600 no problems, but for landscapes or portraits anything above 200 felt intolerable.
12-01-2009, 08:39 AM   #13
Veteran Member




Join Date: Mar 2007
Location: Toronto (for now)
Posts: 1,748
QuoteOriginally posted by kasv Quote
seems you got me wrong (prop. used wrong term; am not native speaker, as you may assume). I set +2 exposure correction ON THE CAMERA itself, not by PP. K20D automatic exposure was way too dim to compare. See pic enclosed, which is w/o exposure correction.
May redo some of the stuff when I have time.
OK I'll cut you some slack on the language thing. It's just getting to the point where we get thread after thread of people doing back to back tests and not a soul seems to do it right.

You need a tripod, you need to set exposure manually, you need to use various apertures, you need to shoot RAW and process through the same converter and apply similar levels of sharpening and NR, you need to up or down sample is two cameras have different pixels counts.
12-01-2009, 09:04 AM   #14
Veteran Member
audiobomber's Avatar

Join Date: May 2008
Location: Sudbury, Ontario
Photos: Albums
Posts: 6,806
QuoteOriginally posted by kasv Quote
. Both were in Av-mode, although I had to give K20D an extra push of +2EV to achieve about the level of brightness of Kx. So actually K20D tends to underexpose under dim conditions heavily, what had been fixed in Kx.
I don't think that's a proper explanation of what's happening. The K20D is exposing to avoid blown highlights. This is the correct way to expose IMO, but Pentax was constantly criticised for "underexposure" by noobs and reviewers. It appears they have changed the processing in the K-7 and K-x to be more newbie friendly, like Canikon.

I agree with you about the grain-like "noise" in the K20D photos. It doesn't bother me most of the time. I'd like to see this test redone in Manual mode as others have stated, but at 3200 and 6400. I don't use ISO 6400 on the K20D because it's just 3200 pushed by the processing engine. 6400 is pretty ugly, I've never used it in real life, only in test shots.
12-01-2009, 09:11 AM   #15
Veteran Member




Join Date: Mar 2007
Location: Toronto (for now)
Posts: 1,748
QuoteOriginally posted by audiobomber Quote
I don't think that's a proper explanation of what's happening. The K20D is exposing to avoid blown highlights. This is the correct way to expose IMO, but Pentax was constantly criticised for "underexposure" by noobs and reviewers.
I disagree strongly, the old Pentax metering would completely PANIC at the slightest hint of a highlight in the frame, it would force you to pull the rest of the image in pping as the mid levels were almost always underexposed. Very few scenes don't have an area that can't be lost to shadows or highlights without damaging the image, this obsession with keeping everything in the frame within clipping points leads to dull images.

My default Ev comp is +0.3 to +0.7 on the 20D as it will otherwise underexpose.
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, color, dslr, images, iso, k20d, k20d comparison, kx, kx / k20d, level, light, photography, scene
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
K-r vs K-x high ISO comparison soalle Pentax News and Rumors 7 11-01-2010 04:55 AM
K7 high iso noise comparison cyy47 Pentax DSLR Discussion 24 08-02-2010 02:19 AM
High ISO comparison, Pentax KX and Canon T1i interested_observer Pentax DSLR Discussion 4 05-07-2010 12:41 PM
K-7 - K20D High ISO comparison Shots Ray Pulley Pentax News and Rumors 56 06-11-2009 01:47 PM
K100D & K10D high ISO comparison switters Pentax DSLR Discussion 3 12-06-2007 07:07 PM



All times are GMT -7. The time now is 06:12 AM. | 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