[Coeurdechene]

Pixel peepers anonymous here...

Unless you are going to make BIG prints, for the normal resolution at wich we look at the images that is a very nasty habit
Even with big prints you are supossed to see them from a distance so you won't notice most of what pixel peepers hunt as "defects".

But for your habit there was a comparison somewhere in the forums with images and all the parafernalia needed to satisfy the pixel addiction.

(Acknowledging it's always the first step)

PDh and even if Kx sensor was better the K7 has some other features that may balance the choice her way, especially if you are not a night shooting fan.

[Wheatfield]

PM me your email address and I'll send you a raw file or two.

[v5planet]

Sure you can fix all sorts of things on post processing, but that takes time and only improves PERCIEVED quality.

I find the latter half of this statement to be an extremely strange notion. Yes, it takes time to remove noise in a satisfactory way from an image with software, and even if the results are good, it may not be worth it to someone to go through the hassle. And in some cases an image is just going to be too noisy to rescue with software without a notable degradation of overall detail.

But what's this business about "Perceived quality"? What does that even mean? If you are willing to take the time to edit a photo at all you are implicitly stating that the raw produced by the camera is just that - a raw material that you will mold into a final image. If your post processing includes noise reduction, and the result is something that you "perceive" as a beautiful image, can you tell me what exactly you are losing out on?

Unless you are a no-PP purist, any other editing you could possibly do to enhance your raw photo smacks of a strange double standard. For example, is playing with curves to make a more striking, contrasty image merely affecting its "perceived quality"?

You are right though in a sense: the quality of an image is in how it's perceived by the viewer. So who is your audience -- the camera itself? Sadly, all it perceives is electron volts.

---

More on the topic of your question: I agree with Wheatfield's assessment. I tend to manually set my sensitivity so I have little experience with the range intermediate between ISO 400 and ISO 800, but somewhere in that gulf there is a large increase in noise. Though I still find ISO 800 quite acceptable because the noise has a grainy quality. By ISO 1600 the noise takes on hideous chromatic characteristics that make it uniquely digital. These images I find to still be rescuable with noise reduction in PP so long as they were exposed well, but there is a noticeable loss of fine image detail by this step.

If you typically shoot under ISO400, I don't think there's much to complain about with the K-7's noise, unless you are trying to pump up the shadows too much, as others have mentioned.

[Pentaxor]

I won't necessarily say that the K-7's High ISO is bad, except that it was underwhelming. people were disappointed for the fact that they want a significant IQ improvement over it's K20D predecessor. unfortunately, Pentax/Samsung rather primarily concentrated on the video and other features on the K-7. point is, the K-7 has a good IQ at lower ISO's and was the first of it's kind to include features that were not offered by other systems during that time. mind you that the High ISO performance of cameras 2 years ago weren't really groundbreaking at all and were slightly just acceptable at best. the only exceptions would be some of the FF cameras like CaNikon 5D and D700. the k-x started the trend of very good High ISO for APS-C system.

anyway, the point is, the K-7 is better overall as a camera and an improvement over the K20D (except the IQ). the k-x has better IQ over the K-7 but not the better camera overall. and the K-5 is refurbished and highly improved K-7. such improvement are in the DR (shadow detail, avoids blowup highlights), AF speed and FPS, Noise Handling (less Chromatic Noises, less blotches and smearing), and High ISO (useful for extreme lowlight). if I were to say this, the K-5 is more of a K-7 on steroids.

personally speaking, I would find the AF speed, DR and Noise Handling to be the main attraction on the K-5. although the High ISOs (3,200 to 25,600) to be useful for extreme lowlight situations with a decent IQ, I would rather stay put at 1,600 and retain a very good level of IQ and avoid any noise. the K-5's DR is extremely good and you could still push it over 2 stops further. thus achieving something around 10,000 ISO equivalent. not bad. for strobists, I don't see any point for High ISO's unless they use it for other uses such as outdoors/indoors candids and street photography. the K-5's DR is useful though for PP (especially extreme recovery) afterwards.

in general, I would prefer a High DR camera over a HigH ISO camera. at the end of the day, the former would give you the best results in terms of IQ and detail.

[conradj]

i agree with both Wheatfield and v5planet for the high ISO behavior of the K-7; ISO800 isn't very good. It's ok down to around the mid tones; anything darker is a mess.

Still it's the best camera i've ever lugged around and once i got it finally had no regrets abt leaving the MX bodies in a closet.

If you've done a fair amount of slide film work then the K-7 isn't hard to use: not a whole lot of DR in either; and where before i had to rely on experience to judge exposure with slide film -- the camera meter usually only got me to the approx reading -- i found the K-7 matrix metering very good indeed.

In long exposures 30s. is abt as much as one can tolerate, and keep it on base ISO since anything higher will pump up sensor noise too much.

i do find that the K-7 LCD image review doesn't let me judge exposures very well. It seems to make underexposed shots look quite acceptable and it's only when i open the image up in my editor that i can see how bad it is. Perhaps i'll have to start getting used to reading histograms rather than looking quickly at the LCD display of the image.

[Pentaxor]

the problem with the K-7 is on the underexposed side since noise would begin to creep up once you try ETTL. the trick usually is when using the K-7 is expose properly or use ETTR. having said that, shadow detail is somehow compromised due to noise. the only thing that you have to contend with when doing ETTR is clipping up the highlights on occasions.

[Ash]

Yes, Pentaxor has got it.
The K-7/K20D excels when images do not need any exposure/brightness boost in PP.
Even ISO 1600 can be quite acceptable if this is the case.
Unfortunately, this falls apart once long exposures are brought into the equation.

[RonHendriks1966]

Well I find that K-7 give more detail then K-x even at a little hi iso-settings. It does give more noise, but I still prefer the details in it (or I just didn't see K-x pictures made with premium glass).

Iso1000
https://www.pentaxforums.com/gallery/images/14650/1__K7O4510HD.jpg

[traderdrew]

Hi Outsider,

You can compare all of the cameras to each other on DXOmark.com

http://www.dxomark.com/index.php/Camera-Sensor/Sensor-rankings/(type)/usecase_sports

The K-7 just didn't cut it in the shady conditions for me. My subjects such as birds were just to fast for it under lower light. I love the K-5.

Pentax if you are listening, please keep up the low light fight. Improve the autofocus if you wish but don't change a thing.

http://www.flickr.com/photos/andrewdalerichardson

[GeneV]

Well I find that K-7 give more detail then K-x even at a little hi iso-settings. It does give more noise, but I still prefer the details in it (or I just didn't see K-x pictures made with premium glass).

Iso1000
https://www.pentaxforums.com/gallery/images/14650/1__K7O4510HD.jpg

For the similar Samsung sensor in the K20d, that was not the case for me. The K-x at ISO 1600 or even 3200 is usable without much PP. Not so with the K20d. I couldn't see much to recommend the K20d image, but that may just be me. Now, at true ISO 200, the extra pixels in the K20d can come out to play. At true ISO 100, I'll take the K10d.

[GeneV]

I find the latter half of this statement to be an extremely strange notion. Yes, it takes time to remove noise in a satisfactory way from an image with software, and even if the results are good, it may not be worth it to someone to go through the hassle. And in some cases an image is just going to be too noisy to rescue with software without a notable degradation of overall detail.

Agreed. The whole concept of PP as separate from digital photography is a bit strange to me. What comes directly off the sensor isn't useful. Some form of "pp" is needed to demosaic and turn the image file into something that works. It is really a question of how much time it takes and workflow rather than whether one is a purist. To me, shooting without post processing means shooting transparency films.

[Hound Tooth]

Well I think that the K-7 beats K-x any time.

Sorry, I had both at the same time, and I couldn't disagree more. As far as low-light ISO performance goes, my K-x blew the pants off my K-7. I still have the K-x, and I got rid of the K-7 in favor of a K-5.

[Pentaxor]

Agreed. The whole concept of PP as separate from digital photography is a bit strange to me. What comes directly off the sensor isn't useful. Some form of "pp" is needed to demosaic and turn the image file into something that works. It is really a question of how much time it takes and workflow rather than whether one is a purist. To me, shooting without post processing means shooting transparency films.

there is a convenience factor involved here. having said that, PP softwares are getting more advanced by the minute nowadays and making it easier to post-process. I can still remember taking a lot of time eliminating noises then recover and sharpen up the image with mediocre results. nowadays, noise elimination can be easily done with set preset values and same with sharpenings, maskings, etc... altogether in a few seconds or within a minute. very scary.

I fear that in the near future when/where camera and lens IQ and rendering would no longer hold much value for most people in general. this maybe a good news as well since there is a strong possibility that lenses would no longer be separated by class and rendering capability. this either may cause the lenses to significantly drop in price or the price difference between lenses would be less significantly apart.

[v5planet]

there is a convenience factor involved here.

I will readily acknowledge the convenience factor. It's not always worth the effort to deal with noise or any other image quality shortcoming, and it's better if you get everything right in camera and that the camera is up to the task than to make adjustments later.

The part that struck me as odd was the denigration of an end result's quality as being merely "perceived", as if the end viewer can somehow look at an image and rewind it back from the final jpeg to what it must have looked like originally and then holds it against the final product.

What is so special about a noiseless raw that a processed image that had noise but no longer does due to processing (and has lost nothing for it) is somehow of a lower quality?

[RonHendriks1966]

Sorry, I had both at the same time, and I couldn't disagree more. As far as low-light ISO performance goes, my K-x blew the pants off my K-7. I still have the K-x, and I got rid of the K-7 in favor of a K-5.

Well if you look at things from different angles: I got rid of my K-7 in favor of a K-5 too. Soo we agree on that topic