[falconeye]

So for some reason Lightroom is only using the JPEG standard preview for the K-x on my computer when in compare mode or even loupe mode at 3:1. When I go to develop, the artifacts go away, but back to loupe, and it goes back. How can I fix that?

I think there is a config option in LR about the JPG quality of cached previews. It is "Preview Cache" under "Catalog Settings" reachable from the genaral settings dialog.

The problem is that LR will tell you when (re)creating its quality preview, except if in compare mode. So, in compare mode and a slow computer, you may easily be watching at a wrong quality. The quality improvements in the develop module should be barely visible only.

conclusion that we could draw here is that the k-x resolution gets better at smaller aperture settings

Can everybody please stop to judge camera resolution from blurred or misfocussed images?

The K-x is already "oversharp" and produces artefacts if used properly, because of a too weak AA filter.

I find it interesting how Pentax puts a better balanced AA filter into the K-7 (meant to be suitable for studio work) while putting an overweak AA filter into their entry level model (meant to appeal to the "masses").

This is similiar to what the music industry does: volume-compressed pop music (the music is loud everywhere) vs. high dynamic range classic music.

And then it is even more ironic to find people here wanting more sharpness.

This is like yelling "help, my music still isn't loud enough, need a CD with louder music" when they miss to find the volume knob or better speakers


ok, ok, my post is a bit over the edge. But testing is hard work, believe me. And time needed is measured in days rather than minutes. And walking into a camera store with the ambition to run a resolution test is, well, how shall I express it, ...

[Pentaxor]

I think there is a config option in LR about the JPG quality of cached previews. It is "Preview Cache" under "Catalog Settings" reachable from the genaral settings dialog.

The problem is that LR will tell you when (re)creating its quality preview, except if in compare mode. So, in compare mode and a slow computer, you may easily be watching at a wrong quality. The quality improvements in the develop module should be barely visible only.


Can everybody please stop to judge camera resolution from blurred or misfocussed images?

The K-x is already "oversharp" and produces artefacts if used properly, because of a too weak AA filter.

I find it interesting how Pentax puts a better balanced AA filter into the K-7 (meant to be suitable for studio work) while putting an overweak AA filter into their entry level model (meant to appeal to the "masses").

This is similiar to what the music industry does: volume-compressed pop music (the music is loud everywhere) vs. high dynamic range classic music.

And then it is even more ironic to find people here wanting more sharpness.

This is like yelling "help, my music still isn't loud enough, need a CD with louder music" when they miss to find the volume knob or better speakers


ok, ok, my post is a bit over the edge. But testing is hard work, believe me. And time needed is measured in days rather than minutes. And walking into a camera store with the ambition to run a resolution test is, well, how shall I express it, ...

haven't you heard of the saying, "dont lay your eggs in one basket". I believe that's how Pentax is playing their game.

[Daemos]

ok, ok, my post is a bit over the edge. But testing is hard work, believe me. And time needed is measured in days rather than minutes. And walking into a camera store with the ambition to run a resolution test is, well, how shall I express it, ...

I never said I was running a resolution test did I?

This is just pictures taken from the KX and the K7 using the same settings.

Then people can subjectively look at the pictures side by side, instead of making statements about one or the other having much better image quality without even seeing two pictures in the same setting.

Honestly most people don't need a resolution chart, or any real measurements with numbers to equate that with image quality. That's just taken in at a simple glance at a picture.

They are nice to have, but if you give me numbers, does that mean I know what the picture looks like?

[UnknownVT]

Note the KX pictures look MUCH better this time (this might be because the previous KX wasn't working properly)

Thank you again.

I think this is mainly due to better lighting and more depth of field - although remember your first set was at f/22 which should have had really good depth of field even though it was considered sub-optimal (can't please everyone)......

My basic take is the same as yours (right from the beginning) -
the K7 at lower ISO is better simply because it is 14.6Mp vs the Kx 12.4Mp
But the Kx takes over at higher ISO somewhere between 1600 to 3200.

I am not too sure if there was anything wrong with the previous Kx's - I think we want to look for differences - side-by-side comparisons can be very critical - we tend to see "big" differences when in real-life and in isolation we might not even notice them.

Not trying to trivialize these differences -
but not many of us truly display images at full size and 100% - mostly they are shrunk in size first to fit the screen - even the highest resolution screens are only 1680 x 1050 or "HD" 1920 x 1080 so we are talking a maximum display size of less than 2Mp!!! ie: 1/6 or 1/7 the fullsize of the Kx or K7.

Or for critical work prints at 10x8 which would mean about 304-428 dpi for the Kx and 388-467 dpi for the K7 depending which dimension one prints for - these all surpass even the most critical (nose-to-paper) viewing.

Yes, I do agree the latest batch of photos do look better - but it's for both the Kx and K7 - and I think it's mainly due to the subject/lighting and good depth of field for the subject.

Thanks for all your effort -
all the photos (all 3 series) were informative and useful.

[txshooter]

Or for critical work prints at 10x8 which would mean about 304-428 dpi for the Kx and 388-467 dpi for the K7 depending which dimension one prints for - these all surpass even the most critical (nose-to-paper) viewing.

This is exactly what I was going to ask about. On prints this size: are the diferences in nose and detail, at lower ISO where the K-7 "shines", going to be noticeable?

[UnknownVT]

This is exactly what I was going to ask about. On prints this size: are the diferences in nose and detail, at lower ISO where the K-7 "shines", going to be noticeable?

At lower ISO my opinion is no - but then Daemos has already provided 3 sets of samples of shots under the same conditions - you can easily print them out to see as only you can truly be the ultimate judge.

Hint to save expense there is no need to print full 10x8 just crop an area (that's the equivalent of a full 10x8) and print to say 6x4 or even better 4 or 6 crops on a single 10x8 and compare.

For full 6x4 that's an area of about 2196x1464 pixels for the Kx and 2565x1710 pixels for the K7
For say 4"x 3" to do multiple images on a single 10x8 that would be 1464x1098 pixels for the Kx and 1710x1283 pixels for the K7 - I've used the mid-point between the two dpi figures for each - ie: 366 dpi for the Kx and 427.5 dpi for the K7)

[Daemos]

This is exactly what I was going to ask about. On prints this size: are the diferences in nose and detail, at lower ISO where the K-7 "shines", going to be noticeable?

Now the question can be re-directed:

At High ISO when printed after some PP will the difference be noticeable?

I don't print, but I do see images on my computer.

With some PP and shrinking it down to something like 1623x1080 (do not remember the exact resolution), then applying some NR to both, even at ISO 3200 and ISO 6400 it is very difficult to see a difference.

Just saying.

[Eruditass]

Now the question can be re-directed:

At High ISO when printed after some PP will the difference be noticeable?

I don't print, but I do see images on my computer.

With some PP and shrinking it down to something like 1623x1080 (do not remember the exact resolution), then applying some NR to both, even at ISO 3200 and ISO 6400 it is very difficult to see a difference.

Just saying.

Resizing tests won't work quite as well as print tests, because the DPI of a printer can be a lot better - versus on a screen when resizing it will throw away or intelligently blend the pixels together.

[zxaar]

Now the question can be re-directed:

At High ISO when printed after some PP will the difference be noticeable?

.

i think yes, and here is why.

after pp when you remove noise you are left with image that has certain amount of detail. If k-x files have noticable less noise to start with after noise removal you have lot more detail than k7 files with same amout of noise but lot less detail. Upon printing the image that has significant higher detail does look better. And yes our eye can pick this.
At lower iso levels detail in both images are already high and difference is not much pronounced.

[Daemos]

Resizing tests won't work quite as well as print tests, because the DPI of a printer can be a lot better - versus on a screen when resizing it will throw away or intelligently blend the pixels together.

ah, I've never printed anything. This is good information to know.

[Daemos]

i think yes, and here is why.

after pp when you remove noise you are left with image that has certain amount of detail. If k-x files have noticable less noise to start with after noise removal you have lot more detail than k7 files with same amout of noise but lot less detail. Upon printing the image that has significant higher detail does look better. And yes our eye can pick this.
At lower iso levels detail in both images are already high and difference is not much pronounced.

Makes sense.

Well from my experience, the K7 looks IMO much better at 100% because of more details being captured. From ISO100-1600.

At ISO3200 even at 100% zoom, with PP with NR on both, the details on the K7 picture are not that much worse than the KX picture, I would say it's still quite close (although not as close as the KX/K7 at lower ISO).

It isn't really until ISO6400 where the KX has a REALLY big difference at 100% even after PP.

I just don't know when or where I would shoot at ISO6400 too often or ISO3200.

Since I really have no experience in printing pictures, I guess I'd need to see hand in hand the differences to appreciate what minor detail differences mean when it comes to print even at a 10x8"

[Eruditass]

I'd agree with what zxaar says - K7 is king for cropping and top of the line lenses, K-x for super high ISO.

With my not as great as your lenses, the 3200+ ISO performance is quite helpful. You should get into prints, there are free / really cheap prints all the time at *************** - The best coupons, lowest prices, and hottest deals. if you search.

[jerrymouse]

The good thing is K-x makes F/2.8 pancakes effectivelly F/2.0 (in terms of low light performance). I was thinking of trading one of my DA 35 and DA 40 for FA 35. Now I don't have to - DA 35 became low light lens with usable ISO 3200 (6400 & 12800 for B/W).

[jacksonpritt]

I'm just wondering, what programs do you guys use for PP and NR?

I have Aperture and Photoshop, but I don't know how to do noise reduction to get rid of high-ISO noise... Is there some program or plugin I need to use to do this?

[rawr]

I'm just wondering, what programs do you guys use for PP and NR?

Post Processing, Printing, Software, and Darkroom - PentaxForums.com
Probably best to ask there.