[JeffJS]

ISO 10000

Very extreme (fill light pushed to 100% in ACR) processing can show banding. It usually shows itself in shadow areas and occurs mainly (for me) when something is severely underexposed and there is virtually no light. I don't know if this is the same thing Nicole is experiencing but this is what it looks like (again, pushed to make the point).



Here is the proper image (it is ice on top of o my car taken last April).

[Rupert]

I know you have all seen it before, but I am still proud of my little ISO 12,800 birdie! I'd be proud of it if no one else had seen it........

[IMG] [/IMG]

Regards!

[philbaum]

ISO 10000

Very extreme (fill light pushed to 100% in ACR) processing can show banding. It usually shows itself in shadow areas and occurs mainly (for me) when something is severely underexposed and there is virtually no light. I don't know if this is the same thing Nicole is experiencing but this is what it looks like (again, pushed to make the point).

Here is the proper image (it is ice on top of o my car taken last April).

Jeff and Nicole:
I occasionally get some badly exposed pictures, nearly black. And some extreme areas of contrast with no light due to my flashing some scenes at night, i'm not seeing any banding, even if I push ACR fill light to 100%. I think you should have your camera looked at.

What i do see in extreme cases, is light being clipped in some color channels leading to color casts, but not banding. But its so much better than my K20, i'm not complaining, and i'm don't think existing technology can do anything about the color clipping any more than they can eliminate clipping highlights entirely.

Nice pic of the ice on roof - well done!

phil

[JeffJS]

Jeff and Nicole:
I occasionally get some badly exposed pictures, nearly black. And some extreme areas of contrast with no light due to my flashing some scenes at night, i'm not seeing any banding, even if I push ACR fill light to 100%. I think you should have your camera looked at.

What i do see in extreme cases, is light being clipped in some color channels leading to color casts, but not banding. But its so much better than my K20, i'm not complaining, and i'm don't think existing technology can do anything about the color clipping any more than they can eliminate clipping highlights entirely.

Nice pic of the ice on roof - well done!

phil

Thank you..

I still have about 28 months to have it looked at and may send it in at some point (it will probably be a backup camera at that point) but for now, I'm quite happy with the results I get.

Here's a better exposed ISO 20000 shot taken a week later (with obvious softness due to NR) that doesn't show any banding.



and here is an ISO 16000, again, no banding.



The main difference was the ambient temperature of the camera (one was below freezing and the other is inside the heated car). So for me to easily repeat this and make a valid claim, I'll have to wait until winter.

---

Last edited by JeffJS; 08-20-2011 at 09:38 AM.

[philbaum]

JeffJS,
20,000iso, cheez, forget sending your camera in, perhaps you'd be willing to exchange yours for mine

I don't go any higher than 12,800.

Below is a picture done at 80 iso with banding, but i think the banding is iron




Lets face it, the K5 is just a whole lot of fun!

[JeffJS]

JeffJS,
20,000iso, cheez, forget sending your camera in, perhaps you'd be willing to exchange yours for mine

I don't go any higher than 12,800.

Below is a picture done at 80 iso with banding, but i think the banding is iron


Lets face it, the K5 is just a whole lot of fun!

[NicoleAu]

I'll take some photos at high ISO later tonight, would doing it in jpg cause any differences? I have a 2 year warranty though and a couple of shoots coming up, so at this stage I'll wait a bit.

[JeffJS]

The photos I show are shot RAW+ JPG (PEF) and converted from the RAW files.

[NicoleAu]

I took some at 12800, I did at 0 comp and again and 1 comp and there is definitely a difference. In the 0 comp shots you can see banding, but not in the +1 shots. So I'm guessing it's underexposed shots, like Jeff mentioned, may be why I have the issue. Still so much to learn about photography! PS Jeff, I couldn't see your example shots

[JeffJS]

I took some at 12800, I did at 0 comp and again and 1 comp and there is definitely a difference. In the 0 comp shots you can see banding, but not in the +1 shots. So I'm guessing it's underexposed shots, like Jeff mentioned, may be why I have the issue. Still so much to learn about photography! PS Jeff, I couldn't see your example shots

I was messing with the settings in my profile and it looks like I had set the albums to Private. Should be visible to registered members now.

[NicoleAu]

Nup, still can't see them

[falconeye]

Preparing for a new project, I was testing shooting handheld in Tungsten light (one ARRILIGHT 800 Watt torch). Here you see the difference between shot on tripod at 80 ISO (pic. #1) and 'handheld' at 10 000 ISO (actually the camera was still on the tripod because I was to lazy and wanted exactly the same framing). For the 10 000 ISO, in LR3, I set the luminance noise at 50 an did a little sharpening for the details.
This a simple setup just a little 'playing' notting serious (yet) , but most of the colours where represented...
K5 + AF 50 mm @ F1:8
Note that the 10 000ISO is a little less saturated (green), I didn't touch that neither...

I've to repeat myself here:

Don't do such tests and then show images resized for the web. Web-sized images aren't demanding at all.

Resizing to 0.67MP (which is what you did) corresponds to a reduction of ISO by a factor 24.4. So, you may have shot at ISO 10,000, but your web sized images show the performance at ISO 400. No wonder they look great, esp. with noise reduction applied.

Don't mistake me: The K-5 takes wonderful high ISO images. The point is that web-sized images cannot make the point. They don't do it justice.

[JeffJS]

Nup, still can't see them

Refresh your browser.

[amoringello]

ISO 10000
Very extreme (fill light pushed to 100% in ACR) processing can show banding. It usually shows itself in shadow areas and occurs mainly (for me) when something is severely underexposed and there is virtually no light. I don't know if this is the same thing Nicole is experiencing but this is what it looks like (again, pushed to make the point).

Just need to make sure that the background is actually flat and there is not a building behind? Does the same banding appear when the lens cap is on?

The K5 holds a LOT of data in the shadows, and pushing up the exposure can show things you would not have expected it to have seen.
You'll see examples on the site where the shot appears almost black, but with pushing up the exposure the resulting image becomes almost "usable". (won't win any contests, but a discernible image of reasonable quality can appear)

I have not seen such banding myself and have to agree it may need servicing if it is that bad.

[philippe]

I've to repeat myself here:

Don't do such tests and then show images resized for the web. Web-sized images aren't demanding at all.

Resizing to 0.67MP (which is what you did) corresponds to a reduction of ISO by a factor 24.4. So, you may have shot at ISO 10,000, but your web sized images show the performance at ISO 400. No wonder they look great, esp. with noise reduction applied.

Don't mistake me: The K-5 takes wonderful high ISO images. The point is that web-sized images cannot make the point. They don't do it justice.

Dear Falconeye,

You are right , I hope this is better?

These are 100% crops, no extra PP applied... But actually these kind of pictures are mostly printed at about A4 size...
Look at the tuning scale of the radio (and the leather, perhaps a little to magenta) and think about an Ektachrome 400 pushed till 1000 ASA from a few years ago...

BTW, what kind of test are better?