Banding issue with K10D

[CapitaineAbitibi]

I'm sure there's a bunch of other threads on the same topic.

After 700 shots with my K10D, mostly on dark backgrounds, I realized that at even ISO 100 if the exposure is not dead on when I post process the pictures there's banding all over the place. I was used to be able to play a lot with exposure with my DS in PP but now I can't do it.

I mostly shoot rock shows and my regular settings were (DS + 70-200 f2.8 EX) -0.7ev ISO 800 1/200 f 2-8-5.6 and with the K10D... I'm still trying to figure it out but I can't do the -0.7ev that's for sure. DS was noisish but without any banding, K10D has less noise but lots of banding.

I'll have to put an example online tonight...

I don't know if it's worth it to send the camera to Pentax Canada or not.

 

[Bronek]

I don't know if it's worth it to send the camera to Pentax Canada
or not.

Probably yes, as you seem to have faulty camera. I compared K10D and DS (own them both) and K10D is visibly much better in shadow area (in their optimal ISO, ie. 100 for K10D and 200 for DS). If you are seeing artifacts where mine K10D shines, it must be the problem with your copy

B.

 

[lol101]

I don't know if it's worth it to send the camera to Pentax Canada
or not.

Yes it it worth because:

a/ I f the camara is faulty you're entitled to a repair/replacment;

and:

b/ If the camera is not faulty (ie if all K10 behave the same), Pentax has to do something about it.

My 2 cents

 

[Thibouille]

Well, K10D being totally different than DS I guess it is normal, the meter doesn't work the same way so you can't use the same exposure than with the DS. Seems to me perfectly normal.

Of course, you may have lemon...

 

[CapitaineAbitibi]

Hopefully I don't have a lemon because it would be the second time with Pentax, I had to send back my DS twice to fix the same problem in the past (Auto Focus didn't work) as for the exposure I know the settings won't be the same

It's just the banding issue at ISO 100, it won't show in jpg but on DNG it's kind of right in your face and I'm always shooting RAW...

 

[guyoliver]

--

There is your problem. Pentax realizes that most people will not use RAW so why bother to make it satisfy that minority.

 

[JensR]

Well, K10D being totally different than DS I guess it is normal,
the meter doesn't work the same way so you can't use the same
exposure than with the DS.

Strange logic.

K10D at f/2.8 and ISO 800 and 1/100 s
should give the same exposure as
DS at f/2.8 and ISO 800 and 1/100 s
should give the same exposure as
Canon 5D at f/2.8 and ISO 800 and 1/100 s
if all manufacturers stick to the rules.

The Captain's problem isn't differing exposure (ISO values can indeed be a bit "off" between sensor types), but a very different noise/banding situation on his K10D. The meter has very little to do with it. Knowing that the DS by default underexposes more than the K10D, bringing an underexposed DS shot to acceptable brightness would require more pushing and hence create/reveal more noise/banding than doing so on the K10D. This is exactly the opposite of what the Captain sees.

Seeing shots straight from his cameras and the post-processed ones will help us determine what is happening here.

Jens

--

'Well, 'Zooming with your feet' is usually a stupid thing as zoom rings are designed for hands.' (Me, 2006)
http://www.jensroesner.de/
--=! Condemning proprietary batteries since 1976 !=--

 

[CapitaineAbitibi]

Here's the resized version of one of the pictures

exif from camera:
ISO 800
1/125
f2.8
98mm
0ev

PP:
-0.5ev



Here's a 100% crop of the top left corner, hard to see but there are some blue and green bands going across the picture.



Now I need to find a good one at ISO 100... 400 pictures to go through. Might be me who's seeing strange things because it's a brand new camera, it's just that I find it odd that if you have to overexpose in PP bands are becoming so apparent. I know the settings won't be the same as my DS, but still I was use to underexpose to get an accurate reading and compensate a little bit for the shake, to gain speed which is no longer an issue with the K10D

 

[CapitaineAbitibi]

Here's a link to the full post processed gallery, you'll see some banding on there even though they were reduced in size.

http://www.pbase.com/dkirouac/vulgaires_1215

 

[lock]

I just can't find it. To me this is plain random noise.

lock

 

[Edvinas]

Can't see any banding on my uncalibrated monitor...

--
Edvinas

 

[K10D]

Nothing visible here; just as a thought, have you tried printing a few of them?

Great shots btw.

--
Bob

Perth, Western Australia

 

[CapitaineAbitibi]

Thanks...

Maybe it's my paranoid eyes that saw things that were not there

 

[JensR]

Maybe it's my paranoid eyes that saw things that were not there

Umm. My eyes are paranoid, too.

It's not too drastic, and in some shots I could mistake it for curtain structures in the background, but in others it's clearly visible vertical pattern noise:

http://www.pbase.com/dkirouac/image/72228983

In the end you have to decide what is acceptable to you .

Cheers
Jens

--

'Well, 'Zooming with your feet' is usually a stupid thing as zoom rings are designed for hands.' (Me, 2006)
http://www.jensroesner.de/
--=! Condemning proprietary batteries since 1976 !=--

 

[Jonas B]

I would second Jens here, it's not that hard to see. In your first picture (man playing guitar with two red lamps in the bottom) the vertical pattern forms a cross pattern together with the rays of the red light, in the picture Jens refers to (singer, pic 0089) there also is a criss cross pattern, easy to see and most easy to dissicate in the blue channel.

The level of this pattern noise is low and my guess is that you can remove it in a few moments with a little work with either levels or curves in Photoshop (or similar software of course).

Now this is at ISO 100... I wonder, where the pictures underexposed and later pushed? Or are they JPGs straight fromthe camera? In the first case there is nothing to worry about, except maybe getting the exposure better. In the second case i would be worried about what ISO 200 and 400 looks like. Did you take any pictures at higher ISO values at the same time?

I wonder why people thinks pattern noise will dissapear magically if pictures get printed. That is either a new art form using color lasers on office papers, or a second stage of denial that is a little fascinating. I still have postcard sized prints of photo_moms high ISO kitchen pictures laying around, pattern noise get printed together with the rest of the data sent to the printer. Uhu?

--
Jonas

 

[JensR]

Hi Jonas!

Now this is at ISO 100... I wonder, where the pictures underexposed
and later pushed? Or are they JPGs straight fromthe camera? In the
first case there is nothing to worry about, except maybe getting
the exposure better. In the second case i would be worried about
what ISO 200 and 400 looks like.

Excellent point. The Captain said he usually pushed his images with his DS, so we really need to know how these images were derived.

I wonder why people thinks pattern noise will dissapear magically
if pictures get printed. That is either a new art form using color
lasers on office papers,

LOL.

Well, you know, there's a truth in that. Not the office paper, but a poor or too-punchy printer profile can indeed lead to noisy black becoming pure black. Sadly, it will smother all nice details that are close to black, too.

I remember Terry's (TMalford's) very nice B&W images and he (experienced he is) had a hard time to get the fine levels printed correctly.

Cheers
Jens

--

'Well, 'Zooming with your feet' is usually a stupid thing as zoom rings are designed for hands.' (Me, 2006)
http://www.jensroesner.de/
--=! Condemning proprietary batteries since 1976 !=--

 

[Rupert60]

I don't really see anything if signifigance here, which at first made me sad that I am such an inferior critic. However, I think maybe it is a blessing, since I will most often think I have a nice shot and be very happy, when in fact, it will have its flaws. Yes, I think I'll take happy. This has other advantages, like my beautiful wife that thinks I am handsome. Now that is pushing it to the limit. LOL

'This is more serious than I thought.....but it is still fun!
http://www.pbase.com/rupertdog Take a look- It's Free!

 

[GordonBGood]

I would second Jens here, it's not that hard to see. In your first
picture (man playing guitar with two red lamps in the bottom) the
vertical pattern forms a cross pattern together with the rays of
the red light, in the picture Jens refers to (singer, pic 0089)
there also is a criss cross pattern, easy to see and most easy to
dissicate in the blue channel.

Yes, there is a "blotchy pattern, but behind it is a systematic vertical pattern noise.

The level of this pattern noise is low and my guess is that you can
remove it in a few moments with a little work with either levels or
curves in Photoshop (or similar software of course).

Well, he could eliminate it if he wanted the background to disappear but not if he wanted to bring out details in the background, which would push process these levels and make the patterns more obvious.

Looks like more of the "vertical pattern noise" which is lurking in all K10D images to catch those who try to brighten the levels far enough to bring it out. Still hoping that Pentax may be able to fix this with a firmware update, but no guarantees.

Regards, GordonBGood

 

[jeffkrol]

--
360 minutes from the prime meridian. (-5375min, 3.55sec) 1093' above sea level.

'The exposure meter is calibrated to some clearly defined standards and the user needs to adjust his working method and his subject matter to these values. It does not help to suppose all kinds of assumptions that do not exist.'
Erwin Puts

 

[Scandalous]

jeffkrol wrote:

It doesn't matter whether the glass is half full, or half empty - it's poisoned anyway......