Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
07-12-2019, 06:47 AM   #1
New Member




Join Date: Apr 2017
Posts: 12
Shadow recovery issue on KP (magenta banding)

Hi everybody,

I'm experiencing a strange behaviour on my KP. Especially on the lower part of the frame, when shooting at high ISO and with severe underexposure, there's often a meganta cast, covering maybe 1/6 of the height.
Investigating more, I saw that it's more visible when shooting with mechanical shutter than with the electronic one. Upon further investigation the issue appears like an uneven sensor noise (from +/- 3200 ISO up, with underexposure). I'm surely going to send the camera in for repair (it has less the a year). I'm adding some screenshots that hope may be helpful to better understand what's happening. The last ones are with the cap on the lens and the finder covered, in a darkroom: it's just noise from the sensor, and the last one is the same but taken with the K5 IIs, which seems to behave normally...
Is anybody able to check her/his KP to see if shooting a completely dark image and recovering in post the exposure leads to something similar?
Many thanks in advance.

Attached Images
         
07-12-2019, 08:42 AM   #2
Otis Memorial Pentaxian
Loyal Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 32,921
QuoteOriginally posted by Éire69 Quote
The last ones are with the cap on the lens and the finder covered, in a darkroom: it's just noise from the sensor, and the last one is the same but taken with the K5 IIs, which seems to behave normally...
Despite not having a KP in hand, my guess is this is not normal. Is there any chance your KP is still on warranty? As you noted, a service visit would likely be in order. The assumption, of course, is that there is nothing in ACR or Photoshop that is at work. Have you tried the same DNGs in the software that came with your camera?


Steve
07-12-2019, 08:45 AM   #3
Journeyman Cat Wrangler
Loyal Site Supporter
SSGGeezer's Avatar

Join Date: Jun 2017
Location: Maine, U.S.
Photos: Gallery | Albums
Posts: 3,742
Definitely time to send that in for service if it shows up in all forms of software. I have not seen that even when shooting at much higher isos than what you are describing. Something is wacky and wrong here. Good luck.
07-12-2019, 09:25 AM - 1 Like   #4
Pentaxian




Join Date: Jan 2011
Location: New York
Posts: 4,326
QuoteOriginally posted by Éire69 Quote
...when shooting at high ISO and with severe underexposure...
That looks a bit like "amp glow". Cameras components might be heating the sensor unevenly. Here's some more information about it from an astrophotography perspective What is Amp-glow and how do I manage it? | ZWO ASI

How many stop are you pushing the exposure?

Does the problem decrease with decreasing temperature? You can test this by sealing the camera in a "ziploc" bag (squeeze out air as you seal it to limit condensation), put it into the refrigerator for a while, then repeat the dark frame test.

07-12-2019, 09:30 AM   #5
Pentaxian
normhead's Avatar

Join Date: Jun 2007
Location: Near Algonquin Park
Photos: Gallery | Albums
Posts: 33,056
Yep, that's not functioning as designed.
07-13-2019, 12:15 PM   #6
New Member




Join Date: Apr 2017
Posts: 12
Original Poster
QuoteOriginally posted by normhead Quote
Yep, that's not functioning as designed.
Well, I don't know how it has been designed, but at the moment at least two collegues confirmed that their Kp samples show exactly the same noise pattern (more on the lower side of the frame (upper part of the sensor) with mechanical shutter, on both the long sides of the frame (but a bit less strong) with electronic shutter.

DeadJohn may be right, there could be components that contribute to a uneven heating of the sensor. Right now in Italy we are having quite a hot summer (over 30°C /86°F on average), I don't know if this may be the cause, but what I saw on the others' KPs is too similar (if not identical) to what happens on mine.

---------- Post added 07-13-19 at 12:20 PM ----------

QuoteOriginally posted by DeadJohn Quote
That looks a bit like "amp glow". Cameras components might be heating the sensor unevenly. Here's some more information about it from an astrophotography perspective What is Amp-glow and how do I manage it? | ZWO ASI

How many stop are you pushing the exposure?

Does the problem decrease with decreasing temperature? You can test this by sealing the camera in a "ziploc" bag (squeeze out air as you seal it to limit condensation), put it into the refrigerator for a while, then repeat the dark frame test.
A collegue used -5 stops of underexposure. With my dark frame shots it is probably more, but I can say that the issue is visible and annoying on the darker areas starting from ISO 3200 and -2 EV, as in the shot attached above.
07-13-2019, 01:35 PM - 1 Like   #7
Otis Memorial Pentaxian
Loyal Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 32,921
QuoteOriginally posted by Éire69 Quote
Well, I don't know how it has been designed, but at the moment at least two collegues confirmed that their Kp samples show exactly the same noise pattern (more on the lower side of the frame (upper part of the sensor) with mechanical shutter, on both the long sides of the frame (but a bit less strong) with electronic shutter.
QuoteOriginally posted by Éire69 Quote
A collegue used -5 stops of underexposure. With my dark frame shots it is probably more, but I can say that the issue is visible and annoying on the darker areas starting from ISO 3200 and -2 EV, as in the shot attached above.
DeadJohn knows his stuff (examples of his astro work) and if he says amp glow, it may well be amp glow.

Send it in for service and find out how they fixed it. Have your friends do the same and post the results back here. Clearly, this is an unexpected and unacceptable result, unless, of course it is neither. A few questions/observations:
  • Is the artifact visible on import to ACR?
  • Is this artifact visible with the PDCU software that came with your KP?
  • Why the intentional underexposure for night shooting and how deep a shadow pull is being attempted?
  • Remember that ACR may be creating the lion's share of the artifact as part of the pull due to a paucity of data for low values. Simply put, ACR makes its best guess based on adjacent pixels. The deeper the pull and the lower the RGB values, the more guesswork involved. A shadow pull may work well for minor adjustments, but to be honest, it is something I do maybe once every couple of years despite regularly shooting subjects having deep shadows.
  • Lastly, just how hot was the sensor? There are two thermal sensors and their readings are included in the makernotes portion of the EXIF of the original DNGs.

Sorry to be so stern, but the example frame at top right is telling. The ghost head by the arrow tells me that there was precious little to pull. The short answer regarding an artifact at high ISO + severe intentional underexposure + high ambient temperature + deep shadow pull is simply, "Don't do that!"

Again, if you feel your camera is defective, I strongly encourage sending it in for service. Your friends having similar problems may wish to do likewise. Your cameras might have all been part of a bad batch from the factory.


Steve
07-13-2019, 02:25 PM   #8
Pentaxian




Join Date: Jan 2011
Location: New York
Posts: 4,326
QuoteOriginally posted by Éire69 Quote
A collegue used -5 stops of underexposure. With my dark frame shots it is probably more, but I can say that the issue is visible and annoying on the darker areas starting from ISO 3200 and -2 EV, as in the shot attached above.
I'll join the "send it for warranty repair" bandwagon. Pushing by 2 stops is a moderate push that the KP should be able to handle.

07-13-2019, 03:09 PM   #9
New Member




Join Date: Apr 2017
Posts: 12
Original Poster
Update: another KP user from Texas confirms the issue. It happens quite clearly between 800 and 6400 ISO, less so from 12800 above.
Below is a shot taken at 800. We are talking about extreme shadow recovery, beware, but the color cast is visible (barely) in normal exposed shots too, if the 'affected' area is very dark.
Attached Images
 
07-13-2019, 03:53 PM   #10
Otis Memorial Pentaxian
Loyal Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 32,921
QuoteOriginally posted by Éire69 Quote
Update: another KP user from Texas confirms the issue.
Link? No link => not even a rumor.

Might I ask why all examples are screen grabs or something other than a JPEG made from the DNG or a download link to the DNG itself? The above is a good example. It may just be a very nice "frame" applied in PP, but if not, it looks an awful lot like a photo of a print. Either way, neither makes much sense.

I, for one, would love to have a copy of an OOC original having this issue, particularly since it seems to not be a problem as of yet for other users on this site and seems to be related to PP flow. I field a lot of troubleshooting requests and a good example for forensics would be useful for diagnosing other similar examples when and if they come in.


Steve

Last edited by stevebrot; 07-13-2019 at 04:03 PM.
07-13-2019, 04:05 PM   #11
New Member




Join Date: Apr 2017
Posts: 12
Original Poster
QuoteOriginally posted by stevebrot Quote
DeadJohn knows his stuff (examples of his astro work) and if he says amp glow, it may well be amp glow.

Sorry to be so stern, but the example frame at top right is telling. The ghost head by the arrow tells me that there was precious little to pull. The short answer regarding an artifact at high ISO + severe intentional underexposure + high ambient temperature + deep shadow pull is simply, "Don't do that!"


Steve
Steve, thanks for all the questions, I'll try to answer tomorrow. Just one thing: those shadow are pulled only to better show the magenta strip; usually I don't underexpose and then pull the shadows (and never to that extent, unless needed, but it should really be on the event of an alien landing in the woods near my house... ), I expose "normally" for my subject: in the ballet scene I could just cut away that part of the frame, there was just an invisible coreographer who could just remain invisible... I only noticed a bit of magenta shifting in the shadows there, but thought it was a faint reflection of the lights on the stage.

---------- Post added 07-13-19 at 04:56 PM ----------

QuoteOriginally posted by stevebrot Quote
Link? No link => not even a rumor.

Steve
Steve, are you stating that I'm a liar or what? She sent me the originals, shot the way I asked (she's a member of the Pentaxians group on FB).

BTW, have you got a KP? Would you be so kind to try and duplicate the kind of setting to see if the noise pattern shows in the same way? Camera in a dark room, finder covered and lens with cap, 1/15s and ISO ranging from 800 to 12800. RAW, of course...

if you like to check for yourself, I've uploaded four dng files here: WeTransfer
There is one of mine (one of those shown above) and three from the Texan KP owner. Sorry if I took so long to reply, but the upload speed is not what I may call fast, here...

PS: I didn't find anything about the temperature of the sensor in the Exif, just the WB Kelvin temperature. Where exactly should I look?
07-13-2019, 05:38 PM   #12
Otis Memorial Pentaxian
Loyal Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 32,921
QuoteOriginally posted by Éire69 Quote
Steve, are you stating that I'm a liar or what? She sent me the originals, shot the way I asked (she's a member of the Pentaxians group on FB).
Sorry...I guess I am getting a little jaded. There have been several recent threads involving strange behavior where the OP has been evasive.

As for the temps, in ExifTool the tag names are "Sensor Temperature" and "Sensor Temperature 2". These may not be supported or exposed by all EXIF readers and may not be written by all bodies. I am looking at your files now.


Steve
07-13-2019, 07:28 PM   #13
Otis Memorial Pentaxian
Loyal Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 32,921
QuoteOriginally posted by Éire69 Quote
if you like to check for yourself, I've uploaded four dng files here: WeTransfer
There is one of mine (one of those shown above) and three from the Texan KP owner. Sorry if I took so long to reply, but the upload speed is not what I may call fast, here...
I got the files and have taken a quick look-see. Here is what I found initially:
  • The file from your camera had an internal temperature of 30C. I would not consider that critically high. The files from Texas showed an internal temp of ~26C. Sensor temp is not written to EXIF for the KP.
  • Only one of the files showed a "glow" on my monitor on initial import to Lightroom 6 or on initial load to PDCU 5. This was true even with the room darkened.
  • All of the files show a full-height spike at far left on the histogram in Lightroom and in PDCU as well as a low contribution "fog" across the full width of the histograms for the rendered DNGs.
  • The one file that showed a faint glow on import (IMGP2239.DNG) had an earlier edit in ACR with application of +3 stops exposure. The details had been written to the DNG XMP and were picked up and applied by Lightroom on import.
  • A 32-bit TIFF generated by LR was unremarkable
  • Playing a little bit with the sliders in LR showed that even a 1 stop bump to exposure is enough to show a glow at the image margins. Additional brightening shows generalized luminance noise across the full frame.
  • I also processed one of the files in Affinity Photo with similar results.
I have a few other tools that are capable of working directly with the sensor data in the DNG without dealing with the rendering converter that I can throw at this. My opinion at this point is that what we are seeing is likely a problem with the image sensor on both cameras and not indication of pushing the limits for capture or temperature or an artifact introduced by a particular RAW processor.


Steve
07-14-2019, 09:28 AM   #14
New Member




Join Date: Apr 2017
Posts: 12
Original Poster
QuoteOriginally posted by stevebrot Quote
I got the files and have taken a quick look-see.

(...)
I have a few other tools that are capable of working directly with the sensor data in the DNG without dealing with the rendering converter that I can throw at this. My opinion at this point is that what we are seeing is likely a problem with the image sensor on both cameras and not indication of pushing the limits for capture or temperature or an artifact introduced by a particular RAW processor.


Steve
Thanks a lot, Steve, for taking the time to check the files.
By now, 6 out of 6 bodies show the same behaviour (edit: seven, there's the sample used by dpreview to test...)

I don't know what to think (except that I'm happy to avoid sending the camera in for repair... )

Gianfranco

Last edited by Éire69; 07-14-2019 at 09:29 AM. Reason: better precision
07-14-2019, 10:45 AM   #15
Otis Memorial Pentaxian
Loyal Site Supporter
stevebrot's Avatar

Join Date: Mar 2007
Location: Vancouver (USA)
Photos: Gallery | Albums
Posts: 32,921
QuoteOriginally posted by Éire69 Quote
Thanks a lot, Steve, for taking the time to check the files.
By now, 6 out of 6 bodies show the same behaviour (edit: seven, there's the sample used by dpreview to test...)

I don't know what to think (except that I'm happy to avoid sending the camera in for repair... )

Gianfranco
I am wondering if a light leak (either through the lens/mount or body) is to blame, does one get a similar pattern at low ISO when hitting the camera with a burst of electronic flash?

BTW...pity that the bulk of this discussion is apparently happening on a closed FB group.


Steve
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!
answer, banding, bit, components, frame, glow, iso, issue, issue on kp, kp, link, noise, pentax kp, sensor, shadow, shadow recovery issue, shadows, shutter, steve, temperature, test, underexposure
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
KP vs K-70 - to KP or not to KP, that is the question OldChE Pentax DSLR Discussion 19 02-04-2018 09:43 AM
Landscape Love the Shadow Recovery on the K3 photosbyspeed Post Your Photos! 2 02-02-2015 03:55 PM
Another K-3 Shadow Recovery Test powasky Pentax K-3 9 01-14-2015 02:00 AM
k-5 - Terrific shadow recovery dexmus Pentax K-5 3 05-27-2011 06:19 AM
K-7 - Banding in DNG but banding gone after conversion to JPG? raider Pentax DSLR Discussion 21 09-27-2010 12:09 AM



All times are GMT -7. The time now is 05:04 PM. | 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