Jump to content

Recommended Posts

Advertisement (gone after registration)

Monochrom, Monochrom...every once in a while you got some issue....

 

I started noticing vertical lines appearing at ISO3200 - this is something that reminds me of my old M8!

I tried at low ISOs and I can't see it coming out...does anyone have knowledge of what it can be?

I had my sensor replaced couple of years ago so it's the new sensor.

Pic attached...I marked in red in case it wasnt obvious.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Link to post
Share on other sites

https://cameraderie.org/threads/leica-m-monochrom.37659/page-37#post-343526

 

I had a line like this show up on my M Monochrom with the new sensor in it. Only happens at High-ISO. I wrote my own DNG processor for the M Monochrom files, so took a look at what was going on. The line was not "dead", it was losing charge while transferring through one pixel. I HACKED a quick piece of code in my DNG code to look for it, and to fix it. I need to know the address of the bad pixel, did not do anything fancy. But- in images taken afterwards, the line is not apparent. It is not averaged out - the software generates a local "fudge factor" to restore the lost charge. I run all my DNG files through this code, ported it to run under Windows. The code applies a Gamma curve, operates over all the files in a directory or on the SD card.

Edited by BrianS
Link to post
Share on other sites

Advertisement (gone after registration)

I had the same issue with my original M Monochrom in February.  Unfortunately after weeks of working with Leica USA tech support, they told me that they could not remap the dead pixels and that the service was only available for M10 series cameras.  I ended up returning the camera to B&H as a result.  Seems as if the answer is different depending on who you speak with at Leica...

Link to post
Share on other sites

8 hours ago, BrianS said:

https://cameraderie.org/threads/leica-m-monochrom.37659/page-37#post-343526

 

I had a line like this show up on my M Monochrom with the new sensor in it. Only happens at High-ISO. I wrote my own DNG processor for the M Monochrom files, so took a look at what was going on. The line was not "dead", it was losing charge while transferring through one pixel. I HACKED a quick piece of code in my DNG code to look for it, and to fix it. I need to know the address of the bad pixel, did not do anything fancy. But- in images taken afterwards, the line is not apparent. It is not averaged out - the software generates a local "fudge factor" to restore the lost charge. I run all my DNG files through this code, ported it to run under Windows. The code applies a Gamma curve, operates over all the files in a directory or on the SD card.

Hey Brian, thanks for your imput. I'm not very practical with codes and I'm also on Mac, otherwise I'd give it a go!
I'm still going to swing by the Leica shop today, just to check what they says.....

man we all love the M9 Monochrom but wow so many problems with it, all the time!

  • Like 1
Link to post
Share on other sites

Hi...just had vertical lines on my prints showing ...are you printing?

if so is your printer connected to the computer wireless ?

if so then it is the wireless connect .try a cable connection like i did and guess what ?

NO more vertical lines

Link to post
Share on other sites

I guess taking it to Jimmy is the first step but I am pretty certain it will need to be sent to Germany, the worst part of that is that it will take 2 months to sort out, the cost would be around £100 I would imagine. I had 3 M240's showing lines, all at ISO 640 (I like that ISO for shooting) they were all new so they were replaced rather than repaired.

The procedure to fix a dead pixel is fairly simple, some cameras actually have this option to self-remap in the menu (My S1R has it), hopefully you get it sorted quickly & cheap'ish...

  • Like 1
Link to post
Share on other sites

  • 5 weeks later...
On 10/20/2020 at 8:39 PM, jonnyboy said:

argh I just got the email from Germany with a £200 repair for that vertical line 😕

Understand the cost, it has to go in for the service, £200 will also cover “free” shipping. Outrageous that you have to send M camera to factory to remap for dead pixel, as far as I know all other makes also have same problem but you never hear about it as the fix is built into the camera.

  • Like 1
Link to post
Share on other sites

17 hours ago, BrianS said:

That much for just remapping the bad pixel!!!!!

One of my Winter projects will be to make my code "just" interpolate bad lines and just fix them in the DNG file. Will make it read in or ask for the bad pixel location.

This would be a great solutions as many users reported this type of issue with M's.

  • Like 1
Link to post
Share on other sites

  • 3 weeks later...

Ok my Monochrom just got back today from Germany. Really excited to get it back to discover there is a new problem now...
The vertical line seems disappeared but now when I click Play to view an image the camera quite often get stuck and i need to remove the battery to be able to use it again. Really annoying...has anyone experienced this sort of things? I really dont want to add another 'quirk' to this camera (mine has already way too many!) but what a pain now to send it back to Germany!!!!

Link to post
Share on other sites

  • 2 weeks later...

Brought my cars in for service- always get some good code written at the Car Shop. I should consider myself a breaking even- the cost to have the car serviced was about the same as sending the camera to have the bad line remapped. And- servicing the car only took a few hours, sending the Camera away would be weeks.

I modified my program to read the bad column(s) in from a file, and made adding the Gamma curve optional. I also found images where the bad column had "hot" pixels, good pixels, and "Cold" pixels. Which just meant taking "Absolute values" for the threshold test when deciding to correct a pixel or not.

 

ANYWAY: when looking at the "bad" column on my camera, the structure is retained- just the level is above or below the neighbors. Rather than just average over the column all the time, my code computes a local offset and applies it when the levels of the bad column is too far above or below the two neighbors. A running average of 5 pixels is used to compute the local offset. The code at present generates a new file, leaves the original intact. It works with DNG files from the M Monochrom, and will not work with those emitted by Photoshop. Turns out that saved DNG files are different from what the camera generates. The same code will optionally add a Gamma curve, it asks for "curve" when starting up. a "0"- Gamma is skipped and a new file is generated starting with a "C" (Corrected). Give a curve from 1-8, get a file starting with "G" (Gamma). The program either works on one file at a time, or processes all the files in a directory. Runs on XP through to WIN10. Also runs on PharLap extended DOS, but different compiler options for that. This is a Windows console program.

 

I did try an algorithm to identify the bad columns automatically. One of the test images: Fabric in a Hat lines up perfectly with the Columns of the image. Broke the algorithm. SO- back to determining bad columns using Photoshop to get the number, and just have the list read in at runtime.

If anyone has a M Monochrom with a similar problem, I could look at a DNG file and see if this code corrects the problem. I can make up the file with the bad locations (trivial) and send the code to try out. I will not be doing a GUI- at least for the foreseeable future- I did this code to get familiar with the new compiler for a work project. I have 100K lines or so to convert to it...and get paid doing it. I love Fortran.

 

Scaled down image, still shows the dark line. 100% crop of the left edge.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Edited by BrianS
Link to post
Share on other sites

The algorithm that fixed the bad line above-

 

C WRITE OUT THE AREA REQUIRING CORRECTION.
C REPAIR 
      DO 1 LINE= 1, BADLINES
      BAD= BADCOLUMN( LINE)
      AVERAGE1= 0.0D0
      AVERAGE= 0.0D0
      AVERAGE2= 0.0D0
C PIXELS IN COLUMN BAD, ROWS 1: BADROW REQUIRE A NON-UNIFORMITY CORRECTION.
C 1) GET AVERAGE
      DO 2 J= 1, BADROW( LINE)
      AVERAGE1= AVERAGE1+ DFLOAT( IMAGE( BAD- 1, J))
      AVERAGE= AVERAGE+ DFLOAT( IMAGE( BAD, J))
      AVERAGE2= AVERAGE2+ DFLOAT( IMAGE( BAD+ 1, J))
2     CONTINUE
      SMOOTH1( 1)= ( IMAGE( BAD- 1, 1)+ IMAGE( BAD- 1, 2))/ 2.0D0
      SMOOTH2( 1)= ( IMAGE( BAD+ 1, 1)+ IMAGE( BAD+ 1, 2))/ 2.0D0
      SMOOTH1( 2)= ( IMAGE( BAD- 1, 2)+ IMAGE( BAD- 1, 3))/ 2.0D0
      SMOOTH2( 2)= ( IMAGE( BAD+ 1, 2)+ IMAGE( BAD+ 1, 3))/ 2.0D0
      SMOOTH( 1)= ( IMAGE( BAD, 1)+ IMAGE( BAD, 2))/ 2.0D0
      SMOOTH( 2)= ( IMAGE( BAD, 2)+ IMAGE( BAD, 3))/ 2.0D0
C MOVING AVERAGE OF 5 VALUES.
      DO 3 J= 3, BADROW( LINE)
      SMOOTH1( J)= 0.0D0
      SMOOTH2( J)= 0.0D0
      SMOOTH( J)= 0.0D0
      DO 4 I= -2, 2
      SMOOTH1( J)= SMOOTH1( J)+ IMAGE( BAD- 1, J+ I)
      SMOOTH2( J)= SMOOTH2( J)+ IMAGE( BAD+ 1, J+ I)
      SMOOTH( J)= SMOOTH( J)+ IMAGE( BAD, J+ I)
4     CONTINUE
      SMOOTH1( J)= SMOOTH1( J)/ 5
      SMOOTH2( J)= SMOOTH2( J)/ 5
      SMOOTH( J)= SMOOTH( J)/ 5
C      WRITE( 7, 100) J, SMOOTH1( J)
3     CONTINUE
C COMPUTE THE DIFFERENCE IN DC OFFSET BY USING AVERAGES OF NEIGHBORS.
      AVERAGE1= AVERAGE1/ DFLOAT( BADROW( LINE))
      AVERAGE= AVERAGE/ DFLOAT( BADROW( LINE))
      AVERAGE2= AVERAGE2/ DFLOAT( BADROW( LINE))
      DCOFFSET= INT( (( AVERAGE1+ AVERAGE2)/ 2.0D0)- AVERAGE)
C APPLY DC OFFSET TO PIXELS NOT CLIPPED.
      IF( IABS( DCOFFSET) .GT. 😎 THEN
          DO 5 J= 1, BADROW( LINE)
          CHANGE= (( SMOOTH1( J)+ SMOOTH2( J))/ 2)- SMOOTH( J)
C CHANGE VALUES THAT REQUIRE IT.
          IF( IABS( CHANGE) .GT. IABS(( DCOFFSET/ 4))) THEN
              IF( IMAGE( BAD, J) .GT. 1 .AND. 
     1            IMAGE( BAD, J) .LT. ( MAXVALUE- CHANGE)) THEN
                  IMAGE( BAD, J)= IMAGE( BAD, J)+ CHANGE
              ELSE IF( IMAGE( BAD, J) .LE. 1 .OR.
     1                 IMAGE( BAD, J) .GT. ( MAXVALUE- CHANGE)) THEN
                  IMAGE( BAD, J)= 
     1                 ( IMAGE( BAD- 1, J)+ IMAGE( BAD+ 1, J))/ 2
              END IF
          END IF
5         CONTINUE
      END IF
1     CONTINUE
 

Link to post
Share on other sites

uploaded, the "non-uniform" column shows up.

 

ISO 10,000- full res

 

Original: Bad Column by fiftyonepointsix, on Flickr

Using program "MONO1FIX" to correct the bad line.

Restored Column by fiftyonepointsix, on Flickr

ISO 5000.

Original: Bad Column by fiftyonepointsix, on Flickr

Restored.

Restored Column by fiftyonepointsix, on Flickr

  • Like 1
Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...