“ canvas effect” during PRNU solved !

[ralphmalph]

Well after much trial and error I got to the bottom of what was causing the “canvas” effect during my PRNU setup.

Just so I am clear when I say LCD block = the round metal plate the LCD drops into

Back story; I have been waiting for a LCD from KIS since last October. During this time I changed the 1.2K resistor out to a 800ohm resistor to get more life out of my current LCD. During this time I purchased a full LCD with PCB and LCD block in almost new condition. ( I was under the impression KIS was not going to make new LCD’s so I found a LCDe that was barley used and the seller through in the LCD block/PCB )

Last week I finally got the new LCD in from France. (So now I have a spare LCD) I took the new LCD from France and put it into the LCD block I purchased. (my goal was to have two LCD blocks  1.2k Resistor & 800ohm resistor. This way I could change the LCD’s very easily with ought soldering boards)

When I installed the new LCD on the new/used LCD block I received I was getting the “canvas” effect during PRNU setup. I was eventually advised to refocus the lens after trying all other options.

This morning just for a last ditch effort I took out the LCD block and PCB. I measured the block that I put in and it was 12mm while the one I had previously installed measured 10mm.  There was a 2mm difference in the LCD blocks which explained the focusing problem.

I then took the PCB with the 1.2 K resistor off this block and attached it to the old ( my original) LCD block.  I then ran a new PRNU and my machine is now uniform! ! ! !

I contacted the person who sold me the LCD , LCD block and PCB. His machine was much older than mine. So somewhere along the line KIS decided to change the thickness of the LCD block from 12mm to 10mm. I was very lucky to catch this and was a absolute last chance guess.

[Beach Photo]

I am glad you worked it out. Thanks for the info ralphie-boy!

[relaxia]

Yeah ! (alc1)

For me that was really impossible to make such harm to the lenses (All lenses) due to the block exchange.

To fix the issue on new block ralphmalph would have to screw the lenses down for 2mm - he would never get out of this calibration to the straight, taking into account that all symptoms were suggesting that he should UNSCREW the lens. Very very lucky guy. And conscious

GOOD lesson for everyone: if you have to - change only LCD - not the LCD block or LCD baseplate - unless it has the same thickness.

3 thicknesses were implemented in the past - 12, 10 (in 2004) and 8,5 mm (in late 2005) - SO BE AWARED.

[andreseverin]

I have a similar defect, but not so wavy pattern, more rigid grid pattern visible in all grey area in test images. also in sky or flesh tones if the are even area.

My technician said as a comment to this knowledge that it should be absolutely no difference if the plate is 8 or 11 millimeters as the LCD is mounted on the upper side and the plate is mounted from below. Does this make any sense?

He changed my LCD block and LCD and did a refocussing. I suggested this wasn´t done properly. He simply dismissed the idea, and stated that this type of machine will always produce a visible pattern in prints and that my expectations are too high!