Great work!
But i am on linux and i don’t know what is magick (i have imagemagick installed, but i can’t call it with magick).
Thank you!
Hi @ggc,
Presumably your imagemagick is too old? Check your version.
Mine is ImageMagick 7.0.9-2 Q16 x86_64 2019-11-06.
Have fun!
Claes in Lund, Sweden
1 Like
I have this one, from ubuntu 18.04 packages…
imagemagick 8:6.9.7.4+dfsg-16ubuntu6.8
Yes, my commands use ImageMagick. I’m using version 7. For version 6, use “convert” instead of “magick”.
In v6, “-evaluate Multiply %%[fx:QuantumRange/1023]” won’t work. Assuming your ImageMagick is Q16 (use “convert -version” to verify), you can substitute “-evaluate Multiply 64.06158357771261”.
In v6, “-size %%[fx:w]x%%[fx:h]” won’t work. Substitute the size of you image, eg “-size 3136x4224”.
I expect this work could be done with G’MIC, but I don’t know how.
An alternative might be to declare the weird pixels to be “dead”, and use dcraw’s “-P” option. I’ve never tried this.
1 Like
Front? Could the inconsistency the OP is dealing with be a front/back lens thing?
Or was one of the units purchased in a different country? Sometimes phones that carry the same “basic” designator (such as Galaxy S3) are VASTLY different internally (and do have a different detailed model number.)
For example, from long ago (but a particularly notable example):
GT-I9300 - International Galaxy S3 (Exynos CPU)
GT-I9305 - International Galaxy S3 (Exynos CPU with ??? LTE modem)
SGH-I747 - AT&T GS3 (Qualcomm Snapdragon SoC with Qualcomm LTE)
I believe some of them had different camera sensors too…
1 Like
@Entropy512 Ho-hum… My Italian was a bit too fast! Here is a more truthful translation:
- a rear 13MP image sensor signed Sony (model IMX278), and
- a front 5MP one
Scusi, Dottore!
Claes in Lund, Sweden
1 Like
There is another possibility: the part has the tech but didn’t pass QA and therefore is processed as if those pixels were dead.
The Sony “Product Brief” for IMX278 (http://cdn.specpick.com/images/photonics/products/IMX278.pdf ) says:
We can pull out all the x0 pixels and make an image. Then the same for x1, x2 and x3.
These are, essentially, lightness. But x0 and x3 are lighter than x1 and x2, by about 10% overall. So these sensels can’t be used in a simple manner for lightness. Further work might establish if the difference is deliberate manufacturing choice or manfactured bugs or whatever.
Demosaicking algorithms for RGBW color filter arrays looks interesting, but is for RGBW CFAs that have a greater proportion of white sensels than the IMX278.
3 Likes
Thank you i’ll try, yes my imagemagick is Q16.
Both phones are been bought in Italy, my friend’s one from Honor store, mine from Amazon.it . Maybe the one from honor store was an international version, i don’t know. Look, software and so on is really the same.
Unfortunatly i have posted a list of phones that should use always this sensor, but on raw.pixls.us there is no other sample from another phone. Iiirc i remember the huawei p8 doens’t create raw files, only jpeg.
Also mine (that demosaics more ok thank this one) has a lot of white dots (see some previous post), that should be used for autofocus, but maybe they also contain some image info.
Maybe ignoring them as a first strategy isn’t a bad idea.
You might be able to find out in the info pane. Often this is hidden but you could access it by a long press or something to that effect. It depends on the software. Or you could crack open the phone and examine its innards but you would risk damaging it and voiding the warranty.
I briefly examined this and others (e.g., ones with yellow pixels) but they are way beyond my ability to implement.
Since this phone’s is sparse and different (x0 and x3 vs x1 and x2), I still think they might be used for something other than straight up RGBW interpolation. Besides focusing, it could be used for depth, transmission or salience mapping. Everything I say is speculation: it is fun to guess what it is. 
1 Like
And beyond mine, unless I spent loads of time on it, which I won’t as I don’t have a RGBW camera.
I haven’t found any online information about how to interpret these weird sensels. A more enthusiastic person might contact Sony. But Sony might mutter about proprietary information.
1 Like
the info panel reports the same things unfortunatly…
thank you anyway for your great study.
I need to find out how to translate your “way” to make dng usable to linux, and it will be a great thing!
For the dcraw part, should be ok. I don’t have a -O option, but i think it won’t matter that much.
But i don’t understand how to translate the exiftool part so i am unable to proceed.
Thank you again.
Use -c to output to stdout, then redirect to a file. In CMD, I could do
dcraw -T -c _DSC0488.NEF > 488.tif
-T Output to *.tiff.
-c Write decoded images or thumbnails to standard output.
command > filename Redirect command output to a file.
With redirecting, notice I could name it any way I want. I like to truncate it to *.tif.
1 Like
Here is a bash script version of my above commands. Tested with Cygwin bash.
CAMERA_SRC=phone/IMG_20191105_134321.dng
dcraw -v -o 0 -6 -r 1 1 1 1 -g 1 0 -D -d -c -T $CAMERA_SRC >cam_gray.tiff
(
exiftool -args -make -model $CAMERA_SRC
echo -DNGVersion=1.4.0.0
echo -DNGBackwardVersion=1.3.0.0
echo -EXIF:SubfileType=Full-resolution Image
echo -PhotometricInterpretation=Color Filter Array
echo -IFD0:CFARepeatPatternDim=2 2
echo -IFD0:CFAPattern2=1 2 0 1
echo -Orientation=Horizontal
echo -BitsPerSample=16
echo -SamplesPerPixel=1
)>exifargs.txt
magick \
-size 16x16 xc:Black \
-fill White \
-draw "point 1,6 point 5,6 point 9,14 point 13,14" \
m.png
KNL=\
0,0,1,0,0,\
0,0,0,0,0,\
1,0,0,0,1,\
0,0,0,0,0,\
0,0,1,0,0
magick \
cam_gray.tiff \
\( +clone \
-define "convolve:scale=!" \
-morphology Convolve "5x5:$KNL" \
\) \
-size %[fx:w]x%[fx:h] tile:m.png \
-compose Over -composite \
cam_gray2.tiff
rm cam_gray.dng
exiftool -@ exifargs.txt -o cam_gray.dng cam_gray2.tiff
dcraw -v -6 -c -T cam_gray.dng >x.tiffThis assumes ImageMagick is version 7. For v6, see my comments above. V7 is better than v6, and I recommend upgrading, unless there are particular reasons for staying with v6.
5 Likes
Great thank you!
I have been able to run the script and to get the picture!
Excellent!
Here’s an idea, if you feel like experimenting further just for kicks:
Open all x1/x2/x3/x4 channels in overlaying windows/layers and try flipping between those quickly. If you see the in focus objects not changing but out of focus ones moving around or changing intensity slightly, then those are indeed phase detect pixels.
Most probably the case of pre-production camera firmware not interpolating those pixels out (some method similar to above) before saving to DNG.
As mentioned above, an alternative method is to tell raw readers that these weird pixels are bad. For dcraw, the bad-pixel list is a text file with each line containing three space-separated numbers: x-coord, y-coord, and timestamp. A timestamp of zero means “we don’t know when this pixel became bad” or “this pixel was always bad”.
Within each 16x16 square, before any image rotation, the four weird pixels are at:
6,10
6,14
14,2
14,6A Windows BAT script that creates badpix.txt, the full list of all weird pixels, is:
setlocal enabledelayedexpansion
echo off
(
for /L %%Y in (0,16,3135) do (
for /L %%X in (0,16,4223) do (
set /A X1=%%X+6
set /A Y1=%%Y+10
echo !X1! !Y1! 0
set /A X1=%%X+6
set /A Y1=%%Y+14
echo !X1! !Y1! 0
set /A X1=%%X+14
set /A Y1=%%Y+2
echo !X1! !Y1! 0
set /A X1=%%X+14
set /A Y1=%%Y+6
echo !X1! !Y1! 0
)
)
) >badpix.txt
echo onA bash script for the same thing is:
for Y in {0..3135..16}
do
for X in {0..4223..16}
do
((X1=$X+6))
((Y1=$Y+10))
echo $X1 $Y1 0
((X1=$X+6))
((Y1=$Y+14))
echo $X1 $Y1 0
((X1=$X+14))
((Y1=$Y+2))
echo $X1 $Y1 0
((X1=$X+14))
((Y1=$Y+6))
echo $X1 $Y1 0
done
done >badpix.txtThe first few lines of badpix.txt are:
6 10 0
6 14 0
14 2 0
14 6 0
22 10 0
22 14 0
30 2 0
30 6 0
:
:Dcraw can then interpolate the missing values before it does any rotation, demosaicing and conversion to sRGB:
dcraw -P badpix.txt -T -6 -O i3.tiff IMG_20191105_134321.dngFor rawtherapee, which doesn’t use a timestamp, see Dark-Frame - RawPedia .
3 Likes
I did it. Seem to work with Raw Therapee.
I’ll test better when my friend will give me more raw images.
Meanwhile, another big thank you!
1 Like