Unable to read RAW image from Agfa Photo DC-833m sensor

Submitted by dsi on

Hi,
I am using libraw 0.20.2 for reading RAW files. I have been handed two files generated from a DC-833m sensor. The user of my application has captured this using an evaluation kit + DC-833m sensor + Software that came with it.

You can find the two RAW files at the link below:
https://drive.google.com/drive/folders/1X1Tdy744XXb2AaMvE9ZIwsNRqjoYbFUe...

libraw is able to read only one of the files 3264x2448.raw. But the information reported appears to be incorrect. The user claims the file is only 10-bit, but the values read in have a max value of 1060. Additional white level reported is 4095. The specific comments were "actual data aligned to MSB that means LSB 6bits are dummy bits, only 10bits MSB is image data."

The second file 1632x1224.raw errors out with the error:
Cannot decode : Unsupported file format or not RAW file

The user claims that ImageJ and ImaTest software are able to read the 1632x1224.raw files as well.

Confirming if this is the expected behaviour or a bug with libraw.

Dinesh

Forums: 

Unfortunately, the link you

Submitted by lexa on

Unfortunately, the link you provided is not for everyone, I've requested access

Alternatively you can attach the samples to E-mail targeted to info@libraw.org

-- Alex Tutubalin @LibRaw LLC

Thank you for the files,

Submitted by lexa on

Thank you for the files, downloaded.

These files are uncompressed 16-bit sensor dumps (for example, 3264*2448 = 7990272 pixels x 2 bytes per pixel = 15980544 bytes, exact size of file provided).
Also, according to manual (easy to find), camera does not provide raw recording, so raw files are some kind on debug mode or something like that.

I'm unable to see any meaningful image in these samples, it looks like flat field shot, or something like that. So it is not possible to determine if white/black level set correctly or not. Please provide something with well known subject.

Sensor dumps are supported via pre-defined table of known file sizes, to enable half-size image support apply this patch:
diff --git a/src/metadata/identify.cpp b/src/metadata/identify.cpp
index 251f97b2..eab078c0 100644
--- a/src/metadata/identify.cpp
+++ b/src/metadata/identify.cpp
@@ -247,7 +247,8 @@ void LibRaw::identify()
{ 10134608, 2588, 1958, 0, 0, 0, 0, 9, 0x94, 0, 0, "AVT", "F-510C" },
{ 10134620, 2588, 1958, 0, 0, 0, 0, 9, 0x94, 0, 0, "AVT", "F-510C", 12 },
{ 16157136, 3272, 2469, 0, 0, 0, 0, 9, 0x94, 0, 0, "AVT", "F-810C" },
- { 15980544, 3264, 2448, 0, 0, 0, 0, 8, 0x61, 0, 1, "AgfaPhoto", "DC-833m" },
+ { 3995136, 1632, 1224, 0, 0, 0, 0, 8, 0x61, 0, 1, "AgfaPhoto", "DC-833m" },
+ { 15980544, 3264, 2448, 0, 0, 0, 0, 8, 0x61, 0, 1, "AgfaPhoto", "DC-833m" },
{ 9631728, 2532, 1902, 0, 0, 0, 0, 96, 0x61, 0, 0, "Alcatel", "5035D" },
{ 31850496, 4608, 3456, 0, 0, 0, 0, 0, 0x94, 0, 0, "GITUP", "GIT2 4:3" },
{ 23887872, 4608, 2592, 0, 0, 0, 0, 0, 0x94, 0, 0, "GITUP", "GIT2 16:9" },

Or use 'custom camera API'

-- Alex Tutubalin @LibRaw LLC