You can of course try a visual assessment but you might have to rely on some actual noise analysis… I have found in DT for instance on some images the default RCD demosaic introduces or at least has way more color artifacts than AMAZE. Also the chromatic aberrations module works really well on color noise so the software and workflow could really influence it as might the default profiles used for color and lenses. It would be nice if there was a DXO mark rating for sensor noise…maybe there is something like that out there…
@Entropy512: Mid-Air Collision ;-(
Thanks to Andy Dodd for your information!
FORC3 : Yes, …
including relying on un-initialized variables,
depending upon the compiler release behaviour …
and so forth.
“replacing the hacked-up dcraw with libraw” : +1 !
[ LibRaw 0.21 (Beta1) | LibRaw ] :
Submitted by lexa on 4 July, 2022 - 10:29
Changes are:
…
Camera support :
…
… OM System OM-1 < ----- <-----< -----
…
+1 Argument:
New format versions spring up like mushrooms,
esp. " Lossless Compression" or “pseudo RAW” files (e.g. l-RAW / m-RAW / s-RAW)
in manufacturer-dependent versions (Canon; Sony; Nikon),
resulting into successive extension issues into LibRaw.
If this wave is being handled sufficiently at one place, we can be grateful 
so that RT can concentrate upon its core: Therapy of images.
Kind regards
Please, find attached the output of exiftool-12.42 applied to an OM-1 ORF example foto:
OM-1__exiftool-12-42__.txt (11.0 KB)
This provides rich information - for example,
WB RB Levels at different temperatures, Color Matrix, Black Level, Crop info, etc. pp.
Compared to an E-M1 Mark II sample, not only these differ, but already structural values like
e.g. Image Width, Strip Offsets, Strip Byte Counts, Preview Image Length in the first place.
I can’t remember such while investigating and implementing RCD, could you share that image?
Away for a few days but I will look some up…
Please, find attached the output of exiftool-12.42 applied to an E-M1 Mark III ORF example foto:
E-M1-III__exiftool-12-42__.txt (11.1 KB)
with similar findings / differences / information supplied from inside the raw file.
Kdiff3 with { mkII | mkIII | OM-1 } exif info was helpful .
SUGGESTION:
Change Title to “Support for Olympus E-M1 Mark III and OM SYSTEM OM-1”.
Providing with pleasure: .JPG and .ORF :
PA160004.ORF (17.2 MB)
HTH !
Cross-reference:
This is the “example foto” from which “OM-1__exiftool-12-42__.txt” was extracted ← c.f. post #44.
This still applies.
@Entropy512 , @bobm , @Thanatomanic , @Beep6581 :
Preliminary Backport from LibRaw 0.21 (Beta1) : olympus_load_raw
proves that OM-1 raw files “*.ORF” can be read without hitting EOF:
→ 3912 rows 5220 columns loaded,
correctly identifying
→ OM Digital Solutions OM-1 + OLYMPUS M.75mm F1.8
During read, a warning results from RawImage::get_colorsCoeff :
→ “Cannot use camera white balance.”
Opened in RT, the resulting images are very dark, but recognizable.
Un-specific for OM-1:
I wondered about the resulting call structure exposed by my debug prints … ? ! ?
rawimage.cc.patch (1.0 KB)
dcraw.cc.patch (5.6 KB)
camconst.json.patch (1.1 KB)
RawTherapee__run.log.txt (1.1 KB)
I’ve been using the local build of RT from the OM-1 branch for several weeks and didn’t encounter any problem. Thanks for the fantastic work!