The result comes out in "TUPLTYPE RGB_ALPHA".

Code: Select all

```
convert pf-269361.red.sp0.5.m1.16.pam pf-269369.red.sp0.5.m1.16.pam -compose CopyOpacity -composite pf-269361_9.red.sp0.5.m1.16.im.pam
head -n7 pf-269361.red.sp0.5.m1.16.pam pf-269361_9.red.sp0.5.m1.16.im.pam | egrep -i "pam|tuplt"
pf-269361.red.sp0.5.m1.16.pam TUPLTYPE GRAYSCALE_ALPHA [16 bits GRAY, 16 bits OPACity]]
pf-269361_9.red.sp0.5.m1.16.im.pam TUPLTYPE RGB_ALPHA [ 16 bits red, green, blue and opacity ]
```

Colorspace: sRGB

Type: GrayscaleAlpha

Base type: TrueColor < GrayScale and RGB completely different datatypes? Uint16 vs Uint[3][16] or ~Uint48?

Channel depth:

Red: 16-bit; Green: 16-bit; Blue: 16-bit; Alpha: 16-bit

And the RGB channels all have the same min/max/mean/sdev/kurt/skew/entr so the image would look gray.

My goal is to do something equivalent to the Photoshop Operations:

- Open PAM1

Select_all

Copy

Open PAM2

Create New Layer

Paste PAM1 on a layer above Pam2

Layer->Merge

Save_as PAM3 (as 16_bit_grayscale with 16_bit_alpha)

OpenSuse: ImageMagick 7.0.8-20 Q16 x86_64 2018-12-25 https://imagemagick.org (latest available)