[PSD Format] Image Data Section vs. Channel Image Data etc.

[_videohead_]

Hi -

Newbie here. Would someone be kind enough to clarify the following points of confusion:

1. On the difference between the Channel Image Data section(s) of a PSD file and its Image Data Section: my guess is that the former contains per layer raster info, wheras the latter holds raster data for the background (pane?).

2. All channel image data section(s) reside at the end of the layer and mask info section, after the adjustment layer info section. The spec (I'm working with the 6.0 doc), vaguely suggests that this is the case (see the description of length attribute of channel length info section).

Thanks,

- Olumide

[qu1j0t3]

Is this not clear from my source?

Btw I'm still considering your suggestion of adding a C++ interface to it.

[chiYeung-ROQCDq]

Hi, I am also trying to decode PSD layers' Image Data Section.

Tried to use the source but unable to build it because of the following missing files with Visual Studio 2008.

Any suggestion ?

1) png.h

2) psdrecover.h

3) sys/erno.h

I also tried  decoding it myself but seem to missing something.

Compression is RLE, then 2 bytes header for each row of run-length iterate over heigth of the image.

Is this correct ? 

CT

[_videohead_]

Sorry Toby, but I did not read your entire source. At the time we communicated, I knew nothing of the Photoshop file format. Now however, I've pretty much written my own -- okay I started from an abandoned project.

You really should consider adding a C/C++ interface to your code. Your code appears to be mature and well tested but it also seems to be doing just one thing i.e., what YOU want it to do.

Can you please help answer my question. I'd rather hear from you than read your source.

Thanks,

- Olumide

[qu1j0t3]

(1) The "image data" section is the merged composite (i.e. unrelated to layers). (In psdparse, this is handled by the call to doimage(), line 727).

(2) Yes, the channel image data comes after the "layer info" section. The layer info section describes each layer (see this loop in dolayers()). Immediately after processing that section, the channel image data is extracted by another looped call to doimage().

Hope this clarifies.

[_videohead_]

Thanks Toby. I didsn't realize that the image data section is the merged composite (of all layers?). If so, I needn't read this section.

[qu1j0t3]

You can't ignore it entirely - a PSD that consists of background image only has just this section and no layers.

A layered file saved without "Maximise compatibility" will have a placeholder image in this section, containing multilingual text, "This layered Photoshop file was not saved with a composite image." Otherwise it is the merged composite.