How to export a multi layer TIF file, keep the original layer name to generate PNG, and not crop the layers

[291598912]

ImageMagick version

7.1.0.10

Operating system, version and so on

windows11

Description

多图层的tif文件，怎么导出图层，保留原来图层名字生成png，且不裁剪图层

[291598912]

How to export a multi layer TIF file, keep the original layer name to generate PNG, and not crop the layers

[snibgo]

Suppose file in.tiff contains images with labels. Then this command will create one PNG file per input image. Each PNG file will be named for the label, with .png extension.

magick in.tiff -set filename:MYFILE %l %[filename:MYFILE].png

If using in a Windows BAT script, double each %.

Does that do what you want?

[291598912]

This Chinese layer name may have garbled characters，But I still cropped the layer. I want to keep the layer in its original background position

[291598912]

Can the png obtained in this way be the same as the original input.tif in synthesis

[291598912]

I would like to consult some teachers about the CMYK color mode multi layer TIF file. Extracting layers will result in color changes. How can this be resolved? I can't even try to convert it to RGB.
Im_.png is the layer obtained by magick splitting tif
Ps_.png is the layer split by Photoshop

The command line is:
magick input.tif %d.png

[291598912]

The original tif file is:
intput.zip

[snibgo]

The file intput.tiff contains 23 CMYK images, each with an embedded ICC colour profile. PNG cannot record CMYK profiles. If you want to convert them to sRGB, use "-profile sRGB.icc" or the filename of whatever sRGB profile you want. If the ICC file is not in your current directory, provide the full path.

[291598912]

The issue of color difference has been resolved, - profile sRGB. icc, but there are still issues with cropping

[snibgo]

You can see what sizes IM thinks the images are:

magick intput.tif info:

intput.tif[0] TIFF 1122x2209 1122x2209+0+0 8-bit ColorSeparation CMYK 12.4279MiB 0.016u 0:00.009
intput.tif[1] TIFF 924x1874 924x1874+10+36 8-bit CMYK 12.4279MiB 0.016u 0:00.012
intput.tif[2] TIFF 924x1874 924x1874+10+36 8-bit CMYK 12.4279MiB 0.016u 0:00.013
intput.tif[3] TIFF 916x1866 916x1866+18+44 8-bit CMYK 12.4279MiB 0.016u 0:00.013
intput.tif[4] TIFF 916x1866 916x1866+18+44 8-bit CMYK 12.4279MiB 0.016u 0:00.014
intput.tif[5] TIFF 917x1867 917x1867+17+43 8-bit CMYK 12.4279MiB 0.016u 0:00.015
intput.tif[6] TIFF 888x1800 888x1800+12+76 8-bit CMYK 12.4279MiB 0.016u 0:00.016
intput.tif[7] TIFF 888x1800 888x1800+12+76 8-bit CMYK 12.4279MiB 0.016u 0:00.016
intput.tif[8] TIFF 524x259 524x259+10+36 8-bit CMYK 12.4279MiB 0.016u 0:00.017
intput.tif[9] TIFF 226x31 226x31+161+41 8-bit CMYK 12.4279MiB 0.016u 0:00.018
intput.tif[10] TIFF 226x31 226x31+554+51 8-bit CMYK 12.4279MiB 0.016u 0:00.019
intput.tif[11] TIFF 226x31 226x31+144+1873 8-bit CMYK 12.4279MiB 0.016u 0:00.020
intput.tif[12] TIFF 31x226 31x226+24+307 8-bit CMYK 12.4279MiB 0.016u 0:00.020
intput.tif[13] TIFF 31x226 31x226+24+703 8-bit CMYK 12.4279MiB 0.016u 0:00.021
intput.tif[14] TIFF 31x226 31x226+24+1094 8-bit CMYK 12.4279MiB 0.016u 0:00.022
intput.tif[15] TIFF 31x226 31x226+24+1470 8-bit CMYK 12.4279MiB 0.016u 0:00.023
intput.tif[16] TIFF 31x226 31x226+897+148 8-bit CMYK 12.4279MiB 0.000u 0:00.023
intput.tif[17] TIFF 31x226 31x226+897+405 8-bit CMYK 12.4279MiB 0.000u 0:00.025
intput.tif[18] TIFF 31x226 31x226+897+801 8-bit CMYK 12.4279MiB 0.000u 0:00.027
intput.tif[19] TIFF 31x226 31x226+897+1192 8-bit CMYK 12.4279MiB 0.000u 0:00.028
intput.tif[20] TIFF 31x226 31x226+897+1568 8-bit CMYK 12.4279MiB 0.000u 0:00.028
intput.tif[21] TIFF 226x31 226x31+570+1873 8-bit CMYK 12.4279MiB 0.000u 0:00.029
intput.tif[22] TIFF 226x31 226x31+162+259 8-bit CMYK 12.4279MiB 0.000u 0:00.030

For example, IM thinks the first image is 1122x2209 pixels. Gimp agrees.

Which of these are wrong?

[291598912]

What I mean is that starting from the first layer, the image size is 1122x2209

[291598912]

The exported layers are all 1122x2209 sized

[291598912]

What I want is the size of the second image

[291598912]

The picture posted above：
intput.tif[8] TIFF 524x259 524x259+10+36 8-bit CMYK 12.4279MiB 0.016u 0:00.017

[291598912]

convert intput.tif -profile sRGB.icc -depth 8 -channel rgba %d.png
convert intput.tif -profile sRGB.icc -depth 16 %d.png
convert intput.tif -profile sRGB.icc -depth 32 %d.png
Regardless of the setting, it is not 32-bit, either 2-bit, 8-bit, or 48,64-bit

[snibgo]

The exported layers are all 1122x2209 sized

That is not what I get.

If we do a simple conversion:

magick intput.tif x.png

... then we get files like this:

x-0.png 1122x2209 pixels
x-1.png 924x1874 pixels

...and so on.

[291598912]

The meaning can only be extracted from the actual size of the layer and the generated image

[291598912]

Then I will obtain the coordinate positions of the layers and generate them myself

[fmw42]

You can crop using the virtual canvas coordinates, e.g. for the last image which is
intput.tif[22] TIFF 226x31 226x31+162+259 8-bit CMYK 12.4279MiB 0.000u 0:00.030
use

magick intput.tif[22] -crop 226x31+162+259 x.png

Does that do what you want?

[291598912]

This is not what I want, it doesn't need to be cropped

[291598912]

Can the png obtained in this way be the same as the original input.tif in synthesis

[CrewS]

Have you done it, composited the png back into the same Tif?

[291598912]

Why do many PNGs have depths of 2 or 8 bits, can they be 32 bits

[snibgo]

Can the png obtained in this way be the same as the original input.tif in synthesis

I don't understand the question. Each output PNG should be the same width and height, in pixels, as the corresponding image from the input TIF file.

Why do many PNGs have depths of 2 or 8 bits, can they be 32 bits

PNG files can have 1, 2, 4, 8, or 16 bits per channel.

[291598912]

The exported RGB image still has color difference
convert intput.tif -profile sRGB.icc %d.png

[291598912]

The color difference has been resolved, but I still don't know how to use commands to generate 32-bit images

[snibgo]

What do you mean by "32-bit images"? PNG files cannot have 32 bits per channel. But they can have 8 bits per channel, with four channels (RGBA), making a total of 32 bits per pixel.

[291598912]

Bit depth is always 8 or 2 bits

[291598912]

It should be using png32 to write files. How can I handle this

[fmw42]

Perhaps you want

convert intput.tif -profile sRGB.icc PNG32:%d.png

the PNG32: preface forces a 32-bit (4*8-bit channels) output

PNG does not support 32-bits/channel only 32-bits per image with each of 4 channels equal to 8-bits