How to save an image with more than 3 channels using PIL?

Question

I need to save a tiff image with 4 channels, specifically R,G,B and A channels.

When I try using Image.save() to save a tiff with 4 channels, the resulting image is an RGBA image but when examining the tiff in Photoshop the only channels are RGB, without the Alpha. Is there any way to merge 4 channels to a RGBA image, with a 4th channel (a separate alpha channel)?

Below is an example of what I've tried

from PIL import Image

# import image A
inputImageA = Image.open(input_imageA_path)

# import image B
inputImageB = Image.open(input_imageB_path)

# split channels
R, G, B, A = inputImageA.split()
alpha_channel = inputImageA.split()[-1]

# merge 4 channels back into a single image
outputImage = Image.merge("RGBA", [R,G,B,alpha_channel])

# save the new image
outputImage.save(ouput_image_path)

In this example the resulting output image only has 3 channels (RGB).

Please see below image for a visual explanation of what I'm trying to do:

Answer 1

Updated Answer

Ok, I think you mean this now:

#!/usr/bin/env python3

from PIL import Image

# Open background image, ensuring RGB
im = Image.open('start.png').convert('RGB')

# Open alpha channel, ensuring single channel
alpha = Image.open('alpha.png').convert('L')

# Add that alpha channel to background image
im.putalpha(alpha)

# Save as TIFF
im.save('result.tif')

Which makes start.png:

plus alpha.png:

into result.tif:

Original Answer

Here's a simple example of creating and saving a 4 channel, RGBA TIFF:

#!/usr/bin/env python3

from PIL import Image, ImageDraw

# Create RGB image full of yellow
w, h = 640, 480
im =Image.new('RGB',(w,h),color=(255,255,0))

# Create alpha channel of white i.e. opaque
alpha =Image.new('L',(w,h),color=255)

# Get drawing context to draw into alpha channel and draw black (i.e. transparent) rectangle
d = ImageDraw.Draw(alpha)
d.rectangle([10,40,200,250],fill=0)

# Add that alpha channel to yellow image
im.putalpha(alpha)

# Save as TIFF
im.save('result.tif')

Answer 2

I figured out a solution to this issue using the OpenCV2 library instead of PIL. See below how I did it:

import cv2

# read original image
original = cv2.imread(original_image_path, cv2.IMREAD_UNCHANGED)

# get dimensions for resizing mask
height, width, channels = original.shape

# read alpha image
alpha = cv2.imread(alpha_path)

# resize alpha image to match original
alpha_resized = cv2.resize(alpha, (height,width))

# split alpha_resized into individual channels
channels = cv2.split(alpha_resized)

# apply to 4th channel of original
original[:,:,3] = channels[0]

# write new image file with alpha channel
cv2.imwrite(output_path,original)