I am getting weird errors when I try to convert a black and white PIL image to a numpy array. An example of the code I am working with is below.

    if image.mode != '1':
        image = image.convert('1') #convert to B&W
    data = np.array(image) #convert data to a numpy array
    n_lines = data.shape[0] #number of raster passes
    line_range = range(data.shape[1])
    for l in range(n_lines):
        # process one horizontal line of the image
        line = data[l]
        for n in line_range:
            if line[n] == 1:
                write_line_to(xl, z+scale*n, speed) #conversion to other program code
            elif line[n] == 0:
                run_to(xl, z+scale*n) #conversion to other program code

I have tried this using both array and asarray for the conversion, and gotten different errors. If I use array, then the data I get out is nothing like what I put in. It looks like several very shrunken partial images side by side, with the remainder of the image space filled in in black. If I use asarray, then the entirety of python crashes during the raster step (on a random line). If I work with a greyscale image ('L'), then neither of these errors occurs for either array or asarray.

Does anyone know what I am doing wrong? Is there something odd about the way PIL encodes B&W images, or something special I need to pass numpy to make it convert properly?