Capture documents in bitonal, or grayscale then downsample

Question

I'm about to embark on a document archival process. I'm going to spend a lot of good money to archive some paper (actually microfiche) to TIFF images. I have a choice of 300-dpi bitonal (2-bit, black/white) or 300-dpi grayscale (8-bit). Cost is the same for either format.

Data volume (and thus image size) is not a factor.

It seems to me that the grayscale, since scanned at the same resolution as the bitonal, would always contain more information and could always be downsampled to the equivalent bitonal image. Are there any downsides to selecting grayscale, and then later downsampling to bitonal if desired? In other words, is it possible that the scanning software will perform a more accurate (or more legible) representation than a grayscale image converted to bitonal?

Answer 1

It's always better to retain as much information as possible for as long as possible.

Scanning greyscale and then converting to bitonal gives you the option to tweak the thresholds and possibly even clean up some of the images (though this would be time consuming) if the results of the conversion aren't good enough.

Scanning to just black and white, while may be good enough in 80-90% of the documents (say), there will always be some that require manual intervention.

I suppose it all depends on how much you value your time and how good you want the results to be.

Answer 2

There is a good chance that, internally, the scanner always works in grayscale anyway, and, when asked for bitonal, just gives black for 0-127 and white for 128-255. So, going for grayscale is by far the best option, since it adds flexibility, in those days of inexpensive storage.