Can a JPEG be truly 1-bit black-and-white?
Asked 6/21/2013
1 views
2 answers
0
I’m trying to understand whether a JPEG can actually store a bitonal image where every pixel is only black or white (1-bit per pixel). In the tools and libraries I’ve tried, even if the image contains only black and white pixels, it still ends up as 8-bit grayscale or 8-bit-per-channel data. Is a true 1-bit JPEG possible, and if not, why not?
Originally by Photography Stack Exchange contributor. Source · Licensed CC BY-SA 4.0
Photography Stack Exchange contributor
13y ago
2 Answers
5
It depends what you mean by JPEG. Commonly, we're talking about JFIF/EXIF image file formats, which specify either one or three bytes per pixel (regardless of the colour space mapping those bytes to bits). Create a file that actually has only two discrete levels present, and it's no longer a JFIF/EXIF file. (And since throwing out high frequency information is the basis for JPEG compression, and all tonal changes in a bitonal image are by definition high frequency transitions, JPEG is a pessimal solution.)
For pure bitonal images using a JPEG-like compression, you'd need to be using JBIG or JBIG2, a format essentially for fax (and not commonly readable). A paletted solution that uses DEFLATE-style compression (redundancy mapping + Huffman or LZW-style bit reduction), such as PNG would be a better solution for most applications, however, since the position (and not merely the distribution) of the bits in the reconstructed image is guaranteed.
Originally by user2719. Source · Licensed CC BY-SA 4.0
user2719
13y ago
0
Generated from our catalog & community — verify before relying on it.
In normal use, no: a standard JPEG/JFIF/EXIF image is not a true 1-bit-per-pixel format. Even if your picture contains only black and white pixels, JPEG typically stores grayscale as 8-bit values (or color as 8-bit per channel).
Why: JPEG is designed for continuous-tone images such as photos, where compression works by approximating tonal detail. A bitonal image has only abrupt black/white transitions, which are exactly the kind of high-frequency detail JPEG handles poorly. So it’s not just uncommon — JPEG is a bad fit for that type of image.
A true 1-bit image is an image where each pixel is only black or white, with no gray levels at all. If you need that, use a format meant for bitonal or indexed images, such as PNG (for broad compatibility) or specialized formats like JBIG/JBIG2 for fax/document-style compression.
So: you can make a JPEG that looks black-and-white only, but it usually won’t be stored as a genuine 1-bit image.
Recommended products
UniqueBot
AI13y ago
Your Answer
Related Questions
What lossy formats can compress 16-bit grayscale images without reducing dynamic range?
How does editing in 16-bit Photoshop relate to 8-bit, 10-bit, and “16-bit” monitors?
Why are image formats different if an image is just pixel values?
Why is white balance adjustment more accurate with RAW than with JPEG?
Does converting an 8-bit image to 16-bit or 32-bit help prevent posterization when editing?