mp3 or jpg to Binary

You would probably end up using hex data for a jpeg file, below is about 10% of a simple jpeg.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You can use HexFiend to look at the Hexadecimal (base 16 instead of 2) code of any file. Then use a hex to binary converter (you can just google that) to see the 1s and 0s.

"...like they say things in movies are not always real."

Right, so for the sake of making the movie, why not just print out a random sequence of 1's and 0's?

So is it possible to open or convert a mp3 or jpg file as binary text which can be printed rescanned as a text file and re opened.

Weird question, and not really appropriate to this forum, but okay, I'll bite. Yes and no.

Why yes? Well, as you say, computers are binary. If you have a jpeg that is 50 kb in size, that translates to 50 x 1024 bytes, and each byte is made up of 8 bits, where a bit is simply a one or a zero. Thus, your 50 kb jpeg file would be made of about 409,600 ones/zeros. A 50 kb jpeg is something you might find on a web site... something large and photographic in nature, such as the images from my 8 megapixel camera, are closer to 3-4 Mb (25,165,824-33,554,432 ones/zeros). An mp3 could easily be larger still.

This brings us to the no. A very simple and small 50 kb jpeg file, converted into binary and printed onto paper, would be many pages. In a 9-point font with the smallest margins allowed by my printer, I could have fit slightly more than 8,200 ones on a page. (Zeros are wider, so the average per page would probably be lower for mixed ones/zeros.) That means your 50 kb jpeg would take up 50 pages (or 25 front and back). A photo from my camera could take more than 4,000 pages, not to mention a large mp3! Not at all practical.

Further, you talk about scanning and converting back to binary, which would presumably involve a scanner with a sheet feeder and some OCR software to convert the scanned page image into text. This would be extremely time-consuming, and would probably involve a significant amount of error introduced into the data in the OCR software.

So, for your idea about people using paper to store data... not very likely at all for this kind of data. Of course, remember that people actually did once store data on paper... boxes of punch cards! (Don't drop the box... card order matters!) Of course, most of the data being stored this way was program code that, if written out in a modern programming language, might take a couple pages.