Watermarking may also be worth looking at.
(First off, I'd like to apologize if I didn't state my question clearly; however, it's a very frustrating problem.)
I guess the appropriate way to state the problem is this
I need to be able to include transaction information inside the file in a format & method that is not readily perceptible to the human. The transaction info that I'm referring to would be something along the line of
Time/date of transaction
Website where the transaction took place
UserID of the seller
UserID of the buyer
Amount that was charged for the product being sold
Payment type that was used for the transaction
UniqueID of the product being sold
If this type of information is connected to the file, is invisible to the human, and is not easily removed, it would then help you protect the content that you've worked so hard to create from either being just handed out by others or sold by someone else.
If you had a way to track your hard work that's being stolen all the way back to the original thief so that instead of suing a ton of people for a little money each, you had the proof of WHO is the thief and then sue that one for a ton of money. Gotta admit… makes perfect sense.
What ticked me off was a story I heard of Disney bringing suit against a little girl for downloading & watching, 'The Lion King' while she was over at her grandfather's house. Sure, the little girl did indeed download that movie; however, wasn't she just unknowingly receiving stolen property? Didn't the original uploader knowingly and willingly commit a ton of copyright infringement? That's what I believe was (and still is) going on and I want to try and help correct it so that it doesn't go stir-crazy again. I just plain can't stomach stories of a big company going after a little girl when a real criminal deserves to be behind bars, WITHOUT internet access, etc…
I've thought of cryptography & steganography, file header info, exif info, drm, etc.. and I haven't found anything that will do the trick.
I would be forever grateful if you have any suggestions on how to do this.
Simply put, what you suggest can't be done. I had hoped that this would be clear to you but let me illustrate what I think is the fallacy of your argument.
EXIF headers can be made to include a great deal of information because that is what they have been designed to do. The various formats which have been used to incorporate copyright and other information into videos have, similarly, been designed for that purpose.
The examples that you point to are the exceptions rather than the rule.
Consider a very simple ASCII text document. It has no headers or footers. It is possible, if it is using only a 7-bit character encoding system, that you could encode some information in the unused bits. But you change that document such that it can accommodate UTF-16 Unicode, then no matter what single bit changes you make to that document, if they are sufficient enough to include the information that you wish to preserve, you are going to affect how that document is rendered.
A great deal of work has been put into the issue of how to protect the authenticity of files. Unfortunately, little of this work can be retroactively applied to the various file types that exist, today.
Consider the case of the automobile. The first implementations were little more than elaborations on traditional horse-drawn carriages but it wasn't long before people figured out that a new means of power required new engineering.
The earliest steamships used to be equipped with masts. The thought was that if they lost power, they could resort to sail. It wasn't long before someone figured out that these vessels were too heavy to propel with sail and the masts were dropped from the design.
You have made it clear what you wish to do but the simple fact is that you can't apply it to the state of digital data, today, at least not in a way that can be generalized to deal with all file types.
If you want to create digital files which can be authenticated and for which you can detect alteration, or theft, those considerations need to be in the design, up front. They will, likely, add to the overhead of processing these files, they would need to be implemented at the byte, block or packet level, but they can be designed to prevent unauthorized alteration and rendering.
But if you want a document that maintains data about every user/system through which it has passed, remember that at each step in the process, something will be responsible for updating that information. And that something will be a defined algorithm or process which means that it will be possible to reverse engineer.
Your best hope is a one-hop protection, such as Macrovision, which allows the user to view a disk/video, but prevents copying.
But if you are talking about the flow of data through the Internet, you need to depend upon enforcement at every hop and that is not something upon which I would want to rely.