[Frooxius]

This is nice, but why is this for images only? Technically you could embed any data into the target image, assuming there's enough space in the LSB's. Although perhaps it should be noted that the image must be compressed using a lossless algorithm, otherwise the hidden data will most certainly get damaged.

However, I don't think that Steganography is useless, especially if you use it for other types of data. You can actually send someone some confidential file hidden inside an image, which adds to the security. If you send just encrypted file, someone might try to crack the encryption to get to the good bits, however with steganography, he won't even know that there's some valuable data, so he won't even make an attempt to get to the good bits.

Here's another question: What about using other types of data as carriers to hide the information? For example sound (although you would need to send a FLAC or WAV, which could be quite huge...) where you could hide some data in each sample. Maybe you could even think of a function that would be able to work with lossy compressions and use more robust way to store the data, so some distortion and some data loss won't affect the hidden data, like using lower frequencies for example (but these ones that don't get cut out because they're unhearable).

Something similar could be used for a video too. It could get interesting with trying to hide something in a plaintext file, because words could be much more sensitive to changes, unless you use other properties and perks.

Well... food for thought :3

[Deque]

Yeah, still the usefulness percentage is really low.

Is it also ridiculous to produce insulin for diabetes, because only a minority of people can actually use it?

I didn't ask you to state my problems.
Just check my sig. Specially dedicated for you.

Very mature reaction.
I didn't state your problems. "first world problems" is just a phrase saying you make a fuzz about something that is not really a problem.

This is nice, but why is this for images only? Technically you could embed any data into the target image, assuming there's enough space in the LSB's.

I am aware of that as you can see in my previous post. The answer is simple: I just started getting interested in steganography, when I created this program. I was so phascinated by this topic and burned to write some code to try it out, that I didn't even think about making this more general. I just wanted to see how it works, see the pictures and so on. This is the result of it.

It was later that I made up my mind about how this could be improved. I have it on my todo-list to create a similar program that hides arbitrary files and provides encryption.

Here's another question: What about using other types of data as carriers to hide the information? For example sound (although you would need to send a FLAC or WAV, which could be quite huge...) where you could hide some data in each sample. Maybe you could even think of a function that would be able to work with lossy compressions and use more robust way to store the data, so some distortion and some data loss won't affect the hidden data, like using lower frequencies for example (but these ones that don't get cut out because they're unhearable).

Something similar could be used for a video too. It could get interesting with trying to hide something in a plaintext file, because words could be much more sensitive to changes, unless you use other properties and perks.

Well... food for thought :3

I already took it into consideration, but there are certain file formats that interest me more than others. Video and audio files don't belong to the interesting ones for me. I am more into image formats, executables and archives (PE, jar, zip). I am especially trying to find new ways to hide data (ways that are not described in the literature). That knowledge also helps to create more advanced binders and crypters.

The plaintext file is interesting too. You can use whitespaces or control-symbols that can't be seen by the naked eye to hide your data. But analysing that is petty easy.

So much interesting stuff that can be done and so less time to actually do it ...

Many thanks for your mindfood and your elaborate feedback.

[Frooxius]

I am aware of that as you can see in my previous post. The answer is simple: I just started getting interested in steganography, when I created this program. I was so phascinated by this topic and burned to write some code to try it out, that I didn't even think about making this more general. I just wanted to see how it works, see the pictures and so on. This is the result of it.

Yes, it was more of a suggestion. Steganography algorithm shouldn't care at all about what the input data are - it should be generic, unless you're trying to optimize it for certain datatypes and use them for an advantage. If you need to lower quality of the input image, it should be done outside of the steganography algorithm.

I already took it into consideration, but there are certain file formats that interest me more than others. Video and audio files don't belong to the interesting ones for me. I am more into image formats, executables and archives (PE, jar, zip). I am especially trying to find new ways to hide data (ways that are not described in the literature). That knowledge also helps to create more advanced binders and crypters.

The plaintext file is interesting too. You can use whitespaces or control-symbols that can't be seen by the naked eye to hide your data. But analysing that is petty easy.

Okay. Though frankly, audio and video are the ones I find interesting, because with them, the problem becomes non-trivial, because they use lossy compressions, therefore you can't expect them to retain all the information, which is why it becomes to me.

I find finding some space in some datafile to hide the information in quite boring, as it's relatively trivial task and if you stuff some data somewhere, where it's ignored, it's relatively easy to detect that there's some additional data.

Embedding them in a non-obvious and basically (almost)-impossible-to-tell way into a lossy formats in a way, that no bits of the embedded data are lost due to the lossy compression is imho more secure and more challenging.

[Deque]

Okay. Though frankly, audio and video are the ones I find interesting, because with them, the problem becomes non-trivial, because they use lossy compressions, therefore you can't expect them to retain all the information, which is why it becomes to me.

I find finding some space in some datafile to hide the information in quite boring, as it's relatively trivial task and if you stuff some data somewhere, where it's ignored, it's relatively easy to detect that there's some additional data.

Embedding them in a non-obvious and basically (almost)-impossible-to-tell way into a lossy formats in a way, that no bits of the embedded data are lost due to the lossy compression is imho more secure and more challenging.

Yes, it is more challenging, but it is also already described in literature. Unless you invent your own algorithms the hardest part of that is understanding them.
I didn't say I just want to find some space to store the data and there it is. I want to make it undetectable too. Like I said, I want to find new ways. That is imho more challenging than implementing what is already done.

[ArkPhaze]

Umm. One question. Why would you want to hide an image in image? Ridiculous.

Keep your porn hidden? lol

Otherwise, there's many uses for stenography. This is touching the bare surface, although you should look it up

[Frooxius]

Yes, it is more challenging, but it is also already described in literature. Unless you invent your own algorithms the hardest part of that is understanding them.
I didn't say I just want to find some space to store the data and there it is. I want to make it undetectable too. Like I said, I want to find new ways. That is imho more challenging than implementing what is already done.

If you invent something new, then understanding it is actually extremely easy. I mean... YOU invented it, so you have to know how it works, every bit of it, because it came from your head.

I never talked about things that are already done, I in fact referred to designing your own algorithms. Following pre-made solution is not challenging at all. Or why do you think that there are no new ways to hide information in the media I described, but there are ways in the ones you did?

However, what I think you're going to encounter as a problem is that if you take some format and find a place that's normally ignored by the usual tools that process it and stuff your data in there, it's going to be quite easy to detect, because you can simply write a tool that will look at the places that are normally ignored by the code that process given format, in other words, it's going to be simple to make the distinction.

With images, the hidden data actually modify the input data, but that doesn't matter for the picture, because the modification doesn't translate to anything a naked human eye would notice.

Take ZIP format for example or EXE. If you change even one bit of the original data, the whole contents of the archive might become corrupted and changing instructions in the EXE file will most probably corrupt the program too.

[LightX]

Just downloaded it, It looks nice Don't know what I will use it for, but nice job!

[Deque]

Just downloaded it, It looks nice Don't know what I will use it for, but nice job!

Thank you.

If you invent something new, then understanding it is actually extremely easy. I mean... YOU invented it, so you have to know how it works, every bit of it, because it came from your head.

I said UNLESS you invented it the hardest part is understanding, which means the premise of my statement was that you didn't invent it.
I think we talk at cross purposes.

I never talked about things that are already done, I in fact referred to designing your own algorithms. Following pre-made solution is not challenging at all. Or why do you think that there are no new ways to hide information in the media I described, but there are ways in the ones you did?

I didn't say there are no new ways. I only think that inventing your own that is better than the ones that are already there is worth a master thesis. Yes, that is more challenging. But out of practical reasons I didn't think that you really suggest doing this or that you want to do this. Where do you get the time for this unless you really do it as master thesis or similar? (I am talking only about image, video and audio formats here, because those are the ones you find plenty of literature and theses; they are already heavily investigated)

Besides: What did or do you actually do in that field?

However, what I think you're going to encounter as a problem is that if you take some format and find a place that's normally ignored by the usual tools that process it and stuff your data in there, it's going to be quite easy to detect, because you can simply write a tool that will look at the places that are normally ignored by the code that process given format, in other words, it's going to be simple to make the distinction.

With images, the hidden data actually modify the input data, but that doesn't matter for the picture, because the modification doesn't translate to anything a naked human eye would notice.

Take ZIP format for example or EXE. If you change even one bit of the original data, the whole contents of the archive might become corrupted and changing instructions in the EXE file will most probably corrupt the program too.

Why do you still think I just want to put the data into some space? I already said that this is not what I want to do. The EXE doesn't have to get corrupted. An EXE file doesn't only contain code to execute. Simple example: the resource section contains arbitrary data (can be images but other stuff too) and you can use them to hide data. But that is just one possibility and finding more is why I study the file formats.