C# mute/unmute specific window/s?

I would like to know if it is possible to mute only a specific window. For example I have got Firefox open and two more windows. I want开发者_JAVA百科 to mute everything related to Firefox but not the whole sound of my computer.

Is this possible? If so, how can it be done?

look, basically, there isn't any relation between the window handle and the sound which something in its code is playing.. the audio card can't tell who wants it to play. theoretically, there is an option to do what you want on web browsers, but it's not easy, and not 100%. it goes like this: most of the audio that is playing from browsers are from known objects like wmp/quick time/vlc/flash/etc.. when the user will choose to mute all audio from firefox, your application will search those known objects in the firefox tabs, and mute/unmute them using their api. in order to do that, you will need to write an extension to firefox, so you could have an access to the tabs memory from your application.

btw, what os? and check this out: http://www.indev.no/?p=projects#flashmute (flashmute) i believe it does what is said - only for flash.

On Vista/Windows 7: I expect there to be some API which can change the volume on a per process basis which the audiomanager uses. Should be relatively straight forward to use.

On XP
I don't think there is any built in functionality for what you want to do. I recommend just not offering that feature on XP. But if you really want to, there are some hackish solutions:

• Usermode API hooking. Intercept the calls to audioapis with your own functions. These change the volume or manipulate the audiosignal so you get what you want. You need to do this differently for any of the several available audio-apis. I guess DirectSound and DirectShow are particularly annoying. And this requires injection of a dll into any process you want to manipulate. And this dll better not require the .net runtime. Search for IAT(import address table) or EAT(export address table) hooking.
• Kernel mode audio hooking. Write a driver which intercepts the audio in the kernel and changes it on a per process basis. No clue how to do that.

But as you can see both solutions aren't good.