OP
- Thread Starter
- #61
It can indeed. This is how it works and why.
In Windows, multiple applications can make sound simultaneously even though you have one audio hardware. You could be playing music while your email program wants to play "bing" to tell you that you have a new message. Your music could be 96 Khz but the bing, 44.1 Khz.
The solution is to convert all incoming audio samples from every application to a fixed sampling rate. This is in the sound control panel for your specific sound device:
The above is the internal DAC for the laptop I am using. As you see, there is a drop down that says, "24-bit, 44100." If you click on that, you will see a list of different bit depth and sampling rates. No matter what you play, Windows will resample to that value.
Windows also has to mix the sound from different applications. To maintain accuracy, it will convert all the integer audio samples to "floating point" and does the mixing in that domain. At the end, before sending it to the sound card, it needs to convert them back to integers. It can't do this by just truncating the fraction as that causes distortion. So it adds a bit of noise to them called "dither." This means that even if you play a file that is at the same sample rate as the above control panel, what goes to the sound card may be a bit different.
Some but not all sound card drivers could have two other interfaces that bypass all of this. One is called ASIO and then other WASAPI. By using an application like Foobar2000 that supports that kind of interface, you have a direct channel to the sound card and avoid any modification of audio samples so your source tones remain pure.
Similar problem exists in iTunes on MacOS but players like Amarra perform a similar function to above, providing "sample accurate" playback.
Don't worry if these things seem complicated. They are complicated
. We are happy to explain. Just ask questions.
