It's difficult to track down detailed information for the iPad's headphone output on the Internet; the alternative is to take empirical measurements.
Before we proceed further, I should point out that my knowledge is a bit fluffy around the edges when it comes to audio signals and systems. I am, after all, a digital man, and audio tends to involve a lot of wibbly-wobbly analog signals. I'm not talking about digital storage or digital signal processing here; I'm talking about the signals being presented to the outside world when the rubber meets the road.
I do know that that there's a difference in the characteristics of line-level signals (used to transmit analog sound between audio components such as CD and DVD players, TVs, audio amplifiers, and mixing consoles) and the signals coming out of the iPad's headphone socket. I also know that you are really not supposed to use a headphone output to drive an amplifier directly, because they tend to have different impedance characteristics.
However, I also know that the creators of products like my Cyber Acoustics subwoofer satellite system are well aware that the world is full of idiots, that a lot of people will use the system in conjunction with some sort of MP3 player, and that a lot of people will do what I'm doing -- drive the system from a headphone output.
The bottom line is that I decided to put my trust in the fates. When I plugged my splitter cable into the iPad's headphone socket and used one side to drive the satellite system, the resulting sound was pretty darned good, let me tell you.
The next step was to look at the signals coming out of the iPad. Actually, I ended up using an iPod, because I wanted to use my iPad as an oscilloscope. (I'm making the not-unreasonable assumption that the headphone outputs from iPads, iPods, and iPhones are all pretty much the same.) I wandered over into the next bay, where they have all the test equipment and set everything up.
Ignore the computer monitor in the background and the mouse in the center of the image. This was just a convenient table to use. The woofer is seen on the left. The two speakers are in front of the monitor. The iPod is resting near the front edge of the table, and the satellite system's control hub with the blue LED on top is seen to the right of the iPad.
My chum Ivan is the master of the oscilloscope (and he won't let anyone else touch his scope). Before you could say "Max truly is magnificent," Ivan had taken the other output from the splitter and informed me that the signal coming out of the iPod was 2 V peak-to-peak with zero DC offset.
It's not that I don't trust Ivan, but I don't like to leave anything to chance, so I next ran the signal from the iPod into the Oscium iMSO-204 oscilloscope running on my iPad. The following screenshot shows just the B channel in blue; I turned the A channel off. The four digital logic signals are seen at the bottom with nothing much happening, which isn't surprising, since they aren't connected to anything.
Hurray. My iPad/Oscium oscilloscope confirms 2 V peak-to-peak with zero DC offset (happy dance). Now I know what values to plug into my frequency generator to drive my BADASS display.
— Max Maxfield, Editor of All Things Fun & Interesting