Rogue Amoeba's Audio Hijack Pro is a great program, but it seems to me that the developer's own Web pages fail to explain exactly why. The conceptual difficulty is that Audio Hijack Pro occupies two niches at once - it does two quite different things. So, in reading about it, if you don't particularly want it for the first thing it does, the second thing it does might not even register upon your consciousness. Yet this second thing is extremely cool and, as far as I can tell, quite unique.
First Things First -- The first thing Audio Hijack Pro does is simple to describe: it records to a sound file any sound that your computer is generating. To see why this is useful, think about sounds your computer generates from time to time that you might like to record to a file. For example, you might be listening to an audio stream via the Internet using RealPlayer - a live radio station webcast, perhaps, or a replay of some earlier show. With RealPlayer, there is no sound file: you download a tiny file to start with, but that's essentially just a URL. The actual sound exists only as it streams. But that sound is coming out of your computer, so with Audio Hijack Pro, you can record it. Similarly you can record the soundtrack from a DVD that you watch with DVD Player. And so forth - if any application on your computer is generating sound, you can record it.
The recording that Audio Hijack Pro generates can be a sound file in any of several standard compressed or uncompressed formats: 16-bit AIFF, 24-bit AIFF, MP3, AAC, or Apple's new lossless (ALAC) format. Also, what's generating the sound doesn't have to an application; it can be a port. So, if sound appears at your built-in microphone, your Line In port, or at your USB port through a "breakout box" such as the Griffin iMic or RadioSHARK, or any of a large number of other more-sophisticated devices, you can record it. For example, if I want to make a quick audio note to myself, I can set Audio Hijack Pro to record from the internal microphone to a highly compressed 32 Kbps MP3 file and just speak directly at my computer. At the other end of the scale of sound quality and file size, I can digitize a cassette tape or vinyl LP of classical music by recording in 24-bit AIFF format from my Tascam USB box, to which my stereo system is hooked up.
Even if this is does interest you, though, you still might not feel that it's worth paying $32 for Audio Hijack Pro. Granted, $32 is not a lot of money; but there are alternatives that are cheaper still. There's Audio Hijack's non-Pro little brother, Audio Hijack, which is only $16, and does the same thing. The main difference is that Audio Hijack records only to 16-bit AIFF files; you can set the sample rate, but that's your only choice. Still, AIFF is the best format for editing your sound files, and you can always use iTunes or any other QuickTime-savvy program to convert from AIFF to one of those other formats afterwards; so you might reasonably feel that the capability to record directly to a compressed format on the fly is no great advantage. Plus, even Audio Hijack has competitors: Ambrosia Software's new WireTap Pro (which replaces their earlier free WireTap) is just $19 and can record to various formats. There is also the free Jack OS X, though this requires a separate application to generate the final sound file, and is more work to set up. And if all you want to do is record the sound coming into your computer, you might be happier with a program such as the $30 Amadeus II, which records the sound, lets you edit it (including click and pop removal and application of many other effects and filters), and can save to more file formats - and costs less - than Audio Hijack Pro.
Plug It In, Plug It In -- This brings us to the second thing Audio Hijack Pro does: it can process the sound digitally as it records it. It does this by means of plug-ins, some of which are included with Audio Hijack Pro, and some of which are already present on your computer (you're free to install others as well). Indeed, one of the best-kept secrets of Mac OS X is that it includes a number of astoundingly powerful digital signal processing plug-ins, including a 31-band graphic equalizer, a compressor, a limiter, and high-pass and low-pass filters. If you've used GarageBand you may have noticed these effects, but they are also available to any AudioUnits-savvy sound application - and that includes Audio Hijack Pro. And AudioUnits are not the only flavor of plug-in that Audio Hijack Pro can handle. It has its own plug-in format, called 4FX, and comes with nearly two dozen functions, good mostly for adjusting gain and balance. Plus it recognizes LADSPA plug-ins, a format originating on Linux; many free LADSPA plug-ins are available, and a handful come with the program. Finally, Audio Hijack Pro accepts VST plug-ins, many of which are both commercial and free.
The great part is that you can apply multiple effects at once. It works like a simple patch board: you can apply effects in series, in parallel, or both. The interface is brilliantly efficient: it's just a grid, which operates as a sequence of columns - all effects in column 1 are applied in parallel before proceeding in series to column 2, and so forth. The sound coming from your computer passes through the effects before your hear it and before it goes to a file; thus you can monitor what you're doing to the sound as you record it. As you're listening, you can turn any effect off and on just by clicking a button, to distinguish what it's doing to your sound; and you can experiment with whatever adjustments in gain or parameters each effect permits.
Bear in mind, furthermore, that you are under no obligation to apply your effects when you originally record the sound. You can just as easily record the sound without effects to start with, then play that recording to make a new recording to which you apply effects. Indeed, you can apply effects to a recording that you didn't create yourself in the first place. This means that Audio Hijack Pro can function as an amazingly inexpensive remastering laboratory. Let's say, for example, that I've obtained an MP3 of one of the old radio Goon Shows. It's good, but I'd like to sweeten up the sound a bit, and I'd like to remove some high-frequency artifacts. So I play the MP3 through iTunes, which I have "hijacked" with Audio Hijack Pro, applying the Excitifier and LowPass effects. As I listen, I tweak the effects parameters until I like the results; then I go back to the beginning of the MP3, set Audio Hijack Pro to record to an MP3 of the same quality as the original, and play through the entire thing. When I'm done, I've got a new MP3 that sounds a little nicer. (Transforming an MP3 to another MP3 usually involves some drop in quality, because MP3 is lossy, but in this case the overall effect is positive because there was no particular fidelity to maintain in the first place.) Again, at the high end of sound quality, I could start with a 24-bit AIFF recorded from a vinyl LP and play that through iTunes, applying a little equalization and compression and ending with a dither effect, and recording to a 16-bit AIFF; presto, I've just used Audio Hijack Pro to master an audio CD. (And to top it all off, I can even burn that audio CD directly from Audio Hijack Pro.)
Now there are, of course, other ways to apply effects, and other ways to master. In a program like Amadeus II, you simply select a stretch of audio and apply the effect directly: the calculations are made, and the audio is rewritten. The disadvantage of Audio Hijack Pro's approach is that the effect is applied live: the only way to pass a sound through an effect is by playing the entire sound, so the time required to apply the effect is the sound's duration. But the advantage of Audio Hijack Pro's approach is the very same thing - the effect is applied live. This means that you can monitor the effect as it is applied, and can even make adjustments in real time (for example, you might wish to turn up the compression for one part of a recording a bit more than for another part). Furthermore, I don't know of any other application at anything like this low price that lets you apply multiple effects in parallel and series so simply as Audio Hijack Pro.
More Than Meets the Ear -- Audio Hijack Pro is full of extra features and capabilities at which I haven't even hinted so far. You can record the sound output of more than one application at once (each being routed separately to its own file), or record one application while listening through your speakers to another. You can set a recording to stop automatically after a certain amount of time - good for preventing a sound file from becoming huge while you're distracted by the phone. You can set up Audio Hijack Pro to record from a certain application at a certain date and time - good for recording the upcoming broadcast of Car Talk via Internet radio while you're out for a run. You can have recordings automatically be split into multiple files as they are created, either at timed intervals or during long silences (so as to separate them into tracks, for instance). You can have Audio Hijack Pro tell the sound application what file to open (or what AppleScript program to run) before it begins recording - thus you might set RealPlayer to switch automatically to the desired Internet radio station, for instance. You can also have it run an AppleScript script after recording is finished, to post-process the recording file (adding it, perhaps, to a particular iTunes playlist).
Seen in the full light of what it can do, Audio Hijack Pro seems a bargain at $32. The program is constantly being improved. The manual could be more complete, but support is excellent, and there are user forums moderated patiently and helpfully by the developers. A clever demo system allows you to try before you buy: the program works normally, but adds some static to any recording longer than about 10 minutes until it is registered. The download is less than 3 MB. Audio Hijack Pro requires Mac OS X 10.2.7 or later.