Audio Hijack

Although Audio Hijack is very accessible, some may need a little time to be acclimated to the interface and especially for those who may be familiar with earlier versions. The one thing to keep in mind is that Audio Hijack 3 uses a grid interface where you assemble audio blocks to build the chain.
One nice thing about this new version of Audio Hijack is you do not need the LineIn app any longer.
Here is an excerpt a friend sent me from a blog post by one of the developers; sorry I do not have the original source. This is just a little info to get you started.
Look for a podcast on the application soon.

1. How do I know where a block will go when I paste it into the grid? Are there keystrokes to more easily move blocks around once pasted?

Good question. When pasting a block copied from the library, it’ll either paste at position 1x, 1y, or alternatively, if a block in the audio grid is already selected, it’ll paste on top of that block and the existing block will then be bumped one position to the right.

One important detail: In the current build, a newly pasted block is not selected as the VO target - making this more confusing than in should be. Use cursor keys or VO navigation keys to jump back and forth after pasting and it’ll be selected. We will fix this.

In cases where you copy existing blocks from the audio grid, they remember the position they were in when copied, and aim for that position again when pasted. The exception again is that pasting one block when there is already one selected block will paste at that selected block’s location, bumping the existing block to the right.

2. I know that blocks have popovers, but how do I access them? Perhaps I just haven't yet used a block with one, but I see no way to do this, and vo-shift-m (the control-click command in VoiceOver) does nothing.

Some blocks don't have a popover, just controls on their face (VO interact with the block to visit them). You can use VO-spacebar or just spacebar to open popovers, escape or spacebar to close them. Some blocks have both a popover and controls on their face.

3. What exactly does the grid signify? I get that it is a visual layout of the audio chain, but how do the numbers VoiceOver reports explain what is where? That is, while audio flows from left to right (thus larger x values are later in the chain), what is the y axis for? Why are numbers reported to the hundredths place? It seems like whole numbers would be faster to listen to and give just as much positional information. This is probably very obvious to look at, but just hearing it, I can't see the logic.

Very good question. I'll explain a few things as best I can and hopefully it'll start to make sense to you. We'll be trying to turn this information into a knowledge base article for VO users soon, so if you wish to provide feedback on this explanation, please do!

The block size corresponds to 1 unit of the grid. So a block at 1x, 1y and another at 2x, 1y are next to each other and as close together as they can be.

The grid allows positioning blocks on quarter unit increments. Say for example 1.25x 1.5y. Most users never see grid coordinates and wouldn't know if their blocks are on whole numbers or not. VO users might prefer to put blocks on whole number coordinate just to avoid hearing decimals all the time, I know I did that during testing!

You can navigate between blocks using just the arrow keys. Left, right, up and down will move the selection to the nearest block in that direction. You can also use VO navigation between blocks, and when doing so you'll visit the connections, so they are described to you.

Blocks can be moved in 1 unit or quarter unit steps with Command-arrow and Option-arrow keys.

Connections are made automatically between blocks. Users don't control connections manually, so moving blocks around is the key to creating the connections you want. Since all audio flows from left to right, its helpful to think of connections as being made from left to right. Connections will head to the nearest block on the right, can span as many as 6 grid units and can be made at angles less than 45 degrees from horizontal. Source blocks (the Application, Input Device and System Audio blocks) don't accept audio connections from their left, but all other blocks will allow connections to and from them.

So, the simplest audio graph is one source block and a recorder on its right, next to each other. If you want to listen to the audio as well, you need an output device - add that to the right of the recorder and you are set. Exploring the template sessions will give you some idea of what is possible, you don't need many blocks to do many regular things.

The Voice chat session is a bit more involved, because the application block is set to split the two sides of the conversation onto the left and right channels, then it records both channels while you only listen to one.

At its most elaborate, the audio grid is designed to support multiple sources, arranged above each other, each with effects and recorders to their right. This connects up to create separate audio chains one above the other, which can then be mixed together by angled connections for final output and recording of the full mix containing all of them. Alternately, audio chains in the audio grid can be kept separate from each other, then they’ll all start and stop in sync, whilst handling audio independently.

4. I'm not sure how to create the chain I'm after, perhaps you can offer suggestions. I want to record both a USB microphone and system audio at the same time, letting me demonstrate using VoiceOver as I explain what I'm doing. I think I would want to have system audio and my mic each going to a record block, but is that right? If it is, how would I monitor or adjust the levels?

This works for me with 4 blocks:

- Input Device block set to your microphone at 1x, 1y

- System Audio block at 1x, 2y

- Recorder block at 2x, 1.5y

- Output device block at 3x, 1.5y

If you move the recorder and output device blocks further to the right (so connections aren't too steeply angled to be allowed), you can put them on the 1y or 2y row so you only hear whole numbers

5. This isn't specific to Hijack, but you guys are audio experts. Using Audio Hijack, is there a way to record the sound output of an iOS device, plus the input from a microphone? I'd rather not record iOS with a mic if at all possible. I know Yosemite's version of Quicktime can record iOS audio and video, but can Hijack perform a similar trick, letting me use my iPhone as just another input source? I'm on a Macbook Air, and I need my audio jack for headphones while monitoring, so a patch cord won't work.

I didn't know Quicktime player could record from the iPhone. Nice!

I just experimented and you can make it work. If you set Quicktime Player to capture the iPhone's output, you can then use Audio Hijack 3 to hijack Quicktime Player's output. You do not need to hit record in quicktime player, it'll play through without recording - just make sure the volume control in quicktime player is not at zero (mine was).

If you substitute System Audio with an Application block in the setup I described above, and set that to hijack Quicktime player, you should be set.