I wanted to create a true ping-pong delay, in which the left and right channels are switched with each delay. So it's not simply panning the signal left and right as Infyuthsion showed in a tutorial.
So I had to solve the problem of bringing audio from the left to the right and vice-versa, which is not a feature that Audiotool has, so I had to do it the tedious way.
I first split the signal into 2: one that would be the dry and the other that would be the wet. I then split the wet into its left and right channels using a splitter and two minimixers (note that all of the minimixers involved in the delay process have their master levels at 0 dB so that they retain the original signal strength), and used a StereoEnhancer on each to make them both mono. Then the mono signals could be panned in the opposite direction and joined using a minimixer. In the process of being converted to mono, each signal went through -6.021 dB, because adding two sounds of that level produces 0 dB. The TubeD conveniently has a maximum gain of 6.021 dB. That then went into the Delay, which I set to 0 feedback and 100% level, because it would undergo a manually-built feedback loop. Feedback is controlled by adjusting the master level of the minimixer that joins the left and right signals. I included an extra filter for damping of frequencies. Dry and wet went through yet another minimixer so that each of their levels could be adjusted.
I encountered a problem that artifacts appear in the feedback loop, so the feedback (master level of the joining minimixer) shouldn't be set above -4 dB. Changing the delay length causes a pitch shift in the delay.
The descriptions of each device are in their names. Descriptions of what's being changed is found in the automation. Full names of devices and clips in the timeline can be viewed by hovering the pointer over the names/clips. Many of the automated parameters are those that you would normally change on the delay unit.
I also have a little melody included. Use it if you want :D
Discovered the problem with the feedback loop: I used a stereo enhancer, which uses a high-pass filter. There's some resonance, so having too high of a feedback level causes that resonant frequency to become louder over time.