Trying to make a loop station guitar pedal with pure data and Makey Makey. I've tried a number of different ways but the behavior is either inconsistent, or I have trouble making sure playback and recording is lined up correctly across different buffers.

Whats the easiest way to make sure all my reading and writing to buffers are synced up?

Poke~ with count~ as an index of table location (in samples) to make sure I'm writing to the correct location in the buffer? Or tabwrite?

What exactly is "interpolation?" (tabread vs tabread4)?

Why do I need to use a phasor object to read through tabread, and not just the output of count?

How are you guys controlling the sync between different buffer loops (with variable buffer size)?

Whats the easiest way to make sure all my reading and writing to buffers are synced up?

Here's an unfinished patch I began a long time. Inspired by Jamie Lidell:
looper-with-master-sync.pd

Check out what I did for the synchronization. I can explain anything in detail if you'd like.

Poke~ with count~ as an index of table location (in samples) to make sure I'm writing to the correct location in the buffer? Or tabwrite?

As far as I can tell they accomplish the same thing. [tabwrite~] is Vanilla compatible.

What exactly is "interpolation?" (tabread vs tabread4)?

Essentially, smoother playback. More noticeable when playback speed is reduced. See here: Section 3.4.1.2 at http://www.pd-tutorial.com/english/ch03s04.html

Why do I need to use a phasor object to read through tabread, and not just the output of count?

You don't. [phasor~] simply generates a ramp from 0 to 1. You can manipulate this function to read through an entire array, but by no means is it the only way. Using [vline~] is another popular alternative.

To understand interpolation and the difference between tabread and tabread4, create a small array with only 5 data points ("size" in the array properties menu). Then use tabread to read through the array between integers (ie. 1.01, 1.02 etc). Then do the same thing with tabread4:

You will find that tabread jumps from one value to the next, whereas tabread4 interpolates, ie. draws a line between the given points. (Notice that this will only happen after you pass point 1, presumably because it needs a point on either side to interpolate).

As for the phasor question, you should read up about control level vs. signal level objects. There is an explanation here, and probably more about it in PD's native help documentation.

You guys are fantastic, thank you so much. I can't believe I haven't found this pd tutorial before now.

rjp9 - Thanks for the patch. How does sending out multiple master_bang 's work? Does whatever loop you start playing first basically become the "master" track that controls all the other ones? Or is there some other global master bang logic that I'm not seeing.

LiamG - Dear god thank you for that explanation. I will eventually want to be able to adjust the playback speed of the loops, so good to know.

Whats the difference between [timer] and [realtime]? Do they essentially do the exact same thing?

I'm uploading my current patch that doesn't sync properly. I'll spend some time with your suggestions and see if I can get it working, and re-upload. I want to apologize for my messy code, Sometimes I try 8 different ways to get something to work, and when it finally does, I'm too afraid to clean it up because it might break.GUITAR LOOPER STATION B.pd

@sglandry said:

Thanks for the patch. How does sending out multiple master_bang 's work? Does whatever loop you start playing first basically become the "master" track that controls all the other ones? Or is there some other global master bang logic that I'm not seeing.

If i remember correctly, I usually just set "track1" to be the "master track". I would record this track first, and however long it was (in ms) would be the length of the master loop. Whatever you decide is going to be the master loop should be the longest recorded sound, because when it repeats any other tracks that are longer will cut off and repeat as well (if you choose to sync them).

When the "master track" repeats, it sends a bang (via [s master_bang]) out to all the other tracks with the "sync_to_master" box checked (well, it gets sent to all tracks regardless, but only tracks that are synched let this bang pass through). The "sync_to_master" box essentially just controls the [multiplex] object, which switches between the sources of the bangs that tell the track when to repeat. You can have a slave track repeat when the master track repeats, otherwise you can just have it repeat itself asynchronously, or just manually when you click on the bang itself.

Whats the difference between [timer] and [realtime]? Do they essentially do the exact same thing?

As far as I know, [timer] measures logical time, and [realtime] measures elapsed real time. Someone smarter than me on here can tell you the difference.

Yeah, loop 3 was just my attempt to test out how poke~ works. I disconnected it when it didn't work.

How does tabwrite~ know to stop recording? I thought I would have to send it a stop message, but it seems to be working just fine without. Does it just stop recording when it reaches the end of the loop1 table, so when I send the message to resize it, it reaches the end and knows to stop writing to it?

Another issue: I'm clicking on the toggle to start recording at the same time I start playing the riff. The very beginning of the riff seems to get cut off or missed. How can I make sure that the signal from the mic is turns on practically instantly, but make sure there's no "click" sound from turning on or off the adc~ signal instantly.

This is my current setup for loop1:

Hokay! So I got the basic 2 buffer loop working OK. (ignore loop3).

I can punch in to loop2 at the appropriate time to stack loops ontop of loop1. Done like this:

And here's the updated patch:
GUITAR LOOPER STATION B.pd

To understand timer vs. realtime, try this:

Change the number of bang objects on the right (or add anything else you want) and you will find that the time readout varies according to how many objects there are, just as you would expect. Then try it with [timer] instead of [realtime]. Now the readout will always be zero, no matter how many objects you put (unless one of them involves a delay).

So logical time is what the time ought to be, assuming that computation happens instantaneously (which it doesn't).

Ahhh very interesting.

I'm assuming since I really want the loop to be the exact "IRL" time that I hold the button down, I should use [realtime], but I'll try it both ways and see if I can even tell the difference.

Any suggestions to why my loop1 recorder ignores the first few ms of mic input from my intended loop (more than just the 75ms I told the line object to take to completely turn on the mic signal)? Could it just be the amount of time it takes my computer to resize, clear, then resize again the table? I think loop2 functions just fine.

Is there any other functionality that a real loop pedal has that I might want to include in my patch? Once I get everything working as is, I might want to add an option to create another loop buffer thats half, 1/4, or 1/8 the size of my original loop for more interesting patterns. I might add some playback speed functionality so I can drop certain loops to half or 2x speed.

Sorry to keep posting, but I finally got the loop station I've always wanted.

1 base loop, and a second stackable loop buffer.
You can clear the second buffer in case you mess up but want to keep loop1.
You can also toggle on and off the second buffer.
And I threw in a distortion "pedal" for funzies, also toggle-able.

I gutted a USB computer keyboard so only the "A", "G", "L", "Enter", "4", and "+" keys remain, making it easy to control this patch with your foot.

Here is the final patch in case anyone would find it useful:

Looping Station 2.0.pd

@rjp9 This is really nice patch.

Does the tabread4 in this patch need adjusting to avoid clicking when looping? I've read and tried a few suggestions to get round the clicks but haven't shaken them off yet.

Did some better searching today and fixed with a little ramp up an then ramp down at end.