I'm reading data from a textfile using [metro 10].
After little logical analysing the data finally the output of [&&] is connected to (the left side of)

|
[0, 66100 1500(
|
[line~]
|
[tabread~ array2]
|
[dac~]

Array2 contains a 88200-Sample-Wav-File.

But always when I build this connection a ground-loop-like sound occurs until I cut this connection again.

Does anybody has an idea how to avoid this loop/such loops?

M.

try using tabread4~ instead of tabread~

should work assuming it's at 44.1kHz. You will have a problem if the 66100th sample in the array is not 0 though since line~ will continue to output 66100 after 1500ms.

you try cutting the output after 1500ms with these:

[1, 0 0 1500(
| |
[tabread4~ array2] [vline~]
| |
[*~ ]
|
[adc~]

i think [hip~ 5] has a similar effect.

I tried saturnos variations, but the feedback loop did not disappear.

On mouseclick at [0, 66100 1500( the generator works well.
The reason of the feedback is outside of the tabread~-Generator.
So a HP-filter too cannot solve the problem....

M.

I found the reason:

At the outlet of the [&&] object sometimes a "1" occurs - as expected.
But 100 times a second the [metro 10] beat triggers the tabread~-Generator through this object.

This is/was the reason for the heared "feedback". Actually it is not a feedback at all.

But how to prevent this [metro 10] - beat ? The input textfile contains data sampled with 100 Hz that should be replayed in real time...

M.

post your patch and it will be much easier to understand the problem.

etap.txt contains measured data, 100 Hz samplng frequency

http://www.pdpatchrepo.info/hurleur/read_write_tabread~.pd

How do you want [&&] to trigger the [line~]? If it's, say, only when the output changes from zero to one, you could do:

[&&]
|
[change]
|
[sel 1]

@Maelstorm said:

How do you want [&&] to trigger the [line~]?

You are right: The output of my[&&] was too "noisy":-(

Now I use

[change]
|
[sel 1]

in combination with a better logic in front of the [&&] and now this happens what I expect.

Exchanging [metro 10] by [metro 100] lowered the pitch of the disturbing "loop", so I assumed the metro-frequency could be responsible for this interference:-(

Thanks.