Getting a start with pd today, and my initial issue is the test after installation is producing just distortion instead of a sine wave. This must be wrong.

On a windows machine with internal soundcard using ASIO, Not at all far into sound manipulation yet, but have software that does involve the ASIO somehow - REW (Room EQ Wizard - used for room acoustic measurements and loudspeaker measurements). Please advise? Fairly non-technical here, as compared to many of you, I'm sure. Thank in advance...

And when using ASIO4ALL you are only having problems with Pd, not with any other audio application?

Make sure that you have a large enough buffer size in ASIO4ALL and PD.

Windows needs a greater delay, somewhere Miller says that it needs about 80ms...which is not a good thing. But I'm not sure what version of Windows he was talking about.

So, go to **Media **-> Audio Settings, and try with the following settings:
sample rate: 44100, delay(ms): 128, block size 64.

If that works, then refine those settings until you can get an efficient playback with a smaller delay, i.e. experiment within a safe range, 80ms to 128. The smaller this delay, the more responsive everything will be.
http://en.flossmanuals.net/pure-data/ch010_configuring/
http://en.flossmanuals.net/pure-data/ch015_trouble-shooting/

By the way, according to Miller, the order of efficiency/responsiveness was 1st Linux, 2nd Mac, 3rd Windows. The really responsive one being Linux...Fedora, in the case of Puckette. The other two are not that good.

@silas said:

80-128

should be 0-1?

@EEight
I guess it depends on your needs.

Achieving zero doesn't seem to agree with the laws of physics, unless you have a very flexible conception of what zero means, .

We are not so sensitive to delays. It seems that we can learn to adjust ourselves to these delays. I've read that we can tolerate up to 200ms of lack of sync between image and sound in a video, like a short note or echo. So, in general, we shouldn't worry too much about these delays.

30ms would be nice, but it's not realistic yet.

I discovered that even though changing the delay isn't removing the distortion, increasing the block size does, improving at each step up from 64. With a sample rate of 44.1K and a delay of 80 (though regardless where I set the delay, it has no impact on distortion) listening at block size 128 and 256, the tone improves but it's still very distorted. But at 512 there's almost distortion / clipping/clicking-sound free. Only at 1024 is it completely clean. So I'm continuing forward running at mostly 1024. What's this maybe indicating? Thanks for the patience in this slow start....

Then your machine likes to compute blocks of 1024 samples instead of the default 64 samples. Maybe it's not very run-time efficient. Once Pd has calculated a 1024 samples block, it outputs sound.

So I see you've fixed the problem. But I cannot really say for sure which are the best settings. I wish I could tell you more about it, but you should keep trying new settings depending on the challenges each patch will present.