Okay so I'm kinda proud of this one, conditional logic in pd has been a bit confusing for me so far, coming from 'normal' programming languages.

Basically I want to send values (from an external source) to a volume-control slider, but instead of having it change immediately, I want it to slide over to the expected value smoothly and automatically. I came up with a solution that works quite well. I might have done some overkill in the logic, I'm not sure, if someone knows a more elegant solution for this, I'd love to see it and learn.

Now, the mechanism works fine, but I've got a strange little quirk that I don't know how to get rid of: the value of the slider keeps dancing around the expected value... I was thinking this shouldn't happen since moses uses < and >= if I understand correctly? And I've used a small enough increment [ + or - .001] with expected values that don't require more precision.... Any ideas?

Oh and one more thing: if I wanted to get a bang for when the value has been achieved, I'm guessing I should just use a [==] with [r slidVal] in the right inlet? I haven't been able to test this yet due to the quirk.

so_smooth.pd

Hi, pd provides some objects to make easier that task. The object [line] can receive a message box composed by two numbers, the first one sets the number you are using and the second number is the time in milliseconds this first number is going to take to pass through the output of the [line] object. This object makes a linear ramp from the last number to the new number automaticaly using the time you want in the second number in the message. I highly recommend to see the help page.

The other objects are [pipe 1000] and [delay 1000], these objects can make a delay that is going to last the number of milliseconds you want (i put 1000, but you can change it). The first one can delay numbers, bangs, messages and many kinds of data, the second object only delays bangs and has a unique and rare behavior if you feed id with numbers.

Hope it heps. Even though your patch is insanely complicated i`d like a lot the efford you do building everithing from scratch, reminds me the time i was learning the basics on pd and a simple clock/calendar patch takes me more than a hundred boxes.

Cheers!

@alex_d Also have a look at the library called "mapping". many many abstractions for playing with values (it's in extended).

Cheers~

@juampablotoledo

OK nice! [line] does pretty much exactly what I want it do to! Funny to see my entire complicated sketch reduced to one box

The only thing that 'bothers' me about this is how a change between 0 and 1 or between 0 and .5 would take the same amount of time (if the same ms interval is used). So I made some additions to make sure it always goes the same speed no matter the difference between the value and the target value. Let me know what you think of this solution.

so_smooth_2.pd

[pipe] and [delay] seem interesting too! I'll keep them in mind for future use.

@EEight Ah thx thats awesome, great to know about such little abstractions to make our life easier. Really starting to love PD!

I had a lot of work lately and i couldn't write you back before. Great solution for the ms problem, i'm going to use it to smooth some arduino/pinguino noisy analog signals from sensors.

PD is love itself

I have what I think is a similar but slightly different issue. I want to use an analog potentiometer to control the argument of a {speedlim] object. I'm using a modified pduino patch driving ableton. I have the analog data coming in and going through the [int] object first so it is sending whole numbers but it is pretty jittery. I would also like to do a similar thing using an analog pot and map it to control tempo in Ableton but again it is too jittery. I've tried [smooth] but it didnt work. I don't quite understand how [line] works. I read the above post and also the help on the [line] object but still don't quite get it. Would this work for my problem or is there a better solution? I'm still learning so I appreciate all advise.
Thanks.
td

@saxdame If you're using pd-extended there's a neat set of abstractions called mapping, maybe [median_7] could help in your case. Or maybe [iir 5] or both!

Cheers~