1.18.02.00 Channel filter not properly reset

[a-lin]

Full Version:
OpenMPT v1.18.02.00

Has the bug occured in previous versions? If yes, please specify version(s): (This option is optional, and doesn't need research. But in case you know...)


Description of the bug:
The reso and cutoff values of an instrument seems to be determined by the instrument played right before.

How often does it happen?:
everytime

How to reproduce (step by step description):
1) load the following song
https://sites.google.com/site/ngeisweirepo2/Home/bug_reso.it?attredirects=0&d=1
2) compare how the instrument 2 plays on the 2 channels, on channel 1 (preceded by an instrument with filter enabled) it plays with the wrong reso and cutoff, on channel 2 it plays correctly
3) mute channel 2, vary the reso and cutoff of instrument 1 and see how it affects instrument 2

[LPChip]

I'm not sure if this actually is a bug or how it supposes to work.

It always has worked like this.

If my memory serves me right, the reason is that when you set this setting in the instrument tab, it changes the channel's setting. It is not a temporary thingy, cus that would be glitchy.

You can use instrument settings to cancel out the settings. Do this for your other instruments, and the problem should be solved. You can also dedicate a channel for a specific filter setting.

[Saga Musix]

It's not a bug - But a common issue with a simple solution:
Enable the Cutoff checkbox for all instruments that share the same channel, and move the cutoff slider to the far right until it says "Off".
Another solution (but not compatible with Impulse Tracker, thus just recommended for MPTM modules) is to use the "local" filter mode, which will reset the cutoff and reso values on every note on a channel (and will thus probably generate unwanted results when using Zxx). This can be enabled by putting a S9D effect on this channel. The lowpass filter can also be reset manually by putting a Z7F effect between to notes, but that just sounds right if no note is currently playing on that channel, of course.

It is not a temporary thingy, cus that would be glitchy

Why would it be "glitchy"? That's exactly the difference between S9C (global filters) and S9D (local filters), and both modes make sense.

[a-lin]

I found very weird that the parameters of an instrument could interfere with another instrument, but if it's not a bug then fine, there are plenty of work around, as you pointed out.

[Saga Musix]

I found very weird that the parameters of an instrument could interfere with another instrument, but if it's not a bug then fine, there are plenty of work around, as you pointed out.

Why would that be weird? Does it also seem weird to you that an instrument/sample that does not have the "set pan" checkbox set inherits the panning of the previously played instrument/sample? It's the same logic. Panning is a channel property which can be set by instruments, and so is cutoff.