Hi Hat Artifacts: Possible Improvement?

[mprinz]

Hi guys,

the cc04 data ist only important to check what samples to play, when th hats are played.
did we think about checking and using the cc04 data only while detecting a stroke?

-> HiHat hit --> check which cc04 value is generated --> play a certain sample.
-> no HiHat notes ---> play last sample until next hit or pedal closed, splash note generation.

defining something like a hihat sample decay time, meaning how long the sample schould be fixed after a hit on the hats.

This can be done in the either in the module or the software.
I dont know. Im wont be the first one with that thoughts.

Best regards

Manuel

[Rob]

I've been thinking about that idea, too. If I play the hats with them ever so slightly open, the position bounces slightly from the impact of the stick (which is exactly what happens when you play a real hihat). I'm not sure if that's something I need to try compensate for or if it's something that programs like Superior Drummer should be handling.

[vandroiy]

I've been thinking about that idea, too. If I play the hats with them ever so slightly open, the position bounces slightly from the impact of the stick (which is exactly what happens when you play a real hihat). I'm not sure if that's something I need to try compensate for or if it's something that programs like Superior Drummer should be handling.

They should be. Exactly right. The part of the chain that is creating the sound should be in charge of which sound to create based on the data you provide.
The way Addictive Drums 2 and BFD does, and nearly all the modules do.

Sadly, with Toontrack an Steven Slate Drums, probably the most popular of those 2 don`t. Especially with SD3 I don`t understand it. They got like 7000 options in their software.

That`s why I (we?) hoped you could fix it for them.

[Rob]

With my VH-10, one big difference is that the cymbal doesn't sway like a real hihat. On something like a VH-13, a lot of the energy from a hit goes into the swaying of the cymbal, so the pedal likely moves a lot less. On the VH-10, that energy goes right into the pedal, causing it to move more. There might be a good reason to implement some heuristics to try and filter that out or tame it. It's something I should probably look into it. If I made a feature like this, what would it be called?

[Sly]

Cymbal move correction ?
top hat move correction ?
Top hat swaying correction ?

[Sly]

But I think that it also could be a choose for the user between two operating modes :
1) normal mode (the current one)
2) "top hat move correction mode" where the CC value is sent only before each hit on the top hat.
Though I'm not sure how well the foot splash could perform in this mode.

[Rob]

Before I do any coding, I'm going to manually edit hihat data to see if there's a real issue in the first place.

[Rob]

Manually editing the hihat data to take out the bounce in hihat position made a significant improvement (cleaner sounds and less artifacts). Definitely warrants further experimentation.

[Rob]

So I played around with this for about an hour today and here's what I came up with. Might have a winner.

Current Implementation with Roland VH-10.

noSteadyHats.jpg (48.98 KiB) Viewed 1215 times

And after I hacked in some code to steady the hihat position.

steadyhats.jpg (43.03 KiB) Viewed 1215 times

[AKA Wayne]

Fantastic. Can’t wait to try it out.