Hi. I greatly appreciate all the variety of Hihat samples I’ve found to be available in my SD3 and EZ3 collection.
Having used sample instruments for awhile, I’m familiar with something that added realism on re-triggers of resonating bodies, which I don’t think is particularly available as a feature to turn on for playback of hi hats in SD3 as it was in BFD3.
IMHO would be fairly quick and easy for SD3 to implement – and since it would be optional what could it hurt : D ?
I’m probably the first person to post about it under hi hat “swell” because I did a search and read every post on that topic – and no one seems to have said this, right? Probably people think of the BFD feature and associate it with cymbals rather than hi hats, and it’s frustrating for me as a fan of funky drumming and breakbeats that people don’t realize how affective this feature can be on acoustic drum sample playback.
In the old days there was a rompler called Sample Cell 2, that ran as an internal card in a PC. The feature in question (a decade prior to BFD3) was implemented for instruments like hi hats, pianos, xylophones and others that continue to resonate or vibrate. Essentially it was a slider that would move the sample retrigger point later in the sample – activated only on a RE-trigger. If you played the instrument freshly it would not behave in this way – but if you RE-triggered the sample while it was playing, this would overlay a new playback instance (as the old instances faded out). So if you were triggering your hi hat and you moved the slider 20 percent in, the sample would bypass that recorded fresh transient sound (and you may want that because you’re not, essentially, affecting a fresh trigger of your virtual instrument). when I play hi hats in SD3 they do sound pretty realistic at times – and the character is there and the mic and room stuff is wonderful. But that one aspect of the hit, while being of different velocity (no complaint there) seems repetitive when it would not be in a real recorded session. If you hit your hi hat again a quarter second after hitting it initially, it wouldn’t trigger the same kind of transient. The energy would not rise from nothing. It would essentially suppress some of the energy of the resonating body, while adding some transient back in because of the impact of the stick (or the closing by pedal. Repeated re-triggers, in the simplest way of mimicking this, would come a bit after the very beginning of the sample. Beyond that, you could apply some compression relative to the amount of resonance that the hit was up against. If the hit were nearly instantaneous, the compression effect would be greatest. If the hit were at a point when the energy was nearly subsided, the compression effect would be negligible.
Fxpansion implemented something like this in BFD3 and called it cymbal swell, affecting cymbals and hihats.
I know it’s hard to believe, but it even affects closed hi hats.
There’s a VU Meter to see the response it creates, and on closed hi hats it’s essentially doing what sample cell did (at least in end result, which is reducing machine gun by reducing the transient on re-triggers by x amount within sample length threshold – it may or may not be changing the start of the sample, but rather downward compressing subsequent transients relative to how recent the previous trigger was. I’m no expert – just let me know if you think it’s doing something else). There are a number of algos a user can select (not that distinguishable to be honest), so I’m not telling you everything in full detail (selecting each algo probably changes different dynamics response parameters like hold and release- possibly more). It may have a slight pitfall in terms of phasing, but it’s hardly that noticeable IMHO.
The sound does help cymbal crashes, but in a different way – it is reducing the transient – but a crash has such a long release envelope that’s so loud, that the relative difference isn’t as great as the closed hi hats (and the less open ones than fully open).
So cymbal swell implementation has its uses, but it’s next to impossible to tell SD3 cymbals from the real thing unless you do a cymbal roll (near instantaneous retriggering). The crash cymbal is quite loud, and hitting it a 2nd time does make it react with a loud new transient that’s very similar. Similarly the ride cymbal is very sustaining and the new transient is somewhat drowned out by this so if it’s a recording of a fresh stroke being used in succession it doesn’t impact realism too much.
On the other hand, the closed hi hat is software and it’s transient is louder in relation to it’s trailing off sound. So it’s “swell” behavior the way BFD3 has it, IMHO can be super important for indistinguishable-from-reality 16th note hi hat playing or fast 8th note playing (where the particular hi hat is still resonating when it’s re-triggered). To my ears, for these cases (closed and quarter open) where there are the sharpest transients it’s much more realistic than just triggering the sample.
I think this could help E-drummers and non-E Drummers. Maybe the E-drummers haven’t noticed the issue I’m describing because they can vary the open amount on their hi hats in a realistic way, and as I tried to explain, triggering a half open hat gives a more gradual envelope, where the quick drop off and sudden transient are the chief offenders. (Because they’re unrealistic behavior on a physical system that continues to resonate – your initial stroke on a hat “breaks the ice” so to speak, whereas subsequent strokes add resonance and certainly add some transient, but the transient that it adds is slightly less sharp (obviously this varies depending on the hi hat one is talking about.)
Anyway, just my opinion. If someone wants a video of the hi hat behavior in BFD3 as witnessed on the dedicated “swell” VU Meter, I can attach one – but it won’t do a lot of good unless you watch the behavior in relation to the hi hat stroke audio (and optimally the waveform being recorded). And unfortunately I can only give the video – my audio setup is conflicting and the videos are coming out silent. I was always hoping SD2-3 had or would add this feature. I’m choosing between BFD3 expansions and SDX’s and this is the primary reason I would consider it (along with tom resonance which is also cool) – I figured if no one mentioned it I may as well be the one.
