Audacity Forum

Hi all

I'm working on a recording of snare drum where, when the drummer hit it with sufficient force, created a noticable ring at 290Hz. When he played quieter, the ring wasn't there.

Is there a way of controlling the amount of EQ applied dependant on level? There's an obvious threashold level the drummer reached that the ring can be heard at.

I could just use a parametric with a reasonably tight "Q" to suck the offending frequency out, but it's not there all the time, especially when he's using brushes or playing quieter.

Are there any EQ plugins with a threashold level for Audacity?

Thanks
Barry

barry.beattie wrote:Are there any EQ plugins with a threashold level for Audacity?

Paul-L's de-esser plugin is capable of that, (although it wasn't designed for the purpose).
It's a multi-band limiter: you could set a narrow band, say 250-350Hz , then lower the threshold till it just took out the loud ring , (assuming unwanted ring is only 290Hz : no harmonics ).
If it's just a few rings, it would be quicker to take them out manually with Audacity's spectral editor tool ...


Demo of Audacity's spectral editor tool.

Are you thinking:
"The ringing occurs when the drum is hit hard, so the solution is to cut the ringing frequency when the amplitude is high."


If so, that probably won't work.
The ringing may well occur as a consequence of hitting the drum hard, but when the amplitude is high, what you will hear is "BANG". Any ringing will only be noticeable after the initial transient - when the level has dropped down. "BANGggggggg"

barry.beattie wrote:I could just use a parametric with a reasonably tight "Q" to suck the offending frequency out, but it's not there all the time, especially when he's using brushes or playing quieter.

So why not apply EQ one type of EQ to loud songs, and a different EQ to quiet songs?

There are ways to apply EQ only to the "ringing" part of a drum hit, by using side-chained gating, but as Audacity does not currently support side-chain effect, it is complex and tricky to do that in Audacity. Briefly, the approach would be:
1) Duplicate the track (track 2)
2) Gate the duplicate track (track 2). The attack time of the gate is quite critical as you will need the gated and non-gated tracks to cross-fade
3) Duplicate the gated track (track 3)
4) Invert track 2
5) Mix tracks 1 and 2 (this gives the "tails" of the beats)
6) Eq the mixed track to reduce the ringing
7) Mix the two tracks (the "mixed" track with the track created in step 3)

steve wrote: The ringing may well occur as a consequence of hitting the drum hard, but when the amplitude is high, what you will hear is "BANG". Any ringing will only be noticeable after the initial transient - when the level has dropped down. "BANGggggggg"

Sort-of. Each snare hit lasts 190ms, from initial transient to the (meaningful) end of the snare rattle. It's within that 190ms that the 290Hz ring can be heard... clear as a bell...

I'm also battling the spill from the bass drum (especially low frequencies) bleeding into the snare mic after that 190ms. There's also high-hat spill there too, and stage noise but a gate set to open and close at threashold slighly louder than the bass drum will also get rid of the high-hats**. That's 90% of the time and the last 10% is managable to manually edit out the occasional rack tom.

It would be nice if the ring was consistent in the loud songs, but within the same song, within the same verse or chorus, it's hitting an acoustic threashold and ringing. There's 90 minutes of music to go through, it's a bridge too far to finesse the solution beat by beat.

The ring (usually) happens when the force has overcome the internal stiffness of the shell. It takes so much and then gives and rings like a bell. Careful tensioning of the skins can provide stiffness to overcome this (or change the note) but this recording has been made and there's no going back.

Yes, 290Hz is the annoying frequency - close enough for a mid-D note. There are no noticiable harmonices (580Hz, etc) that can be heard above the rest of the drum noise (the much sought-after snare rattle that's difficult to not truncate with gates).

** why get rid of the bass drum sound in the snare mic? What the snare mic picked up sounds disgusting and is detrimental when adding it to the now-tidy bass drum channel. I've also found the mic under the snare is unusable - it's just all rattle, set off by the bass amp only a metre away - so all my snare sound will be from this mic and whatever was picked up by the overheads (also full of the stage noise, vocals, bass, guitar, crowd noise). At least I was able to clean up the tom mics with 98% only toms in those channels.

This looks interesting

http://www.tokyodawn.net/tdr-nova/

So at least I know the concept does exist and it's not just me being pedantic.

I know it's too late, but what's the microphone and can you hear the damage in real life?

Drums are killer hard to mic. You may be listening to the element inside the microphone overloading. The SM57 worked so well with drums because it's almost completely impossible to overload it. Was this a condenser? I wonder if that's what it sounds like when the diaphragm hits.

Koz

well, the overhead mics can hear the ring, although not anywhere near as obvious as the close mic'd Shure Beta 56A.

That's why I'm trying to do as much as I can on indivudual channels - there's spill all over the place that I'm trying wrest some control over. A bit of crap here in this channel, plus a bit more in another channel... and it ends up sounding like a big bucket of poo...

I can see where the same damaged hit appearing on multiple microphones can be a challenge, but that's not quite what I wanted. If a human was standing there, could they hear it?

Koz

I've got no idea. I can't remember - I was too busy singing at the time.

Regardless, it's not the microphone - the Shure Beta 56A's are designed for drums, the main thing is what was captured on the recording, that's the task at hand.

Yes, but...

If the drum really sounds like that, then the microphone is working perfectly correctly.

Koz