Live acoustic band recording

[TravGRO]

Hey there. Short time reader, first time poster.
I record my bluegrass bands shows and wonder what I should be doing different.
Here's my setup.
12 channel mixer, everyone direct plug with D/I box except for bass, an extra SM-57 for the banjo for reinforcement (make it sound like a banjo, not a 5 string electric guitar. the pickup sucks) and a big old radio vocal mic placed directly in front of the bass cab. From there I do one of two things. Take the Tape Out RCA output and run that into the Line In on my laptops 16 bit soundcard or I take a mono Sub Out and run that into the same input on the laptop recording with Audacity.

Everything to there is pretty good unless I overdrive the signal. Then it's the wonderful world of digital distortion.

Being unproccessed, the recording sounds very thin. I know I need to work with the EQ and add some compression. It's the compression that I don't know how to work with. Once I export the file as a .wav I then open it in Cool Edit Pro 2.0. Great program but not very helpful if you don't know how to reset the presets. I used to use a light vocal compressor but as my ears learn more I can hear that it's too much compression. That leads me to my question.

What do I want to do to add just a little compression to my stereo mix? Just to bring up the low signals and even out the spikes. I just don't want it to sound muddy and over proccessed.

If you want a reference point on what I've done to overprosses a recording check out http://www.archive.org/audio/etree-details-db.php?id=29884
I can hear sounds drop off when when everything is quiet then someone comes in with a loud note.

Anyone work with Cool Edit Pro 2.0 or Audacity and know what the different compression settings mean and do?

Any help and advice would be greatly appreciated. Thanks.

 

[TravGRO]

btw, I just picked up the Izotope plugin for Cool Edit. It looks great but I have no idea how to use it correctly.

 

[scrubs]

Some thoughts:

To me, bluegrass is all about the feel of a group of people playing together in a room (often with a crowd). With that in mind, I would try just a pair of mics in the room (or a stereo mic) to pick up more vibe. You could still track your other stuff direct (although close miking would be better) if you need to beef something up, but I would probably go mostly with the room mics. A pair of small condensers in XY, or even a spaced pair would be good, I imagine.

As for compression, it sounds like you are applying too much. Ideally, this style of music is not compressed much at all. If you need it, it should be applied to individual tracks rather than the whole mix. That being said, you would need a soundcard with more inputs, so you can record each track individually and mix them together later (live sub-mixing is always a challenge).

There are good articles about compression to be found at:
http://www.theprojectstudiohandbook.com/directory.htm

 

[AGCurry]

Amplifiers? For bluegrass? Bill Monroe is spinning in his grave.

I agree with scrubs. I've seen quite a bit of true live bluegrass. True bluegrass does not involve pickups and amplifiers. The PA does all the amplification if it is needed.

Use spot mics for instruments that need it, plus a room mic.

 

[mixsit]

...I can hear sounds drop off when when everything is quiet then someone comes in with a loud note.

The pumping from the compression/limiting is immediately noticeable on these songs. I caught 'Riptide's intro, it almost sounds like it has mild gating.

..Being unproccessed, the recording sounds very thin.
..I just don't want it to sound muddy and over proccessed.

Let's talk about thin'. Right now you have your 'thick thing going. That nails the 'muddy and over processed' so we can skip that one.

..Just to bring up the low signals and even out the spikes.

This is almost invariably two separate dynamic functions. A fast attack/release, even at mild ratios, goes after the spikes. You job is to choose how much (attack time) transient is let through, and how far you are willing to chop them off (threshold). This is going to be job one for acoustic music.
A slow very low ratio can help from the other direction. You're going for a little gentle gain riding of the volume in this method.
Know also that what ever might be 'too hot' in the mix 'drives' the compression. Balanced playing and mixing = smoother and more natural sound.

 

[TravGRO]

I suppose I should have mentioned that we are not your typical old timey bluegrass band. I come from a hard rock/metal background. Our mando player used to front a punk band. The second guitarist is a hand drummer. Bassist also plays dirty mean 'smack your momma' bass in my funk band. Think more rock band with grass instruments. Over the past few months we've gotten this very fast driving in your face sound going. Those are a couple reasons I don't use just a couple mics in the room. I'd love to do a sbd/aud matrix but budget has yet to be my friend.
About multitracking the recording. I'm recording these shows for our own archives and every once in a while I want to spruce one up to give away to our fans and upload to archive.org. No need to spend a few dozen hours on something that won't be officially released. We have our guy with his ADAT setup to do multi track recordings of any shows we think we might want to sell in the future.
yadda yadda. I just woke up, looking forward to another midnight shift.

Would I be better off using only some EQ work to work the sound? With enough time I can work a graphic EQ but I'm just now learning paramatic EQ and other tools. Basicly I'm a rookie who can run a sound board while playing a show and who's teaching himself the ins and outs of live recording. The equipment I use is a definate factor in the sound. I completely understand that but I know I should be able to get good sounding final recordings from a stereo recording directly from the soundboard.

btw, thanks for the reply mixsit. You brought up a couple of things for me to do some research on.

 

[TravGRO]

This is almost invariably two separate dynamic functions. A fast attack/release, even at mild ratios, goes after the spikes. You job is to choose how much (attack time) transient is let through, and how far you are willing to chop them off (threshold). This is going to be job one for acoustic music.
A slow very low ratio can help from the other direction. You're going for a little gentle gain riding of the volume in this method.
Know also that what ever might be 'too hot' in the mix 'drives' the compression. Balanced playing and mixing = smoother and more natural sound.

I kinda sorta understand what you're saying here. The current settings I used are...

compress 5.45:1 above -24 dB
expand 1.26:1 below -24 dB

output compensation 19.6 dB

Gain Prossesor
attack: 1 ms
release: 250 ms

Level Detector
attack: .5 ms
release: 300 ms
set on RMS

Lookahead time: 3 ms

Just a preset that sounded the best at the time. Now I heavily dislike the sound.

Are you saying that I should increase the time of both attack and release settings? I'd just like to make the overall sound slightly fuller.

If I've got it all wrong then I'm very confused again.

 

[Harvey Gerst]

Okay, simple solution:

Lose all the compressor crap. You're not using it right, and it's gonna take a long time to explain it right, especially since we can't hear what you're doing.

Run everything into the ADAT flat (no EQ), with the peaks hitting NO HIGHER that -12dBFS.

Worry about making it all sound good later, when you go to mix down.

 

[TravGRO]

you misunderstand. The ADAT will be used in the future. I have to pay for the guy and his equipment to come to the show to record. What I'm currently talking about is the recording I do myself via Tape Out > Laptop Soundcard. Just the stereo output, not multitracked. I'm sorry if I worded things in such a way to create confusion.

 

[treymonfauntre]

if you just want to bring the levels up you want a limiter. if your software doesn't have one you can try a compresser with a fast attack and release and only apply a bit, not 24db of reduction/gain

 

[Harvey Gerst]

Sorry about that, but the main thrust of my advice still holds:

Don't use compression, limiting, or expansion during recording. Try to set the eq as neutral as possible, and don't let the peaks exceed -12dBFS.

If you feel that you MUST use compression, set the controls to no lower than -5dB threshhold, no more than 4:1 ratio, no less that 100ms attack, and no less than 500ms release. At least, those numbers won't do too much damage.

 

[mixsit]

I kinda sorta understand what you're saying here. The current settings I used are...

compress 5.45:1 above -24 dB
expand 1.26:1 below -24 dB

output compensation 19.6 dB

Gain Prossesor
attack: 1 ms
release: 250 ms

Level Detector
attack: .5 ms
release: 300 ms
set on RMS

Lookahead time: 3 ms

Just a preset that sounded the best at the time. Now I heavily dislike the sound.

Are you saying that I should increase the time of both attack and release settings? I'd just like to make the overall sound slightly fuller.

If I've got it all wrong then I'm very confused again.

I would say in general, and from the songs, you're in deep dodo (nothing personal there ); Very fast attacks, fairly slow release, and deep threshold, on transient rich sources; pump city.
Back to the two general ideas; Follow Harvey's guidelines for a bit of slow gain control.
As a second pass you might get away with a small amount off the top in the fast attack/fast release approach. Last, anything like .5-1ms will kill all life very quickly. 10-20 ms is still pretty 'fast'. You're not clipping snare hits here.

 

[TravGRO]

thank you guys very much. I did find the source .flac files during practic at another bandmate's house tonight. All is not completely lost with the show I posted for you. I can easily replace that if/when I churn out a better sounding copy. Since I can't find a Post Production for Supreme Idiots guide on the net I guess experimentation is the key. Thanks for helping me understand a littke bit more.

One more note, any altering of the sound is done after I record flat from the soundboard. It's kinda tough to moniter levels and adjust any settings while I'm trying to play a Cripple Creek too fast cuz our mando player got a bug up his butt to play it really fast. Just to clarify.