Recording with a Pickup - Demo

[Doug Young]

Warning - LONG post.

We've had a few discussions about recording with a pickup lately, and in one thread I promised a demo of some approaches. Let me say right upfront, that it's a losing battle for the most part - even a bad mic beats the best pickup easily, but I do realize there are people who can't use mics due to noise, or other issues. I wrote an article for AG many years ago about some techniques for at least getting a pickup recording to sound passable, but I've learned a few things since then, and also that article had no sound clips, and is long out of print. So I thought I'd post some examples. These are far from definitive - just a few things I threw together. Your results may vary depending on pickup, what techniques you use, what kind of sound you want, etc. If you have other ideas, maybe add them to this thread to share with others.

If you just want to hear the best I've been able to do (and I make no claim it's actualy "good" :-) )- go directly to the last link - that's my best stab. Also, maybe a way to decide if listening to all the rest of these examples is worth your time! Otherwise, check out these incremental steps and alternate techniques.

All these samples (except the 1st two) are from a single recording, using a Ryan Mission Grand Concert with a &K Pure Western Mini and a DPA 4061 internal mic. I start with just the K&K, then add the mic. The tracks were recorded thru a Grace Designs Felix preamp, directly to an Apogee Ensemble interface to Logic on a Mac.

First, here's an initial recording of the K&K thru the Felix, with all EQ set flat.



And here's the internal mic, thru the Felix, all controls set flat.



For this exercise, I also recorded the guitar with real mics, and I compared the raw un-EQd sound of the pickups, and then tweaked the EQ on the Felix to try to get the K&K to sound more like the mics. I also tweaked the mic track just by gut feel, to be far less low end. We'll use that in a bit.

here's the re-recorded K&K track with EQ added on Felix



and here's the EQd mic track



OK, these sound fairly lousy! What can we do?

My first stab is just to add some reverb to the K&K track. Reverb can add a bit of space and maybe compensate a bit for the directness of the pickup. No mic in this mix:




Maybe the mic will help a bit? Here, I added in the mic. The mix is mono, but with some reverb added:



Going Stereo

Pickups usually sound rather one-dimensional on a recording. Where it makes sense, adding some space with some stereo techniques may help. The rest of these examples explore "fake stereo" techniques.

First off, let's just pan the K&K and mic a bit to each side:



Another technique is to make a copy of a track, pan left and right, and EQ each side differently. Logic has a plugin that will do this all automatically, creating a series of cuts and boost on one side, and doing the opposite on the other. You could do this manually with a graphic EQ, but here's how the Logic plugin sounds on just the K&K pickup. No mic in this mix:



So what if we combine this idea with the internal mic? Here's the stereoizer effect applied to the K&K, with the mic mixed in the middle:



or maybe the other way around? Here's the stereo-izer applied to the mic, with the K&K in the middle:



Another approach is to use a slight delay on one side. Here's the K&K+mic, mixed to mono, but with a copy of the same track delayed by 10 ms on one channel:



Another way to create a fake stereo effect is to apply a pitch shift to one channel. So again, K&K+mic, mono mix, but with a copy on one channel with a slight pitch shift applied:



Someone mentioned a UAD plugin called Woodworks, that appears to be sort of a stereo Aura. I checked out the demo and applied it to the K&K signal alone - no mic. There are lots of different settings and mix options on this plugin, and frankly I hated most of them. Maybe it works better on a different pickup (I also tried it with a Barbera UST - no better luck). Everything sounds distant, chorusey, and pretty fake. But with some playing, I got this track:



Saving my favorite stereoizing technique for last - this runs the K&K and mic thru an MS decoder, treating the K&K as if it was the "mid" and the "mic" as if it was the side. The result is a mix with the K&K centered in the mix, and the mic providing air and stereo spread. Here's a raw mix using a Voxengo MS decoder:



And finally, here's an all-out stab at my best mix. This uses the MS decoder, some compression, more EQ, reverb, a UAD tape simulator and an exciter plugin (also from UAD).



Notice that we still have a big dose of the unrealistic pickup sound, to me, there's a sort of hardness to the sound that I haven't found a way to get rid of, but it's perhaps acceptable for someone in a harsh home environment, or for demos, etc.

So that's a grab-bag of ideas and techniques. What else have you tried that works for you?

[runamuck]

Thanks Doug. What a service.

Jim

[ljguitar]

Hi Doug…

What a crazily infomative post…thanks for the work invested.

[Doug Young]

Originally Posted by ljguitar

Hi Doug…

What a crazily infomative post…thanks for the work invested.

It was less effort to do than it looks like, just a bunch of mixes of 1 short recording. I didn't realize it was going to look so overwhelming until I started to put together the post!

[RedJoker]

Awesome! Thanks. I learned a ton.

[scripsit]

Very informative, thanks Doug. The direct comparisons are really useful.

Perhaps an indication of the different possibilities and parameters when playing live (amplified), too.

Kym

[Doug Young]

I'd actually love to be able to use the MS trick live - I've only been able to once, and didn't get a great feel for how well it worked. But it should, in theory be pretty cool. If you normally pan a dual source, like, say a K&K+mic in stereo thru a PA, half the audience will hear the K&K and half just the mic. But with MS, you'd get like 1/3 that hear the K&K+mic in a mono blend, another 1/3 also hear in mono, with just the phase of the mic inverted - should sound no different, really, and the 1/3 or so in the middle would hear very nice stereo.

I suppose the delay and pitch could work somewhat the same, tho the MS seems safest, since in the worst case, it just collapses to mono and you just hear the K&K.

unfortunately, almost no one runs live sound in stereo.

[Vancebo]

Wow! All I can say is that was an entire workshop wrapped up in one little thread. I am glad that my favorite sounding clip was your last one. Even through my iPad speakers I was hearing subtle differences in your clips.

When I was running the same DPA mic you are. I never got as good of a raw sound as you did. I was surprised.

[Mobilemike]

Wow Doug, thanks so much for all this detailed info and examples! Amazing!

I agree completely with you when you wrote that no matter what you do to the pickup, there's an inherent "hardness" to the sound of a pickup that is a bit unpleasant. Reverb helps a bit with this, but one other thing that I've found helps is "worldizing" the sound a bit: playing your pickup recordings through a decent set of speakers and recording them in the room with a pair of mics. Then you mix a bit of this worldized sound in underneath the direct signal.

This still works if your room is too noisy or untreated to record with mics in the first place - because you are only mixing a bit of this signal in with the direct sound, you can get away with a lot more "bad" sound.

-Mike

[The Colonel]

Thanks Doug. You proved that every SoundCloud has a silver lining. <groan>

[pieterh]

Originally Posted by Doug Young

I'd actually love to be able to use the MS trick live - I've only been able to once, and didn't get a great feel for how well it worked. But it should, in theory be pretty cool. If you normally pan a dual source, like, say a K&K+mic in stereo thru a PA, half the audience will hear the K&K and half just the mic. But with MS, you'd get like 1/3 that hear the K&K+mic in a mono blend, another 1/3 also hear in mono, with just the phase of the mic inverted - should sound no different, really, and the 1/3 or so in the middle would hear very nice stereo.

I suppose the delay and pitch could work somewhat the same, tho the MS seems safest, since in the worst case, it just collapses to mono and you just hear the K&K.

unfortunately, almost no one runs live sound in stereo.

I do though it depends on the venue.

You are way more of an expert than I am but surely the basis of MS recording is the side signals are created by mic with a figure 8 pattern, duplicated on two tracks where one track is polarity inverted. These two channels/tracks are then panned hard left-right together with the mono channel and the three together give a stereo reproduction that is usually very rich and spacious (depending on the distance from source).

As I said, I'm not the expert here but those times I have done MS (Mono/side) recordings they have turned out pretty well, using two u87s, one with cardioid pattern, the other (directly underneath) with figure-8 pattern where the mic is turned 90 degrees relative to the first.

[Doug Young]

Originally Posted by pieterh

You are way more of an expert than I am but surely the basis of MS recording is the side signals are created by mic with a figure 8 pattern, duplicated on two tracks where one track is polarity inverted. These two channels/tracks are then panned hard left-right together with the mono channel and the three together give a stereo reproduction that is usually very rich and spacious (depending on the distance from source).

Absolutely, that is MS recording - I love to use it.

What I'm doing here is leveraging (abusing) the MS decoding process and applying it to 2 signals that weren't produced by a figure-8 setup. All it does is exactly what you describe. I have a pickup that I'll pretend is the mid mic (M), and a mic that I'll pretend is the side (S). So I run it into the MS decoder, and what comes out is:

Right = M+S
Left = M-S

if applied to a live sound situation, the audience on one side would hear a mono blend of the two (M+S), or in other words the same sound they'd get if the PA was mono and I used a mono mix of the two sources. The audience on the other side of the room would also get a mono blend, but with the mic signal phase inversed (M-S). Since an internal mic is all over the place phase-wise, it shouldn't matter much at all. The people in the center would get a stereo spread, like what you hear in the recording. And if the sound guy hits the mono button, the sides cancel and everyone just hears M - the K&K, which is not a disaster. Is it "MS"? No, it's leveraging the principles of MS, or at least the implementation, to manufacture a fake stereo image.

[pieterh]

Ok I read you

In my inexperience I hadn't quite realised what you were talking about but now I get it.

Yeah I like genuine MS too. I recorded a music school concert a few weeks ago with most instruments coming off the desk (Yamaha M7CL) either direct or sharing a channel (the soloists never played at the same time so at mix down I extracted each instrument and gave them their own tracks). I then had two more inputs with the u87s, duplicating the figure 8 and inverting it on mix down. Of course the u87s were at FOH which is about 20 metres from the stage so the first thing I had to do was to delay the MS mics to time-align them with the direct tracks. A little volume envelope here and there where needed and it turned out really well!

[pieterh]

" And if the sound guy hits the mono button, the sides cancel and everyone just hears M - the K&K, which is not a disaster."

That's the advantage of MS and X-Y stereo techniques - they're always mono compatible.