Mixing Irish Music in the Studio

Hello everyone!

I have a pretty good feel for Irish music, the frequency ranges of the instruments (which are essentially the same), and compression techniques. I was wondering if anyone who has recorded or mixed down ITM tracks had any thoughts on how to go about creating some separation between the instruments that occupy the same frequency ranges. Fiddle, flute, and pipes sound great in unison, but as you know it can be hard to really narrow in on one instrument when they all play a tune together.

What are your thoughts, techniques, and hiccups when mixing ITM? An obvious solution I am looking at is a LCR approach (pan instruments left, right and center in a meaningful amount). However, I’m wondering if this would cause the main attraction, which is always the melody, to become lost in mono. For instance, playing the tracks back on a phone or laptop speakers (while not truly mono) can make the stereo image become lost. Has anyone else run into this issue?

Another solution is to use a small amount of reverb to push things back in the mix, but this can overly smooth out the sound even in small amounts making the music lose its distinctive rhythmical transients, which is counter-productive for the genre. Of course this can be mitigated with a small amount of pre-delay on the reverb. Alternatively, a small amount of delay with a low feedback and quick reflection like a haas effect to push things perceptively back in the mix. The latter effect becomes a problem in the phone/laptop scenario mentioned above when that depth can no longer be heard.

As instrumentalists, we spend a lot of time perfecting our tone and technique. The classic unison sound of traditional instruments together is a distinguishing factor of Irish music, but shouldn’t the individual performances also be discernible?

- Matt

I just finished mixing my second album in a studio in San Francisco a couple weeks ago. The general approach we take is like this:

1. EQ each instrument individually (and occasionally add some compression, where needed. The only thing on this album that we needed to compress was a concertina, to smooth it out a bit, and then we put some “dynamic compression” on things like guitar string squeaks and concertina button clicks, etc)

2. Devise a stage plot. We use a couple of different techniques for this. Depending on the instrumentation, we will spread the melody instruments out left and right. If there are only two, we’ll usually put them about 30-40 degrees off center. If there are more than two, then we’ll spread them fairly evenly in the mix. We deal with accompaniment differently sometimes. If there are only two melody instruments, we’ll often take (or make) a stereo version of the accompaniment and spread it all the way left and right, to make a “bed” for the melody to lie in. If there are two accompanists (like guitar and bouzouki), we’ll often spread them pretty wide as well. If there is only one melody player and one accompaniment, we’ll just spread them like two melody instruments. But stereo separation is the easiest way to differentiate instruments in the same range. If things get really muddied up in the same frequency range, you can EQ different instruments in the same range a bit differently to make them stand out, but that can make them sound worse if you’re not careful.

3. Determine the mix volumes. In general, we shoot for things to all peak around the same range, but there are times when you want something a bit softer, like harmonizing lines, etc.

4. Determine what type of room this group would be playing in, and build reverb based on that. It’s not uncommon for us to use slightly different reverb on different instruments to help differentiate them, but we are (almost) always shooting for something that sounds real.

The thing that frustrates me the most is that it’s pretty much impossible to make it sound great in all situations. Something might sound great in headphones and on high end audio equipment, but really lose something when played on a bluetooth speaker or car stereo, for instance. So I highly recommend that, while you’re in the mixing process, get everything sounding exactly the way you want on the studio monitors. But then take that audio and play it everywhere that you normally listen to music. Your car, your home stereo, your phone, your ear buds, your computer… And then if you find things that you don’t like, work to fix them before you call it a wrap!

Best of luck!

The first thing I do is to do all the mixing with everything panned dead centre, only panning them apart at the very end of the process. If you pan them early on you give yourself the illusion of separation, but to anyone listening in mono or sub-optimal conditions they will lose that separation.

When it comes to EQ, I use a technique called mirror EQ - up in the overtones end of the spectrum (between about 2-8Khz) each instrument is allocated a frequency band where it is boosted by a few dB and the other instruments are cut by a similar amount. The actual frequencies have to be chosen to suit the specific instrument as there is an effect on their tonality, but by having an area of the spectrum where each instrument is dominant the brain is able to focus on it. You’ll find you can clearly identify each instrument, even when they are playing in unison and panned together.

When it comes to panning, instead of just turning the knob, if you duplicate each track so that you have three identical copies, leave one dead centre and pan the other two hard left and hard right. You can now pan by adjusting the level of the two outer tracks, and if you add a short delay (between 10-30mS) to the quieter side you’ll get much tighter stereo positioning. The central track copy reduces phasing and makes things work for mono playback. It’s not quite the same as LRC conversion,. as you are not inverting one of the channels). Read up on the Haas effect for more details.

Compression is tricky with this music and has to be kept very subtle - I usually use additive compression, where you let the whole dry signal through, but add a compressed signal with a low threshold and fairly steep roll-off, so that a lot is added to the quiet bits and little or nothing to the loud peaks, that sounds more natural than cutting the peaks, and is much less likely to pump.

Reverb also has to be treated with respect. I would never use it on an individual track, in fact when the recording has been made in a lively room I often use a de-verb plugin to reduce the natural reverb. Only add reverb at the final mastering stage, so that everything gets the same reverb, if individual tracks are treated differently it sounds very unnatural.

Thanks Mark that was some of the most useful information I have heard on recording/mixing even if some of it is a little beyond my capabilities

Ditto Donough. Some of that is probably fiddly (if even possible) to do with basic studio software like Audacity, but I will certainly try out some of those ideas next time I am mixing a recording.

Best mixing advice I ever had was “use a reference track”
If you can find an album you like the sound of, with similar instrumentation, import the MP3 into your project and regularly compare your mix to it.

Obviously that’s a lot easier to do with rock music or singer-songwriter type stuff, as the instrumentation is fairly standard: might be harder to find impressive instances of well recorded music with the exact same instruments in ITM…

Regarding getting a mix that sounds good everywhere, it might be worth considering paying for someone to master it. I mean, ‘making it sound good in a range of listening environments’ is generally considered the remit of mastering rather than mixing. Online mastering is offered by so many people, so cheaply, that if you don’t have an ideal listening situation at home (i.e. an untreated room) then it might be worth spending a little money to get your album mastered professionally.

…also, personally I’m not too bothered about that whole ‘does my stereo mix translate well to mono’ quandary. Some of the best sounding albums I’ve ever heard were captured live in the studio using a stereo pair of mics. Equally many of my favourite albums used quite extreme panning that doesn’t translate to mono. I personally listen to most music on headphones or directly in front of a computer or in the car or off the stereo in the sitting room. Que sera sera.

Interesting discussion, I’ve no experience with music production, but I notice that there seems to be a different approach in the older albums. I like the balance of the instruments in the 70s era Chieftains albums for instance, it just sounds more authentic to me. A lot of more recent recordings I have heard ( 2000s onwards) seem to have a completely different feel, more like popular music recordings, with thicker reverb, higher compression, and also a lot more bass. This later sound seems to make the music sound more intense, and I personally have mixed feelings about it.

I agree with Matt Milton regarding ‘it might be worth considering paying for someone to master it.’ My wife and I recorded some O’Carolan tunes on harp and mandolin, and then I paid the professionally trained son of one of my coworkers to master the tracks. We think he did a great job, and the price was reasonable.

I used good equipment to record the tunes, and I think the room has good acoustics, but I could not achieve the quality of the final sound that the pro got. As part of the transaction, I paid him a little additional $ to provide a description of exactly what he did on each track - this was money well spent, as I learned a lot, and my own recordings are now sounding better as a result of his advice.

If anyone wants to hear the final result of the professional mastering if some home recordings, you can have a listen here:

www.banishmisfortune.net

Yes, a lot has changed since the ’70. Technology has advanced, studio time has got much cheaper, which means producers more likely to spend a lot more time multi-tracking and fiddling about, seeking perfection but in the process loosing the vitality and life of those old one-shot recordings.

But the biggest change has been in the way we listen to music. Back in the day you would put a record on the deck in your quiet living room, then sit and listen to it. Today you are much more likely to just have it on in the backgound while you drive, work or do other noisy things. For that to work the music has to be quite heavily compressed and bass loaded so that you can hear it over the background noise.

> I personally have mixed feelings about it.

I don’t disagree with this, I have to say. Sometimes you can improve something so much it goes all the way around and ends up on the wrong side.

One other thing that drives me absolutely bonkers is impossible dynamics. If you have a GHB playing, you should not be hearing finger percussion noises from the tin whistle or the clack of accordion keys.

I wouldn’t give any consideration to what it sounds like on a phone or laptop speaker. That is no a proper way to hear music.