The Wexford Carol medley with The Holly and the Ivy is complete. I described the writing and arranging in my previous post. See that post for a description of how and why I put the music together the way I did. This post is about recording and then mixing the final product, but first, here’s the finished piece:
The recording consists of a combination of live recording and sampled instruments. As a refresher, a sampled instrument is where the individual notes of an instrument are recorded one at a time and saved in the computer in a program called a sampler which can then be played back and manipulated to create a performance that can sounds very much like a live performance. The performance is, in essence, programmed. This is a very different skill from playing an instrument, and maybe not as difficult to master, but still takes a lot of time and effort to get right.
- Synth Pad (from the Kontakt Factory Library)
- Piano (Addictive Keys)
- Hammered Dulcimer (I’m sorry I don’t remember the provenance of this library, but it was a free download sometime in the past)
- Recorders (alto and soprano, played live)
- Cello (the discontinued Garritan Gofriller Cello layered with the Pocket Blakus cello – more about why I layered them later in the article)
- Mandolin (recorded live)
- Bodhran (a soundfont from FMJ Software – I sprang for the full version)
- Accordion (Kontakt Factory Library)
- Pennywhistle (Embertone’s excellent Shire Whistle)
- Strings (Vienna Symphonic from the Kontakt Factory Library)
- Bansuri Flute (from the Kontakt Factory Library), and
- Yanggeum (from Seoul National University Center for Arts & Technologies)
These instruments weave in and out throughout the arrangement. Details are in the previous post, but generally speaking the piano, synth pad and recorders play together, then the mandolin, accordion, bodhran and pennywhistle, then the piano and dulcimer with string and cello, the yanggeum and bansuri for the ambient mash-up section, and the finale with seven of the instruments all playing together.
As mentioned, the pads and keyboard instruments were simply played in with a keyboard controller. I used my trusty Yamaha WX-5 to enter the basic data for the pennywhistle and bansuri. This allows me to use the pressure of my breath to control the volume of the sound in a very natural manner. The pennywhistle sample library uses the note velocity (so named from measuring the velocity, or strength, one plays a piano key) to trigger the ornaments (initial note flip). Since I set the WX-5 to only transmit a velocity value of 100 (to retain more control of the sound later) I went back and edited the attacks so that the notes I wanted to “trill” would do so.
Odd bug in the Shire Whistle library: if I began a really long note with the ornament at the beginning, at some point within the note it would trill again. My guess is that is where the sample loops, and there’s a bust in the programming. I mentioned it to Embertone so we’ll see if they can reproduce/fix it.
Recording the recorders was fairly straightforward. I generally have to use multiple instruments to cover the entire range I want to play since my modest skill limits me to roughly an octave of range. To play this I go back and forth between the alto and soprano instruments to cover the higher and lower parts of the melody. Back when I recorded This is My Father’s World a couple of years ago I used four different recorders to cover the entire song!
Recording the mandolin was a bit more of a challenge. I don’t have a wealth of experience with fretted strings, having grown up with a baritone ukulele in the house and that’s about it. I’ve messed with guitar a little and own an appalachian dulcimer (which will appear on the next track I’m planning on posting), but mostly I’m just not that good at it. However, I’m not aware of a virtual mandolin instrument that can do a convincing strum, so I set out to borrow one and play the part myself.
One thing about mandolins – they have really narrow necks and the strings are comparatively closer together. As can be seen in the screen shot I needed at least 8 takes to get everything (this isn’t counting the takes that were so bad I just deleted them, unable to use any part). Looking at the take lanes there are plenty of places where a used segment of the recording follows silence. That was because I needed that time to carefully finger the chord. In the end I was able to put together a convincing performance and the feel is exactly what I was looking for.
When everything was recorded, the arrange window looked like the screenshot to the left. The top and bottom tracks are the acoustic tracks, and everything in between was MIDI recording of sampled instruments. The two brightly colored MIDI tracks are the two wind instruments – the pennywhistle and bansuri. What you are seeing is the wind control data.
If you look closely several of the tracks have additional automation lanes. Usually this is just volume automation, allowing natural swells and diminuendos. A couple of tracks have additional lanes, such as the celli. They have volume and vibrato.
I promised earlier that I would explain the use of doubled celli for the single cello line being played. The Garritan Gofriller cello is a very deeply sampled and powerful instrument, but is limited to a very classical style of playing. The other instrument, the Blackus Pocket Cello, was specifically recorded for slow, deep, very rich lines – by necessity having a slower, mellower attack. By combining the two I get the attack, fluidity and virtuosity of the Garritan instrument with the rich, vibrant tonality of the Blackus. This is not the last time I will use this method.
With all the track recorded and edited to perfection, I rendered each MIDI track to an audio track (and re-rendered the comped and edited audio recordings) I saved everything into a new project and removed the virtual instruments to save memory (my computer is beginning to strain under the load – I would like to replace it this spring depending on how things go).
The tracks to the left are in the same order as the MIDI tracks above, with the exception that the two recorder tracks have been combined into a single track at the top of the screen. The strings, pad and piano have all been rendered as stereo tracks, and everything else as mono.
Under most of the tracks I broke out the automation lanes. This time it is almost exclusively volume automation, with the exception of the recorder tracks. Zooming in the on the screen shot will reveal four additional automation lanes (one is overlaid on the waveforms – not sure why Sonar sometimes does that). These lanes control parameters on the processing and effects for this track – three for EQ and the fourth for reverb,
The last screenshot from this project, above, is a composite of the entire mixer. I didn’t add much processing above and beyond the Sonar ProChannel (remarkable piece of kit, really). Some highlights include the Kush Audio Clariphonic on the recorders – this is an incredible, if limited in scope, EQ that I won from Kush last year. I can’t believe this is only the first mix I used it on. Look for it in the future on everything! The recorders also have a limiter to help control some of the piercing overtones on certain notes – and yes, they are extra piercing because I have the highs boosted so much – it’s all a balancing act.
The only other non-ProChannel processors include a multiband compressor on the pennywhistle evening out some of its shrillness, a compressor and delay on the mandolin, a compressor on the bodhran and a chorusing effect on the accordion.
There are three reverbs, each on their own sub-mix. The Main Verb is simply a nice medium length, medium duration plate reverb from Valhalla. The ambiance is similarly from the ValhallaRoom plug in, and the “ethereal,” used during the mash-up section, is the Warm Orchestra preset from Sonar’s included Breverb plugin.
Some gentle mid-side compression was applied to the mix bus, and I used the bx-Limiter processor to make sure the signal never got in to digital clipping. Finally, the Utility plug in is my in-depth metering plug in. I’m experimenting with different ones, so my next mix article will feature another, which I have to say I am impressed with – more to come.
So that’s it – the Wexford Carol. This was much fun to work on and I found it quite inspirational. My only regret was missing a Christmas release date for it. Well, there’s always next year, and with some luck and hard work, I’ll publish 3 or 4 more in 2018.