well, I stated my goal for this past weekend as doing the needed repairs on the one unfinished song from “gone native”, “flying solo”, and, dealing with the album artwork.
while I can’t say that both are finished, I made significant progress on both, and I a pretty certain that “flying solo” is complete, whereas the artwork, well, we did photos over the weekend, and sketched out some cover ideas, but that just needs more time – one possible idea for a cover has been conceived, front and back, so that just needs more work during the coming week.
as far as repairing “flying solo” goes, well, it’s been a really interesting experience. first, I don’t believe I have ever tried to “reverse engineer” a guitar part that I had previously recorded, which is a challenge to begin with, however, in this case, it was made even more difficult by the fact that said guitar part was originally completely improvised, take one, off the top of my head.
in the event, it proved to be quite, quite difficult to “learn” an improvised guitar part – more difficult than I would have imagined, a two-hour session last saturday, with 8 takes made, yield no clear winners but it did mean that after two hours, I had pretty much mastered the form of the part – which is a solo guitar, with no accompaniment at first, so about a minute playing utterly on my own, and then the drums and existing guitar solo come in, and I then had the beat to play along to – and then another brief solo section at the end.
returning to the piece on sunday, I did a further ten takes, with take ten marked “best” – and, it does have a clearly superior (and new) ending. in fact, in many ways, it’s entirely better than the original track. however, there are two notes in the introduction that went slightly out of tune, so something will need to be done to fix that – probably “fly in” those notes or that phrase from another take would be simplest, and take 6 from sunday seems to have a “good” intro, so it’s a likely candidate.
I did do some very rough tests last night, using 95% of take ten as the take, and flying in part of take 6 for the questionably-tuning notes, and it seems like it is a winner with that slight adjustment. of course, I really was trying to get the entire take live, but after 18 tries, I think 95% live might be all I can manage – sure, I could do many more takes, until I got one that was more perfect, but – I doubt I would get the beautiful ending that I improvised on take ten again…
It’s disappointing though, because on a couple of occasions, I got VERY near to nailing the entire thing in one live take, which of course I would have preferred. I found it strangely difficult to “play” this part, it was like learning someone else’s guitar part, and it took a lot of work to make it sound natural, as if it were improvised – which strangely, originally – it was…but now, it’s a re-creation of an improvised part.
the fun part of this though was the middle section, when the drums and guitar solo come in, I just completely improvised the rhythm guitar that went behind the central solo, in the original version, I had just taken a section of the improv guitar, and “looped” it as a backing, so now, the song has a proper rhythm guitar part, including a lot of really strange harmonics and bending – strange because of the synth voice I used, “flying tremolo” – which does some amazing things to the sound of your guitar, especially when you play harmonics. I thoroughly enjoyed playing a sort of adrian-belew-vamp behind the solo, and I think it has improved the song greatly to have a new intro, backing rhythm for the solo, and outro – nearly captured in one very good take.
the two notes in question are right at the top of the neck, and you guitarists will know that if you “miss” hitting these notes with extreme accuracy, they WILL go out of tune. it’s a shame, because those two notes are “right” on most of the other 17 takes, but never mind.
I am also tempted, now that I know the part like the back of my hand, to possibly have one more go at it – it depends, if I can find an intro that I love, or that can be used to repair my near-perfect take ten, then I would just use that, but if I can’t…I may end up playing this two-minute guitar part yet again…I want it to be right, and if possible, to be perfect.
re-recording the part has changed the character of the piece a little bit, but not in a bad way, I actually think that piece is stronger now, especially since I totally improvised the middle section EACH time, on every take, I did it differently each time, “learning” parts and then discarding them, replacing them with better ideas, learning what NOT to do – and by the time I got to take ten on the second day, I actually had constructed, almost by accident, a rhythm guitar part that really works with the piece. I knew exactly how to handle it, and I even introduced a chord change in two spots that did not exist in the original. why not?
the new introduction, well, it’s very nearly the same as the old one, but hopefully this time, with no inherent distortion (levels are being kept on the low side just in case). the intro was made even more difficult in that I had to calculate the timing “blind” – I had to play for a certain amount of time, play all the notes, octaves and chords planned, and end up on a descending riff of G, F#. E, D# JUST as the snare drum comes in… so that is a feat of extreme estimation – start too fast, you end up early, start too slow – late – and it was not easy ending up at JUST the right spot every time. to my credit, I did make it there at the right time on the majority of the takes – you just “get it” after a while, you know just how long you have…
much to my everlasting astonishment, on most takes, I ended up there at precisely the right moment – basically because I was playing the part correctly – but a couple of times, my tempo would stray, and I would end up either early or late – not that that is a huge issue, because of course in SONAR I could just move the intro to where it needed to be – if I detached it from the rhythm guitar that follows. but it was and is my preference to play the part right, from start to finish, and not resort to any moving of guitar parts, and I think my current solution, using almost ALL of take ten, plus a tiny bit of take six (or another one if I find a better candidate) – take ten went extraordinarily well, and I really liked the ending I played, it was very gentle at the end, ending on a barre e minor on the 12th fret – strangely, a couple of beats AFTER the drummer hit his final cymbal crash, but it was always meant to be some solo guitar after the drums finished – and I held the chord briefly, and then stopped it – which has the strangest, most wonderful effect of doing a sort of detune-then-fade out – thanks to the very odd “flying tremolo” patch – it gave me the perfect ending – very pretty followed by a briefly detuned, dissonant farewell – the sweet followed by the sour – which worked out perfectly. if the synth hadn’t of done that, the ending would have been too sickly sweet, so I am really glad it did!
so my first experience of learning a “dave stafford guitar part” – an improvised, take one part was – that it was bloody difficult to learn, and nearly as difficult to play! that surprised me, because I thought “it will be easy” – but in the end, it turned out to be difficult enough that it took me four hours of practice to play one minute of guitar “well” – because once I got to the end of the intro, I just “wung it” (that being the past tense of “winging it” I assume) and made up the entire thing from middle section onwards through the ending – and that was really fun, playing along to this very outrageous solo, a very loud and “in your face” solo, and trying to make it work in the “reverse engineered” fashion – it was a blast..
I’m sure if it had been any other part, it would have been much easier to “re-learn” but by chance, this particular little piece of music is quite tricky – it starts with an e minor ninth chord at the root, which then jumps up to a high melody beginning on the 15th fret; then, back down to the root for another e minor chord, with a 2 note trill; then, jump back up to the 12th fret, a climb up the e minor scale across the strings, and then a really high melody including an incredibly hard to replicate top-string bent D to E at the 22nd fret followed instantly by a two note trill on the second string, and then a couple more difficult trills on the third string…then a reverse strummed e minor chord at the 12th fret which then turns into a riff that climbs up from the 12th fret e on the low string, resolving on the third string in an F# (to re-iterate the e minor ninth theme present throughout) – then, a section involving some octaves played on the fifth and third string, that eventually climb up to a final strummed chord e minor chord with a set of descending half-step note pairs, descending down across the 12th fret; then a jump back down from the 12th fret to the root, and at last to that final four-note ascending pattern, coming off an e minor chord…into the middle section.
written like that, above, it sounds as difficult as it was, I really had to analyse it to truly become familiar with every nuance, but now it’s committed to memory, I could play it live if I had to – and maybe I will, as it’s a very distinctive piece indeed.
I assume that I will resolve this eventually with a new mix that I am happy with, and I love the “new” middle part and ending – even if I am never 1000% happy with every note in the impossibly-designed introduction – it’s just bloody difficult, so I definitely challenged myself there…but then I never dreamed when I was originally playing it, on march 6, 2011, that I would have to come back a year and a bit later, and try to re-play it – if I had known, I would have played something easier to learn!
it doesn’t surprise me, however, the lengths I will go to, to try and “save” a song that has problems, especially if I think that the song is a good song – I will spend a disproportionate amount of time “fixing” a song that I spent three minutes making, originally – because I care about the music being “right” – and for me, “right” means, basically, perfect, or as perfect as it can be given my current faculty on the instrument in question…
pureambient / dave stafford 