AR Roots: A Whiter Shade Of Pale – cover by AndyR

[AndyR]

The perennial paean to sex under the influence. (Although it's only as I get older - and I have fond memories of being in my 20s and going through this sort of thing - that I've learnt to appreciate that this is what the song's really about ).

I can't quite believe I'm posting this, or that I've even recorded it. But I've played it like this sat on the sofa in the living room for many years, and I needed something to check out how to record a live "one vocal and one guitar" song.

You would not believe the hastle it was capturing this!!!

Anyway, here we go, me, guitar, one take, three mics, a BR1600, warts and all...

Time:

0:00

Volume:

50

0

 Stop   Play   Pause   Vol   Vol   Mute   Repeat   Download

Music and Lyrics: Brooker/Reid (although, since 2009, it appears that Fisher is now officially credited as a co-writer, about time, I guess)

[Hilary]

 

[Oldrottenhead]

my oh my, this was well worth the hassle andy. FMGWABP

[kenny mac]

This is fantastic,your voice is very good and this recording is also very good,I love it when someone comes accross with a stripped back performance that really works,and this does the job for me.10 out of 10

[cuthbert]

Very introspective and soulful, Andy. I love the spareness and the spontaneity of it.

Your rendering shows this song can stand very well without the J.S. Bach organ melodies (which can sometimes feel a little gimmicky, although I have nothing against Johann Sebastian or their use in the original!).

[launched]

I am so very pleased to have subjected myself to this wonderful cover! Phenomenal changeup, that's the way it should be done!

[Geir]

This is now my favourite version of this song! I can't see how it could be done any better.




oh .... please do tell more about the setup. What kind of mics, placement of mics, compressor settings ... etc.

[Flash Harry]

Wow! I'd say you had captured the guitar and the voice brilliantly.

It's a fabulous interpretation too. Giving me goose bumps. It sounds remarkably intimate.

Bravo.

Like Geir, I'd be keen to know how you did this.

[AndyR]

Like Geir, I'd be keen to know how you did this.

Funnily enough, I've been working on it for the last hour or so... You even get an "honourable mention" in it...

I'm checking it at the moment - should get posted shortly - It's a bit long though!

[AndyR]

'ere tis then...

==================
Thanks folks

(Great big waffly post alert – I thyink I might even have broken some sort of record... Geir the stuff you want, or what I can remember of it, is second half!)

Seems to be getting (as far as I'm concerned) a surprising amount of approval around various places. And there was I thinking it was a generally loathed song by "those in the know" - ie musicians!!

I've always loved this ever since I first heard it on the radio. I can remember being in the back seat of the family car going "what's this!?!?! Turn it UP!". It was probably already about 10 years old by then, and my parents would have been familiar with it and couldn't see what the fuss was about. But I was hearing it for the first time. The film "The Boat that Rocked" almost manages to get that feeling/emotion across when it plays it (as it does some other classics).

In the early 80s, my band used to get asked to play it at a particular residency we did. It's the one thing we refused to learn "for next week" – mainly because the hippy bass-player and keyboard-player classed it as "rancid" (and the drummer and I thought it would be too difficult).

A few years later I investigated why this guy (my age now) always used to ask us 20-somethings to play it. And I found out why: apart from the fact that our line-up and approach fitted, it's a perfect song for me to sing (I have others like that, that I know of, I might strip them down as well - it's partly what this "AR Roots" thing is about).

I discovered that it's just a blues song with a LOT of melancholy and yearning for lost opportunities in it. That's how they wrote it, without Matthew Fisher's organ part. The argument for many years has been that the organ part added just before the recording made it into a hit... but he didn't get a writing credit (nor did JS Bach! – but it's the bass line that suggests the Bach piece for me, try playing anything over it and you get Air on the G string ). Apparently Mr Fisher or his estate finally won in 2009 and he is now credited.

I'm not so sure that it couldn't have been a hit without this hook though, though. EXCEPT, they would have been playing it as a band, not stripped down like this. You start adding arrangement but avoid the Bach-iness and it gets stodgy very quickly.

Does anyone remember Annie Lennox's version? She left the organ signature out as well. To be honest, as much as I adore her voice, I thought it was a bit of a fail... it didn't really do her or the song justice. Two problems (for me) in her case - it had a fullish arrangement, and... she's a girl. Whether all the people who have it played at weddings realise it or not, it's about an inebbriated and (possibly) failed, but certainly regretted, sexual encounter told from the male's perspective. Even if you change it to "his face" it doesn't quite have the ring of unreached but everlasting truth to it (not to a male listener, anyway) when a girl sings it, even when you don't realise what the song is about (I didn't for many years, and I didn't when I heard Annie Lennox's version).

The writers, last time I looked into it, openly admit that this is what inspired it. They're happy to say "it might mean something else more worthy as well, but it's about that...". There are another two verses (you can find them on the web). One was dismissed by Gary Brooker as unsingable, the other was omitted to keep it down to single length.


So, when it came to do this, I had a variety of decisions. Do I play the organ part? I can do it, on organ (too cheesy when someone copies it), or guitar. Sounds great as slide. But I never do it when I'm sat on the sofa... Do I sing the missing verse(s)? Apparently Procul Harum do them live. But the thing is, I'm doing it a lot slower, I think the song needs to be slower when stripped down, and it would be monstrously long if I did the intros and the extra verses.

Next, do I do a bass part (on a bass)? It's a bit arse without the bass part – if you just play chords and sing it drags a lot. But if I want to add a bass part I'll need a click track, probably end up with drums and a choir and wotnot... And that's not how I sing it on the sofa. But keeping the bass part going on the guitar is hard work when you're singing as well.

Last was - do I reprise the last chorus? During the many many takes I did of this thing, I discovered that the chorus reprise is really only there so you get a second chance at doing an impressive last chorus – you've usually done a not so good one just before it! So, if you get the original last chorus right (ish) then the song's better without repeating it.


OK, Geir – technicalities and logistics...

This one was the first take when I got home from work on Tuesday. Power up, set mics, cough, sing, stop – about 20 minutes. It then took an (expected) hour or two of post-processing, and decision making in the mixing and mastering.

However, this does not take into account the five hours spent the evening before on aborted and failed takes. It was mixtures of poor performance and not having three separate tracks I could work on. It was not being able to find a bluddy seat in the place that doesn't creak. It was "my studio is too small to record an acoustic guitar easily without it banging into the piano just before the last chord" (Flash Harry told me the 3rd bedroom might look big, but it'll soon be too small... ).

Nor does it take into account the weekend spent trying to make the previous version (recorded with much angst the weekend before) work in some way.

I nearly gave up and went for a click so that I could record repairable guitar and vocal separately. But there is "something" about the "wandering troubadour" thing, when the singer accompanies his/herself. That's missing when you record them separately, unless you put a LOT of work into rehearsing. And also, with a click, unless you spend hours working on a tempo map, the performance won't breathe (tempo-wise) like a guitar/vocal performance would in real life.

So at the end of Monday night I'd conquered how to mic the thing, and how to post-process the three resulting tracks if I ever got them down ok. I decided I'd spend a few evenings of just "come home, record a take or two and then forget them til the weekend".

The very next take, first one Tuesday was it

It's recorded with three mics:

Vocal - Rode NT1000, 5 inches to the left of my mouth (pointing down, tip about an inch above my mouth). It's positioned so that I can see the fingerboard without moving my head too much (and getting too much vocal into the other mics). It's on the left because that's where there's least sibilants coming out of my mouth. This mic has the BR1600's Vocal Comp 1 (first mic preset) on it. I use this for all my vocals – the noise gate is quite important, but I had to make sure that not too much guitar gets to the mic. Noise gates and acoustic guitars don't mix in my experience

Guitar 1 – Shure SM57 close-mic'd, pointing at guitar soundboard behind and below bridge (away from soundhole). And pointing AWAY from my mouth.

Guitar 2 – Shure SM57 pointing at the same place on the guitar, but 2 feet back and from about a foot higher. Again, pointing away from mouth. (On earlier takes, this was pointing at the 12th fret, but try as I might it just picked up too much vocal).

The Vocal mic was recording as I high as I could. The Guitar mics were recording conservatively (for later compression trickery).

Each went to its own track in the BR1600. Once I'd captured the performance I wanted, this is how it went:

Guitar 2 cloned, the original panned fully left. The clone slipped back just a smidgeon and panned fully right. Both were EQ'd for bottom and as much mids and brightness as they could handle and were given max reverb.

Guitar 1 sent down the middle. It had a high-ish frequency ripped off it (I think it was a peak rahter than a shelf EQ) to deaden the vocal it was carrying. It also had it's bottom reduced to make room for the vocal. It had a light chorus slapped on it, and very little reverb. During the mix, I rode the fader on this one to bring out any tentatively played bass notes. This guitar was for body, the Guitar 2 tracks were for stereo spread, twinkles (caused hiss problems later), and warmth. I experimented with chorus on both, but discovered it gave me better stereo without losing the acoustic sound by only putting chorus on Guitar 1.

I then clobbered the guitars (especially the stereo ones) with heavy-ish compression. Because the guitars weren't that hot, I could see when loud vocal bits were intruding on their tracks. So I set a low-ish threshold and heavy compression – this left the guitar part untouched until the vocal got loud, then the whole thing got ducked. You don't notice the affect this has on the guitar because the lead vocal track is so loud, and it means the stray vocal from these tracks doesn't intrude on the lead vocal track itself. Over all, it meant I could get the guitar tracks up loud without causing strange room reverb effects on the lead vocal when I got louder

It did add to my hiss problems later though. In future, I would record the guitar signal hotter, and just set the compression with a higher threshold.

The lead vocal track was cloned twice. First clone slipped back a smidgeonette and panned left, second clone a full smidgeon (don't ask how much, it was a little bit!), and panned right. These were EQ'd soft, just to add body to the main voice.

The original lead vocal was put down the middle with some light compression and standard EQing for my voice (I look to boost/reduce as follows: 200Hz peak to add or remove fullness, 3000Hz peak to add or remove atttack – eg reducing softens backing vocals and makes them more breathy, 5000Hz to add or remove vocal presence – in this case it needed removing!).

The guitar tracks were bounced to a submix first. This meant I could simplify fader rides for dynamics (do the guitars first, then the vocal and master second), and I was able to review the overall EQ of what became a single stereo guitar part during the final mix (I gave it a bit more body and tinkle – the later being a big mistake!).

Then I created the final mix by adding the vocal tracks to the guitars. I rode the original lead vocal fader to bring out the ends of lines and other interesting emotion (breathing etc. – I have some takes with a far better vocal than this, but the guitar performances weren't good enough, or there was too much creaking and bashing going on), and I rode the master fader to add dynamics to the performance.

Before mastering, I normalise the stereo mix. This means the track gives the mastering algorithm maximum opportunity to do what it's designed for. It does mean, however, that if you tweak the mastering settings you must not boost too much anywhere, especially before the multiband compressor – if you do, you squash the mix.

When I normalised, I found two nasties now that the mix was loads louder. Nasty one was the hiss. Nasty two was some plosives on the vocal that the pop shield hadn't caught. It meant I couldn't use the "Live Mix" mastering algorithm which the song cried out for – this one has the effect of sticking a load of top and bottom on a mix.

As a solution I plumpled for the Pop Mix mastering algorithm (always a safe option!) and tweaked it by turning on the bass cut (at 60 or 70 Hz) to reduce the plosives, and turning on the initial EQ to take out the top (can't remember whether this was peak or shelf EQ). I also tweaked the mix settings after the compressor to compensate a bit. Both the guitar and vocal could handle this, just about (I've noticed since that some of the diction that I know is there has gone), but the final thing has ended up sounding warmer than I intended, just to lose as much hiss as possible.

The original plan was why I picked the reverb I did. I'd toyed with no reverb on the previous version, but it didn't do anything for me. I tried my usual warm Hall stuff. I tried various Room, and Plate options. What I ended up liking was a cold "morning after" vibe caused by a very short Hall (0.6) with a lot of early reflections. Then I set it as dense and dark as I could and didn't bother putting any bass cut on it. I then piled as much of this onto the vocal as I could, and enough on the guitar to make it seem like they were part of the same game. I actually had it in my cans when I was performing so that I could work the "Hall" while I was singing – cutting off words to get the reverb effect, that sort of thing.

The effect on the original mix was a bit more "shivery" or "I've got a hangover and various bits seem sorer than last night". The final master had to lose some of that, especially on the guitars, just to lose the hiss.

If it had worried me enough, I'd have gone back to the mix – it's not like it's one of my multi-layered things where it's buried miles down. But the thing did the job, and the missus thought it was better than the earlier version (lots better!) so I posted it thinking "right, now I can get on with my life"