Double tracking - natural sounding options?

[hotmitts]

Thanks for all the ideas. Going to have fun trying stuff this weekend. The thing I'm trying to avoid is those phasey chorus type effects so I'll experiment

[hotmitts]

I'm not really sure what "natural-sounding" double-tracking would be anyway. The classic studio "thickening" effect of overdubs is not the real-world sound of multiple singers in unison.

What it means to me is thickening the sound without having too much of a chorus or phase effect

[hotmitts]

Well I had a play around with a few of the suggestions. As far as all the plug-ins I tested go, I found the Doubler from Nembrini to have the most pleasant sound to my taste (least of the phasey/boxy/ chorus type feel)

Then I did some experimenting with parallel channels, mostly EQing with the Voxengo free EQ. So here's an example of the original signal from the Piezo input from my acoustic guitar, then with parallel channels, EQing, some panning to create stereo, and a small amount of the doubler plus a tiny amount of reverb. Apologies for the shaky guitar playing, it's bloody cold here and I'm not used to playing the steel string.
i don't know how much improved it is but I did a lot of tweaking to get the best EQ I could.
Thought I'd post the results for posterity anyhow

https://soundcloud.com/hotmitts/sets/example-files

[imrae]

I'm not really sure what "natural-sounding" double-tracking would be anyway. The classic studio "thickening" effect of overdubs is not the real-world sound of multiple singers in unison.

What it means to me is thickening the sound without having too much of a chorus or phase effect

What natural phenomenon does that correspond to, though? Is it supposed to sound like multiple singers, or is it supposed to sound like the original singer is different in some way?

(I'm not having a go! This has always bothered me about the word "thickening". See also people describing synthesisers as "thin" when there is nothing abnormal about the low-frequency range.)

[Watchful]

What natural phenomenon does that correspond to, though? Is it supposed to sound like multiple singers, or is it supposed to sound like the original singer is different in some way?

All it really does is detune the original so that your ears detect "more sound" than a single voice by itself. I'm cautious with chorusing, and prefer to hear it done to a parallel track that's lowered in level behind the original vocal track. That provides a really full sound.

The term itself goes back decades, before synths even: a single sound by itself was "thin." A detuned sound played with it was "fatter." The term "thickening" refers to ways to put more detuning into a track for a richer sound.

If you apply chorusing as an insert, it makes the original singer sound different. In some cases, this works; in many cases, it does not. If chorusing is applied to backing tracks or a parallel track, it can create a nice, rich sound that doesn't detract...and therefore can sound like multiple singers.

Of course, nothing sounds like multiple singers like...multiple singers. But a couple of backing tracks in a submix fed into a chorus? That can sound really nice.

[hotmitts]

i knew I was going to get in trouble with the word 'natural' lol

[jens]

You really need to record twice. Effects will just get you something that doesn’t really emulate the random timing, tone and volume fluctuations of a true second performance.

The afore-metioned Sonnox plugins work perfectly and are a huge time-saver.

They work in a completely different way to the ADT and chorus stuff that has been mentioned here...

the only drawback is that they have a huge latency so you need to disable it whenever you record more takes/parts on the same track.

https://youtu.be/kSQrvnsSpAM

[hotmitts]

Those sonnox plug-ins do look interesting , will test the demo.

[meloco_go]

I am surprised nobody did an intellectual double tracking app for guitar, or at least I am not aware of one. IMO it should be a standalone app, or an ARA plugin, that is, it should work as an offline process. The algo should analyze the performance, identify notes and transients, and apply a combination of time stretching, pitch shifting, and different degrees of regular delay on a per-note basis. To pull it off the app must analyze the performance as a whole.
The same one can do himself, just slice the performance and shift the notes around manually. REAPER (and probably other DAWs) allows to put effects directly onto audio items, so each note can have its own pitch and formant shift and stretch.
It will take time, but will sound better than any "simple" doubletrackig. I know, I did that (luckily for a short part I wanted stereo in a mix).
OTOH, if one is ready to spend this much time editing, maybe its easier and faster to record a real double?
Also, if the artistic direction calls for a sloppy feel, but one still wants stereo, why not go triple? A sloppy and expressive track will take a center stage and be loud and proud, while a more reserved and precise performance (and possibly a more basic version of the part) can be doubled and panned hard left and right, perhaps with a darker tone.

PS If anybody wants to write this after reading my musings, please give me a free copy =)

[HAL76]

I am surprised nobody did an intellectual double tracking app for guitar, or at least I am not aware of one. IMO it should be a standalone app, or an ARA plugin, that is, it should work as an offline process. The algo should analyze the performance, identify notes and transients, and apply a combination of time stretching, pitch shifting, and different degrees of regular delay on a per-note basis. To pull it off the app must analyze the performance as a whole.
The same one can do himself, just slice the performance and shift the notes around manually. REAPER (and probably other DAWs) allows to put effects directly onto audio items, so each note can have its own pitch and formant shift and stretch.
It will take time, but will sound better than any "simple" doubletrackig. I know, I did that (luckily for a short part I wanted stereo in a mix).
OTOH, if one is ready to spend this much time editing, maybe its easier and faster to record a real double?
Also, if the artistic direction calls for a sloppy feel, but one still wants stereo, why not go triple? A sloppy and expressive track will take a center stage and be loud and proud, while a more reserved and precise performance (and possibly a more basic version of the part) can be doubled and panned hard left and right, perhaps with a darker tone.

PS If anybody wants to write this after reading my musings, please give me a free copy =)

I think you should look closer at what Eventide has made over the past decades.

[meloco_go]

I think you should look closer at what Eventide has made over the past decades.

Well, I guess so, and not to disrespect them, but all their products I am aware are intended as "online" effects and this limits the options available.

E.g., the best "fake" doubletrack is if the part is repetitive, just flying one bar onto the other. Can't do that with an "online" effect.

[kevinsparks]

Boz Provocative might be a good option. You can adjust the bandwidth of the doubling effect so it's not overbearing.

[Bombadil]

Always used a slight delay. You can pan tha bus differently to the voice, do whatever you want to it, really.
Back in my 4 track cassette and earlier days, I'd occasionally capture a take that had me do something spontaneously in my phrasing, only to find I couldn't replicate it for the DT. That usually led to doing the entire thing over and losing whatever it was I/we thought was better.
With the 4 track, I could control the digital delay the right amount to thicken and enliven my voice.

[donkey tugger]

I am surprised nobody did an intellectual double tracking app for guitar, or at least I am not aware of one. IMO it should be a standalone app, or an ARA plugin, that is, it should work as an offline process. The algo should analyze the performance, identify notes and transients, and apply a combination of time stretching, pitch shifting, and different degrees of regular delay on a per-note basis. To pull it off the app must analyze the performance as a whole.
The same one can do himself, just slice the performance and shift the notes around manually. REAPER (and probably other DAWs) allows to put effects directly onto audio items, so each note can have its own pitch and formant shift and stretch.
It will take time, but will sound better than any "simple" doubletrackig. I know, I did that (luckily for a short part I wanted stereo in a mix).
OTOH, if one is ready to spend this much time editing, maybe its easier and faster to record a real double?
Also, if the artistic direction calls for a sloppy feel, but one still wants stereo, why not go triple? A sloppy and expressive track will take a center stage and be loud and proud, while a more reserved and precise performance (and possibly a more basic version of the part) can be doubled and panned hard left and right, perhaps with a darker tone.

PS If anybody wants to write this after reading my musings, please give me a free copy =)

Melodyne does the pitch and timing parts with the 'random deviations' function.

[metalifuxx]

They’ll all sound just like a chorusing effect. Just practice and get better at double tracking. You’ll thank yourself later. I gave up on plugins to do that years ago, and now double tracking or even quad tracking is second nature to me.

If you cant get the timing just right on same chord parts, learn all the chord inversions and harmony chords and stick those in here and there. It will sound more orchestral and symphonic as an alternative to double tracking the same bar or power chords. Unless your doing tight hard rock and metal type genres that call for the twin guitar thing.