Is there such a thing as Photoshopped audio?
I find this comparison shot scary. Heaven knows, the model is as attractive as anyone could reasonably aspire to be, yet in the Photoshopped version no fewer than eighteen 'imperfections' have been 'corrected'.
This made me think about what we commonly do as a matter of routine in audio to correct similar imperfections. We could call it 'Photoshopping the audio'. Let's start with...
The microphone and close miking
If you have a great singer, then why not record them with a super-accurate mic, such as for instance the DPA 4006-TL, which is designed to capture a signal that is extremely faithful to the original acoustic sound.
But who does that? You can get a much bigger, fuller and richer sound with a vacuum tube microphone such as the Neumann M-147.
Oh, and by the way, don't forget that close miking will create a much more present and upfront sound, something a singer couldn't do naturally unless they sang right into your ear.
You could record your great performer through your super-accurate mic connected to a precision transistor preamplifier. Or you could pump up the sound from your tube mic with a tube preamp, with the gain turned up high, output level low.
Even very good singers vary in level, particularly at the low and high ends of their vocal range. Modern digital recording equipment is fully capable of capturing this dynamic range. But why not squash it up so that it comes out more evenly in level? Just plug-in your favorite compressor.
Now that we have four ways to Photoshop a vocal straight from the mic, let's see what else we can do with clever recording techniques...
Comping and punch-in
Now if our singer were truly attractive enough already, one take should do it. No, let's be realistic - even an unshopped photo shoot would burn through a lot of film before capturing the best image. So go for a lot of takes, and choose the best. That seems a reasonable thing to do.
But hey, it isn't likely that there will be one take that is the best all the way through. So why not edit out the best bits of as many takes as you like and compile (comp) them together. And if there are still a few problem areas, then get the singer back in for punch-ins.
You might say that there isn't anything like this that would happen in photography. Well think again. It isn't at all unusual to replace hands with those of a specialist hand model. Legs too. And while medical science hasn't yet perfected the full body transplant, the advertising industry certainly has.
Oh, and how do you think wedding photographers capture all of the guests in one shot with none of them blinking? Scan Photoshop-era wedding photos for the scary eyes!
I don't see any analogy with photography here, but it is very common to enhance a singer's voice by double tracking, i.e. recording the same vocal twice. In an earlier era of recording this was done perfectly unselfconsciously, and it can be heard clearly. These days, the ease of editing that we have allows double tracking to be done to a much higher standard - unnoticeable apart from sounding better.
Reverb is something that happens in the natural non-sound-engineered world. The singer sings, the voice reverberates around the room, everything sounds good. Well, it may sound good for classical music, but for popular music we want something even better.
So why not separate the dry vocal from the reverb? Record with a close mic in a very dry acoustic, then add the reverb exactly the way you want it. And why would you want to do that? Because it sounds better than real life, that's why. The same applies to delay and spin echo, even though you would never hear these effects in a purely acoustic performance.
I mean Auto-Tune used for pitch correction, not as a T-Pain soundalike effect, which is a different thing entirely. Prior to digital sampling becoming commonly available in the mid-1980s, there was no practical way to retune vocals. Before then, singers had to be able to sing in tune. If you couldn't sing in tune, you couldn't be a singer. Simple as that.
Vocal too thin? Then you can put it through a delay, modulate the delay time, and mix it back in with the original. Or you can make a copy that is pitch shifted slightly, about 8 or 10 cents or so, and mix that in. Or mix in a chorused version. There are a variety of ways to do this but the end product is a thickened vocal, albeit often obviously so. A bit like a bad Photoshop in fact.
This links with double tracking above. If your singer isn't providing as rich a vocal sound as you would like, then you can double track as I said. Or you can bring in a really good session singer and they can do the double track! Or two singers. Using different voices can provide much more richness than doubling up the same voice. Attention to detail is required, and you can expect a lot of editing.
Of course you could just use the voice of the session singer alone. Technically this wouldn't be an audio Photoshop because it would be analogous to using a different model. But you wouldn't want to get caught out doing it. On the other hand, you could get your background vocals done a lot faster using experienced session singers rather than a boy band's or girl group's members. I wonder if anyone has ever done that...? ;-)
I've said a lot about vocals, but we commonly Photoshop audio more generally too. In the days before multitrack recording, the whole band had to be able to play the song all the way from the beginning to the end with no mistakes. And the engineer had to capture the whole thing in stereo (or previously mono). Even in the 1950s however, vocals were being overdubbed to the backing track, played in from a second tape recorder. In the 1950s photographs were altered using an airbrush, which was the Photoshop of its day.
Multitrack recording allows instruments to be recorded individually, vocals too, so that each player and singer can perfect his or her part, to a much greater extent than would be possible in a simultaneous performance.
It wouldn't be hard to think of other examples of Photoshopping in audio, but I wonder whether it is such a problem? I have to say that using Auto-Tune for pitch correction does seem to have encouraged a breed of singer that can't really sing but, other than that, I think that all of the creative techniques we use in audio equate more to lighting, makeup and overall styling in photography, not to Photoshop-style manipulation.
At the end of the day, it isn't an either/or situation. If you want to make a natural recording, then get out your matched pair of super-accurate microphones and take it from there. And if you want to be creative, then Photoshop your audio to your heart's content.
P.S. Credit where it is due, the photo is courtesy of Debenhams, a very well-known chain of stores in the UK. I admire their stand and, if you do too, why not buy a pair of socks from them in support?