There is something that music production “in the box” (meaning completely inside your computer) makes possible, that simply wasn’t available to recording artists of a previous age. I’m not sure if I will take this route or not, but I might.
When you record a piece and you need to improvise a melody or a guitar solo, or something that requires you to actually play your instrument well, you’re doing three things at once. You’re trying to write something that works, trying to play it well and trying to capture it as a recording. Three things at once. Even the best multitaskers would have to admit that this is a stretch. If you are also self-engineering and producing, it’s five things at once.
What’s the compromise? Well, in trying to come up with a line and get it down as a recording, your playing might not be perfect. You might have some slight intonation problems, timing issues or the flow could have been better. As you are probably playing your improvised melody for the very first time, your rendition of it might not be as good as what you would have liked, had you rehearsed it. In the studio, it’s quite hard to play exactly the same thing twice from memory, especially if the piece covers more than say eight bars. You could write it out as a part on paper, but that’s time consuming. It’s far better to get a take and then use that as an audible reference.
Even when you have reference, if you are trying to play the same thing as you recorded immediately after you played it (as a second take, for example), that’s quite hard to do. People just don’t learn that quickly, as a rule. To really play the part again, exactly as you first played it, would take some practice, some time and some dedication to learning how. To substantially improve on it definitely takes longer than a next take would. It might take you a few days or a week to perfect being able to play a repeatable rendition of something you already recorded as an improvisation, on demand.
The nice thing about a DAW is that, even once the track is mixed and mastered, you can play the track back as many times as you need and learn to actually play the line you initially wrote as improvisation, properly. You can take your time to learn the notes exactly and develop the dexterity and muscle memory required and to get all the articulations right. After you spend some considerable time with the piece you played in the studio, you can go back into the studio and replace it with a performance that is much better than your initial one.
The advantage here is that this time, you don’t need to be inventing and improvising as you record. You have learned the thing you created and now you can play it, with ease, only better than when you first came up with it. That’s the theory, anyway.
In practice, the later performance might lack some of the freshness and danger of the first performance. You can always compare them. DAWs let you keep all versions and select the best one. It might be that the first take was the best, but it’s also possible that after having practiced playing what you wrote for a few weeks, you can play it really well now. You also have the option of creating a composite from takes played weeks apart, especially if you used amp simulators and software that gives you a repeatable sound. You just choose the phrases that work best from all the takes you have.
Recording artists of the analogue tape and big studio age didn’t have this luxury. Firstly, studio time was expensive. Coming back to correct things was prohibitive. Even if you could, if the recording was on tape, then where were you going to put the new take? Maybe there was a spare track available, but usually not. You’d have to exactly recreate your amplification and effects rig and set it up to sound the same as the last time, including mic positions and where the rig was situated in the tracking room. Then, there was the problem with punch in recording, in the old days. It was notoriously difficult to get right. It was hard to get clean in and out points and if you didn’t, some of the first take would still be a part of the new one. If you punched in cleanly over the first take, it was gone forever, too. Finally, to recreate the mix and mastering process usually involved a set up that was both elaborate and that nobody had written down. It involved dozens of patch cables and setting of knobs to exactly the right settings. No two mixes were ever the same, even on automated consoles. The outboard gear was not automated.
In the DAW, replacing a part is trivial and it doesn’t destroy the part you’re replacing. You can go back to it any time you like. If you were careful with your file management, the remix is trivial. It’s all still remembered as part of your project file. Finally, mastering, if you kept your files in good order, is also pretty straight forward. You just recall the mastering project and replace the mix down audio file. It’s all incredibly simple, fast and repeatable.
I’m not aware if anybody else composes and records this way – writing and recording parts, then studying how to play them well, then re-recording better played performances of the part you first wrote on the fly. It might prove to be not worth the candle. On the other hand, it could be a way to get much better performances.
No matter what, you have to admit that it is an intriguing possibility for the modern music producer.