I guess for many amateurs this balance isn’t obvious. Beginners likely don’t do much sound design at all, rather relying in presets and sequence samples. Over the last couple of years sound design has become a far bigger part of my process. First, to make sure I’m not being misunderstood, I’ll define what I’m talking about. Sound design is fairly obvious, this is making sounds at a low level. Designing synth presets, building drums and programming loops. Then there is composition, which I define as what you do when you turn sounds into music. However you do that.
The first balance you need to make is simply a matter of time. You have to give some time to each. How much depends on, fundamentally, how much you value the need to make your own sounds from scratch. For some genres, or some… musically career strategies… no-one cares whether sounds are made by you or anyone. For them presets are fine, so they may as well use them. Many can even get away with using 10 presets for their entire musical career. If you are an amateur, you make the rules. If you have the time I think you should definitely start to think about making your own sounds. I say if you have time as it may well seem time consuming. The advantage is that you can truly say your music is your own.
Anyway, I wanted to point out a more pragmatic balance that needs to be made. If you set about making a sound from scratch, there are two extremes of ways of doing it. At one extreme is making a sound with a very definite use in mind, to the extent that you have a precise place in a specific track for it to go. The other extreme is making a sound for its own sake, with no specific use in mind.
Making a sound for a very specific use is hard. This is because you will likely have a very small frequency space for it to live. The exception is when the track you want it to live in is bare. In this case you have all the room you like for your new sound, there are no other sounds for it to compete with. The problem with this is that if a sound like that, one that sounds good on its own, you will have a hard time fitting other sounds into ‘its’ space (i.e. the track). I think many tracks are made with some random hardware synth noise that was captured from hours of playing with the synth, looped into a hook. In this case that feature will often be a primary feature.
Anyway, back to the point. Making a sound for a specific purpose is harder than just turning on a synth and playing with it until it makes a sound you like. The trouble is that it can be equally difficult to make a sound for no reason, with no purpose in mind. If you want to do undirected sound design you tend to have to produces many sounds, create your own preset library, that you can come back to when you’re in composition mode. If you have loads of time to spend you can spend it making a huge custom library. For more experimental types of music you can probably make more bizarre combinations of sounds work. The more ‘nonstandard’ your sounds become the harder it will by to mix them together well (one of the reasons so many mainstream producers use the same sounds over and over: they can just recall the same presets, processing and effects) I think you will find yourself making a mental balance. You will have a rough idea of the type of sounds you might want for some projects, but won’t be so fixed as to make sound design too difficult. How you balance that is something you’ll have to figure out.