There's nothing wrong with using limiters and compressors, so long as you don't overdo it
Sure it's usually a good idea not to heavily compress thing
Let me explain this more clearly. The first step of making a song sound good is mixing. Mastering comes afterwards. Forget about the mastering for now and concentrate on the mixing. The second step of mixing (after choosing good sounds), is getting volumes right across the tracks, but also along the entire song. If there is one instrument which is peaking its balls off, or some automation that spikes the volume at some point in the song, that is the thing to correct. Compression or limiting might fix this, but at the expense of the dynamics of the song. No good producers would blindy apply compression to the master track and squeeze the shit out of the whole song. That's just nuts.
Some explanation of compression is probably necessary. This is from wikipedia:
Dynamic range compression, also called DRC (often seen in DVD and car CD player settings) or simply compression reduces the volume of loud sounds or amplifies quiet sounds by narrowing or "compressing" an audio signal's dynamic range
In other words, compression reduces wild spikes in volume. Compression is great when say for example you have a sample where the player struck some notes too heavily, and other notes too lightly. But, it should be pointed out that you wouldn't apply compression to the entire song, you would only apply compression to that track to reduce harsh volume differences.
If in fact there are wild volume differences between tracks, or between parts of songs, this is the first thing to address. The first thing you need to do is find where this is happening, and reduce the volume spikes with the plain old volume fader. Compression can then be added later to smooth things out further but not to magically make a mix sound professional.
Compression is obviously used a lot in electronic music for various effects, but don't make this confusing. First and foremost, get the volume right across the tracks, and across the song.
so he should do what he's trying to avoid doing twice as much, to get the following results?:
No. The point of exporting it the first time is to figure out where the volume spikes are. If the volume is consistent across the song, the normaliser will do a better job of pumping out a WAV file and you won't have to guess which level the Master volume should be at to get the loudest sound without clipping.
Sorry I should have addressed this:
I'm not to hot on "normalizing" my song because I feel it destroys the carefully planned mix I've made. It outputs a rather bland product...
That's not quite right. I'm sorry I can't comment on Ableton's normaliser but different DAWs do normalisation in different ways. Some DAWs use compression in exactly the same way mentioned here: it merely reduces the dynamic range of the song and you end up with the bland product you are talking about (compression). However, the better kind of normalisation simply multiplies the volumes of the track in relation to the loudest peak in the song. If there is a massive peak somewhere in the song, the song will end up being very quiet. Sorry I can't comment on which method Ableton uses. You'd have to read up or experiment. Normalisation is not necessarily bad if you are just starting out - just make sure your mix is mooth before you bounce.
