A bit of explaining might be needed as per my previous post.
-12dBFS means that we are two bits down from top as one bit is about 6dB. We still have about 18 left in the AD. Good AD converters today gives about 20bits S/N regardless of the claim for 24 bit fame. Compare to a CD that is max 16 bits, and we still have quite a lot to work with. If you run a digital recorder very close to 0dBFS you are letting the electronics work as a limiter, there are always short strong sounds, transients. Some of the boxes are made to sound good "overdriven", some not, but it is definitely not a true representation of the recorded signal anymore. My guess is that the V3 mentioned is made to sound good in "overdrive" distorting in a pleasant way. The 722 circuits are not, they distort ugly (according to my ears). With the 722 limiter in circuit the "distortion" is again nice to hear, it starts at -6dB though so still a good idea to be a bit lower.
In post I might do the following. Remember I do acoustic music, open recordings, on request by the choir or orchestra. In "industry" terms it is more like what a mastering engineer would do (and I am really a beginner at this).
1 - normalize, just beeing lazy and to get things up to a known level. (I use Samplitude and the normalize is done virtually, no files are written, just parameters set).
2 - high-pass, removing everything below, say, 30 Hz. Only subways and buses down there anyway. Depending on the source, sometimes as high as 150 Hertz (girls choir). If I use cardioid mics these start falling of at something like 100-150 Hz anyway when used at a distance, one of my omni pairs goes down to 10Hz (or even a bit lower).
3 - if it improves things, very careful EQ, maybe 2 or 3 db down on offending room modes. Some rooms sound very "boxy" and a little eq can improve things quite a lot without beeing heard. Disturbing though if it is not EQ-d off.
4 - sometimes, a very slight compression. Often a factor of 1.1 or 1.2 (that is just a slight, slight touch). The idea is to "sweeten" things just a tiny bit, not changing the character, just making it easier to listen to. Not every source is improved by this though.
5 - adjust volumes of all songs to get good sequencing between them. Typically a request in order for the record to be playable at all in a car -- the client is always right.
6 - and finally, often a limiter set at -3dB. The volumes before this should be set so that it will only do anything in a small parts in a few of the songs. It could be a single drum hit or a single very percussive note somewhere. The effect is not to be heard, but a listener is often disturbed by the recording if the transients are left in. The exact value is depending on several things, but -3dB seems right most of the time. Mp3 coders as example struggle to get up to the last dB-s and may sound much better with a little headroom left. Some CD-players also struggle at the very top, best to leave them the room.
Sometimes i do a bit of noise removal, audience coughs, things dropped, perhaps a screaming baby. All to get it down to a level where it does not remove focus from the music performed. Another thing is on amateur choirs the sopranos can be a bit strong on resonance tones. A tad of multiband compressor can be used, all bands except the "soprano band" set to no compression. Has to be done with taste.
Sorry, a bit of rambling in a thread really about other things.
Gunnar