Not exactly, because I whish to consider the sustain of an ending cymbal for example. It could work like that, for two songs chained :
* play backing track, automation ON (midi fx control essentially)
* pré final of the 1st song with stop on a cymbal
* the singer sing freely the ending lyrics, during that the audio is still playing and the automation at the end of the cymbal tell to the audio machine (Fostex D160 in my case) "fostex pause, fostex locate to this point (beginning of the next song), automation pause".
* the singer sing the very last syllabe of ending lyrics, and on this very last syllabe I restart automation and play audio located on the beginning of the 2nd song.
...and then my automation is still in time with the 2nd song independantly from the time taked by the singer to sing the end of the first song. By the way, one thing very important that I haven't mentionned : these 2 songs chained are considered as just one song in SLM.