Assuming you’re using a 24bit capable interface then it’s advisable to set the project to 24bit also, it won’t really affect performance, obviously the disk file size is bigger. I personally notice a bigger difference between 16bit to 24bit than 44.1 to 48 or even 96!
IMHO,
The more dynamic the music, the more you may notice the diff.
I find that if I am recording a jazz combo or string ensemble
or a single guitar/piano or flute etc, I can really notice it.
But if I am recording a ‘beer drinkin’ rock band or a dope smokin’
reggae band or an ‘in yo face’ hiphop group etc, the difference
is not as noticeable.
Haha, its weird cause people used to discuss this endlessly a few years back but now it rarely gets a mention - I guess that’s because we’ve all got much faster machines and oodles of disk space so using 24 or 32 bits doesn’t really have an overhead anymore. But, yep, 24 or 32 bit is better quality than 16bit so I’d advise 24 or 32. I don’t think there’s much of a change between 24 and 32 though. I use 24 or 32 indiscriminately, and in fact sometimes I find I’m using 16
But one of the main advantage points of using 24 bits is that you can record at a low level thus completely avoiding digital peaks (which are very very bad), and yet you’ve still got enough bits to give you a really good quality sound. Whereas if you record at low levels in 16bit then you’re going to lose quality because you may well be only using 12 bits, which would start to become a little crunchy.
Playing around with a free bit reduction tool is a good way to find out what effects the bit size has. And, bit reduction is a really good tool too for crunchyfying all sorts of things, excellent fun!