I've read on here that it's fine to record at 44.1k because it's going to be rendered down to that anyway. With some A B testing I've found that 96khz absolutely sounds far better than 44.1k for guitars using my hardware. I recorded tests and exported out to 44.1 .wavs and listened back. There's an obvious difference in texture, the 96k track is much more together and smooth, everything seems to be more cohesive across the frequency spectrum. It's just more pleasant and realistic, more like my cab sounds in person. I'm going 100% 96k for now unless it causes problems somewhere. It could be that I'm using cheap hardware and it does a better job of converting at that sample rate or maybe there's less jitter. I'm using the Presonus audiobox 22vsl. The 22VSL uses the CPU clock. Is this typical? What can I expect when I upgrade to RME? The RME fireface UC has a clock I/O but I don't understand how it works. Is there a clock built in or does it just support an external clock? What advantages will there be for someone like me who uses only software monitoring at around 3ms latency? I'm trying to google all of this and learn but I'm finding interpreting what's in front me me difficult.