View Single Post
Old 08-21-2018, 10:28 AM   #46
Jack Winter
Human being with feelings
 
Jack Winter's Avatar
 
Join Date: Aug 2007
Location: Luxembourg/Spain
Posts: 1,854
Default

Quote:
Originally Posted by karbomusic View Post
When I first ran into it around 2011, it was explained as not being able to tell "who was first out of the gate". There is some initial spin up of the stream or similar of each device that are unaware of each other's start position, then something like WC or ADAT simply sync that and it stays perfectly locked after that point, it just may not be sample-aligned perfectly.
That sounds similar to what I was thinking of in terms of being in sync, but slightly out of phase. Oh well I'm not really clever or qualified enough to comment in an authoritative way anyways I'll ask around to see if any of the smart people I know can give a satisfactory answer.

Or who knows, maybe there actually is a reason to buy a high end clock like a big ben or similar.

As a matter of interest with what hardware have you seen this? I'm gonna run some tests with my RME and ADAT converters when I get home to see what I can discover. I hate stuff like this...

I'll let you in on a small secret.. With USB on Linux (but I've heard of this from Windows users too), each time you connect to the soundcard you might get a slightly different latency. The difference is not great, maybe +- 20 samples. What is somewhat worse is that the bigger the buffersize the bigger the hidden latency is, and at 1024 samples it is really significant...

What is important when recording is to tell reaper to compensate for the big "hidden" latency so that newly recorded tracks are aligned with the playback tracks, but if you want to get anal about it, you'd better do a loopback test each time you start up reaper, as otherwise you might be +-20 samples off perfect alignment..

I am much more satisfied with my HDSP card as there the hidden latency is exactly 95 samples, no matter what the buffer size is. It's composed of 32 samples of hardware buffer in each direction and 15/16 samples for the conversion itself.

As a somewhat nerd, things like this gives me a stomach ache and I always wonder why so many things are so crappily implemented. It's like the designers say this is good enough, let's sell it, instead of going the extra mile to make sure that things actually work as one would expect..:S
__________________
Reaper for Linux Documentation (WIP). Software: Archlinux/KDE, Fabfilter FX, Komplete 8, Nebula, Schwa/Stillwell, T-racks Max/Amplitube/SVX, etc. Gear: i7-2600k/4700HQ/16GB, RME Multiface/Babyface, Behringer X32, Genelec 8040, etc. :)
Jack Winter is offline   Reply With Quote