I have a request from a customer to make an audio loop that plays a sound at the top of the hour for 12 hours, then silent for 12 hours. He did it in Audacity, and was asking me if I could do it and make it NOT drift.
Seems that every 24 hours after the loop, it will drift. I’m assuming that audacity may not be as precise as Wavelab. I have WL8.5. What are your suggestions for file format and/or compression for such a large file vs. keeping timing precise?
It’s a sound file that’s exactly 24 hours long? But it drifts after 24 hours because whatever’s playing it can’t loop back to the start quickly enough? What player? Can’t the delay be calculated and compensated by chopping secs or ms off the end of the file for that particular player? But how many days would that last and not drift? Wouldn’t you have to re-sync to a world master clock every day? Or do I have it all wrong?
(Can Wavelab make a 24 hour file?)
In theory yes, if you use the RF64 option. But I don’t test that everyday…
Concerning drifting, it’s more a matter of the player than the file. What player is used?
It would certainly be wiser to use a 12 hour files and have an application triggering the player everyday. In that way, triggering will start with the computer clock, that can be sync to an atomic clock. And you won’t drift.
I just made a 22 hour FLAC file 44.1 24 bit in Wavelab with 15 seconds of program every hour. It was 32MB, which is pretty great. I don’t think I’d try this with WAV because the file would be huge.
I started to open the FLAC in Wavelab and it estimated 10 minutes to open. Then I realized Wavelab would be converting to temp WAV (?) and stopped it after 5 minutes.
Ogg, MP3, AAC ? To try next. But I still don’t think I’d want to open any of those very often in Wavelab if huge WAV temp files are generated to open them. (?)
Interesting problem. As others have suggested, I’d think that timing accuracy is a function of the playback device’s clock rather than the file itself. A loop would progressively get more and more out of sync whereas a triggered file would at least reset itself each day. This link is one of many that explains how to make a computer sync to an internet-based time server. If playback was from a computer you could automatically trigger playback each day with a high degree of accuracy. You could even create 12 files and trigger them exactly at the top of the clock for each hour if you want to use wav files.
Or for a low-budget solution use a mobile phone for playback and load your files as the alarm sound files and set 12 alarms- I’m guessing they’re short files at the top of each hour. The phone would get its time from the nearest cell tower and might not even need a SIM card to get that data. I have found mobile phones to be be only accurate to within a couple of seconds (apps could improve this) and the audio quality might not be as good as you’re looking for.