Hello everyone!
We're a team of 15+ users and our SFX are stored on a server. Using CodeMeter for everybody and Ultra for the soundbank keeper (that's me). Everything worked fine for years, but we've recently added ~300 000 new sfx, around 4Tb. I have imported all of our 150+ new collections, created some groups and everything turned out clean and tidy.
On my end, the new sounds took just a bit more time than the older ones to load, which was ok since the waveforms were not generated yet. For some users, they were not playing at all. Some sounds might play, others don't, the other one plays after several seconds. We've compared our options and everything was pretty similar and we've never pinpointed what was causing this issue with our new sounds. We've kinda found a workaround for the time being though. Everybody copied my "User.config" and everything seemed good, but the new sfx were still a tad slow on their end compared to the ancient sfx. So I began the titanic task of generating the waveforms for the 300 000 new sounds, which should take around 200 hours non-stop if my calculations are correct (good thing I'm in vacation next week, I'll let this run).
Just note that for security reasons, only I have the writing rights on the sfx folders in order to avoid unintentional sfx deletion.
I have read the FAQ on how to speed up BaseHead, but I was wondering if the community had other advices on how to gain some speed in a setup like this. The search is pretty fast for everybody, but we'd really like the file reading to be almost instantaneous.
Also, is BaseHead looking for an available waveform cache even if "Use Waveform Caching" is unticked? Or should it be ticked even if the user cannot write in these folders in order to read them? Or should I have generated all of the waveform cache locally and then shared with everybody in order to be even faster?
Thanks a bunch! 🙂
V
Would probably need to know what "A tad slow" actually is compared to your older files but the suspicion is either the file sizes are much bigger or the fact BH needs to generate that waveform cache file is causing the delay.
A test would be if you click a file and it is slow to load then you go and click a few other files and then go back to the first, is it faster the second time (or even after you restart BH)? If it is only slow the very first time you access it then this will be due to the WF cache file being generated and once that is done it is much faster. That would be my first guess since it takes time to draw the waveform and hence the reason a cache option is available to speed things up.
Also, is BaseHead looking for an available waveform cache even if "Use Waveform Caching" is unticked?
No, if the option is unchecked BaseHead will generate it on-the-fly.
If you do not mind the sub folders in your archive, just have one machine generate the WF files and these should be loaded when others access them, otherwise if you want a local cache folder you will need to then copy that to every machine.
If the issue is more related to large files causing the delay this can likely be improved in the latest BH version by checking this option in the main options page:
This is designed for remote cloud files but can also improve network server based files also. It disabled the need for BH to download the whole file first before playing (which is how it usually works) and instead streams the file and starts playing the moment enough data is downloaded but a side affect is that it is not possible to then use the "Reverse Playback" feature as the file is not available locally for that to work.
Hey Jason,
I'll run some speed test after the waveforms are all generated. I'll also test that new feature, looks like the thing we need!
Thanks a bunch for the reply, really help 😀
