Mofleaker on 22/12/2014 at 12:18
I'm having trouble creating a seamless audio loop in Audacity and converting it properly. Following (
http://www.wearytaffer.com/tutorials/tut_loops.html) this tutorial by Yandros, I was actually capable of making a working seamless loop in the program itself.
Problem is, once the project is compiled down into Thief's pitiful audio codec, the loop is totally defiled and there's a huge noticable blip on every iteration. Can anyone tell me how they keep fragile loops unscathed by all the compilation? I'm running out of tufts of hair to rip from my scalp.
EDIT: Looking through some of my installed fan missions, most if not all custom sounds are Ogg Vorbis files! From what I'd previously read, custom schemas require "
IMA ADPCM 22Khz 4 bit Mono format" to work properly... is this actually the case? If you say no I might kiss you.
gigagooga on 22/12/2014 at 14:07
The "tap" sound between loops is most likely because of this:
[ATTACH=CONFIG]2070[/ATTACH]
If the sine cuts out at different point than where it begins, it will make that sound.
At the green line on that picture you can see both sines at right about 0, so there would be better spot to cut, or from that spot onward to the end make the fade out effect.
If the fade out is shorter that 0,004 seconds it might not make the tap go away completely, so from 0,005 to 0,01 might work the best, depending a bit from what kind of sound you have.
Quote Posted by Mofleaker
EDIT: Looking through some of my installed fan missions, most if not all custom sounds are Ogg Vorbis files! From what I'd previously read, custom schemas require "
IMA ADPCM 22Khz 4 bit Mono format" to work properly... is this actually the case?
I doubt that's your problem. The Ogg Vorbis format is just compressed format which works too.
Yandros on 22/12/2014 at 14:53
You don't have to use IMA ADPCM any more, but I still do so that's not the problem here. You can use OGG or uncompressed PCM WAV too if you want to. MP3 is not recommended any more, use IMA ADPCM or OGG if you want compression.
You say the loop works fine when just exported from Audacity as a PCM WAV, right? I would suspect your conversion utility of introducing that issue then. I still mostly use the old Windows XP version of Sound Recorder, which you can get here: (
http://www.wearytaffer.com/storage/SoundRecorder.exe)
It has no batch functionality, but if you're just working with a few at a time it's not bad. I recommend saving profiles for IMA ADPCM 44kHz mono and stereo to access quickly.
Speaking of those, you should know when to keep stereo and when to save as mono. Stereo is only needed for music which will be used as an environmental ambient (or menu music). Anything else - voice lines, sound effects or even music played from a victrola - should be mono, which makes them half the size. That's because in the game they are positional, i.e. broadcast from a single location like an AI, a machine, a victrola etc., as opposed to being environmental.
Finally, for the record, that link goes to the tutorial on the old version of my website, and may not work in the future. The new link is (
http://www.wearytaffer.com/?tul=tut-loops). The content is the same, I just redid the site.
Mofleaker on 22/12/2014 at 15:17
EDIT: Oh! You posted while I was typing this out, Yandros. And you also happened to answer every question I had in the process, too! Thank you both.
Quote Posted by gigagooga
If the sine cuts out at different point than where it begins, it will make that sound.
I know that this is what is causing the blip, the problem is that I seem to have no control over it. Before I compile the sines at the beginning and the end are in synch, but AFTER I compile they are in disarray no matter what I do.
Quote Posted by gigagooga
If the fade out is shorter that 0,004 seconds it might not make the tap go away completely, so from 0,005 to 0,01 might work the best, depending a bit from what kind of sound you have.
Just tried this as you recommended, and it seems to have eliminated the sharp click between loops. However it still seems to have a bit of a noticeable discontinuity. I'll mess around with it a bit more.
Quote Posted by gigagooga
I doubt that's your problem. The Ogg Vorbis format is just compressed format which works too.
I tried exporting as an Ogg Vorbis file and the loop is working perfectly.
Can I simply use Ogg Vorbis instead? If so, is there a special quality (Hz / kbps) I have to save them to?
Thanks for your detailed reply :)
Yandros on 23/12/2014 at 00:59
Glad to help. I've never used OGG so I'm not sure what settings are best.
