Stereo Tool

in all atention to tuning ST .. i recently found some problems about buffers.
i still use Realtek Soundcard (till Juli@ comes) to transmit Stereo and RDS and testing (FM Transmitter output) but i notice that buffer after some time ~1 hour .. drops all down .. and than start very small sound droping.
Buffer size is ~0.50 found that most stable..
Input sound buffer is on 0.
CPU is ~70% (affinity for one core)
Sound comes from Virtual cable.

so, is it because Sound card or what ?

_________________
control point
control point2

Then apparently the sound card is playing things faster than VAC... (Say vac uses 44100 Hz and the sound card plays 44101 Hz, then it is falls one sample short every second. After an hour that would be 3600 samples... Of course the difference can be bigger.

For the next version I'll add a "minimum buffer filling" check, if it falls below that the output will stop until the buffer is full again. Not a perfect solution, but at least after this short drop it will start playing properly again. (Actually this check is already there, but it doesn't always work because the Windows audio output has a delay. Which means that I can send data to it before the buffer was really empty, but you still get a hiccup because of the delay in Windows.)

As a temporary solution, increase the buffer size so it takes a very long time before this occurs.

i use 192000, so if i set 192001 in VAC, maybe that wiil solve problem ?

edit:

no.. this did not help.

_________________
control point
control point2

I have made a new version that should work better:

Winamp DSP plugin: http://www.stereotool.com/download/dsp_ ... r_BETA.exe
Stand alone version: http://www.stereotool.com/download/ster ... r_BETA.exe
Command line: UNAVAILABLE for now
VST version: http://www.stereotool.com/download/vst_ ... l_BETA.dll

Instead of the current behavior (continuously stuttering sound when the buffer gets empty), this version stops the playback and waits until the buffer is full again when it gets below 10% of the buffer size. So you'll still get a hiccup, but afterwards it should continue without hiccups.

Note: I've also made a change to the audio processing - the volume should be more stable (loud highs have less effect on the rest of the sound) and the bass sounds are a bit louder again - but some loud highs MAY sound a bit less good (haven't heard it anywhere yet, but theoretically this could be the case...).

ok testing new beta last 3 hours .. and now this is strange. (in positive meaning)

Buffer still drops but now drops at ~20-25% .... and stays there.

i'll monitoring more ... but if this 100% true .. this is better then previous ver.... strange but better.

_________________
control point
control point2

Actually that doesn't make any sense at all.....

i know .. and that is why i said "strange"

start:


after 2 hours:

_________________
control point
control point2

Actually, that's rather close to the minimum buffer filling below which I'm halting the playback for a (short) while. Are you sure that there have been no hiccups? They should be less than 0.15 seconds, so you might not have noticed them!

Rigt now i don't here hiccups, but buffer stays there and no buffer restarting as you said.
Erlier buffer was half smaller.
now testing with more buffer ~0.50s .. i'll report in ~4horus

_________________
control point
control point2

Ok, I've now uploaded a new version that stops when the buffer contains far less than 10%. Also the sound quality should be slightly better.