Re-playing a stopped jingle, reproduce a part of audio before playing

[Couin]

Hi,

An user (the same of the other topic I just opened), reported me that if she stop a jingle before its end, when she play again this jingle, she can ear (I ear it too, on its recording) a part of audio, around 100 ms (the suit of where she stopped the first playback) before the beginning of the jingle.

In my "Play" function code, I even add a Stop even before recalling the start position (customizable).

Code Select Expand

Call BASS_ChannelStop(chan)
Call BASS_ChannelSetPosition(chan, position, BASS_POS_BYTE)
Call BASS_ChannelSlideAttribute(chan, BASS_ATTRIB_VOL, 1 10)
Call BASS_ChannelPlay(chan, BASSFALSE)
In the "Stop" function code, I use:

Code Select Expand

Call BASS_ChannelSlideAttribute(chan, BASS_ATTRIB_VOL Or BASS_SLIDE_LOG, -1, FadeOutDuration)
So the playback should stop at the end of FadeOutDuration (I should not have to add the Stop event befor SetPosition in the "Play" function).

I can not reproduce its problem at all.

Are my procedures correct?

Is there a way to "purge" any resting data of "chan" or something else, before playing it?

Thanks

[Ian @ un4seen]

That code looks fine. The BASS_ChannelSetPosition call should clear the channel's playback buffer, so that old data isn't heard. What output device is the user using? Perhaps there's a buffer in that that isn't being fully drained before stopping. You could try enabling non-stop output (before BASS_Init) and see if that helps:

Code Select Expand

BASS_SetConfig(BASS_CONFIG_DEV_NONSTOP, 1)

If you aren't already using the latest BASS.DLL build, please also try that to see if it makes any difference:

    www.un4seen.com/stuff/bass.zip

[Couin]

Hi Ian

I think there is something weird with its setup, because I can't reproduce its problem at all.
She uses virtual audio cables, but I don't know which one, I will ask her.

About the DLL, is it a stable version ?  Version number v32.65.18.15 instead of 2.4.17 of the download section, is it normal ?

[Ian @ un4seen]

I think there is something weird with its setup, because I can't reproduce its problem at all.
She uses virtual audio cables, but I don't know which one, I will ask her.

OK, I guess that's where the old data is probably buffered. Please check the BASS_INFO "latency" value with BASS_GetInfo.

About the DLL, is it a stable version ?  Version number v32.65.18.15 instead of 2.4.17 of the download section, is it normal ?

That's a strange version number. Where are you seeing that? You can use BASS_GetVersion to check what version of BASS.DLL is loaded by your app.

[Couin]

Hi Ian

I will check to add BASS_INFO latency return so she could report me the value.

About BASS version, I get it with this code:

Code Select Expand

        Dim GetBassVer As Variant
        Dim BassVer As Variant
       
        GetBassVer = Format(Hex(BASS_GetVersion), "00000000")
        BassVer = ""
        For i = 1 To 7 Step 2
            BassVer = BassVer & "." & val("&H" & Mid$(GetBassVer, i, 2))
        Next i
        BassVer = Mid(BassVer, 2)It gaves me the good version number (for example 2.4.17 for dll version in the downloads section), but with the last dll of "stuff" directory, it returns the value I reported above.

If I try (with some VB6 conversion) the first method of your message https://www.un4seen.com/forum/?msg=120623, so

Code Select Expand

Debug.Print CStr(HiByte(HiWord(BASSVERSION))) & "." & CStr(LoByte(HiWord(BASSVERSION))) & "." & CStr(HiByte(LoWord(BASSVERSION))) & "." & CStr(LoByte(LoWord(BASSVERSION)))
I get 0.0.2.4 instead of 2.4.17 and same (0.0.2.4) with stuff DLL.

But with this easier way

Code Select Expand

    ' Requires reference to Microsoft Scripting Runtime
    Dim fso As FileSystemObject
    Set fso = New FileSystemObject
    Debug.Print fso.GetFileVersion("bass.dll")
    Set fso = NothingI get better result (2.4.17.0 for wbesite's DLL, 2.4.17.47 for "stuff" DLL).

[Couin]

Hi,

Her latency returns 25 ms (I get the same value).

So, after several messages exchanges, we finally spotted the problem. It occurs on jingles for which we appy some reverb. I can reproduce the problem.

Of course, I'm trying to reproduce the problem in a small test project and I can't for the moment, but this spotted problem reveal another question (I will open a new thread for this).

Edit : I get no problem on versions before implenting mixers/splitters.

Edit 2: Adding BASS_FXReset before re-play the jingle, looks remove the "not played" reverb datas.