Topic: BASS_FX 2.4.12.1

[workoutDJ]

I have copied the DLL files to both the BIN folder and the DEBUG folder.  Which is the best version of VB for me to code with:  VB6 or VB15 ?    Having a lot of problems attempting to get the libraries referenced and available.  Thanks

[Ian @ un4seen]

VB6 doesn't support multithreading well (eg. callbacks) so I would suggest using VB.Net instead. I'm not a .Net user myself, but I think you only need to add a reference to BASS.Net, not the BASS/add-on DLLs. What is the exact error message that you are seeing?

[workoutDJ]

I get:  'BASS_Mixer_StreamCreate' is not a member of 'Un4Seen.BassMix     I have all of the DLL files using the latest versions, in both the BIN and DEBUG folders.   Funny thing is that this was working about four days ago.  I am having trouble with the Mixer code.   Thanks.

[Ian @ un4seen]

Strange, I'm not really sure what could be causing that. Are you only having the problem with that one function, and none of the other BASS/BASSmix/BASS_FX functions? I notice in the code you posted that you omitted the BassMix and BassFx prefixes in some of the calls.

[workoutDJ]

I will prefix all my calls.   ALSO, I am going to create a new, fresh, environment of code.  Again, thanks for your help.

[teq]

Hi there! I'am trying to make pitch shifting effect like in "Music Speed ​​Changer" (Android) with BASS, but the result is not even so good as if this app. "Music Speed ​​Changer" uses Superpowered, may be this lib just better, than BASS built-in? Or I should tweak settings and it's 100% possible to achieve the same quality

[teq]

Hi there! Are you going to update BASS_FX to the latest version? 

[artistXenon]

may I please request for a x86_64 version compilation of BASS_FX on android?  It seems only arm64, arm, x86 are supported 😢

[Ian @ un4seen]

An update including an x86_64 build is available here:

   www.un4seen.com/files/z/0/bass_fx24-android-beta.zip

[fmcoder]

There appears to be a bug in the Peak EQ implementation in the 64-bit Windows version (does not happen on 32-bit). After running a while, sometimes a day, sometimes more, the right channel becomes silent. Does not happen always, but regular enough to be a problem.

The code that sets the DSP:

Code: [Select]

var   EQ: BASS_BFX_PEAKEQ;
...
        HFX := BASS_ChannelSetFX(Chan, BASS_FX_BFX_PEAKEQ, 49);
        with EQ do
        begin
          fBandwidth := 2.5;
          FQ := 0;
          fGain := 0;
          lChannel := BASS_BFX_CHANALL;
          for I := 0 to 10 - 1 do
          begin
            lBand := I;
            fCenter := PluginSettings.EqBands[I].CenterFreq; //values are in 31-16000Hz range
            BASS_FXSetParameters(HFX, @EQ);
          end;
        end;

And updating it the following way

Code: [Select]

  for I := 0 to 10 - 1 do
  begin
    EQ.lBand := I;
    BASS_FXGetParameters(HFX, @EQ);
    EQ.fGain := ...//actual gain value, -15...+15;
    BASS_FXSetParameters(HFX, @EQ);
  end;

[Ian @ un4seen]

Is it only ever the right channel that's affected, never the left? That seems strange if so, as the BASS_FX_BFX_PEAKEQ processing is identical on all channels. Is it definitely caused by the BASS_FX_BFX_PEAKEQ effect, ie. it never happens without that?

Does the problem begin while sound is playing or only after a period of silence? Does it go away if you call BASS_FXReset or BASS_ChannelSetPosition? If the stream is floating-point, please check the sample values when it's happening, eg. are they 0s or NaNs?

Also confirm that you're using the latest versions of BASS_FX (2.4.12.6) and BASS (2.4.16.7).

[fmcoder]

I use bass_fx 2.4.12.6 and bass 2.4.16.16.

The EQ is set on the mixer channel, and from what I've seen myself and user reports, it's always the left channel that disappears. The mixer channel is floating point and decoding. If the EQ effect is removed, the playback returns to normal. It never happens if the EQ was not applied. I'm not sure if there's silence but I suppose it happens regardless.

[Ian @ un4seen]

To confirm what the silence is (eg. 0s or NaNs), please set a WAV writer on the same stream as the EQ when the problem happens. For example, like this:

Code: [Select]

BASS_Encode_Start(stream, "rightsilence.wav", BASS_ENCODE_PCM | BASS_ENCODE_AUTOFREE, 0, 0);

And then upload the written file to have a look at here:

   ftp.un4seen.com/incoming/

[fmcoder]

OK, it'll take a while though as it only happens occasionally and after several days of operation.

[fmcoder]

As usual in such cases, the bug still not reproduced, but I have seen many reports from users so I'm sure it's there Ian, can you please check the EQ implementation code to see if there are any potential issues that could lead to such a behavior in 64-bits?

[Ian @ un4seen]

I can't see any reason in the BASS_FX code for the problem you describe, so unfortunately I'm not sure what it could be. The most puzzling thing is why only one channel would be affected. Is the "lChannel" parameter sometimes not set to BASS_BFX_CHANALL? Btw, I just noticed that you first wrote "the right channel becomes silent" and then later wrote "it's always the left channel that disappears". Does it vary?

To confirm, when you wrote "If the EQ effect is removed, the playback returns to normal", does that mean if BASS_ChannelRemoveFX is called on the EQ effect while the problem is happening then the problem immediately stops? If so, is it only a BASS_ChannelRemoveFX call or are other changes applied too? If there are other changes then please try disabling those to see if the BASS_ChannelRemoveFX call alone stops the problem.

Are you using BASS_ATTRIB_PAN at all?

[fmcoder]

Is the "lChannel" parameter sometimes not set to BASS_BFX_CHANALL? Btw, I just noticed that you first wrote "the right channel becomes silent" and then later wrote "it's always the left channel that disappears". Does it vary?

It's applied for all channels: https://www.un4seen.com/forum/?topic=2181.msg138328#msg138328
Only right channel disappears, left is never affected - I'm sorry for the mistake.

To confirm, when you wrote "If the EQ effect is removed, the playback returns to normal", does that mean if BASS_ChannelRemoveFX is called on the EQ effect while the problem is happening then the problem immediately stops? If so, is it only a BASS_ChannelRemoveFX call or are other changes applied too? If there are other changes then please try disabling those to see if the BASS_ChannelRemoveFX call alone stops the problem.

Yes, after calling BASS_ChannelRemoveFX to remove the peaking EQ, the problem disappears.

Are you using BASS_ATTRIB_PAN at all?

No, it's never used.

[AndyMK]

Hi, will we get a universal binary that supports apple silicon natively?

[Ian @ un4seen]

Support should already be included in the latest BASS_FX release (2.4.12.6). Is it not working when you try it?