Is BMax fast enough for realtime sound synthesis?

I've been playing around a bit with your code and it seems to work reasonably well like this:

SuperStrict

Const freq:Int=11025
Const updatespeed:Int=10
Const buffersize:Int=(freq/updatespeed) * 32
Print "Buffer Size: "+buffersize
Const numBuffers:Int = 2
Const numChannels:Int = 40
Const numSounds:Int = numChannels'20

Global my:Int ' mouse y position

Local iterator:Int
Local currentbuffer:Int
Local currentChannel:Int
Local t:Int

Graphics 640 , 480

Global buffer:TAudioSample[numBuffers]
	For iterator = 0 Until numBuffers
		buffer[iterator]=CreateAudioSample(buffersize,freq,SF_STEREO16BE)
	Next
	
Global channel:TChannel[numChannels]
	For iterator = 0 Until numChannels
		channel[iterator]=AllocChannel()
	Next
	

Global sound:TSound[numSounds]
	For iterator = 0 Until numSounds
		sound[iterator] = LoadSound(buffer[0])
	Next

Local timer:TTimer=CreateTimer(updatespeed)
calcnew buffer,currentbuffer

Repeat
	WaitEvent()
	
	If EventID() = EVENT_TIMERTICK
	
			currentChannel = getAvailableChannel()
			
			PlaySound sound[currentbuffer],channel[currentChannel]
			currentbuffer = currentbuffer Mod numBuffers
			Cls
			calcnew buffer,currentbuffer
			sound[currentChannel]=LoadSound(buffer[currentbuffer])
	
		

		SetColor 255,255,255
		For t=0 Until Min(buffersize,640)
			Plot t,240+buffer[currentbuffer].samples[t]
			'DrawLine t-1, 240+buffer[currentBuffer].samples[t-1], t,240+buffer[currentbuffer].samples[t]
		Next
		DrawText my,0,0
		Flip
	EndIf
	
	my=MouseY()
	
	If KeyDown(KEY_ESCAPE) End
Forever


Function calcnew(buffer:TAudioSample[] Var,index:Int)
	Local ratio:Int=8'Rnd(15)+1
	Cls
	Local a:Int
	For Local t:Int=0 Until buffersize
		a=Sin((t*ratio))*my
		buffer[index].samples[t]=a
		SetColor 255,128,64
		Plot t,240+a
	Next
End Function

Function getAvailableChannel:Int()
	For Local i:Int = 0 Until numChannels
		If Not ChannelPlaying(channel[i]) Then Return i
	Next
	Print "no channel available"
	Return 0
End Function

Part of the stuttering problem appears to be that the buffersize is incorrect. I discovered this with my sound generation code above, in that the produced sample played for about half the calculated buffer duration.

Also added channel switching, the numbers need tweaking, but it's going somewhere...