Soundin works fine on my PPC running 3.55.

However, I was able to duplicate the reported follow bug using both 
soundin and by generating a trivial signal within the instrument itself. 
In the following example culled from the manual,   
  1. a signal that ramped up worked fine
  2. a signal that ramped down was never followed down
  3. a signal that went up a bit, stayed, up some more, and then down
     was followed fine while ramping up and holding, but couldn't be 
     followed back down

I hope this is of some help tracking down the problem!

--Dan Hosken

;follow-test.orc

sr = 44100
kr = 4410
ksmps = 10
nchnls = 1

instr 1

;k1	line 	0, p3, 30000
;ramp up to max
;k1	line 	30000, p3, 0 	
; ramp down from max
k1	linseg	0, .25*p3, 10000, .25*p3, 10000, .25*p3, 30000, .25*p3, 0
; ramp up, stay, up to max, ramp down

a1 	oscil 	k1, 1000, 1 	; Make a simple signal using k1
ak1	follow 	a1, .02 		  ; ak1 is now like k1
a2 	oscil 	ak1, 1000, 1 	; Make a simple signal using ak1
	out 	a2 				            ; Both a1 and a2 are the same

;    display     ak1, 0.5
    display     k1, 0.5
	
	endin

;follow-test.sco

f1	0	512	10	1	;sine wave
i1	0	10

e

>A closer look at the reported bug in follow seems to sugest to me that
>follow is behaving exactly as I expect.  However the soundin is
>generating silly input.  Taking an orchestra like
>
>instr 1
>  ifrq      =           cpspch(p5)
>  ain       soundin     "lemur.aiff" 		;input soundfile
>            out         ain
>            display     ain, 0.5
>endin
>
>displays rubbish which could be byte order, but I get junk with both
>AIFF and WAV files.  I am totally at a loss to know what I have broken
>or when.  Anyone with tehsources understand this stuff?  Does anyone