linuxaudio.org

--_527175eb-567d-49f5-a995-ffadb56ed1d9_

Content-Type: text/plain; charset="Windows-1252"

Content-Transfer-Encoding: quoted-printable
> Could you tell us more about the VibraChorus ? From what you've

quoted text
> told already it seems to use 'mechanical' variable capacitors...

And I did add that my knowledge of the vibrachorus was limited=2C the gearb=

ox itself was more fun.
The vibra had a single rotating armature with multiple stacked horizontal f=

ans that formed plate capacitors with 16 fixed plate segments=2C each fan c=

overs about 20 degrees of arc - would you call these 'plate capacitors'=2C =

no? As the rotating armature moved between the fixed plates/fans=2C whateve=

r=2C they fead more signal into it depending on its position. The 16 fixed =

segments were from an 8 stage delay line (including 0 delay) going up then =

coming back down through a single rotation - each of the 8 stages was a cas=

cade of RC delay lines so the larger delays had a darker/warmer sound. The =

output was roughly a vibrato due to scanning through consequetively larger =

phase shifted delays from the RC filters - the movement from one RC delay t=

o the next was quite smooth since the rotating fans would pick up signal fr=

om adjacent delay sections=2C gradually mixing them as it moved. Chorus was=

 generated by mixing this signal back in with the direct signal. I could no=

t find any information on the phase shift per section of the cascaded delay=

 line by which I mean the actual RC values did not seem to be documented an=

ywhere online.
The rotation speed was fixed as the arm was connected directly to the main =

gearbox spindle=2C the different speeds/depths that made up V3/V2/V1/0/C1/C=

2/C3 purely depended on which of the rotating fans was tapped to the output=

.
You can find a block diagram on wikipedia and if you search further you can=

 find a diagram of the build=2C they are quite enlightening as to how it wo=

rked but lacked detail of the RC circuits and naturally emulating its compl=

exities is left up to you. I have not looked at how Beatrix emulates this. =

A lot of synths implement a more simple multitap chorus which is a reasonab=

le extraction. Bristol uses an 8 stage delay line with a different filter c=

utoff on each delay to emphasise the phase shift and taps off up to 4 diffe=

rent positions depending on the C/V selection - it previously also used a c=

horus circuit which had very different characteristics. The current algorit=

hm is also not optimal=2C only the C3 position sounds half reasonable=2C V1=

/V2/V3 are weak. I think this algorithm=2C even with just C3=2C sounds bett=

er than a multitap chorus but then I would naturally have a preference.
Regards=2C Nick.
_________________________________________________________________

Windows Live=99: Keep your life in sync. Check it out!

http://windowslive.com/explore?ocid=3DTXT_TAGLM_WL_t1_allup_explore_0120...
--_527175eb-567d-49f5-a995-ffadb56ed1d9_

Content-Type: text/html; charset="Windows-1252"

Content-Transfer-Encoding: quoted-printable
.hmmessage P

{

margin:0px=3B

padding:0px

}

body.hmmessage

{

font-size: 10pt=3B

font-family:Verdana

}
&gt=3B Could you tell us more about the VibraChorus ? From what you've&=

gt=3B told already it seems to use 'mechanical' variable capacitors...<=

br>And I did add that my knowledge of the vibrachorus was limited=2C the ge=

arbox itself was more fun.The vibra had a single rotating armature =

with multiple stacked horizontal fans that formed plate capacitors with 16 =

fixed plate segments=2C each fan covers about 20 degrees of arc - would you=

 call these 'plate capacitors'=2C no? As the rotating armature moved betwee=

n the fixed plates/fans=2C whatever=2C they fead more signal into it depend=

ing on its position. The 16 fixed segments were from an 8 stage delay line =

(including 0 delay) going up then coming back down through a single rotatio=

n - each of the 8 stages was a cascade of RC delay lines so the larger dela=

ys had a darker/warmer sound. The output was roughly a vibrato due to scann=

ing through consequetively larger phase shifted delays from the RC filters =

- the movement from one RC delay to the next was quite smooth since the rot=

ating fans would pick up signal from adjacent delay sections=2C gradually m=

ixing them as it moved. Chorus was generated by mixing this signal back in =

with the direct signal. I could not find any information on the phase shift=

 per section of the cascaded delay line by which I mean the actual RC value=

s did not seem to be documented anywhere online.The rotation speed =

was fixed as the arm was connected directly to the main gearbox spindle=2C =

the different speeds/depths that made up V3/V2/V1/0/C1/C2/C3 purely depende=

d on which of the rotating fans was tapped to the output.You can fi=

nd a block diagram on wikipedia and if you search further you can find a di=

agram of the build=2C they are quite enlightening as to how it worked but l=

acked detail of the RC circuits and naturally emulating its complexities is=

 left up to you. I have not looked at how Beatrix emulates this. A lot of s=

ynths implement a more simple multitap chorus which is a reasonable extract=

ion. Bristol uses an 8 stage delay line with a different filter cutoff on e=

ach delay to emphasise the phase shift and taps off up to 4 different posit=

ions depending on the C/V selection - it previously also used a chorus circ=

uit which had very different characteristics. The current algorithm is also=

 not optimal=2C only the C3 position sounds half reasonable=2C V1/V2/V3 are=

 weak. I think this algorithm=2C even with just C3=2C sounds better than a =

multitap chorus but then I would naturally have a preference.Regard=

s=2C Nick.Windows Live=99: Keep your life in sync. Check it out!

=
--_527175eb-567d-49f5-a995-ffadb56ed1d9_--