Synth-DIY Yahoo! Groups Archives

Thread

Re: [Fairlight-CMI] Sound Character Comparion between 8bit (II/IIx) and 16bit...

2007-08-08 by NormLeete@aol.com

In a message dated 08/08/2007 06:40:18 GMT Standard Time, yuen_paul@yahoo.com writes:

Does it make a difference
to its sound, let say, if we playing the same sample?

Slightly off topic but here is a true story. It has always intrigued me that the Series IIx has such character to the sounds it produces so with this in mind I tried the following experiment(s).

Experiment 1.

Sample IIx into (borrowed) top grade 16bit sampler. Observation - single notes are identical but chords on the top end sampler are thin.

Experiment 2:

Sample Mellotron (I have one of those as well and have been involved in the development of the new M4000) into another sampler, every note (full 8 seconds) same result - single notes are identical but chords on the sampler are "thin",

Experiment 3:

Record each note of IIx for 8 seconds and get the guys at Streetly to transfer sounds to tape so they can be played by a Mellotron. Result - sounds that rival the original when played as chords. Streetly were so pleased with the results that they now offer a Fairlight frame for the Mellotron featuring ARR1, SWANEE and SAXY.

Conclusion: the character of the IIx comes not only from the 8 bit resolution but from the architecture that takes 8 monophonic voice cards and mixes their output in the analogue domain.

Probably this info is of no practical use, but it is an interesting illustration that specifications aren't everything you use your ears to decide. For that reason alone the new Streetly M4000 still uses tapes!

Regards,

Norm

Re: [Fairlight-CMI] Sound Character Comparion between 8bit (II/IIx) and 16bit...

2007-08-08 by Marcin 'Rambo' Roguski

ahhh, there we go- the superiority of Christmas over Easter :D

Re: Sound Character Comparion between 8bit (II/IIx) and 16bit...

2007-08-08 by yuen_paul

--- In Fairlight-CMI@yahoogroups.com, Marcin 'Rambo' Roguski 
<rambo@...> wrote:
>
> ahhh, there we go- the superiority of Christmas over Easter :D
>

Thanks for your interesting reply. :)  Guess I am not trying to find 
out which one is better per se, but just to understand their different 
characters.  III has more "modern" spec, but I->II/IIx kicked off the 
Fairlight "legacy".  Since members of this group represent probably the 
best group of current global fairlight users, I just want to gain 
better understanding.

That's all. :)

Paul

Re: Sound Character Comparion between 8bit (II/IIx) and 16bit...

2007-08-08 by Steve Rance

The main differance in sound quality when samples are not played at the 
native pitch is due to the design of the tracking filters used in all 
versions of the CMI. Other samplers in the era used different 
(cheaper!) methods to change the pitch of samples.

[Steve]

Re: [Fairlight-CMI] Re: Sound Character Comparion between 8bit (II/IIx) and 16bit...

2007-08-11 by Megacurve

Steve,

It isn't just the tracking filters. The CMI's used a
variable samplerate, with a huge range. There isn't
any digital sample interpolation happening. Was it 7
octaves of shift possible? I don't remember If you
upshifted the sample you can get harmonics that are
higher than in the original sample rates 1/2 of the
input filter cut off frequency. Much Higher.

A digital version of playing an audio tape faster or
slower.

Other samplers, like Roland, used a fixed sample rate,
with only a 2 octave shift range.  
The method of digital sample interpolation with a
fixed sample rate also makes a huge difference, and
will contribute a lot to the resultant sound, but in
all cases you never got any harmonics above the
original sample filter cutoff...

But, it was just a cooler bunch of people that made
it..

Adrian S. Bruce
www.artandtechnology.com.au

--- Steve Rance <steve@fairlightus.com> wrote:

> The main differance in sound quality when samples
> are not played at the 
> native pitch is due to the design of the tracking
> filters used in all 
> versions of the CMI. Other samplers in the era used
> different 
> (cheaper!) methods to change the pitch of samples.
> 
> [Steve]
> 
> 
> 
> 

____________________________________________________________________________________
Sick of deleting your inbox? Yahoo!7 Mail has free unlimited storage.
http://au.docs.yahoo.com/mail/unlimitedstorage.html

RE: [Fairlight-CMI] Re: Sound Character Comparion between 8bit (II/IIx) and 16bit...

2007-08-11 by Peter Vogel

Yes, Adrian's got a good point there. The variable sampling is generated by
"rate multiplier" chips which get the inbetween frequencies by stickingin
extra samples in strange places so that on average the freqnacy is exatc but
it jitters all ovber the place.  This is the cause of what we called "the
birdies" which are apparently random notes that appear if you hit a
frequency that results in artifacts in the audibe range.  
 
Peter


  _____

Show quoted textHide quoted text

From: Megacurve [mailto:megacurve@yahoo.com.au] 
Sent: Saturday, 11 August 2007 5:15 PM
To: Fairlight-CMI@yahoogroups.com
Subject: Re: [Fairlight-CMI] Re: Sound Character Comparion between 8bit
(II/IIx) and 16bit...

Steve,

It isn't just the tracking filters. The CMI's used a
variable samplerate, with a huge range. There isn't
any digital sample interpolation happening. Was it 7
octaves of shift possible? I don't remember If you
upshifted the sample you can get harmonics that are
higher than in the original sample rates 1/2 of the
input filter cut off frequency. Much Higher.

A digital version of playing an audio tape faster or
slower.

Other samplers, like Roland, used a fixed sample rate,
with only a 2 octave shift range. 
The method of digital sample interpolation with a
fixed sample rate also makes a huge difference, and
will contribute a lot to the resultant sound, but in
all cases you never got any harmonics above the
original sample filter cutoff...

But, it was just a cooler bunch of people that made
it..

Adrian S. Bruce
www.artandtechnology.com.au

--- Steve Rance <steve@fairlightus. <mailto:steve%40fairlightus.com> com>
wrote:

> The main differance in sound quality when samples
> are not played at the 
> native pitch is due to the design of the tracking
> filters used in all 
> versions of the CMI. Other samplers in the era used
> different 
> (cheaper!) methods to change the pitch of samples.
> 
> [Steve]
> 
> 
> 
> 

__________________________________________________________
Sick of deleting your inbox? Yahoo!7 Mail has free unlimited storage.
http://au.docs. <http://au.docs.yahoo.com/mail/unlimitedstorage.html>
yahoo.com/mail/unlimitedstorage.html

RE: [Fairlight-CMI] Re: Sound Character Comparion between 8bit (II/IIx) and 16bit...

2007-08-11 by Steve Rance

Oh yeah, I remember the B Flat birdies on a series I/II/IIX - The series III
was never the same!

[Steve] 

 

  _____

Show quoted textHide quoted text

From: Fairlight-CMI@yahoogroups.com [mailto:Fairlight-CMI@yahoogroups.com]
On Behalf Of Peter Vogel
Sent: Saturday, August 11, 2007 17:49
To: Fairlight-CMI@yahoogroups.com
Subject: RE: [Fairlight-CMI] Re: Sound Character Comparion between 8bit
(II/IIx) and 16bit...

Yes, Adrian's got a good point there. The variable sampling is generated by
"rate multiplier" chips which get the inbetween frequencies by stickingin
extra samples in strange places so that on average the freqnacy is exatc but
it jitters all ovber the place.  This is the cause of what we called "the
birdies" which are apparently random notes that appear if you hit a
frequency that results in artifacts in the audibe range.  

Peter

  _____  

From: Megacurve [mailto:megacurve@yahoo.com.au] 
Sent: Saturday, 11 August 2007 5:15 PM
To: Fairlight-CMI@yahoogroups.com
Subject: Re: [Fairlight-CMI] Re: Sound Character Comparion between 8bit
(II/IIx) and 16bit...

Steve,

It isn't just the tracking filters. The CMI's used a
variable samplerate, with a huge range. There isn't
any digital sample interpolation happening. Was it 7
octaves of shift possible? I don't remember If you
upshifted the sample you can get harmonics that are
higher than in the original sample rates 1/2 of the
input filter cut off frequency. Much Higher.

A digital version of playing an audio tape faster or
slower.

Other samplers, like Roland, used a fixed sample rate,
with only a 2 octave shift range. 
The method of digital sample interpolation with a
fixed sample rate also makes a huge difference, and
will contribute a lot to the resultant sound, but in
all cases you never got any harmonics above the
original sample filter cutoff...

But, it was just a cooler bunch of people that made
it..

Adrian S. Bruce
www.artandtechnology.com.au

--- Steve Rance <steve@fairlightus. <mailto:steve%40fairlightus.com> com>
wrote:

> The main differance in sound quality when samples
> are not played at the 
> native pitch is due to the design of the tracking
> filters used in all 
> versions of the CMI. Other samplers in the era used
> different 
> (cheaper!) methods to change the pitch of samples.
> 
> [Steve]
> 
> 
> 
> 

__________________________________________________________
Sick of deleting your inbox? Yahoo!7 Mail has free unlimited storage.
http://au.docs. <http://au.docs.yahoo.com/mail/unlimitedstorage.html>
yahoo.com/mail/unlimitedstorage.html

RE: [Fairlight-CMI] Re: Sound Character Comparion between 8bit (II/IIx) and 16bit...

2007-08-12 by Megacurve

There were several differences between the I/II/III in
the output filter and envelope side.

A Series I had a "switched capacitor" lowpass filter,
and that contibuted to the B flat Birdies. This
machine also had linear envelopes.  

The Series II cards had resonant VCF chips and a dbx
VCA and exponential (i.e. 'normal') envelopes.

But the way the envelopes worked meant that you always
had more than '8 bits', as the volume control stuff
was an 'additional' 8 bits scaling the sound.  

Other Sampler makers didn't do that, so the sound got
rougher as the envelope made it fade away.

The III had pretty much the same as the III, or the
stuff I was involved with anyway, except the sound Dac
was now 16bits, and the sample rate was way higher, as
there was just so much more memory. And as much as
could fit on the cards, tried to make it low
distortion and noise (not with standing no 'dac cliche
S/H') as we could. 
More Hi-Fi always seems to mean it lacks
'musicality'...  

Would anyone guess to say that the Series I 'sound'
made more ground "breaking music" than the later
machines? 

And that is probably all I'll have to contribute here
again for another 2 years... 8^)

Adrian S. Bruce
www.artandtechnology.com.au 

--- Steve Rance <steve@fairlightus.com> wrote:

> Oh yeah, I remember the B Flat birdies on a series
> I/II/IIX - The series III
> was never the same!
> 
> [Steve] 
> 
>  
> 
>   _____  
> 
> From: Fairlight-CMI@yahoogroups.com
> [mailto:Fairlight-CMI@yahoogroups.com]
> On Behalf Of Peter Vogel
> Sent: Saturday, August 11, 2007 17:49
> To: Fairlight-CMI@yahoogroups.com
> Subject: RE: [Fairlight-CMI] Re: Sound Character
> Comparion between 8bit
> (II/IIx) and 16bit...
> 
>  
> 
> Yes, Adrian's got a good point there. The variable
> sampling is generated by
> "rate multiplier" chips which get the inbetween
> frequencies by stickingin
> extra samples in strange places so that on average
> the freqnacy is exatc but
> it jitters all ovber the place.  This is the cause
> of what we called "the
> birdies" which are apparently random notes that
> appear if you hit a
> frequency that results in artifacts in the audibe
> range.  
> 
>  
> 
> Peter
> 
>  
> 
> 
>   _____  
> 
> 
> From: Megacurve [mailto:megacurve@yahoo.com.au] 
> Sent: Saturday, 11 August 2007 5:15 PM
> To: Fairlight-CMI@yahoogroups.com
> Subject: Re: [Fairlight-CMI] Re: Sound Character
> Comparion between 8bit
> (II/IIx) and 16bit...
> 
> Steve,
> 
> It isn't just the tracking filters. The CMI's used a
> variable samplerate, with a huge range. There isn't
> any digital sample interpolation happening. Was it 7
> octaves of shift possible? I don't remember If you
> upshifted the sample you can get harmonics that are
> higher than in the original sample rates 1/2 of the
> input filter cut off frequency. Much Higher.
> 
> A digital version of playing an audio tape faster or
> slower.
> 
> Other samplers, like Roland, used a fixed sample
> rate,
> with only a 2 octave shift range. 
> The method of digital sample interpolation with a
> fixed sample rate also makes a huge difference, and
> will contribute a lot to the resultant sound, but in
> all cases you never got any harmonics above the
> original sample filter cutoff...
> 
> But, it was just a cooler bunch of people that made
> it..
> 
> Adrian S. Bruce
> www.artandtechnology.com.au
> 
> --- Steve Rance <steve@fairlightus.
> <mailto:steve%40fairlightus.com> com>
> wrote:
> 
> > The main differance in sound quality when samples
> > are not played at the 
> > native pitch is due to the design of the tracking
> > filters used in all 
> > versions of the CMI. Other samplers in the era
> used
> > different 
> > (cheaper!) methods to change the pitch of samples.
> > 
> > [Steve]
> > 
> > 
> > 
> > 
> 
>
__________________________________________________________
> Sick of deleting your inbox? Yahoo!7 Mail has free
> unlimited storage.
> http://au.docs.
>
<http://au.docs.yahoo.com/mail/unlimitedstorage.html>
> yahoo.com/mail/unlimitedstorage.html
> 
>  
> 
> 

____________________________________________________________________________________
Sick of deleting your inbox? Yahoo!7 Mail has free unlimited storage.
http://au.docs.yahoo.com/mail/unlimitedstorage.html