issue16

EMUSIC-L Digest                                      Volume 58, Issue 16

This issue's topics:
	
	Roland E86
	ROLAND SCC-1
	Roland Sound Canvas (5 messages)

Your EMUSIC-L Digest moderator is Joe McMahon .
You may subscribe to EMUSIC-L by sending mail to listserv@american.edu with 
the line "SUB EMUSIC-L your name" as the text.
 
The EMUSIC-L archive is a service of SunSite (sunsite.unc.edu) at the 
University of North Carolina.
------------------------------------------------------------------------
Date:         Mon, 1 Nov 1993 10:05:50 +0200
From:         Sam Safran 
Subject:      Roland E86

Greetings:

I have recently bought a Roland E86, after having owned several previous E
models (those with auto accompaniment styles).  I am interested in an E86
users group if it exists, or in getting and eventually sharing files with
voice patches, original and modified styles etc.  I can be reached at the
internet address listed above.

Sam Safran

------------------------------
Date:         Tue, 2 Nov 1993 19:01:54 GMT
From:         "Steven D. Bramson" 
Subject:      Re: ROLAND SCC-1

Shawn writes:

...So here I am cash in hand
about ready to finally put that sound canvas card
in my pc when I open up the October issue of electronic
musician. In their new product spot they have a card by Media Vision
(PAS16xl I think) that is their sound acard with a Korg
chipset onboard(OW/3 I thunk) ... NOW I HAVE A CHOICE?
well has anyone heard it ? I want to compare sounds...I was getting the SCC-1
bundled with some medium generic software but the PAS comes with
TRAX and some other Win 3.1 stuff. At about the same price! If anyone's
heard anything (SOUNDS) especially if you've heard both. let me know...

------------------

I have a sound canvas CM-300 which has exactly the same sounds as the SCC1.  To
 my ears they are a
bit variable.  Some are rather crude (eg various woodwind) and others are very
 good (eg the
harpsichord).  It is not a sythesizer so you don't have much control over the
 sound other than stereo pan,
chorus, reverb etc.  It is great for playing MIDI song files that you can get
 from internet and from
various companies.

Just to add confusion, Future Music Nov 93 has a news item about a Kurzweil
 K2000 in a box at 399 UK
pounds.  It is called the Summit K2K and has 300 sounds, 6Mb of wave ROM, 32
 note polyphony and 48
effects.  It is GM compatible.  It says you need an existing sound blaster style
 card to plug into (where?,
what other cards?).  I have sent off for details.

Steven Bramson

sdb@jet.uk

------------------------------
Date:         Sun, 21 Nov 1993 10:43:51 GMT
From:         Safran Shmuel 
Subject:      Roland Sound Canvas

Can anyone help clarify the specs on the samples used in the Sound Canvas
family of modules and synths.  Roland advertises "newly developed sound source"
and calls them CD quality.  What does that mean?

    --- Are they 16 bit ?
    --- What is the sampling frequency?
    --- Are any of the sounds multisampled?
    --- How do they compare, say to the U20/U220 quality?


This may be helpful information for many people.

------------------------------
Date:         Tue, 23 Nov 1993 16:05:09 +0200
From:         bpsafran@WEIZMANN.WEIZMANN.AC.IL
Subject:      Re: Roland Sound Canvas

Thanks for your detailed reply.  I am afraid I do not follow all the
technicalities, but I take it that although it does come out at 10 bits or
so per sample, it is not linear so it may be equivalent to 16 bit sampling.


In fact, I recently bought a Roland E86.  I belive that that sounds are
essentially sound canvas sounds (there is NO technical info in the manual
or from the dealer, so I had to rely on my ears); you can edit the sounds
as far as attack, sustrain, release, vibrato (rate, depth, delay) and
filter and resonance.  They sound OK but the pianos are not quite up to the
U220 in my opinion (my previous instrument was a Roland KR500 which had the
sound machine of the U20).  The reason I went with the E86 was because it
has a fantastic pattern programming capability.  Not only does it have
"auto accompaniment" which is very realistic, one can program
accompaniments and each style can have up to 36 patterns which change with
major, minor, 7th chording.  Also it has a built in disk drive which plays
and stores standard midi files.

Anyway, thanks for the info.  If I find out any more I will let you know.
If you want any info on the E86, drop me a line.

Sam Safran
Dept.of Materials and Interfaces
Weizmann Institute
Rehovot, ISRAEL 76100
Fax: 972-8-343362 or 972-8-344138
Phone:  972-8-343362
Internet: BPSAFRAN@WEIZMANN.WEIZMANN.AC.IL

------------------------------
Date:         Tue, 23 Nov 1993 14:48:07 CET
From:         Nicola Larosa 
Subject:      Re: Roland Sound Canvas

>Can anyone help clarify the specs on the samples used in the
>Sound Canvas family of modules and synths.  Roland advertises
>"newly developed sound source" and calls them CD quality.
>What does that mean?
>
>    --- Are they 16 bit ?
>    --- What is the sampling frequency?
>    --- Are any of the sounds multisampled?
>    --- How do they compare, say to the U20/U220 quality?

  Well, we had a close look at the patent, and ... (since it's
patented, it's public info, right?  Then it's up to you how you
use it - and up to them how they sue you :-) ) we found that the
samples are memorized in a modified kind of delta-modulation
coding system.

  That is, each group of three samples are memorized in four bytes,
where the first three are the differences from the previous sample,
and the fourth one contains some shifts for the previous three.

  In the end, it amounts to 32 bits for three samples, i. e. about
10.3 bits per sample.  :-) Of course, since it is not linear PCM,
this is almost meaningless for comparison.  Nonetheless, the sonic
result is, I should say, remarkable.

  The sampling frequency should be 32 KHz, but I'm not quite sure.

  All the sounds are multisampled, even the ones that don't seem to
be.  This storing system cannot encode a range greater than a few
octaves.

  Switching from objective to subjective first-hand evaluation, :-)
the sounds are clearly derivative of the same old Roland sound
bank, but many loops seem shorter.
  Nonetheless, the more powerful voice structure, comprehensive of
a good second-order filter with resonance, allows much more
versatile programming and, above all, much more modulation (i.e.
real-time) control.
  Just have a look at the sysex: how I wish that many hi-end
synths would have a matrix modulation which includes the
pitch-bend as a source capable of addressing any destination!

Another nicety: try setting a channel to monophonic and switch
portamento on, keeping its speed at maximum: if you play legato
notes (better from a MIDI guitar or a Yamaha WX7 MIDI sax), you
will hear a single, non-retriggering note that changes the sample
while you go!  Try that on your Wavestation!  (I tried, and was
bitterly disappointed).

Finally, be careful with Paolo Pizzi: sometimes he forgets he's
such a nice guy ... :-) Ciao, Paolo!

Nicola Larosa - mc3211@mclink.it

------------------------------
Date:         Wed, 24 Nov 1993 14:53:50 EST
From:         Brian Good 
Subject:      Re: Roland Sound Canvas

Nicola Larosa writes:

> Another nicety: try setting a channel to monophonic and switch
> portamento on, keeping its speed at maximum: if you play legato
> notes (better from a MIDI guitar or a Yamaha WX7 MIDI sax), you
> will hear a single, non-retriggering note that changes the sample
> while you go!  Try that on your Wavestation!  (I tried, and was
> bitterly disappointed).

I think you're confusing two separate issues here.  The Wavestation
does, in fact, have a usable unison-legato mode, which allows you
to do slurs using a wind controller or whatever.  As long as you don't
re-articulate, you can move from note to note without retriggering.
What you *can't* do is a lot of sliding around from note to note using
portamento.  How important this is depends, I suppose, on what kind of
style you're playing in.

brian good

------------------------------
Date:         Wed, 24 Nov 1993 23:16:19 +0100
From:         Hmeljak Dimitrij 
Subject:      Re: Roland Sound Canvas

brian good writes:

>I think you're confusing two separate issues here.  The Wavestation
>does, in fact, have a usable unison-legato mode, which allows you
>to do slurs using a wind controller or whatever.  As long as you don't
>re-articulate, you can move from note to note without retriggering.

Tell me more about this, please.
I'm always on my voice-to-MIDI project and I send Note On first, then
Note Off for the previous note. What I try to do is a non-restarting-note
when I sing legato. I know MIDI is a keyboard-oriented protocol
(as dr.Metlay wrote in man-machine-interface...) but, there is no other
way to control a synth these days (...ok, I did'nt want to start another
discussion...). Now. I use a K1000PX as output. I have some TX81, Rolands
D-20 or something, and an Akai S900. But I find easyer to have simply
good programmed sounds and not bothering for them, for my work.
So I try to reorganize my question:
I have a continuous flow of pitch (Hz or MIDI key numbers) and
amplitude coming out from my pitch tracker software (and I'm verrry proud
of having it working on a Quadra 900), at a rate of let's say 50 samples
per second. I make a simple smoothing on them doing a mean of a few
elements and I have to create MIDI commands out of them.
I started with Note On - Note Off at every sample and I found it
not very smart ;-). Now I'm trying to build a simple algorythm
that triggers a Note On at a certain volume of the voice,
then back Note Off when it gets down.
If the pitch changes during a sustained sequence of samples (volume
remains constant) I send Note On - Note Off as writen before.
It sounds not very good.
Whew. hard to explain all this in English on a line-editor.
Any suggestions? Any comments on the project?
Thanks in advance

Dimitrij Hmeljak
hmeljakd@uts340.univ.trieste.it

------------------------------
End of the EMUSIC-L Digest
******************************