From bil at ccrma.Stanford.EDU  Mon Sep  2 04:27:51 2019
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Mon, 02 Sep 2019 04:27:51 -0700
Subject: [CM] Snd 19.7
Message-ID: <262a1e60f13716c16fe7b00a149a223f@ccrma.stanford.edu>

Snd 19.7:

in clm, Anders fixed a bug in mus.lisp.
in s7, I mostly worked on optimization

checked: sbcl 1.5.6

Thanks!: Anders Vinjar, Kenneth Flak, David O'Toole



From deeteeoh1138 at gmail.com  Tue Sep  3 17:24:36 2019
From: deeteeoh1138 at gmail.com (David O'Toole)
Date: Tue, 3 Sep 2019 20:24:36 -0400
Subject: [CM] undefined functions related to MOVING-SPECTRUM
Message-ID: <CAAz1J51b1GLwhm0VL3PmK8eA6K2F_9aCckHPQn-Z+oU1G4LVRA@mail.gmail.com>

The functions mentioned in the documentation, "moving-spectrum-amps,
moving-spectrum-phases, and moving-spectrum-freqs" appear to be
undefined. Grepping /usr/local/share/snd/*.scm doesn't show it, and
calling the functions fails with "moving-spectrum-amps: unbound
variable". MOVING-SPECTRUM actually works, though. I just can't get
the data out.


From bil at ccrma.Stanford.EDU  Wed Sep  4 03:57:43 2019
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Wed, 04 Sep 2019 03:57:43 -0700
Subject: [CM] undefined functions related to MOVING-SPECTRUM
In-Reply-To: <CAAz1J51b1GLwhm0VL3PmK8eA6K2F_9aCckHPQn-Z+oU1G4LVRA@mail.gmail.com>
References: <CAAz1J51b1GLwhm0VL3PmK8eA6K2F_9aCckHPQn-Z+oU1G4LVRA@mail.gmail.com>
Message-ID: <38e9c25d79e8952b34c5ac1ffb13cb27@ccrma.stanford.edu>

The moving-spectrum generator is a "let" (in s7 terminology),
so you access its fields via (sv 'amps) or (sv 'freqs).
The test-sv function in generators.scm has examples.
("sv" is the generator in this case).

Thanks for pointing out the obsolete documentation --
I'll fix it later today.



From deeteeoh1138 at gmail.com  Wed Sep  4 05:13:42 2019
From: deeteeoh1138 at gmail.com (David O'Toole)
Date: Wed, 4 Sep 2019 08:13:42 -0400
Subject: [CM] undefined functions related to MOVING-SPECTRUM
In-Reply-To: <38e9c25d79e8952b34c5ac1ffb13cb27@ccrma.stanford.edu>
References: <CAAz1J51b1GLwhm0VL3PmK8eA6K2F_9aCckHPQn-Z+oU1G4LVRA@mail.gmail.com>
	<38e9c25d79e8952b34c5ac1ffb13cb27@ccrma.stanford.edu>
Message-ID: <CAAz1J51CqwnEwV9FPJ6DmuShGTuen6GXcXOtC4bJKhavK29gRQ@mail.gmail.com>

Thanks! I've made progress but I believe I had encountered a bug so, so I
will send a separate message

On Wed, Sep 4, 2019, 6:57 AM <bil at ccrma.stanford.edu> wrote:

> The moving-spectrum generator is a "let" (in s7 terminology),
> so you access its fields via (sv 'amps) or (sv 'freqs).
> The test-sv function in generators.scm has examples.
> ("sv" is the generator in this case).
>
> Thanks for pointing out the obsolete documentation --
> I'll fix it later today.
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20190904/7bbbf4ae/attachment.html>

From deeteeoh1138 at gmail.com  Wed Sep  4 05:33:38 2019
From: deeteeoh1138 at gmail.com (David O'Toole)
Date: Wed, 4 Sep 2019 08:33:38 -0400
Subject: [CM] strange assertion failure and negative pitches with
 MOVING-SPECTRUM in snd 19.6
Message-ID: <CAAz1J50noWtZsbyHN6TDbmcPguVgbzu+pxktaQYMiO9q7b6s1w@mail.gmail.com>

I'm experiencing strange troubles with detecting the loudest pitch in
a region via MOVING-SPECTRUM. (I am assuming that freqs[i] is the
loudest frequency when amps[i] is the loudest amp.) It outputs
negative pitches and also I get a strange assertion failure where the
loop counter becomes non-integer somehow:

(define (find-max-amp-index amps)
  (let* ((amp0 (find-max-amp amps))
(len (length amps))
(index 0))
    (do ((i 0 (+ i 1)))
((= i len))
      ;; how can this assertion possibly fail?
      (assert (integer? i))
      (when (>= (vector-ref amps i) amp0)
(set! index i)))
    index))

The repository includes the sound file and tests.
https://gitlab.com/dto/mosaic-el/
Full information is at the bottom of the source file:
https://gitlab.com/dto/mosaic-el/blob/master/mosaic.scm

I would greatly appreciate any help you can offer.


From bil at ccrma.Stanford.EDU  Wed Sep  4 07:20:21 2019
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Wed, 04 Sep 2019 07:20:21 -0700
Subject: [CM] strange assertion failure and negative pitches with
 MOVING-SPECTRUM in snd 19.6
In-Reply-To: <CAAz1J50noWtZsbyHN6TDbmcPguVgbzu+pxktaQYMiO9q7b6s1w@mail.gmail.com>
References: <CAAz1J50noWtZsbyHN6TDbmcPguVgbzu+pxktaQYMiO9q7b6s1w@mail.gmail.com>
Message-ID: <e36781ef30a07aa8ee5db1e92a9f1a5e@ccrma.stanford.edu>

You get the assertion failure because (format () ...)
sends its output to the current output; I think you
want (format #f ...) which just returns a string.
There is an assert macro in s7.html.

The "freqs" are probably negative because they are
phase increments ("instantaneous frequency") (in radians).
As in the phase-vocoder, it's using a sliding fft
and returning data for reconstruction as the
windows go by.  pins in clm-ins.scm shows one way
to use this kind of data.  The test-sv function
in generators.scm shows the parallel.  sine-bank
below that function shows a direct use.



From bil at ccrma.Stanford.EDU  Wed Sep  4 07:22:28 2019
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Wed, 04 Sep 2019 07:22:28 -0700
Subject: [CM] strange assertion failure and negative pitches with
 MOVING-SPECTRUM in snd 19.6
In-Reply-To: <CAAz1J50noWtZsbyHN6TDbmcPguVgbzu+pxktaQYMiO9q7b6s1w@mail.gmail.com>
References: <CAAz1J50noWtZsbyHN6TDbmcPguVgbzu+pxktaQYMiO9q7b6s1w@mail.gmail.com>
Message-ID: <4322b4b86cd83f6b72ff589408afbdfb@ccrma.stanford.edu>

I forgot to say, you get the nominal in Hz frequency
by using the bin number (i in this case), the fft size
and the srate.  I'm probably forgetting something else.




From deeteeoh1138 at gmail.com  Wed Sep  4 07:55:09 2019
From: deeteeoh1138 at gmail.com (David O'Toole)
Date: Wed, 4 Sep 2019 10:55:09 -0400
Subject: [CM] strange assertion failure and negative pitches with
 MOVING-SPECTRUM in snd 19.6
In-Reply-To: <4322b4b86cd83f6b72ff589408afbdfb@ccrma.stanford.edu>
References: <CAAz1J50noWtZsbyHN6TDbmcPguVgbzu+pxktaQYMiO9q7b6s1w@mail.gmail.com>
	<4322b4b86cd83f6b72ff589408afbdfb@ccrma.stanford.edu>
Message-ID: <CAAz1J53SxanfJbPDNNhcaL3jMkUgE2Bf+hwFXYBrC+ywQ=F8FQ@mail.gmail.com>

Thanks again :).
I'm working to build a solution out of the SINE-BANK and SV-TEST
functions you mention. I also may need to borrow a little of PINS.

On Wed, Sep 4, 2019 at 10:22 AM <bil at ccrma.stanford.edu> wrote:
>
> I forgot to say, you get the nominal in Hz frequency
> by using the bin number (i in this case), the fft size
> and the srate.  I'm probably forgetting something else.
>
>
>


From deeteeoh1138 at gmail.com  Thu Sep 12 15:59:11 2019
From: deeteeoh1138 at gmail.com (David O'Toole)
Date: Thu, 12 Sep 2019 18:59:11 -0400
Subject: [CM] snd 19.7 gtk-effects menu not working
Message-ID: <CAAz1J51ahTt7x1st7LANS06GL75pCUydEAi7=sgtw6-8rKiDBw@mail.gmail.com>

None of the menu options work, I get this:

;gtk_box_pack_start: unbound variable
;    (gtk_box_pack_start (GTK_BOX...
;    >gtk-effects-utils.scm, line 111
; add-sliders: (gtk_box_pack_start (GTK_BOX... ; dialog: #<GtkWidget_
0x561e45912b50>
;                                              mainform: #<GtkWidget_
0x561e3ca18820>
; add-sliders: ((gtk_widget_show mainform) ... ; use-hbox: #f
;                                              table: #<GtkWidget_
0x561e3c838870>
;                                              sliders: (("delay time"
0.0 0.5 2.0 ...)...
; ((add-target (gtk_dialog_get_content_area... ; echo-dialog:
#<GtkWidget_ 0x561e45912b50>
;                                              echo-target: sound
;                                              echo-truncate: #t
; ((activate-dialog echo-dialog))


From bil at ccrma.Stanford.EDU  Fri Sep 13 04:11:01 2019
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Fri, 13 Sep 2019 04:11:01 -0700
Subject: [CM] snd 19.7 gtk-effects menu not working
In-Reply-To: <CAAz1J51ahTt7x1st7LANS06GL75pCUydEAi7=sgtw6-8rKiDBw@mail.gmail.com>
References: <CAAz1J51ahTt7x1st7LANS06GL75pCUydEAi7=sgtw6-8rKiDBw@mail.gmail.com>
Message-ID: <10a0251ea60313b0e3bd0432bdac5cf5@ccrma.stanford.edu>

The gtk version of Snd is caught in-between versions.
I thought gtk 4 was "right around the corner" years
ago and started trying to port everything to it.
Since the changes between version 3 and 4 are enormous,
and (it turned out) gtk 4 really isn't ready yet,
I got trapped.  Snd itself works (though I am not
happy with it) in gtk 2/3 but does not work in 4,
and the *.scm files are not in any consistent version.
I use motif which is still (thank goodness) available
from debian and fedora.  I suppose I could return to
the gtk3 scheme files.  The real problem is that I
think the gtk interface is ugly, so it's hard to
get up the energy to work on it.



From cm at jrigg.co.uk  Sat Sep 14 01:35:36 2019
From: cm at jrigg.co.uk (John Rigg)
Date: Sat, 14 Sep 2019 08:35:36 +0000
Subject: [CM] snd 19.7 gtk-effects menu not working
In-Reply-To: <10a0251ea60313b0e3bd0432bdac5cf5@ccrma.stanford.edu>
References: <CAAz1J51ahTt7x1st7LANS06GL75pCUydEAi7=sgtw6-8rKiDBw@mail.gmail.com>
	<10a0251ea60313b0e3bd0432bdac5cf5@ccrma.stanford.edu>
Message-ID: <20190914083535.GA26158@localhost.localdomain>

On Fri, Sep 13, 2019 at 04:11:01AM -0700, bil at ccrma.Stanford.EDU wrote:
> The gtk version of Snd is caught in-between versions.
> I thought gtk 4 was "right around the corner" years
> ago and started trying to port everything to it.
> Since the changes between version 3 and 4 are enormous,
> and (it turned out) gtk 4 really isn't ready yet,
> I got trapped.  Snd itself works (though I am not
> happy with it) in gtk 2/3 but does not work in 4,
> and the *.scm files are not in any consistent version.
> I use motif which is still (thank goodness) available
> from debian and fedora.  I suppose I could return to
> the gtk3 scheme files.  The real problem is that I
> think the gtk interface is ugly, so it's hard to
> get up the energy to work on it.

Another motif user here. It's much nicer IMO than the GTK
interface. The ICS version of motif as used by Debian is
still actively maintained (some Linux distros eg. Void
still use the Open Motif version, which is unmaintained
and buggy).

John


From deeteeoh1138 at gmail.com  Sat Sep 14 05:40:21 2019
From: deeteeoh1138 at gmail.com (David O'Toole)
Date: Sat, 14 Sep 2019 08:40:21 -0400
Subject: [CM] snd 19.7 gtk-effects menu not working
In-Reply-To: <20190914083535.GA26158@localhost.localdomain>
References: <CAAz1J51ahTt7x1st7LANS06GL75pCUydEAi7=sgtw6-8rKiDBw@mail.gmail.com>
	<10a0251ea60313b0e3bd0432bdac5cf5@ccrma.stanford.edu>
	<20190914083535.GA26158@localhost.localdomain>
Message-ID: <CAAz1J53DBSfq6Av22r+++_NFo2d1VWxBBLe8FH+g9vnur8C1nA@mail.gmail.com>

Hey there. Thank you, I want to use the motif version too because a lot of
the fancy extensions only work in motif, unfortunately I tried it out and
ran into an issue where suppressing graphical updates does not work and
causes the user interface to misbehave because my software scheme mosaic
opens and closes a lot of sounds. The gtk version seems fine on this task
but the motif versions malfunctions and hangs. I'm looking for a work
around. I thought I would send this message to the list because my
follow-ups with Bill were off list.

On Sat, Sep 14, 2019, 4:47 AM John Rigg <cm at jrigg.co.uk> wrote:

> On Fri, Sep 13, 2019 at 04:11:01AM -0700, bil at ccrma.Stanford.EDU wrote:
> > The gtk version of Snd is caught in-between versions.
> > I thought gtk 4 was "right around the corner" years
> > ago and started trying to port everything to it.
> > Since the changes between version 3 and 4 are enormous,
> > and (it turned out) gtk 4 really isn't ready yet,
> > I got trapped.  Snd itself works (though I am not
> > happy with it) in gtk 2/3 but does not work in 4,
> > and the *.scm files are not in any consistent version.
> > I use motif which is still (thank goodness) available
> > from debian and fedora.  I suppose I could return to
> > the gtk3 scheme files.  The real problem is that I
> > think the gtk interface is ugly, so it's hard to
> > get up the energy to work on it.
>
> Another motif user here. It's much nicer IMO than the GTK
> interface. The ICS version of motif as used by Debian is
> still actively maintained (some Linux distros eg. Void
> still use the Open Motif version, which is unmaintained
> and buggy).
>
> John
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> https://cm-mail.stanford.edu/mailman/listinfo/cmdist
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20190914/a6cb77cf/attachment.html>

From koen.deturck at gmail.com  Sat Sep 14 05:47:02 2019
From: koen.deturck at gmail.com (Koen De Turck)
Date: Sat, 14 Sep 2019 14:47:02 +0200
Subject: [CM] snd 19.7 gtk-effects menu not working
In-Reply-To: <CAAz1J53DBSfq6Av22r+++_NFo2d1VWxBBLe8FH+g9vnur8C1nA@mail.gmail.com>
References: <CAAz1J51ahTt7x1st7LANS06GL75pCUydEAi7=sgtw6-8rKiDBw@mail.gmail.com>
	<10a0251ea60313b0e3bd0432bdac5cf5@ccrma.stanford.edu>
	<20190914083535.GA26158@localhost.localdomain>
	<CAAz1J53DBSfq6Av22r+++_NFo2d1VWxBBLe8FH+g9vnur8C1nA@mail.gmail.com>
Message-ID: <CAKCgp29EU7tWyJNkbgpxkeXmLrQp5T0r+S+E47kgibNmEciENA@mail.gmail.com>

Does anyone use the combo arch+motif? on openbsd I use motif and it works
great, but on my arch machine I tried to get motif to work (with different
versions
iirc), but I failed so I settled with gtk instead, thinking that this was
the standard option
anyway (and forgot to document my different attempts so I can't recall
exactly which version of
motif I tried)
Koen

On Sat, Sep 14, 2019 at 2:40 PM David O'Toole <deeteeoh1138 at gmail.com>
wrote:

> Hey there. Thank you, I want to use the motif version too because a lot of
> the fancy extensions only work in motif, unfortunately I tried it out and
> ran into an issue where suppressing graphical updates does not work and
> causes the user interface to misbehave because my software scheme mosaic
> opens and closes a lot of sounds. The gtk version seems fine on this task
> but the motif versions malfunctions and hangs. I'm looking for a work
> around. I thought I would send this message to the list because my
> follow-ups with Bill were off list.
>
> On Sat, Sep 14, 2019, 4:47 AM John Rigg <cm at jrigg.co.uk> wrote:
>
>> On Fri, Sep 13, 2019 at 04:11:01AM -0700, bil at ccrma.Stanford.EDU wrote:
>> > The gtk version of Snd is caught in-between versions.
>> > I thought gtk 4 was "right around the corner" years
>> > ago and started trying to port everything to it.
>> > Since the changes between version 3 and 4 are enormous,
>> > and (it turned out) gtk 4 really isn't ready yet,
>> > I got trapped.  Snd itself works (though I am not
>> > happy with it) in gtk 2/3 but does not work in 4,
>> > and the *.scm files are not in any consistent version.
>> > I use motif which is still (thank goodness) available
>> > from debian and fedora.  I suppose I could return to
>> > the gtk3 scheme files.  The real problem is that I
>> > think the gtk interface is ugly, so it's hard to
>> > get up the energy to work on it.
>>
>> Another motif user here. It's much nicer IMO than the GTK
>> interface. The ICS version of motif as used by Debian is
>> still actively maintained (some Linux distros eg. Void
>> still use the Open Motif version, which is unmaintained
>> and buggy).
>>
>> John
>> _______________________________________________
>> Cmdist mailing list
>> Cmdist at ccrma.stanford.edu
>> https://cm-mail.stanford.edu/mailman/listinfo/cmdist
>>
>> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> https://cm-mail.stanford.edu/mailman/listinfo/cmdist
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20190914/3a49c0b2/attachment.html>

From cm at jrigg.co.uk  Sun Sep 15 07:43:51 2019
From: cm at jrigg.co.uk (John Rigg)
Date: Sun, 15 Sep 2019 14:43:51 +0000
Subject: [CM] snd 19.7 gtk-effects menu not working
In-Reply-To: <CAKCgp29EU7tWyJNkbgpxkeXmLrQp5T0r+S+E47kgibNmEciENA@mail.gmail.com>
References: <CAAz1J51ahTt7x1st7LANS06GL75pCUydEAi7=sgtw6-8rKiDBw@mail.gmail.com>
	<10a0251ea60313b0e3bd0432bdac5cf5@ccrma.stanford.edu>
	<20190914083535.GA26158@localhost.localdomain>
	<CAAz1J53DBSfq6Av22r+++_NFo2d1VWxBBLe8FH+g9vnur8C1nA@mail.gmail.com>
	<CAKCgp29EU7tWyJNkbgpxkeXmLrQp5T0r+S+E47kgibNmEciENA@mail.gmail.com>
Message-ID: <20190915144351.GA11605@localhost.localdomain>

On Sat, Sep 14, 2019 at 02:47:02PM +0200, Koen De Turck wrote:
> Does anyone use the combo arch+motif? on openbsd I use motif and it works
> great, but on my arch machine I tried to get motif to work (with different
> versions
> iirc), but I failed so I settled with gtk instead, thinking that this was
> the standard option
> anyway (and forgot to document my different attempts so I can't recall
> exactly which version of
> motif I tried)
> Koen

I'm not an Arch user but I notice they currently use motif-2.3.8
(which appears to be from ICS despite the package name openmotif).
That version is also in OpenBSD 6.4 and 6.5.

My DAW systems are running Debian 9 with motif-2.3.4. Debian 10
uses motif-2.3.8 but I haven't tried it yet.

I don't know if choice of window manager is relevant. I used
Window Maker for many years and recently switched to cwm. Both
have worked well with snd and motif. Also of unknown relevance:
none of my systems run dbus or systemd.

BTW I was incorrect about Void using Open Motif; it uses lesstif
0.95, released in 2009. IIRC that was used in earlier versions
of Debian and was quite crash-prone when used with snd.

John 


From deeteeoh1138 at gmail.com  Mon Sep 16 10:52:56 2019
From: deeteeoh1138 at gmail.com (David O'Toole)
Date: Mon, 16 Sep 2019 13:52:56 -0400
Subject: [CM] error loading lint.scm in snd 19.7
Message-ID: <CAAz1J51vhu1GCupe5CsR_6VSyz8PUoR+aivv=bftKNUh6PG3Yg@mail.gmail.com>

Hi, I am adding support to Emacs Flycheck-mode for the Snd-scheme
linter, but cannot load Lint anymore:

loading lint.scm
let-set!: *pretty-print-cycles* is not defined in #1=(inlet)
"/usr/local/share/snd/lint.scm"[396]: (set! (lint-pp-funclet
'*pretty-print-cycles*) #f)

loading lint.scm
hash-table got an odd number of arguments: ((car) (caar . car) (cdar .
cdr) (caaar . caar) (cdaar . cdar) (cddar . cddr) (cadar . cadr)
(caaaar . caaar) (caadar . caadr) (cadaar . cadar) (caddar . caddr)
(cdaaar . cdaar) (cdadar . cdadr) (cddaar . cddar) (cdddar . cdddr))
lint: unbound variable


From bil at ccrma.Stanford.EDU  Mon Sep 16 11:42:37 2019
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Mon, 16 Sep 2019 11:42:37 -0700
Subject: [CM] error loading lint.scm in snd 19.7
In-Reply-To: <CAAz1J51vhu1GCupe5CsR_6VSyz8PUoR+aivv=bftKNUh6PG3Yg@mail.gmail.com>
References: <CAAz1J51vhu1GCupe5CsR_6VSyz8PUoR+aivv=bftKNUh6PG3Yg@mail.gmail.com>
Message-ID: <2da9c49453a37a90842eea598ce3ca81@ccrma.stanford.edu>

That variable is in the pretty-print function in write.scm.  There's an
earlier (require write.scm) to load it. To see if it is being loaded,
you can set *load-hook*:

(set! (hook-functions *load-hook*)
        (list (lambda (hook)
                (format () "loading ~S...~%" (hook 'name)))))

*pretty-print-cycles* is somewhat new, so the version of write.scm
and lint.scm need to match.  I'd check /usr/local/share/snd for
an obsolete version of the file (I'm not sure how that would
happen -- the makefile puts *.scm in the scripts directory
if you use "install").  It might depend on your *load-path*.
Now that I look at it, the *load-hook* argument does not always
include the directory -- maybe it should if *load-path* is in use.
strace probably could tell you which write.scm is being loaded,
but it's very verbose.

If you can start snd, and (require write.scm), print
(outlet (funclet pretty-print))
and you should see *pretty-print-cycles*.



From deeteeoh1138 at gmail.com  Mon Sep 16 16:16:38 2019
From: deeteeoh1138 at gmail.com (David O'Toole)
Date: Mon, 16 Sep 2019 19:16:38 -0400
Subject: [CM] error loading lint.scm in snd 19.7
In-Reply-To: <2da9c49453a37a90842eea598ce3ca81@ccrma.stanford.edu>
References: <CAAz1J51vhu1GCupe5CsR_6VSyz8PUoR+aivv=bftKNUh6PG3Yg@mail.gmail.com>
	<2da9c49453a37a90842eea598ce3ca81@ccrma.stanford.edu>
Message-ID: <CAAz1J51fXtAOqgNs1LB1-R=tCezrjpYG-PAeeFV0C1wzehfSmw@mail.gmail.com>

This was indeed caused by a version mismatch Problem by having multiple
versions installed. Thank you so much again for your help!

On Mon, Sep 16, 2019, 3:19 PM <bil at ccrma.stanford.edu> wrote:

> That variable is in the pretty-print function in write.scm.  There's an
> earlier (require write.scm) to load it. To see if it is being loaded,
> you can set *load-hook*:
>
> (set! (hook-functions *load-hook*)
>         (list (lambda (hook)
>                 (format () "loading ~S...~%" (hook 'name)))))
>
> *pretty-print-cycles* is somewhat new, so the version of write.scm
> and lint.scm need to match.  I'd check /usr/local/share/snd for
> an obsolete version of the file (I'm not sure how that would
> happen -- the makefile puts *.scm in the scripts directory
> if you use "install").  It might depend on your *load-path*.
> Now that I look at it, the *load-hook* argument does not always
> include the directory -- maybe it should if *load-path* is in use.
> strace probably could tell you which write.scm is being loaded,
> but it's very verbose.
>
> If you can start snd, and (require write.scm), print
> (outlet (funclet pretty-print))
> and you should see *pretty-print-cycles*.
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://cm-mail.stanford.edu/pipermail/cmdist/attachments/20190916/f544af29/attachment.html>

From deeteeoh1138 at gmail.com  Tue Sep 24 16:11:08 2019
From: deeteeoh1138 at gmail.com (David O'Toole)
Date: Tue, 24 Sep 2019 19:11:08 -0400
Subject: [CM] issues with graph->ps
Message-ID: <CAAz1J53CKOkAywyZnPd0wp0hGHmHWrt1Wu7jMfGf41A9LkV3oQ@mail.gmail.com>

I'm trying to render a full waveform preview of a sound file using the
GRAPH->PS function, but encountering issues. With (SET! (CHANNEL-STYLE
SOUND) CHANNELS-SEPARATE) i get an output graph, but only of the first
snippet of sound. Even if SHOW-FULL-DURATION is set to #t before the
sound file is opened.

With CHANNELS-SUPERIMPOSED or CHANNELS-COMBINED, I get an EPS file
that is just a white square.

Here is the code I'm using to try to make the waveform preview:

(define^ (mosaic-render-waveform sound-file output-file)
  (set! eps-size 1.0)
  (let ((old-show show-full-duration))
    (set! show-full-duration #t)
    (let ((sound (open-sound sound-file)))
      (select-sound sound)
      (set! (channel-style sound) channels-separate)
      (graph->ps output-file)
      (set! show-full-duration old-show)
      (close-sound sound))))


From bil at ccrma.Stanford.EDU  Tue Sep 24 16:25:34 2019
From: bil at ccrma.Stanford.EDU (bil at ccrma.Stanford.EDU)
Date: Tue, 24 Sep 2019 16:25:34 -0700
Subject: [CM] issues with graph->ps
In-Reply-To: <CAAz1J53CKOkAywyZnPd0wp0hGHmHWrt1Wu7jMfGf41A9LkV3oQ@mail.gmail.com>
References: <CAAz1J53CKOkAywyZnPd0wp0hGHmHWrt1Wu7jMfGf41A9LkV3oQ@mail.gmail.com>
Message-ID: <1bb1d78cb7bb6139273712e566a6d486@ccrma.stanford.edu>

eps-size and show-full-duration are functions, so
use (set! (eps-size) 1.0) for example.  You can also
use the variable forms: *eps-size* and *show-full-duration*
as (set! *eps-size* 1.0).