audacity-devel

[Audacity-devel] click5.ny

[David R. S. <dav...@sh...>]

Hi Gale and Richard,

The attached click5.ny gives a different error report. Erroneous
entered values (except for MIDI pitches at the moment) are
explained, with the next line below each value explanation showing
the entered value, and "error:" before the value if there's an
error - for fast visual scanning and correction.

At this stage I have not yet included error MIDI pitch values in
error-checking since they can even be below 0 - negative infinity
is the MIDI value of 0hz.

if there's a range of MIDI pitches outside of which you'd like to
have an error thrown, let me know. Current lower and upper MIDI
pitch values are 48 through 96.

Many of the input fields are int (integer) inputs. In response to
a series of errors in these plug-ins many years ago, Dominic
patched up the Nyquist bridge so that when you enter eg. 4.4, plug-
ins now "see" this as 4.

;info line fixed up a bit.

Thanks

David

-- 
David R. Sky
http://www.shellworld.net/~davidsky/

[Audacity-devel] click5.ny

[Gale A. <ga...@au...>]

Hi David 

Re click5.ny which you posted on 22 September and I sent back to 
you with modifications to make the error dialogue look better for 
sighted users, I noticed something has gone wrong with the actual 
click generation. If you ask for 60 beats per minute, each click is
generated with a 917 ms interval between them instead of 1000 ms.  
It isn't an accidental edit I made as the problem seems to be with the
original click5 you posted as above.  Perhaps it is something to do
with the new time offset? 



Gale      


Outbound message virus free.
Tested on: 9/24/2007 5:58:28 PM

Re: [Audacity-devel] click5.ny

[David R. S. <dav...@sh...>]

Hi Gale,

Thanks very much for this specific feedback!

Since clicks should be 10ms in duration, the gap between clicks at 60bpm 
should be 990ms. (1000ms from start of one click to start of next click).

I have an idea where to start checking in the plug - once 
I've got my energy back. (Off-topic details)

Thanks

David


-- 
David R. Sky
http://www.shellworld.net/~davidsky/


On Mon, 24 Sep 2007, Gale Andrews wrote:

>
> Hi David
>
> Re click5.ny which you posted on 22 September and I sent back to
> you with modifications to make the error dialogue look better for
> sighted users, I noticed something has gone wrong with the actual
> click generation. If you ask for 60 beats per minute, each click is
> generated with a 917 ms interval between them instead of 1000 ms.
> It isn't an accidental edit I made as the problem seems to be with the
> original click5 you posted as above.  Perhaps it is something to do
> with the new time offset?
>
>
>
> Gale
>
>
> Outbound message virus free.
> Tested on: 9/24/2007 5:58:28 PM
>
>
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Audacity-devel mailing list
> Aud...@li...
> https://lists.sourceforge.net/lists/listinfo/audacity-devel
>

Re: [Audacity-devel] click5.ny

[David R. S. <dav...@sh...>]

Hi Gale,

Gale:I notice that if you enter an alphabetical character in a
field it treats the error as if you entered zero.

David: Confirmed here. I assume it's due to the fact it's an
"artifact" of the edit field being an integer (rather than a
string) input field.

I'd prefer to leave it "as is", since someone can use the track bar
to move the value up and down, which you don't have in a string
input.

Gale: If you enter a combined number and alphabetical character
e.g. "3X" it shows in the error dialogue e.g. "ERROR - 3 beats per
minute" which is of course valid as it stands.

David: I discovered early on that very low tempo rates caused the
click generator to grind on and on and on, necessitating a
deliberate crash of Audacity. It's the way Dominic originally
programmed the generator. For example, even a tempo of 1 bpm, one
beat per measure and 1 generated measure caused this grinding. 
(Technical explanation if you want it, not here.)

I can rework the code to allow such low tempo values, which would
mean that most of Dominic's original code would disappear.

The range he originally programm was 30-300 bpm, which is a large
enough range I think - have you ever received a request for having
a click once every 4 seconds, for instance? (That's tempo of
15bpm.)

Gale: I think you can extend the allowable {MIDI value} range
significantly as long as you disallow something that is going to
crash Audacity (as for example 922 will). I have the attached set
to minimum 18, maximum 116 and it sounded quite good to me at that
setting (the 116 actually sounded a bit like a jingle bell that
some people seem to want).

David: *chuckling* There's no MIDI pitch setting which would
generate "hee ho ho ho hee ho ho ho hee ho ho ho hee ho ho ho" lol!
- that would be funny.

Yes, that makes sense - I think the highest frequency of human
hearing 20khz is MIDI pitch somewhere above 120, which _would_ give
an aliasing error from Nyquist.

{from a later post}

Gale: Re click5.ny which you posted on 22 September and I sent back
to you with modifications to make the error dialogue look better
for sighted users, I noticed something has gone wrong with the
actual click generation. If you ask for 60 beats per minute, each
click is generated with a 917 ms interval between them instead of
1000 ms. It isn't an accidental edit I made as the problem seems to
be with the original click5 you posted as above. Perhaps it is
something to do with the new time offset?

David: I got no such results, testing in both 1.2.4 and 1.3.3.

There's a slight "weirdness" though, which I attribute to rounding-
off in the Nyquist real value edit field:

when I enter :

tempo 60bpm
beats per measure 1
measures 60
starting time offset 1.0 seconds

the first time I do this the duration is exactly 1m 1s, as it
should be.

When I clear the track I created, then generate the click track
using same values as I previously put in the edit fields without
changing any of them, the duration of the generated track is 
1m 0.99s.

No big deal, it's just a time offset rounding-off error (or
whatever you want to call it), I've found this recently in another
plug-in.

In real value input fields, the track bar is divided into 1001 
(from 0 to 1000) increments.

When you input a value into a real value edit field without using
any of the cursor keys when inside the edit field, the _precise_
value is used.

If you use any cursor keys after entering a value (such as I do
with my screen reader, to check I input a correct value), somehow
the plug-in "rounds" to the nearest value, according to the track
bar increments.

I doubt this is important for this plug-in, since the click track
would assumably be the first track for a multitrack session.

One final note: I was thinking (for the list of error and okay
values) that I'll indent non-error/okay lines so the error/okay
messages are at the left margin.

I won't bother about listing error values at the top for screen
readers, this would make the error-checking code much more complex
than the already-existing error-checking and click track generation
code.

Will get this to the list soon.

Thanks

David

-- 
David R. Sky
http://www.shellworld.net/~davidsky/

Re: [Audacity-devel] click5.ny

[Gale A. <ga...@au...>]

    | From "David R. Sky" <dav...@sh...> 
    | Tue, 25 Sep 2007 06:25:10 -0700 (PDT)
    | Subject: [Audacity-devel] click5.ny
    | Gale: If you enter a combined number and alphabetical character
    | e.g. "3X" it shows in the error dialogue e.g. "ERROR - 3 beats per
    | minute" which is of course valid as it stands.
    | 
    | David: I discovered early on that very low tempo rates caused the
    | click generator to grind on and on and on, necessitating a
    | deliberate crash of Audacity. It's the way Dominic originally
    | programmed the generator. For example, even a tempo of 1 bpm, one
    | beat per measure and 1 generated measure caused this grinding. 
    | (Technical explanation if you want it, not here.)
    | 
    | 
    | I can rework the code to allow such low tempo values, which would
    | mean that most of Dominic's original code would disappear.
    | 
    | The range he originally programm was 30-300 bpm, which is a large
    | enough range I think - have you ever received a request for having
    | a click once every 4 seconds, for instance? (That's tempo of
    | 15bpm.)

Hi David 

Sorry I was not thinking quite clearly there - 3 BPM as you say is not valid
and if you enter 30O (the last character being alphabetical), it will give 
you 30 BPM and disregard the alphabetical character.  That will have to 
do. 

I've never received requests for higher or lower BPMs than are currently 
allowed, so as you say it's probably not worth extending the range of 
valid BPMs if it means a lot more work.  


    | Gale: I think you can extend the allowable {MIDI value} range
    | significantly as long as you disallow something that is going to
    | crash Audacity (as for example 922 will). I have the attached set
    | to minimum 18, maximum 116 and it sounded quite good to me at that
    | setting (the 116 actually sounded a bit like a jingle bell that
    | some people seem to want).
    | 
    | David:
    | Yes, that makes sense - I think the highest frequency of human
    | hearing 20khz is MIDI pitch somewhere above 120, which _would_ give
    | an aliasing error from Nyquist.


OK so I will leave that to you to allow a greater (sensible) MIDI range than 
now, but add in the error throwing outside that range. 

    | {from a later post}
    | 
    | Gale: Re click5.ny which you posted on 22 September and I sent back
    | to you with modifications to make the error dialogue look better
    | for sighted users, I noticed something has gone wrong with the
    | actual click generation. If you ask for 60 beats per minute, each
    | click is generated with a 917 ms interval between them instead of
    | 1000 ms. It isn't an accidental edit I made as the problem seems to
    | be with the original click5 you posted as above. Perhaps it is
    | something to do with the new time offset?
    | 
    | David: I got no such results, testing in both 1.2.4 and 1.3.3.
    | 
    | There's a slight "weirdness" though, which I attribute to rounding-
    | off in the Nyquist real value edit field:
    | 
    | when I enter :
    | 
    | tempo 60bpm
    | beats per measure 1
    | measures 60
    | starting time offset 1.0 seconds
    | 
    | the first time I do this the duration is exactly 1m 1s, as it
    | should be.
    | 
    | When I clear the track I created, then generate the click track
    | using same values as I previously put in the edit fields without
    | changing any of them, the duration of the generated track is 
    | 1m 0.99s.

I still have a problem here in 1.3.3 (both a build from a few days ago and 
the released 1.3.3 from May). If I enter your settings above, the first click 
starts at about 0.92 seconds, each subsequent click starts some 
0.92 seconds later, and the total track length is 56.043 seconds. So, both 
the offset and the spacing of the clicks is wrong.  But If I do the above in 
1.2.6 using the exact same version of click5.ny, it's correct (total length 
1 min, 1 second). I'm on Windows XP SP2. This will need more testing as 
it's more important it works consistently in 1.3.x than 1.2.x (as 1.3.x is the
next stable version).  
   

    | One final note: I was thinking (for the list of error and okay
    | values) that I'll indent non-error/okay lines so the error/okay
    | messages are at the left margin.

I think any changes e.g. indentation should probably be so that the
ERROR lines are more distinct from the OK lines, not to make the 
ERROR and OK lines more distinct from the lines giving the correct 
values. But even now, (to me) with everything lining up on the left,
and ERROR a longer word than OK, it looks very clear to me where 
the error line(s) are. 

In my suggested error message, I noticed "field" should be "field(s)".

Thanks 


Gale   
 



Outbound message virus free.
Tested on: 9/25/2007 8:01:56 PM

Re: [Audacity-devel] click5.ny

[David R. S. <dav...@sh...>]

Hi Gale,

On Tue, 25 Sep 2007, Gale Andrews wrote:

>>  I still have a problem here in 1.3.3 (both a build from a few days ago 
and
> the released 1.3.3 from May). If I enter your settings above, the first click
> starts at about 0.92 seconds, each subsequent click starts some
> 0.92 seconds later, and the total track length is 56.043 seconds. So, both
> the offset and the spacing of the clicks is wrong.  But If I do the above in
> 1.2.6 using the exact same version of click5.ny, it's correct (total length
> 1 min, 1 second). I'm on Windows XP SP2. This will need more testing as
> it's more important it works consistently in 1.3.x than 1.2.x (as 1.3.x is the
> next stable version).

David: It would seem to me that something has changed in Audacity itself, 
not that there's something wrong with the Nyquist plug-in.

I've reported here some time ago several strange effects I see in 1.3.3, 
the latest of which when I created audio in a single track, moved to the 
end of the track, and appended what I wanted: a pluck sound of 3s 
duration, the generated pluck was exactly 8.5s in duration.

Which is one of the several weird things that have been happening in 1.3.3 
on my 233Mhz Win98 computer which results in me continuing doing most of 
my plug-in and audio work in 1.2.4.

The click generator works just fine in 1.2.4 and 1.3.3 on my computer as I 
reported earlier.

David

Re: [Audacity-devel] click5.ny

[Gale A. <ga...@au...>]

    | From "David R. Sky" <dav...@sh...> 
    | Wed, 26 Sep 2007 07:26:36 -0700 (PDT)
    | Subject: [Audacity-devel] click5.ny
    | On Tue, 25 Sep 2007, Gale Andrews wrote:
    | >  I still have a problem here in 1.3.3 (both a build from a few days ago 
    | > and the released 1.3.3 from May). If I enter your settings above, the first
    | > click starts at about 0.92 seconds, each subsequent click starts some
    | > 0.92 seconds later, and the total track length is 56.043 seconds. So, both
    | > the offset and the spacing of the clicks is wrong.  But If I do the above in
    | > 1.2.6 using the exact same version of click5.ny, it's correct (total length
    | > 1 min, 1 second). I'm on Windows XP SP2. This will need more testing as
    | > it's more important it works consistently in 1.3.x than 1.2.x (as 1.3.x is
    | > the next stable version).
    | 
    | David: It would seem to me that something has changed in Audacity itself, 
    | not that there's something wrong with the Nyquist plug-in.
    | 
    | I've reported here some time ago several strange effects I see in 1.3.3, 
    | the latest of which when I created audio in a single track, moved to the 
    | end of the track, and appended what I wanted: a pluck sound of 3s 
    | duration, the generated pluck was exactly 8.5s in duration.
    | 
    | Which is one of the several weird things that have been happening in 1.3.3 
    | on my 233Mhz Win98 computer which results in me continuing doing most of 
    | my plug-in and audio work in 1.2.4.
    | 
    | The click generator works just fine in 1.2.4 and 1.3.3 on my computer as I 
    | reported earlier.

Hi David 

I've raised an issue in Watching Brief on the Checklist as this isn't acceptable. 
For me, at least there is consistency:

ClickTrack5:  Ask for 60 bpm but clicks occur every 0.918 seconds
Tuning Fork: Ask for 5 seconds but get 4.593 seconds
Pluck: Ask for 3 seconds but get 2.756 seconds

In other words I am always being sold short by around 8.13% or 
a bit more. 

David, can you remind us are the other weirdnesses you see in 1.3.3 all 
concerned with Nyquist generation, or what else is there?
 
Can others please try those generation examples when you are doing other
things in Audacity and let us know? Any ideas what would be causing these
discrepancies in 1.3.x?      


Gale
 


Outbound message virus free.
Tested on: 9/26/2007 6:47:47 PM

[Audacity-devel] click6.ny was Re: click5.ny

[David R. S. <dav...@sh...>]

Hi Gale et al,

Sorry for the delay - my Sunday dialysis run was botched. Gradually
recovering.

2 things (newest mod of click track generator, feedback on 1.3.3
problems)

1. Thanks Gale for your suggestions. Attached is click6.ny with
added MIDI error messages and what I hope is a clearer error screen
for sighted users.

Screen reader users don't need to listen to the entire screen - we
can simply cursor up and down the screen (numlock off, use the
numeric keypad - 8 is up arrow, 2 is down arrow), 'ok' and 'error'
are on the left margin so it's simple to hear what's wrong.

The error message code is longer than the click track generation
code, I'm not going to 'fancy it up' beyond what I've done.

2. I've been testing in 1.3.3 with increasing chagrin - I don't
know how many times my computer crashed: I lose all screen reader
output, and I have to push the reset button to reboot. (Sometimes
this seems to have been triggered by me pressing a button {such as
'y' to confirm a process} before either the message has been fully
read, or before I have temporarily shut off speech before pressing
the key.)

Apart from that - as per at least two reports I gave perhaps 2
months ago or more, the following is consistant behavior in 1.3.3
(this is May 5 or later build):

create some audio using Nyquist

(this audio is always the correct duration ONLY IF Audacity has
just been newly opened - if other Nyquist activity has been done in
Audacity previously and cleared out of Audacity, Nyquist-generated
output is ALWAYS longer)

either move to end of above track, or open new track

generate new audio using Nyquist

second generated audio is ALWAYS much longer in duration than it
should have been.

(Built-in tone generator had no problem after creating above
audio.) And for some reason none of the click track generators has this 
problem.

Using binaural tones with surf 2 generator
http://www.shellworld.net/~davidsky/bitone2.zip 
after previously-generated Nyquist audio, and setting adjustment
time to 1 minute, it didn't stop chugging away until I pressed the
reset button again. (This sounds like a similar problem reported to
me several months ago by someone using 1.2.6, but I didn't and
don't get this problem in my 1.2.6.)

Escape, control+c and control+x don't work. How can I interrupt
Audacity without rebooting?


After posting the 2 reports of longer-than-should-be Nyquist audio
as above back when I did, someone responded in a way which implied
to me that it's my problem since I'm using obsolete equipment. So
I didn't bother to report this any more until Gale's recent posts
about recent incorrect durations of the click track generator.


After extensive crashing and other problems, I've had enough of the
computer for at least a few hours. sorry for the glum.

David
On Wed, 26 Sep 2007, Gale Andrews wrote:

>    | David: It would seem to me that something has changed in Audacity itself,
>    | not that there's something wrong with the Nyquist plug-in.
>    |
>    | I've reported here some time ago several strange effects I see in 1.3.3,
>    | the latest of which when I created audio in a single track, moved to the
>    | end of the track, and appended what I wanted: a pluck sound of 3s
>    | duration, the generated pluck was exactly 8.5s in duration.
>    |
>    | Which is one of the several weird things that have been happening in 1.3.3
>    | on my 233Mhz Win98 computer which results in me continuing doing most of
>    | my plug-in and audio work in 1.2.4.
>    |
>    | The click generator works just fine in 1.2.4 and 1.3.3 on my computer as I
>    | reported earlier.
>
> Hi David
>
> I've raised an issue in Watching Brief on the Checklist as this isn't acceptable.
> For me, at least there is consistency:
>
> ClickTrack5:  Ask for 60 bpm but clicks occur every 0.918 seconds
> Tuning Fork: Ask for 5 seconds but get 4.593 seconds
> Pluck: Ask for 3 seconds but get 2.756 seconds
>
> In other words I am always being sold short by around 8.13% or
> a bit more.
>
> David, can you remind us are the other weirdnesses you see in 1.3.3 all
> concerned with Nyquist generation, or what else is there?
>
> Can others please try those generation examples when you are doing other
> things in Audacity and let us know? Any ideas what would be causing these
> discrepancies in 1.3.x?
>
>
> Gale
>
>
>
> Outbound message virus free.
> Tested on: 9/26/2007 6:47:47 PM
>
>
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Audacity-devel mailing list
> Aud...@li...
> https://lists.sourceforge.net/lists/listinfo/audacity-devel
>

Re: [Audacity-devel] click6.ny was Re: click5.ny

[Gale A. <ga...@au...>]

    | From "Samuel Humphreys" <sam...@gm...> 
    | Sat, 29 Sep 2007 00:57:54 +0100
    | Subject: [Audacity-devel] click6.ny was Re: click5.ny
    | > david Sky:  "Escape, control+c and control+x don't work. How can I 
    | > interrupt Audacity without rebooting?"
    | 
    | sometimes this works:  'control alt delete' to get up the task manager,
    | cursoring to the audacity process and then selecting the 'end task'
    | button...(thats how you do it in win98 if I remember correctly). hmmm 
    | might be troublesome for the screen reader... 

Also if you context menu over audacity.exe in Task Manager this may give
you an additional option "End Process Tree". Often if "Dr Watson" (the MS
error logging tool) is has appeared that's the only way to kill Audacity - End 
Process won't do it. 

      
Gale 




Outbound message virus free.
Tested on: 9/29/2007 2:08:49 AM

Re: [Audacity-devel] click6.ny was Re: click5.ny

[David R. S. <dav...@sh...>]

On  Sat, 29 Sep 2007, Gale Andrews wrote:

> Also if you context menu over audacity.exe in Task Manager this may give
> you an additional option "End Process Tree". Often if "Dr Watson" (the MS
> error logging tool) is has appeared that's the only way to kill Audacity - End
> Process won't do it.

Thanks Gale for this additional tip.

David

Re: [Audacity-devel] click6.ny was Re: click5.ny

[Gale A. <ga...@au...>]

    | From "David R. Sky" <dav...@sh...> 
    | Fri, 28 Sep 2007 12:59:27 -0700 (PDT)
    | Subject: [Audacity-devel] click6.ny was Re:  click5.ny
    | 2 things (newest mod of click track generator, feedback on 1.3.3
    | problems)
    | 
    | Attached is click6.ny with added MIDI error messages and what
    | I hope is a clearer error screen for sighted users.

I'm afraid I had to remove the spaces between each value in the error 
dialogue, otherwise it's off the screen even on 1024 x 768. I also tried 
to improve the three lines of explanation above the error/OK lines.
   

    | 2. I've been testing in 1.3.3 with increasing chagrin - I don't
    | know how many times my computer crashed: I lose all screen reader
    | output, and I have to push the reset button to reboot. (Sometimes
    | this seems to have been triggered by me pressing a button {such as
    | 'y' to confirm a process} before either the message has been fully
    | read, or before I have temporarily shut off speech before pressing
    | the key.)

Do you have DMA enabled on the drive? 
http://audacityteam.org/wiki/index.php?title=DMA_mode


    | Apart from that - as per at least two reports I gave perhaps 2
    | months ago or more, the following is consistant behavior in 1.3.3
    | (this is May 5 or later build):
    | 
    | create some audio using Nyquist
    | 
    | (this audio is always the correct duration ONLY IF Audacity has
    | just been newly opened - if other Nyquist activity has been done in
    | Audacity previously and cleared out of Audacity, Nyquist-generated
    | output is ALWAYS longer)
    | 
    | either move to end of above track, or open new track
    | 
    | generate new audio using Nyquist
    | 
    | second generated audio is ALWAYS much longer in duration than it
    | should have been.
    | 
    | (Built-in tone generator had no problem after creating above
    | audio.) And for some reason none of the click track generators has this 
    | problem.
    | 
    | Using binaural tones with surf 2 generator
    | http://www.shellworld.net/~davidsky/bitone2.zip 
    | after previously-generated Nyquist audio, and setting adjustment
    | time to 1 minute, it didn't stop chugging away until I pressed the
    | reset button again. (This sounds like a similar problem reported to
    | me several months ago by someone using 1.2.6, but I didn't and
    | don't get this problem in my 1.2.6.)

At the moment using this build from yesterday:
http://mjshaw.at-uclan.com/audacity.zip

I am getting all Nyquist generation exactly length-correct (even after many
tries). I don't yet believe it, as I don't think anyone changed anything 
relevant (?), but see how you get on with the above. 

When you are finally happy with click6, rename it as clicktrack.ny and 
ask Richard to commit it (assuming he has no comments himself 
about it). 


Thanks 

Gale 

[Audacity-devel] No longer Re: click6.ny was Re: click5.ny

[David R. S. <dav...@sh...>]

Hi Gale,

No longer clicktrack.ny.

I got into checking DMA in Win98, am taking a break from it. At least 2 
out of 3 DMA settings were unchecked, should they all be checked?

I grabbed the audacity.zip file fine. 7-zip gave me the message 
"unrecognized archive", so re-installed 7-zip (I defragged 2 days ago). 
Same message, even though I've 
unzipped zip files before using 7-zip.

Thanks

David


-- 
David R. Sky
http://www.shellworld.net/~davidsky/


On Tue, 2 Oct 2007, Gale Andrews wrote:

>
>    | From "David R. Sky" <dav...@sh...>
>    | Fri, 28 Sep 2007 12:59:27 -0700 (PDT)
>    | Subject: [Audacity-devel] click6.ny was Re:  click5.ny
>    | 2 things (newest mod of click track generator, feedback on 1.3.3
>    | problems)
>    |
>    | Attached is click6.ny with added MIDI error messages and what
>    | I hope is a clearer error screen for sighted users.
>
> I'm afraid I had to remove the spaces between each value in the error
> dialogue, otherwise it's off the screen even on 1024 x 768. I also tried
> to improve the three lines of explanation above the error/OK lines.
>
>
>    | 2. I've been testing in 1.3.3 with increasing chagrin - I don't
>    | know how many times my computer crashed: I lose all screen reader
>    | output, and I have to push the reset button to reboot. (Sometimes
>    | this seems to have been triggered by me pressing a button {such as
>    | 'y' to confirm a process} before either the message has been fully
>    | read, or before I have temporarily shut off speech before pressing
>    | the key.)
>
> Do you have DMA enabled on the drive?
> http://audacityteam.org/wiki/index.php?title=DMA_mode
>
>
>    | Apart from that - as per at least two reports I gave perhaps 2
>    | months ago or more, the following is consistant behavior in 1.3.3
>    | (this is May 5 or later build):
>    |
>    | create some audio using Nyquist
>    |
>    | (this audio is always the correct duration ONLY IF Audacity has
>    | just been newly opened - if other Nyquist activity has been done in
>    | Audacity previously and cleared out of Audacity, Nyquist-generated
>    | output is ALWAYS longer)
>    |
>    | either move to end of above track, or open new track
>    |
>    | generate new audio using Nyquist
>    |
>    | second generated audio is ALWAYS much longer in duration than it
>    | should have been.
>    |
>    | (Built-in tone generator had no problem after creating above
>    | audio.) And for some reason none of the click track generators has this
>    | problem.
>    |
>    | Using binaural tones with surf 2 generator
>    | http://www.shellworld.net/~davidsky/bitone2.zip
>    | after previously-generated Nyquist audio, and setting adjustment
>    | time to 1 minute, it didn't stop chugging away until I pressed the
>    | reset button again. (This sounds like a similar problem reported to
>    | me several months ago by someone using 1.2.6, but I didn't and
>    | don't get this problem in my 1.2.6.)
>
> At the moment using this build from yesterday:
> http://mjshaw.at-uclan.com/audacity.zip
>
> I am getting all Nyquist generation exactly length-correct (even after many
> tries). I don't yet believe it, as I don't think anyone changed anything
> relevant (?), but see how you get on with the above.
>
> When you are finally happy with click6, rename it as clicktrack.ny and
> ask Richard to commit it (assuming he has no comments himself
> about it).
>
>
> Thanks
>
> Gale

Re: [Audacity-devel] No longer Re: click6.ny was Re: click5.ny

[Gale A. <ga...@au...>]

    | From "David R. Sky" <dav...@sh...> 
    | Mon, 1 Oct 2007 23:41:49 -0700 (PDT)
    | Subject: [Audacity-devel] No longer Re:  click6.ny was Re:  click5.ny
    | I got into checking DMA in Win98, am taking a break from it. At least 2 
    | out of 3 DMA settings were unchecked, should they all be checked?

You may well have problems if you enable DMA on CD ROM drives on 
Windows 98. If in Device Manager you click on the + sign next to the Disk 
Drives Icon, then you should be able to select your IDE Disk Drive (hard 
drive) then double-click it to display the Properties Window. If you click on 
the Settings Tab located at the top of the screen, that should have an 
Options section in which there is a single white box marked "DMA" to enable
DMA mode for that drive. At least that's what appeared to be true when I saw
screenshots for Windows 98 when I wrote that.

    | I grabbed the audacity.zip file fine. 7-zip gave me the message 
    | "unrecognized archive", so re-installed 7-zip (I defragged 2 days ago). 
    | Same message, even though I've unzipped zip files before using 7-zip.

If the later 7-zip still gives problems then I can post the unzipped exe for 
you to download and you can just launch it from your Audacity 1.3.3 folder
instead of the .exe you have in there now. The only problem is that unzipped,
it weighs in at 8.6 MB. 



Gale 
 



Outbound message virus free.
Tested on: 10/3/2007 7:59:05 PM

Re: [Audacity-devel] No longer Re: click6.ny was Re: click5.ny

[David R. S. <dav...@sh...>]

On Wed, 3 Oct 2007, Gale Andrews wrote:

> If the later 7-zip still gives problems then I can post the unzipped exe for
> you to download and you can just launch it from your Audacity 1.3.3 folder
> instead of the .exe you have in there now. The only problem is that unzipped,
> it weighs in at 8.6 MB.
>
Hi Gale,

The executable itself is 8+megs? Go ahead and post it please.

thanks

David>

>
> Gale
>
>
>
>
> Outbound message virus free.
> Tested on: 10/3/2007 7:59:05 PM
>
>
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a browser.
> Download your FREE copy of Splunk now >> http://get.splunk.com/
> _______________________________________________
> Audacity-devel mailing list
> Aud...@li...
> https://lists.sourceforge.net/lists/listinfo/audacity-devel
>

-- 
David R. Sky
http://www.shellworld.net/~davidsky/

Re: [Audacity-devel] No longer Re: click6.ny was Re: click5.ny

[Gale A. <ga...@au...>]

    | From "David R. Sky" <dav...@sh...> 
    | Sat, 6 Oct 2007 00:31:15 -0700 (PDT)
    | Subject: [Audacity-devel] No longer Re:  click6.ny was Re:  click5.ny
    | On Wed, 3 Oct 2007, Gale Andrews wrote:
    | > If the later 7-zip still gives problems then I can post the unzipped exe for
    | > you to download and you can just launch it from your Audacity 1.3.3 folder
    | > instead of the .exe you have in there now. The only problem is that
    | > unzipped, it weighs in at 8.6 MB.
    | >
    | The executable itself is 8+megs? Go ahead and post it please.

We do have the 1.3.4.rc0.0 compressed installer now. Want to try that
and see if you can get Audacity to launch after you have installed it?


Gale 



Outbound message virus free.
Tested on: 10/6/2007 6:11:25 PM

[Audacity-devel] clicktrack.ny

[David R. S. <dav...@sh...>]

Hi Richard,

Gale and I are satisfied with the attached clicktrack.ny. If you are 
satisfied with it too could you commit it.

Thanks

David

On Tue, 2 Oct 2007, Gale Andrews wrote:

> When you are finally happy with click6, rename it as clicktrack.ny and
> ask Richard to commit it (assuming he has no comments himself
> about it).
>
>
> Thanks
>
> Gale

Re: [Audacity-devel] clicktrack.ny

[Richard A. <ri...@au...>]

On Mon, 2007-10-01 at 21:54 -0700, David R. Sky wrote:
> Gale and I are satisfied with the attached clicktrack.ny. If you are 
> satisfied with it too could you commit it.

Done. The only thing I changed was to set the menu item name to "Click
Track" rather than "Click Track 6".

Richard

Re: [Audacity-devel] clicktrack.ny

[David R. S. <dav...@sh...>]

On Tue, 2 Oct 2007, Richard Ash wrote:

> On Mon, 2007-10-01 at 21:54 -0700, David R. Sky wrote:
>> Gale and I are satisfied with the attached clicktrack.ny. If you are
>> satisfied with it too could you commit it.
>
> Done. The only thing I changed was to set the menu item name to "Click
> Track" rather than "Click Track 6".
>
*grin* Ooops! Thanks Richard.

David

Re: [Audacity-devel] No longer Re: click6.ny was Re: click5.ny

[Martyn S. <mar...@go...>]

David

I can unzip it fine here on XP SP2 using 7-zip 4.53 Beta.  What version 
of 7-zip do you have?

Martyn

David R. Sky wrote:
> Hi Gale,
> 
> No longer clicktrack.ny.
> 
> I got into checking DMA in Win98, am taking a break from it. At least 2 
> out of 3 DMA settings were unchecked, should they all be checked?
> 
> I grabbed the audacity.zip file fine. 7-zip gave me the message 
> "unrecognized archive", so re-installed 7-zip (I defragged 2 days ago). 
> Same message, even though I've 
> unzipped zip files before using 7-zip.
> 
> Thanks
> 
> David
> 
> 

Re: [Audacity-devel] No longer Re: click6.ny was Re: click5.ny

[David R. S. <dav...@sh...>]

Hi Martyn,

Thanks for your note - I have 4.42 installed, just grabbed 4.55beta from 
the 7-zip website. Will go at it again.

Thanks

David


On Tue, 2 Oct 2007, Martyn Shaw wrote:

> David
>
> I can unzip it fine here on XP SP2 using 7-zip 4.53 Beta.  What version
> of 7-zip do you have?
>
> Martyn
>
> David R. Sky wrote:
>> Hi Gale,
>>
>> No longer clicktrack.ny.
>>
>> I got into checking DMA in Win98, am taking a break from it. At least 2
>> out of 3 DMA settings were unchecked, should they all be checked?
>>
>> I grabbed the audacity.zip file fine. 7-zip gave me the message
>> "unrecognized archive", so re-installed 7-zip (I defragged 2 days ago).
>> Same message, even though I've
>> unzipped zip files before using 7-zip.
>>
>> Thanks
>>
>> David
>>
>>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Audacity-devel mailing list
> Aud...@li...
> https://lists.sourceforge.net/lists/listinfo/audacity-devel
>

-- 
David R. Sky
http://www.shellworld.net/~davidsky/

Re: [Audacity-devel] click6.ny was Re: click5.ny

[Samuel H. <sam...@gm...>]

"Escape, control+c and control+x don't work. How can I interrupt
Audacity without rebooting?"

sometimes this works:  'control alt delete' to get up the task manager,
cursoring to the audacity process and then selecting the 'end task'
button...(thats how you do it in win98 if I remember correctly). hmmm might
be troublesome for the screen reader...

hope that helps,
samuel



On 28/09/2007, David R. Sky <dav...@sh...> wrote:
>
> Hi Gale et al,
>
> Sorry for the delay - my Sunday dialysis run was botched. Gradually
> recovering.
>
> 2 things (newest mod of click track generator, feedback on 1.3.3
> problems)
>
> 1. Thanks Gale for your suggestions. Attached is click6.ny with
> added MIDI error messages and what I hope is a clearer error screen
> for sighted users.
>
> Screen reader users don't need to listen to the entire screen - we
> can simply cursor up and down the screen (numlock off, use the
> numeric keypad - 8 is up arrow, 2 is down arrow), 'ok' and 'error'
> are on the left margin so it's simple to hear what's wrong.
>
> The error message code is longer than the click track generation
> code, I'm not going to 'fancy it up' beyond what I've done.
>
> 2. I've been testing in 1.3.3 with increasing chagrin - I don't
> know how many times my computer crashed: I lose all screen reader
> output, and I have to push the reset button to reboot. (Sometimes
> this seems to have been triggered by me pressing a button {such as
> 'y' to confirm a process} before either the message has been fully
> read, or before I have temporarily shut off speech before pressing
> the key.)
>
> Apart from that - as per at least two reports I gave perhaps 2
> months ago or more, the following is consistant behavior in 1.3.3
> (this is May 5 or later build):
>
> create some audio using Nyquist
>
> (this audio is always the correct duration ONLY IF Audacity has
> just been newly opened - if other Nyquist activity has been done in
> Audacity previously and cleared out of Audacity, Nyquist-generated
> output is ALWAYS longer)
>
> either move to end of above track, or open new track
>
> generate new audio using Nyquist
>
> second generated audio is ALWAYS much longer in duration than it
> should have been.
>
> (Built-in tone generator had no problem after creating above
> audio.) And for some reason none of the click track generators has this
> problem.
>
> Using binaural tones with surf 2 generator
> http://www.shellworld.net/~davidsky/bitone2.zip
> after previously-generated Nyquist audio, and setting adjustment
> time to 1 minute, it didn't stop chugging away until I pressed the
> reset button again. (This sounds like a similar problem reported to
> me several months ago by someone using 1.2.6, but I didn't and
> don't get this problem in my 1.2.6.)
>
> Escape, control+c and control+x don't work. How can I interrupt
> Audacity without rebooting?
>
>
> After posting the 2 reports of longer-than-should-be Nyquist audio
> as above back when I did, someone responded in a way which implied
> to me that it's my problem since I'm using obsolete equipment. So
> I didn't bother to report this any more until Gale's recent posts
> about recent incorrect durations of the click track generator.
>
>
> After extensive crashing and other problems, I've had enough of the
> computer for at least a few hours. sorry for the glum.
>
> David
> On Wed, 26 Sep 2007, Gale Andrews wrote:
>
> >    | David: It would seem to me that something has changed in Audacity
> itself,
> >    | not that there's something wrong with the Nyquist plug-in.
> >    |
> >    | I've reported here some time ago several strange effects I see in
> 1.3.3,
> >    | the latest of which when I created audio in a single track, moved
> to the
> >    | end of the track, and appended what I wanted: a pluck sound of 3s
> >    | duration, the generated pluck was exactly 8.5s in duration.
> >    |
> >    | Which is one of the several weird things that have been happening
> in 1.3.3
> >    | on my 233Mhz Win98 computer which results in me continuing doing
> most of
> >    | my plug-in and audio work in 1.2.4.
> >    |
> >    | The click generator works just fine in 1.2.4 and 1.3.3 on my
> computer as I
> >    | reported earlier.
> >
> > Hi David
> >
> > I've raised an issue in Watching Brief on the Checklist as this isn't
> acceptable.
> > For me, at least there is consistency:
> >
> > ClickTrack5:  Ask for 60 bpm but clicks occur every 0.918 seconds
> > Tuning Fork: Ask for 5 seconds but get 4.593 seconds
> > Pluck: Ask for 3 seconds but get 2.756 seconds
> >
> > In other words I am always being sold short by around 8.13% or
> > a bit more.
> >
> > David, can you remind us are the other weirdnesses you see in 1.3.3 all
> > concerned with Nyquist generation, or what else is there?
> >
> > Can others please try those generation examples when you are doing other
> > things in Audacity and let us know? Any ideas what would be causing
> these
> > discrepancies in 1.3.x?
> >
> >
> > Gale
> >
> >
> >
> > Outbound message virus free.
> > Tested on: 9/26/2007 6:47:47 PM
> >
> >
> >
> >
> >
> -------------------------------------------------------------------------
> > This SF.net email is sponsored by: Microsoft
> > Defy all challenges. Microsoft(R) Visual Studio 2005.
> > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> > _______________________________________________
> > Audacity-devel mailing list
> > Aud...@li...
> > https://lists.sourceforge.net/lists/listinfo/audacity-devel
> >
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft
> Defy all challenges. Microsoft(R) Visual Studio 2005.
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> Audacity-devel mailing list
> Aud...@li...
> https://lists.sourceforge.net/lists/listinfo/audacity-devel
>
>
>

Re: [Audacity-devel] click6.ny was Re: click5.ny

[David R. S. <dav...@sh...>]

On Sat, 29 Sep 2007, Samuel Humphreys wrote:

> "Escape, control+c and control+x don't work. How can I interrupt
> Audacity without rebooting?"
>
> sometimes this works:  'control alt delete' to get up the task manager,
> cursoring to the audacity process and then selecting the 'end task'
> button...(thats how you do it in win98 if I remember correctly). hmmm might
> be troublesome for the screen reader...
>
Hi Samuel,

Thanks - as you say, speech stops at that point. However, once I've 
learned which key to press, and how many times, it will be simple. :-)

Cheers

David

Re: [Audacity-devel] click5.ny

[Gale A. <ga...@au...>]

    | From "David R. Sky" <dav...@sh...> 
    | Sat, 22 Sep 2007 02:03:56 -0700 (PDT)
    | Subject: [Audacity-devel] click5.ny
    
Hi David 

Thanks -  returned herewith with some suggested modifications aimed 
mainly at making it look better for sighted users (change all/any if you 
don't agree).

On the dialogue itself, tidied up some missing square brackets and 
slightly different ;info line wording including explanation of start time 
offset. 

On the error dialogue:

* made clearer instructions for user

* capitalised ERROR (when shown) against a field to make it more 
   visible and give it a hyphen instead of colon 

* give OK when input is OK (oddly, I find that makes it much easier
  to identify the line with the error).

I am not 100% sure though if you might be better only to show the line
with the error, as even trying to improve it visually it is quite a lot to read
(and more of a pain with a screen reader if the error is in the last line)? 
OTOH it does tell them what they are allowed to input for each field. 
On balance I'd leave it as is, as long as there are no more fields than 
there are now. Or can you give them a button to press so they can read
what the valid inputs are if they want to know, then just display the line 
with the error?   

I notice that if you enter an alphabetical character in a field it treats 
the error as if you entered zero. If you enter a combined number and 
alphabetical character e.g. "3X" it shows in the error dialogue e.g. 
"ERROR - 3 beats per minute" which is of course valid as it stands. 

    | At this stage I have not yet included error MIDI pitch values in
    | error-checking since they can even be below 0 - negative infinity
    | is the MIDI value of 0hz.
    | if there's a range of MIDI pitches outside of which you'd like to
    | have an error thrown, let me know. Current lower and upper MIDI
    | pitch values are 48 through 96.

I think you can extend the allowable range significantly as long as you 
disallow something that is going to crash Audacity (as for example 922 
will). I have the attached set to minimum 18, maximum 116 and it 
sounded quite good to me at that setting (the 116 actually sounded a bit 
like a jingle bell that some people seem to want).     


Gale