Re: [PD] Pd double precision testing.

[William Huston]

This makes me very happy!

I will be happy to test a Windows version of this! :)




--
William Huston:  williamahus...@gmail.com
Binghamton NY

*Public Service Mapping / Videography / Research / Education / Safety
Advocacy*
Blog  -- Facebook
 -- Twitter
-- Youtube

* -- Podcast Blog *
*Document collections*: VirtualPipelines
 -- BHDCSDimockArchive

*Please support my work! -- *TinyURL.com/DonateToBillHuston





On Fri, Dec 6, 2019 at 5:08 AM Dan Wilcox  wrote:

> There is a double precision test build for macOS here:
> http://docs.danomatika.com/pdbuilds/0.51/
>
> As development continues, I will throw additional test builds up there.
>
> On Dec 6, 2019, at 12:13 AM, pd-list-requ...@lists.iem.at wrote:
>
> Hi all,
>
> There is an experimental branch[1] on Pd-double-precision[2].
>
> So far we know a couple of things not working yet like writing audio
> files with [soundfiler] or [writesf~].
>
> If anyone is willing to test and find something else not working we
> welcome and appreciate it.
>
> Here's a Windows64 version for tests :
> http://lucarda.com.ar/x/Pd-double-w64-testing.zip
>
> Tests should be done with "vanilla-only" patches. 99% of externals were
> not compiled for double-precision so they don't work with this Pd.
>
> If you find something not working write back here to the list or issue
> it to github (prepend "double-precision: " on the name of the issue).
>
> [1] https://github.com/pure-data/pure-data/tree/double-precision
> [2] http://www.katjaas.nl/doubleprecision/doubleprecision.html
>
>
> 
> Dan Wilcox
> @danomatika 
> danomatika.com
> robotcowboy.com
>
>
>
> ___
> Pd-list@lists.iem.at mailing list
> UNSUBSCRIBE and account-management ->
> https://lists.puredata.info/listinfo/pd-list
>
___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] store a symbol and send on bang

[Dan Wilcox]

How about [value -s] instead? :)

I just whipped up a [value] update which changes the backing value storage to 
use a t_atom and allows setting the value type based on a flag: -s makes a 
symbol value, otherwise a float value. There are some checks against type 
mismatches between the same name.

This allows for sending & receiving symbols via [value -s] or [v -s] objects. 
Existing [value] and [v] objects are float types.

The only issue I can see, other than a little more memory usage and type 
checking, would be possibly breaking patches which use "-s" as a value name. In 
this case, sending a float to the inlet is silently ignored and a default 
"symbol" symbol would be output on a bang which would cause a trackable error.

I can put up a PR soon when I clean up this change a bit. In the meantime, 
macOS people can try out a test build: Pd-0.50.2-value-symbol.zip 


> On Dec 6, 2019, at 5:50 PM, pd-list-requ...@lists.iem.at wrote:
> 
> From: IOhannes m zmölnig mailto:zmoel...@iem.at>>
> To: pd-list@lists.iem.at 
> Subject: Re: [PD] store a symbol and send on bang
> Message-ID:  >
> Content-Type: text/plain; charset=utf-8
> 
> Am 6. Dezember 2019 15:27:02 MEZ schrieb Christof Ressi 
> mailto:christof.re...@gmx.at>>:
>> This would just complicate the interface. Just use [symbol] + [send] and 
>> you're done.
> 
>  
> +1
> 
> also, [symbol -s] would make it quite complicated to create a symbol "-s"...
> 
> 
> 
> mfg.hft.fsl
> IOhannes


Dan Wilcox
@danomatika 
danomatika.com 
robotcowboy.com 



___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] Pd double precision testing.

[Alexandre Torres Porres]

done

Em sex., 6 de dez. de 2019 às 14:32, Lucas Cordiviola 
escreveu:

> Alex,
>
> Tell the facebook people that we know there are troubles writing 16 or
> 24 bit audio files and for the linux users that they have to compile the
> branch with:
>
> ./autogen.sh
> ./configure CPPFLAGS=-DPD_FLOATSIZE=64
> make
> make app
>
>
> --
>
> Mensaje telepatico asistido por maquinas.
>
> On 12/6/2019 1:49 PM, Alexandre Torres Porres wrote:
> > awesome, just posted binaries on the facebook group so people can test it
> >
> > Em sex., 6 de dez. de 2019 às 07:07, Dan Wilcox  > > escreveu:
> >
> > There is a double precision test build for macOS here:
> > http://docs.danomatika.com/pdbuilds/0.51/
> >
> > As development continues, I will throw additional test builds up
> > there.
> >
> >> On Dec 6, 2019, at 12:13 AM, pd-list-requ...@lists.iem.at
> >>  wrote:
> >>
> >> Hi all,
> >>
> >> There is an experimental branch[1] on Pd-double-precision[2].
> >>
> >> So far we know a couple of things not working yet like writing audio
> >> files with [soundfiler] or [writesf~].
> >>
> >> If anyone is willing to test and find something else not working we
> >> welcome and appreciate it.
> >>
> >> Here's a Windows64 version for tests :
> >> http://lucarda.com.ar/x/Pd-double-w64-testing.zip
> >>
> >> Tests should be done with "vanilla-only" patches. 99% of
> >> externals were
> >> not compiled for double-precision so they don't work with this Pd.
> >>
> >> If you find something not working write back here to the list or
> >> issue
> >> it to github (prepend "double-precision: " on the name of the
> issue).
> >>
> >> [1]https://github.com/pure-data/pure-data/tree/double-precision
> >> [2]http://www.katjaas.nl/doubleprecision/doubleprecision.html
> >
> > 
> > Dan Wilcox
> > @danomatika 
> > danomatika.com 
> > robotcowboy.com 
> >
> >
> >
>
___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

[PD] 64-bit accumulator external

[Charles Z Henry]

The double precision testing thread leads me to comment on an external
I've been writing.  I decided I needed an external (with 32-bit
signals input) that would add up blocks of signals over time with an
internal 64-bit accumulator to avoid losing precision during this
step.

But as I thought about other options for it to perform (sample-wise
accumulation, sum of squares, whether the output should be
normalized), it started turning into a real swiss army knife, one
object with a ton of configuration choices, outputs, and control
options.  It could be used to read long tables with tabread4~~ for
instance, or turn distributions into normalized cumulative
distribution functions.  There would be the option of retrieving the
full 64-bit value on the output through pairs of signals/messages or
truncating the precision back to 32-bits with a single output.

When signals fan in to a inlet they are summed together in 32-bits, so
that has to be avoided.  Should I have it use a variable number of
inlets or make new objects that have multiple inlets?

Yikes... what am I doing?  Is there a different question here?  What
set of 64-bit externals could do all the things a double-precision Pd
*could* do?  You only really lose precision when doing certain things.
Is there a better way to organize the functions for usage among
several objects?  Which ones are essential?  What set would make it
complete?  Yikes...

In the same manner as the PdCUDA signal memory space and perform
routine duplication (I designed but never published an
implementation), you could control whether objects used a 32-bit
signal memory space and dsp functions vs a 64-bit version of the same.
You'd control the behavior through the dspcontext struct and add an
object to switch it on/off in subpatches, thus using the double
precision only when you need it.  Array design and implementation
remains problematic.  This is a terrible idea.

I'll go ahead and build something for my immediate usage (after my
classes conclude in 2 weeks) that does just a few things, but I have
to wonder if there's something better there that should be done
instead.  Any advice?

Chuck



___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] Pd double precision testing.

[Lucas Cordiviola]

Alex,

Tell the facebook people that we know there are troubles writing 16 or 
24 bit audio files and for the linux users that they have to compile the 
branch with:

./autogen.sh
./configure CPPFLAGS=-DPD_FLOATSIZE=64
make
make app


--

Mensaje telepatico asistido por maquinas.

On 12/6/2019 1:49 PM, Alexandre Torres Porres wrote:
> awesome, just posted binaries on the facebook group so people can test it
>
> Em sex., 6 de dez. de 2019 às 07:07, Dan Wilcox  > escreveu:
>
> There is a double precision test build for macOS here:
> http://docs.danomatika.com/pdbuilds/0.51/
>
> As development continues, I will throw additional test builds up
> there.
>
>> On Dec 6, 2019, at 12:13 AM, pd-list-requ...@lists.iem.at
>>  wrote:
>>
>> Hi all,
>>
>> There is an experimental branch[1] on Pd-double-precision[2].
>>
>> So far we know a couple of things not working yet like writing audio
>> files with [soundfiler] or [writesf~].
>>
>> If anyone is willing to test and find something else not working we
>> welcome and appreciate it.
>>
>> Here's a Windows64 version for tests :
>> http://lucarda.com.ar/x/Pd-double-w64-testing.zip
>>
>> Tests should be done with "vanilla-only" patches. 99% of
>> externals were
>> not compiled for double-precision so they don't work with this Pd.
>>
>> If you find something not working write back here to the list or
>> issue
>> it to github (prepend "double-precision: " on the name of the issue).
>>
>> [1]https://github.com/pure-data/pure-data/tree/double-precision
>> [2]http://www.katjaas.nl/doubleprecision/doubleprecision.html
>
> 
> Dan Wilcox
> @danomatika 
> danomatika.com 
> robotcowboy.com 
>
>
>
___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] Pd double precision testing.

[Alexandre Torres Porres]

awesome, just posted binaries on the facebook group so people can test it

Em sex., 6 de dez. de 2019 às 07:07, Dan Wilcox 
escreveu:

> There is a double precision test build for macOS here:
> http://docs.danomatika.com/pdbuilds/0.51/
>
> As development continues, I will throw additional test builds up there.
>
> On Dec 6, 2019, at 12:13 AM, pd-list-requ...@lists.iem.at wrote:
>
> Hi all,
>
> There is an experimental branch[1] on Pd-double-precision[2].
>
> So far we know a couple of things not working yet like writing audio
> files with [soundfiler] or [writesf~].
>
> If anyone is willing to test and find something else not working we
> welcome and appreciate it.
>
> Here's a Windows64 version for tests :
> http://lucarda.com.ar/x/Pd-double-w64-testing.zip
>
> Tests should be done with "vanilla-only" patches. 99% of externals were
> not compiled for double-precision so they don't work with this Pd.
>
> If you find something not working write back here to the list or issue
> it to github (prepend "double-precision: " on the name of the issue).
>
> [1] https://github.com/pure-data/pure-data/tree/double-precision
> [2] http://www.katjaas.nl/doubleprecision/doubleprecision.html
>
>
> 
> Dan Wilcox
> @danomatika 
> danomatika.com
> robotcowboy.com
>
>
>
>
___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] store a symbol and send on bang

[IOhannes m zmölnig]

Am 6. Dezember 2019 15:27:02 MEZ schrieb Christof Ressi :
> This would just complicate the interface. Just use [symbol] + [send] and 
> you're done.

 
+1

also, [symbol -s] would make it quite complicated to create a symbol "-s"...



mfg.hft.fsl
IOhannes


___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] store a symbol and send on bang

[Alexandre Torres Porres]

Em sex., 6 de dez. de 2019 às 07:56, Christof Ressi 
escreveu:

> > [value] and [float] also have a built in send function,
>
> [value] doesn't have a [send( method - yet!
> https://github.com/pure-data/pure-data/pull/606
>

Oh yeah, I'm jumping ahead of time here :) I thought this had been merged
already, sorry ;)
___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] store a symbol and send on bang

[Christof Ressi]

This would just complicate the interface. Just use [symbol] + [send] and you're done.

 

OTOH, I have a made a PR to add more methods to [list store] and it includes a [send( method: https://github.com/pure-data/pure-data/pull/594

 

Christof

 
 

Gesendet: Freitag, 06. Dezember 2019 um 14:49 Uhr
Von: "José de Abreu" 
An: "PD list" 
Betreff: Re: [PD] store a symbol and send on bang


One idea for [symbol] to have send method is to give it some flag, maybe? something like [symbol -s] so it understand "send something"?
 

Just a noob idea, don't know if it makes sense

 


Em Sex, 6 de dez de 2019 07:54, Christof Ressi  escreveu:




> [value] and [float] also have a built in send function,

 

[value] doesn't have a [send( method - yet! https://github.com/pure-data/pure-data/pull/606

 
Christof

Gesendet: Donnerstag, 05. Dezember 2019 um 21:18 Uhr
Von: "Alexandre Torres Porres" 
An: "Csaba Láng" 
Cc: Pd-List 
Betreff: Re: [PD] store a symbol and send on bang


the [symbol] object stores symbols, you can connect it to a [send] object.
 

[value] and [float] also have a built in send function, maybe we could include one too for [symbol], but it would break things as giving it a message "send x" these days just turns "send" to "symbol send".

 


Em qui., 5 de dez. de 2019 às 15:48, Csaba Láng  escreveu:



Dear list,

 

is there a way to store a symbol and send it on a bang as the [value] or [float] objects do it with floats?

 

Popesz

___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list

___ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list





___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list

___ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list






___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] store a symbol and send on bang

[José de Abreu]

One idea for [symbol] to have send method is to give it some flag, maybe?
something like [symbol -s] so it understand "send something"?

Just a noob idea, don't know if it makes sense

Em Sex, 6 de dez de 2019 07:54, Christof Ressi 
escreveu:

> > [value] and [float] also have a built in send function,
>
> [value] doesn't have a [send( method - yet!
> https://github.com/pure-data/pure-data/pull/606
>
> Christof
> *Gesendet:* Donnerstag, 05. Dezember 2019 um 21:18 Uhr
> *Von:* "Alexandre Torres Porres" 
> *An:* "Csaba Láng" 
> *Cc:* Pd-List 
> *Betreff:* Re: [PD] store a symbol and send on bang
> the [symbol] object stores symbols, you can connect it to a [send] object.
>
> [value] and [float] also have a built in send function, maybe we could
> include one too for [symbol], but it would break things as giving it a
> message "send x" these days just turns "send" to "symbol send".
>
> Em qui., 5 de dez. de 2019 às 15:48, Csaba Láng 
> escreveu:
>
>> Dear list,
>>
>> is there a way to store a symbol and send it on a bang as the [value] or
>> [float] objects do it with floats?
>>
>> Popesz
>> ___
>> Pd-list@lists.iem.at mailing list
>> UNSUBSCRIBE and account-management ->
>> https://lists.puredata.info/listinfo/pd-list
>
> ___ Pd-list@lists.iem.at
> mailing list UNSUBSCRIBE and account-management ->
> https://lists.puredata.info/listinfo/pd-list
> ___
> Pd-list@lists.iem.at mailing list
> UNSUBSCRIBE and account-management ->
> https://lists.puredata.info/listinfo/pd-list
>
___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] writing to cyclone/coll takes a long time

[Jakob Laue]

Hi all!
@IOhannes, msgfile looks promising, we testet it and it writes really fast.

Regarding iemmatrix: It looks like it is not possible to change the size of matrices dynamically. But we definitely need this functionality.

Hence, our first idea now is to implement a wrapper-abstraction around [msgfile] that works like [coll], so that we hopefully can more or less replace our current [coll] objects with our abstraction within our main patch. One issue that comes into our minds here is that [coll] offers a way of referencing one buffer at different places in a patch. E.g. you can have different [coll myMatrix] objects and they all refer to the same collection of data. We do have various [coll] objects in our patch as well, all of them referencing to the same data matrix. In your opinion, what would be the best approach to implement this referencing-mechanism with [msgfile]?

 

Thanks, Hajar and Jakob
 
 
 

Gesendet: Donnerstag, 05. Dezember 2019 um 13:37 Uhr
Von: "IOhannes m zmoelnig" 
An: pd-list@lists.iem.at
Betreff: Re: [PD] writing to cyclone/coll takes a long time
On 30.11.19 14:00, Jakob Laue wrote:
> Do you have any hint why it takes that long to write 6 rows into coll? Maybe we do something wrong or have missed some smarter way to write our rows into the coll matrix.

i don't know what you want to do, but if this is about "writing" (to
file), you could also use zexy's [msgfile].


some numbers (on an i7-7700, running Debian/GNU linux, 64bit)

- adding 700 entries: <1ms
- adding 6000 entries: <11ms
- adding 6 entries: <50ms
- writing 6 entries to disk: <150ms
- reading 6 entries from disk: <130ms
- output next entry: <1ms
- seek to entry #2 <3ms

outputting the "next" entry is O(1).
seeking to a position, resp, searching an entry is O(n).


dfmasdr
IOhannes

PS: there's also a memory leak that ed.kelly discovered recently, that
occurs when dealing with long lines. need to put a new version online

___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list



___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] store a symbol and send on bang

[Christof Ressi]

> [value] and [float] also have a built in send function,

 

[value] doesn't have a [send( method - yet! https://github.com/pure-data/pure-data/pull/606

 
Christof

Gesendet: Donnerstag, 05. Dezember 2019 um 21:18 Uhr
Von: "Alexandre Torres Porres" 
An: "Csaba Láng" 
Cc: Pd-List 
Betreff: Re: [PD] store a symbol and send on bang


the [symbol] object stores symbols, you can connect it to a [send] object.
 

[value] and [float] also have a built in send function, maybe we could include one too for [symbol], but it would break things as giving it a message "send x" these days just turns "send" to "symbol send".

 


Em qui., 5 de dez. de 2019 às 15:48, Csaba Láng  escreveu:



Dear list,

 

is there a way to store a symbol and send it on a bang as the [value] or [float] objects do it with floats?

 

Popesz

___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list

___ Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list






___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] Pd double precision testing.

[Dan Wilcox]

There is a double precision test build for macOS here: 
http://docs.danomatika.com/pdbuilds/0.51/ 


As development continues, I will throw additional test builds up there.

> On Dec 6, 2019, at 12:13 AM, pd-list-requ...@lists.iem.at wrote:
> 
> Hi all,
> 
> There is an experimental branch[1] on Pd-double-precision[2].
> 
> So far we know a couple of things not working yet like writing audio 
> files with [soundfiler] or [writesf~].
> 
> If anyone is willing to test and find something else not working we 
> welcome and appreciate it.
> 
> Here's a Windows64 version for tests : 
> http://lucarda.com.ar/x/Pd-double-w64-testing.zip 
> 
> 
> Tests should be done with "vanilla-only" patches. 99% of externals were 
> not compiled for double-precision so they don't work with this Pd.
> 
> If you find something not working write back here to the list or issue 
> it to github (prepend "double-precision: " on the name of the issue).
> 
> [1] https://github.com/pure-data/pure-data/tree/double-precision 
> 
> [2] http://www.katjaas.nl/doubleprecision/doubleprecision.html 
> 

Dan Wilcox
@danomatika 
danomatika.com 
robotcowboy.com 



___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] Knowing if Pd is in Edit mode when instantiating an object (was Query if Pd is in Edit mode for a GUI)

[IOhannes m zmölnig]

Am 6. Dezember 2019 09:16:19 MEZ schrieb Alexandre Torres Porres 
:
>answer: just ask for "canvas"->gl_edit
>


how to that on pd-list (as opposed to pd-dev)?


mfg.hft.fsl
IOhannes


___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list

Re: [PD] Knowing if Pd is in Edit mode when instantiating an object (was Query if Pd is in Edit mode for a GUI)

[Alexandre Torres Porres]

answer: just ask for "canvas"->gl_edit

Em qui., 28 de nov. de 2019 às 11:09, oliver  escreveu:

> IOhannes m zmölnig wrote:
> > Am 28. November 2019 14:18:22 MEZ schrieb Alexandre Torres Porres <
> por...@gmail.com>:
> >>
> >> Sure, but there are other ways to instantiate an object, like when
> >> opening
> >> a patch or with dynamic patching, where PD won't switch to edit mode.
> >>
> >
> > but with dynamic patching
> > - you are breaking a lot of interaction expectations
> > - you can control the edit-mode state at your will
> >
>
> i.e. you can "force" edit mode with a "editmode 1" message to the
> corresponding namecanvas, and also disable it with a dedicated message
>
> (see attached examples)
>
> best
>
> oliver
> ___
> Pd-list@lists.iem.at mailing list
> UNSUBSCRIBE and account-management ->
> https://lists.puredata.info/listinfo/pd-list
>
___
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> 
https://lists.puredata.info/listinfo/pd-list