back to list

Re: Cents to ratios

🔗Robert Walker <robertwalker@...>

6/4/2003 12:26:13 AM

HI Paul,

BTW I just looked again at the code for the javascript page - it simply
goes through the quotients and finds the nearest approximating ratio above and
below for that quotient as I described it. But that isn't the best thing
to do because it could miss out ratios that are a bit away from the
nearest integer for that quotient - more likely to happen when numbers
involved are large.

Looking at the code for FTS, I see that it goes through the possible denominators first
and then for each one works out
the range of possible values for the denumerators from the tolerance, then goes through and
tests each one.

It also uses the lattice method too, choosing one or the other depending on which is
likely to be fastest. So the FTS coding is more sophisticated and faster and
won't miss any ratios out.

So FTS will find ratios that will get missed out by that javascript page, though the
javascript will be faster becuase it doesn't do so much seraching. I will update it
presently to do the search through all denumerators in range for the given tolerance as
an option - not hard to do at all.

Thanks,

Robert

🔗Porres <decuritiba@...>

6/4/2003 8:01:17 AM

Gee, I just had a look on the applet page by Robert Walker... and
that's the sort of thing I was thinking about develop with my
database, I have no mathematical background so I have no clue about
the different ways of getting ratio proportions from cents, (farey
series, ect...), I'll try and check this theories better, but then my
first idea of a use for my database blows away, I guess the only use
extremely related to a database should be the names of several
intervals, that would explain a table, but then one should spend a
great deal of time copuing the names from a book to the form of an
eletronic database...

I know that you can check names from the databases of the Scala
software, but I guess that database has other main interests than the
names of intervals, is that correct? And also, is there a separate
database with interval names, cents and ratio from Scala? Or one
should get each scala file and compile them into a database?

cheers

Alex Porres!

--- In MakeMicroMusic@yahoogroups.com, "Robert Walker"
<robertwalker@n...> wrote:
> HI Paul,
>
> BTW I just looked again at the code for the javascript page - it
simply
> goes through the quotients and finds the nearest approximating
ratio above and
> below for that quotient as I described it. But that isn't the best
thing
> to do because it could miss out ratios that are a bit away from the
> nearest integer for that quotient - more likely to happen when
numbers
> involved are large.
>
> Looking at the code for FTS, I see that it goes through the
possible denominators first
> and then for each one works out
> the range of possible values for the denumerators from the
tolerance, then goes through and
> tests each one.
>
> It also uses the lattice method too, choosing one or the other
depending on which is
> likely to be fastest. So the FTS coding is more sophisticated and
faster and
> won't miss any ratios out.
>
> So FTS will find ratios that will get missed out by that javascript
page, though the
> javascript will be faster becuase it doesn't do so much seraching.
I will update it
> presently to do the search through all denumerators in range for
the given tolerance as
> an option - not hard to do at all.
>
> Thanks,
>
> Robert

🔗Robert Walker <robertwalker@...>

6/4/2003 11:39:21 AM

Hi Alex,

Ah well, hope you find the page useful!

Thanks,

Robert

🔗Robert Walker <robertwalker@...>

6/4/2003 11:39:20 AM

Hi there,

Just to say I uploaded a new version of that javascript page with the bug fix:

http://tunesmithy.netfirms.com/japplets/ratios_with_factors.htm

Also while at it - added a new field to show the factorisation.

Robert

🔗Jonathan M. Szanto <JSZANTO@...>

6/4/2003 2:13:11 PM

Robt,

Sorry for being public: did you receive a note from me, about a week or so ago, about retrieving Yahoo archives? If not, would you send me a short note offlist and I'll resend, just to make sure the mail is getting to you...

TIA,
Jon