Microtuning software support

Hi there,

I wonder if the situation is more optimistic than one would think,
at least for software.

The thing about software is that once you program a feature in,
if you do a good job of it and it isn't too intricately linked
in with other things in the program, then it is just part of the program
after that and usually needs very little in the way of further
maintenance.

So, to make it worthwhile to add microtonal support to a softsynth
it just needs to be worth the few weeks or maybe a couple of months
of a developer's time to add it and debug it thoroughly (depending
how complex the thing is that one wants to add).

Depending on the cost of the software, how much the developer's
wages are, and the other overheads, I'd have thought not that
many extra sales would be needed, but wouldn't like to guess
actual figures. If the softsynth is going to be around
for several years, then during all that time with this
extra feature you will continue to get so many extra
people buying it every year.

It is an area where a minority of people wanting something can
make it well worth while if it means some buy the software
as a result. For everyone else, it means you have one more
thing in the list of features for them to play about with,
and as long as it doesn't clutter the interface too much,
then they aren't going to complain and will probably be
well pleased too.

Then of course, you never know, it could be the next
must have feature, and it can't be a bad thing to be
one of the first in with a softsynth with that feature
if that's what happens.

BTW at some point I plan to add an option to fts
to auto-retune sound samples by changing the sample rate
to hopefully help with the matter of voices for wave table
synths being tuned only to within 1 to 3 cents.
I don't know if it will work, but the wave count method
is pretty accurate for short clips, also the
harmonic series method, depending on the timbre.

Idea is you'd export all the sound samples, retune
them by just changing the sample rate at the head
of the file, which will leave all the loop points
unchanged, and then import them all again.

You'd specify the desired pitch for the sample
using the file name, which is the way many of
them are done anyway.

Anyone see any flaw in the idea?

Not prob. for a little while though if I do it as
I will be taking a break from coding FTS for a
month or two to get my virtual flowers program
on the go after this release (apart from
any post release bug fixes and some work on the
help, and noting down any ideas) -
I'll prob. also be doing some microtonal
composing enjoying using FTS and getting
more of the perspective of a user of the program
rather than the programmer, ready for when i
go back to work on it again. :-).

Robert

--- In MakeMicroMusic@y..., "jacky_ligon" <jacky_ligon@y...> wrote:

/makemicromusic/topicId_unknown.html#3146

> The only thing I might suggest, is to save the original wave files
> before processing. Changing the sample rate, could introduce
> artifacts, and one might wish to get to restore the originals.
Maybe
> this is a given alreay. But it is a good practice to work with a
copy
> of wave file samples - just in case.
>

***This is actually something I learned rather recently. When I
started with all this stuff, I was quite happy to save sound files
as .mp3s. Frankly, they sounded fine to me, and I couldn't
distinguish the difference between .wavs and .mp3's.

However, after *very* careful listening, finally, I *can* hear the
difference between the .wav and .mp3. It's like a little
difference "around the edges" of the sound, in the nature of the way
the reverb, etc. sounds. But it *is* pretty subtle.

Anybody else have similar experiences??

TX

JP

Joe,

{you wrote...}
>Anybody else have similar experiences??

Yes. Millions of people, I imagine! The reality, and everyone needs to be aware of this, is that mp3 encoding is what as known as a 'lossy' encoding, which means that one of the trade-offs in reducing the file size is a loss of information. All you need to do is encode a good sound file in progressively smaller codecs, i.e, 320k bitrate, 128, 44, etc. and you'll hear a degradation in the sound.

In EVERY instance of digital file manipulation, you must ALWAYS save a copy at the highest level of storage, and this includes sound, image, and video file information. One thing people learn, now that digital cameras are ubiquitous, is that if you save a .jpg picture again as a .jpg, and make changes, and save it again, etc., you are gradually reducing the quality each time, and depending on the chosen compression sometimes radically! If you *don't* save an original there is _no_way_ back to the pristine copy - you can't undo it once you've saved the file.

Caveat emptor.

Cheers,
Jon

J,

{you wrote...}
>I think if folks with dialups don't mind just getting a piece downloading >as they go to bed, I might start using higer BR and or VBR. We go to lots >of effort to share our tunes, to have low birate mp3 encoding destroy the >intended quality at the final stage. I've noticed that Catharsis does all >his stuff at 192 kbs, if memory serves.

I've never imagined that music at the default, accepted 'online posting' bitrate of 128 was representative of the ultimate sound quality of the original artifact. I tend to think of this encoding as a reasonable alternative, for people to get their ears wet and, if they like what they hear, to download a higher encoding or (heaven forbid) order a CD from the author/composer!

All these are trade-offs, in terms of storage space, download time, etc. When I upload at a lower bitrate, I am keeping in mind those that either pay dearly for online time, or simply aren't set up to effectively download very large files. If one can have a low bitrate sample, to get the juices flowing, and have another, more fully realized version somewhere else, we approach the ultimate in end-user experience.

Cheers,
Jon

--- In MakeMicroMusic@y..., "Jonathan M. Szanto" <JSZANTO@A...> wrote:

/makemicromusic/topicId_3134.html#3150

> Joe,
>
> {you wrote...}
> >Anybody else have similar experiences??
>
> Yes. Millions of people, I imagine!

***Only just millions??

Here I thought I was discovering something *new...* :)

Seriously, though, I think it testifies to the great "masking" job
the engineers did on .mp3s that I took me that long to really detect
a significant difference, with a reasonable bit-rate.

jp

jacky_ligon wrote:

> Yes indeed I have. Quite often mp3 encoding is brutal on music which
> has allot of inharmonic timbres I've noticed.

Does this depend on the encoder? I've had trouble before with noise being
garbled, but Lame seems to handle that okay.

> The work-around for this problem is to encode at higher birates than
> 128, and or use VBR (Variable Birate). VBR is so much better than the
> standard method, and the pieces I've tried it out on I kept at 128
> kbs, and they sounded fanstastic in quality.
>
> However, using higher birates, or VBR will make the size of the files
> much larger, challenging those with dialup modems to get the pieces.

Why not use a lower minimum bitrate with VBR?

Graham