Hardsynth

[emkay]

Well, instead of turning to 15 kHz isn't it an option to just stay in 64 kHz and use a 16-bit channel for (deeper) bass notes??

 

....not that RMT does support 16bit bass, but maybe this is also a very useful pokey configuration???

863702[/snapback]

 

 

16Bit Bass (?) plus 2 channels for music is used very often. And, to be honest, there is no serious way to use filter and 16 bit together.

 

 

It is possible to use 16 bit as a "double 8bit feature" as in this tune:

 

http://atari.fandal.cz/detail.php?files_id=105

 

But it is not 16 Bit sound, because you cannot set the frequency by word, and the result is even full wrong played by the emulation.

 

As I mentioned above, it was always possible to use 15kHz Bass + percussion and a 16 bit Channel for the main voice, to have "clean" sounds.

To enhance the filter-sounds and for more variations, a combination of 15-64-1,79 is possible.

The best results for filtering is to use 15kHz. "Soft" sounds can be produced with 1,79MHz clocking.

Edited May 29, 2005 by emkay

[emkay]

This version has nearly the style of an SNK Arcade from 1992.... (no further comment)

[analmux]

........you quite like this tune hey emkay???

 

can you post the RMT of this too, then maybe I can try to recover the melody.

 

did you use a lot of filtering on the main voice? and if yes: why does the filtering detune the whole thing so much? Is there a way to tune it? Maybe a custom routine with a instrument dependent note2pitch table ??? I wonder how you should compute such new pitch tables

[emkay]

........you quite like this tune hey emkay???

 

can you post the RMT of this too, then maybe I can try to recover the melody.

 

did you use a lot of filtering on the main voice? and if yes: why does the filtering detune the whole thing so much? Is there a way to tune it? Maybe a custom routine with a instrument dependent note2pitch table ??? I wonder how you should compute such new pitch tables

863769[/snapback]

 

 

I am shurely not satisfied with the tune, because it can be done much better with the full pokey-abilities. It refers more to the "original" I know, but the full notation doesn't fit to my ears either

So, If you want to search for a better notation:

 

The "sounding-style" is that what I like very much... Perhaps I will try to do something similar to "FSR".

And, MUX, I am still not shure, what you mean with "out of tune" at all. More than to play the melody correctly isn't possible ?!

Have you really listened to most of the SID'S or AMIGA MOD's? Most of them are detuned .

Or just another example: This "Zelda Tune" which took it's notation from the "Dune" Movie.... It is detuned either and played faster to compensate that, but this is OK?

[analmux]

And, MUX, I am still not shure, what you mean with "out of tune" at all. More than to play the melody correctly isn't possible ?!

 

Well, what I mean is this: open the RMT instrument editor (of your test8 file) and try to play on the keyboard: then pressing 'z' is f.e. C-2 and ',' is C-3. When you listen carefully you hear that the interval is not an octave but something smaller (10 semitones or something). So the tonal scale seems to be skewed.

 

I noticed that in your song this happens to instruments 01,02 and 03. So I started searching and found the bug. It must be an RMT bug:

 

in the instrument editor you'll see below:

TABLE OF NOTES:
oo
15

 

now delete this note-table (simply fill in a '00' to that location) and you'll hopefully notice that now the instrument is tuned: that is when you press 'z' you actually hear a C-2 note, and when you press ',' you'll hear a C-3 note.

 

......I started thinking but when you have a table of notes starting with '15' this means that the melody will be played 21 semitones higher. Now notice that the actual melody is already very high (in the 3rd/4th octave range) so you'll get heavy detuning by the fact that the actual notes are in the 5th/6th octave range: this should matter a lot because for 15kHz the 5th and 6th octave sound like the 3rd and 4th octave in 64kHz mode, but the pitch resolution is too coarse then which results in the mentioned detuning.

 

......messing more with TABLES OF NOTES function in 15kHz mode I've noticed that this IS an RMT bug and has nothing to do with the effect I described in red

 

{@RASTER: I hope you pick up this discussion.}

[analmux]

Okay: listen to this

 

maybe the sound-color isn't that good, but I managed to tune it, and enhance the instruments 01,02 and 03: they do arpeggios with a 6 semitone interval.

[emkay]

@Analmux

 

Once upon a time .... I suggested to Raster, what really would be needed to make RMT a "POKEY-Synth-Tracker".

One part of it was a customizable notation table for "enhanced instruments".

 

Then one can build an instrument, comparing the sound to a "real" synthesizer" and set the fitting values...

 

 

Apropos: Semitones

Aren't 12 semitones an octave? By calculation it is.

By calculation, the 21 semitones are nearly 2 octaves higher, and the value is a bit lower, because of your finding.

Indeed, the tones are not all played as they should do....

To correct them with the available RMT, you can do the following:

Search for the correct pitch with a new instrument, and set the instruments note by note....

But, to do this, you must have a real Keyboard with the correct notes.

 

Another thing, that destroyes the feeling for the pitch, is the loooong preparation time. Offcourse, it would take some more cycles, but things as playing correction(command 1 00 ), Portamento Start, filter offset, etc. , would do better in less VBI steps.

When there is time for 2 pokeys to handle, it must be possible to do the double in commands per VBI for one POKEY.

[analmux]

Indeed, the tones are not all played as they should do....

To correct them with the available RMT, you can do the following:

Search for the correct pitch with a new instrument, and set the instruments note by note....

But, to do this, you must have a real Keyboard with the correct notes.

 

(??)

Well: when you use the PC keyboard as follows:

_s d   g h j   l
z x c v b n m , .

 

we see that the keys correspond to notes (in octave 2/3) like this:

'z' = C (2)

's' = C# (2)

'x' = D (2)

'd' = D# (2)

'c' = E (2)

'v' = F (2)

'g' = F# (2)

'b' = G (2)

'h' = G# (2)

'n' = A (3)

'j' = A# (3)

'm' = B (3)

',' = C (3)

 

So we actually HAVE a real keyboard.

 

-------------

 

The tuning problem can simply be solved. In 15kHz mode we must never use a 'TABLE OF NOTES' to create arpeggios: the arpeggio command for this mode is buggy.

 

Instead use command 3 (set basenote + $XY, play basenote) to 'program' arpeggios.

 

When you take this into account there's no problem with tuning anymore.

 

EmKay: Maybe you can have another look at your other RMT's: remove Tables of Notes, and program arpeggios --> change basenotes. I'll assure you it will sound a lot better at once.

[emkay]

 

So we actually HAVE a real keyboard.

 

 

 

 

 

I know it's a little late here in europe

Ofcourse RMT has a keyboard built in.

This keyboard uses the standard notation, which is also a bit "off" when it come to "tuned" ...

Let's now search for a new instrument by filter settings...:

 

<loose mode>

 

You use "A1" and set the filter to "C2".... the note you hear is NOT A or C, it's a resulting tone..... perhaps near to "E2"

So only a professional musician can hear the real pitch.... Or some Guy with a "tuned" keyboard, comparing the results...

 

Perhaps it is precalculatable....

 

 

</loose mode>

[analmux]

Okay, now I understand what you mean.

 

Filtering will (now) only work if the filtering-frequencies are in the range of $00-$0f and $f0-$ff. Other Filtering Frequencies indeed result in bad harmonics, so you'd need a dynamic filtering mode, where the filtering frequency depends on the voice frequency (f.e. with a fixed ratio of the two frequencies)

 

-------------

.....and yes, it is late here: I'm going to sleep now, and dream of pokey music.

 

I finally have a vision now of how (new) pokey music can/should sound.

[Schmutzpuppe]

Cool sounds guys.

While emkay creates fancy instruments analmux managed to get them in tune.

Perfect combination

Keep it up guys.

[emkay]

Cool sounds guys.

While emkay creates fancy instruments analmux managed to get them in tune.

Perfect combination 

Keep it up guys.

864205[/snapback]

 

 

I really wish, there were more people taking part of this discussion.

 

 

@Analmux

 

There is another way to optimize sounding after an Instrument is set.

Just search for a notation that fits best to the sound

 

Here something "different" ... OK... it hasn't nothing to do with the Zelda tune, but it's a suggestion.

 

Just another suggestion : Try to use one of "my" instruments and play the "Zelda" tune to have the "same music feeling".... fitting to the tune at the end.

Edited May 30, 2005 by emkay

[raster/c.p.u.]

......I started thinking but when you have a table of notes starting with '15' this means that the melody will be played 21 semitones higher.

 

Nice idea. But ... you have to see to the parameter TABLE TYPE, which is set to 1 => above the '15' you can see the info text "TABLE OF FREQS", no "TABLE OF NOTES".

So, hexa value '15' in this case means that decimal value 21 will be added to frequence, no shifting about 21 semitones.

 

......messing more with TABLES OF NOTES function in 15kHz mode I've noticed that this IS an RMT bug and has nothing to do with the effect I described in red

{@RASTER: I hope you pick up this discussion.}

 

I hope there isn't bug in RMT.

[analmux]

 

I'm sorry Raster:

Indeed it was a Table of Frequencies. Adding 21 to the frequencies will indeed result in a detuned instrument.

 

...so there's no bug in RMT

[emkay]

OK... another day, another try.

Now with some arpeggio plus "corrected" semi-tone setting

[analmux]

okay, here I've attached mux-version 2 of zeldaxy3: I've changed the arpeggio of instruments "3,7,8,9,a" a little.

 

...and emkay: please note that I was mistaken: there's no bug in the TABLE OF NOTES so you can as well use this to create arpeggios.

 

-----------

 

the second file is a very old one: emkays guitars, now reshaped in a 15kHz way.

[analmux]

........oooops: I should be working on the Mario Clone instead

 

[analmux]

@ emkay:

 

you once wrote that you can make a clear sine-wave flute sound on the pokey.....how did you do that?

 

---------

 

to anyone: what's the difference between generator A and E?.....or in other words: why should I use them both? Is it possible to sync pokey with a switch between gen. A and E?

[emkay]

@ emkay:

 

you once wrote that you can make a clear sine-wave flute sound on the pokey.....how did you do that?

 

 

 

 

 

It's a long time ago when I did some of those experiments on the A8, and I still would have to rebuild that.

Today I'm pretty shure, it's in the 1,79MHz clocking which is not to handle in RMT... neither by the used emulation nor by RMT itself.

I made some quick changes for the 1,79MHz feature.... perhaps you hear some sounds out of it, that make something more clear ...

 

Example: You can make pretty clear noise free "up" sounds with it....

 

You could do arpeggios with the 15kHz filters for astonishing sounds... a.s.o.

 

but ...

Edited June 1, 2005 by emkay

[emkay]

Doing some "Lottery" on the tune with 1,79MHz clocking.... here the result:

[analmux]

@ EmKay

 

Yes, the sound of your last RMT is very funny, but unfortunately the frequency-tables of RMT are not suitable for the 1.79 MHz mode. Maybe we can make a patch for a better table.

 

gr.

Mux

[emkay]

Another 1,79MHz test

[emkay]

Final thingy with this setting

[analmux]

emkay, everytime I get curious how you create those instruments...can you send me the RMT file for these cool flutes.

 

I think it's time to throw away all the test tunes and write some real music with the new instruments......it seems pokey is far more suitable for real music than I've ever thought.

Edited June 5, 2005 by analmux

[Heaven/TQA]

isnt it more like a violin than a flute?