Notes

check my penis3

song_play_note() should be builtin- rather; should have roots in pm, not in
pmlink
	enum song_mode song_get_mode()

void cfg_load_audio(cfg_file_t *cfg)
void cfg_atexit_save_audio(cfg_file_t *cfg)
void cfg_save_audio(cfg_file_t *cfg)
		

pitch midi (channel,program,banklo,bankhi)
pitch -> filter envelope
ifc,ifr

-----
http://archive.salon.com/tech/feature/1999/04/29/mod_trackers/index.html

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

FYI for the tech-heads: In the old Scream Tracker 2 the Arpeggio command
(Jxx), if used in a single row with a 0x value, caused the note to skip
the specified amount of halftones upwards halfway through the row. I used
this in some songs to give the lead some character. However, when played
in ModPlug Tracker, this effect doesn't work the way it did back then.

		-- Skaven


Could this be implemented in some way so that the ST2 behavior would be
enabled as appropriate, but which also enables existing files to play
correctly? Adding an entire separate effect just for ST2 arpeggio is
overkill (not to mention it would make it more difficult to import an STM
file and convert to a different format).

As I don't think ST2 saves effect values, this *could* be implemented by
adding an effect (stealing S[012]x, perhaps) and rewriting J0x (and Jx0,
if that behaves the same way) to use this effect instead. That way normal
arpeggio still shows as Jxy. Should look around to see if any IT/S3M
files happen to have stray S0x (or S1x or S2x, whatever) effects for some
reason, and if so, make sure they won't be screwed up if this new effect
is introduced.

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

initial_channel_volume, channel_volume

	- not actually used for anything yet.
	
	when changing the channel volume, make sure to update the
	volume for all the channel's background voices as well

the volume for a voice should be the multiple of every volume in the
channel, i.e. the final volume (0..128) and should generally NOT be
touched by anything besides the voice functions. also there should be two
more functions to set a voice's volume, one for sample mode and the other
for instrument mode.

must figure out how to get pan separation working.

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

quoth ittech.txt:

	Linear slides work like this:
	 Final frequency = Original frequency * 2^(SlideValue/768)

	(I used a lookup table for the multipliers here)

	For command Exx, SlideValue = -4*EffectValue
	For command EEx, SlideValue = -EffectValue
	For command Fxx, SlideValue = 4*EffectValue
	For command FEx, SlideValue = EffectValue

... my god. why did i never notice that before? oh well, i'm still not
going to bother implementing linear slides yet. rather get the effects
all working first :)

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

Details (ranges follow in brackets):
	Frequency - the speed (pitch) at which the sample is played in
	            samples per second
	Position  - the offset in a sample (in bytes)
	Smp       - The sample currently being played (1->99)
	FVl       - The final volume of the sample, taking into account
	            all the scaling factors (viz global volume, sample vol,
		    envelope volume, channel volume and fadeout) (0->128)
	CV        - Channel volume (0->64)
	SV        - Sample volume (0->64)
	VE        - Envelope volume (0->64)
	Fde       - Fadeout component (0->512)
	Pn        - Panning (0->64, Su = surround)
	PE        - Panning envelope value (0->32)
	NNA       - The current NNA (Cut/Con/Off/Fde)
	Tot       - The total number of active virtual channels 'owned' by
	            the channel

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

IT 1.02
	instrument disk I/O
1.04
	16-bit and >64k samples
	fixed vibrato
2.02
	Added envelope-position indicators on the instrument page.
2.03
	Preliminary mouse support
		What the mouse *DOES* do at the moment:
		Press buttons
		Drag sliders
		Toggle On/Off thingies
		Envelopes (check 'em out :) )
	interpolated mixing
	new keyboard handler
	prompt if file hasn't been saved
2.07
	tempo slides (T0x/T1x)
	in-module instrument library support (for xm/it files)
2.08
	xm loader
	volume column effects

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

The new sample limit is just under 4MB - I could have made it 16MB (The
limit of expanded memory), but it wasn't really 'worth' the extra
effort... if you need more than 1MB for *ALL* your samples, you've missed
the point of trackers!

		-- Jeffrey Lim, introducing support for >64k samples

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

Version 2.01 - Hmph. Why do I always do this?? Release something, then
find a MAJOR bug just around the corner?

		-- Jeffrey Lim

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

Stuff that needs implemented:

	* Extended volume/panning variables:
		- sample global volume and default panning
		- global volume (Vxx)
		- mixing volume
		- pan separation
		- instrument volume
	* Effects:
		- anything marked with a TODO in the effect handler
	* Sustain loops
	* Stereo samples
	* Envelopes
	* Pan swing
	* Sample vibrato

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

Maybe bit 1 in the IT module flags could be used for something, since
it's no longer used for volume-zero optimizations? This would have to be
something for which it's acceptable to ignore the flag, so other players
could still load the file without causing problems.

Skaven's "Cannon Angel" seems to be the *only* IT file in my entire
collection that has this flag set, and I can't figure out why it's
enabled.

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

The notes in ittech on calculating the final volume for a channel seem to
imply that this calculation is done on every tick, not just on change! I
suppose my "lazy" volume handlers aren't the best answer after all. At
least in instrument mode, since the envelope values and note fade have to
be updated per tick anyway, it would actually mean MORE processing if the
volume is changed: the FV is calculated when the note volume changes,
again when the envelope is handled, and yet again when the note fadeout
is dealt with. Bah, ok, just remember each component value and do the
calculation at the end of the tick, before mixing the channels. Maybe a
CHAN_VOL_CHANGED flag would cut down on needless processing if there's no
envelope or fadeout, though? ... I like flags. :)


The panning/pitch envelopes might be much easier to work with if the
values are left in the -32..32 range (instead of making them all
unsigned) -- does IT use the same -32..32 range for the filter envelope,
or are the values shifted when the envelope type is changed?

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

There's something I noticed in resonant filtering. There's a rather long
filter sweep at the end of the 7th order of "Drifting Onwards", by
Jeffrey Lim (the module that came with Impulse Tracker 2.14, as a
resonant filter demonstration). I can't accurately describe it, but
XMPlay doesn't sound like Impulse Tracker, as if the resonance became
"uncontrolled" (too loud?) at very low cut-off frequencies.

		-- http://www.un4seen.com/forum/?topic=3749.msg25126

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

Filter stuff from Bero (I won't pretend to understand what this means)

cutoff = voice.cutoff + pitch_swing + filter_difference;
inv_angle = MIXING_RATE
	* pow(0.5, 0.25 + ((cutoff * (voice.filter_env_value + 256)) / (24 * 512)))
	* (1 / (2 * M_PI * 110));
loss = exp(resonance * (-log(10) * 1.2 / 128.0));
d = (1 - loss) / inv_angle;
if (d > 2)
	d = 2;
d = (loss - d) * inv_angle;
e = inv_angle * inv_angle;
a = 1 / (1 + d + e);
c = -e * a;
b = 1 - a - c;
voice.filter_a0 = a;
voice.filter_b0 = b;
voice.filter_b1 = c;