Hextracker (updated to 0.849B + BPM preview)

[ljbk]

Hi !



In one month, there will be 21 years (!) since the day i did release the Overscan Demos back in 1990.
For sure at that time, i had no idea i would still be releasing some ST application that long time after ...

Well, 19 years, almost to the day, after a preview of this program was shown to some guys at the
Megaleif party in Uppsala(Sweden) during Easter 1992 (http://www.demoparty.net/megaleif1992/) where i was invited to by Michael "Daryl of TEX" Raasch, i am (finally) releasing Hextracker to the ST users.

This program was built during 1991/92 (the last date on the original source file points to July 1992)
but the core virtual mix/dma is from 1991.
The program idea was to break some of limits existing at the time allowing:
- fast and quality replay routs;
- up to 16 voices mixing on STF with full volume control;
- sample size up to 24 bits; (only 23 bits from 0.846 on !)
- 255 samples;
- 256 patterns;
- variable sample base frequency to allow better quality samples;

Initially the idea was to make it a commercial product.
But i was told by several editors, one of them was Marc Rosocha (if i remember right), that there was too much competition from the Amiga and PC trackers and so there was no market chance. Some guys at the Megaleif also told me the same thing in behalf of Marc.
So after a few months, as i started working for a telecom firm, i decided to put this project on ice ...
It remained there for about 12 years.

In 2004, with the cheap internet boom, i discovered the ST community on the net and started having a look at the MOD evolution with the years.
So at the end of 2004, i implemented all the missing Protracker commands that could fit on the replay strategy from 1991/92 and built YM50K.
With a PC, during 3 months, i managed to measure the YM 2149 output levels in order to generate more accurate replay tables that allowed me to release YM50K in March 2005.
But still there was no will to finish Hextracker ... so, after including the direct import of Fast Tracker
modules, it returned to the ice.

Now in March 2011, again i started looking for some Fast Tracker MODs to at least ear how they would sound on my STF with my program because of course back in 1991/92 there was no Fast Tracker. The maximum i tested were some Oktalyser MODs up to 8 voices: a 6 voices example was shown to some guys in Upsalla.
As i found some little bugs with the Protacker digisynth routine i updated YM50K and published it again here.

That is when i found out that there are some guys like "yerzmyey" still doing some MOD work on STF !
So i decided to try to release something about this project.
So, in less than 2 weeks, i did:
- write a quick reference manual (if you can call it that );
- kill one or two known bugs (don't worry there are probably still more );
- debug the STE output with STeem as i have no STE (not sure if it works on a real STE);
- write a small replay program for anyone to use;


Anyway, here it is now.


Please remember that:

- this is a "FOR EVER BETA" product provided "AS IT IS";
- it should work with a 512 KB STF (hopefully E) but 1 MB is better and can handle from 1 to 16 voices in colour or high res mono 640x400 / 71 Hz;
- the STE output was tested with STeem (with a virtual ST machine up to 32 MHz );
- the virtual mix/dma core was STF built, with the STE mixing on top of it, so it is not near as fast as
Lance's one thas was STE built;
- some Protracker features/commands: BPM, finetune, set finetune ($E5), and funk it ($EF) are not
suported;
- the included code can be used by anyone to replay MODs in Hextracker or supported formats;
- please have a look at the reference text file;
- have a test at it on real HW, if you can;

Some emulator/technical notes:

- The application has some problems on STE mode with SainT 2.14 but Leonard is taking care of that bug ... Thanks ! ;
- The small replay program has some problems with STeem 3.2, both with STF and STE outputs, when synching to the VBL or to one specific scan line but works synching on a 50 Hz timer (like for mono mode replay): this is a STeem bug as the main application (using the same code) works and as it works on real STF HW (EDIT: This NOT a STeem bug: simply STeem wakes up in NTSC 60 Hz (that i did not know !) and so the problem is solved with version 0.832);
- The application was not tested on other emulators;
- You can push the application in STeem 3.2 to have 16 voices with all options on in stereo at 50 KHz
with a virtual 32 MHz STE;



That is all for this version 0.825 ! Will there be any one more ...



Enjoy it,
Paulo.


PS: Edited with attachment of last version available !

From version 0.849 on, ALL previously saved files in Hextracker format have to be updated changing the reference samples frequency (was FSAMPL and is now C2 REF). The new value has to be 0.8408964 times the old one.
This is because the reference note has changed from D#2 to C-2 !!!
But keep calm, for Amiga type samples, the new value is 8287 if you want to emulate an Amiga working in PAL or 8363 if you want to emulate an Amiga working in NTSC.
For Octalyser compatible values, use 10000 for 20 KHz base and 12500 for 25 KHz base.
This change was done for clarity reasons as most of the other programs use a C note as reference.
If you never saved a file in Hextracker format then forget this.

[nativ]

Cheers

I write 4/6/8 channel mod's in Protracker f030 octalyser flextracker and ace.

I have ST FM / E AND Falcon (possibly under Backward) to test, I will let you know how it works!

nativ

[ljbk]

Cheers

I write 4/6/8 channel mod's in Protracker f030 octalyser flextracker and ace.

I have ST FM / E AND Falcon (possibly under Backward) to test, I will let you know how it works!

nativ

Thanks !

The YM output should not work on Falcon or above due to the shadow addressing of the YM2149 but you can change that in the replay source example.
But if the STE ouput works, then maybe that one will work on newer models.
As the prog is from the early 90s, there are also some other output modes for dedicated HW existing at the time like Prosound, ST Replay, Replay Pro, MV 16, ST Replay Stereo and FAFB Stereo but they were not tested with the respective specific HW. I remember to rip the code from several places at the time (shame on me ).

Bye,
Paulo.

[nativ]

Paulo,

MV16 very nice! Monster Stereo too? lol!

I have Stereo replay cart, replay16 and replay8?

good to know about shadow addesses! There is a ST emulator on the Falcon that might run it especially if as you say it's only a smallish fix.

My mobile won't download at the moment will grab it tomorrow and look over the weekend!

thanks again for the release!

Shame you were put off earlier, An Cool (tcb tracker) had a different trouble!

Cheers

nativ / Ashley

[dma]

Wow 16 voices ST soundtracker, this looks really interesting!
Thanks for releasing it, i can't wait to try it.

[spiny]

cool ! I'm no musician, but have pointed 'rhino' to the thread

[ljbk]

About the 16 voices ...

Please keep in mind that it is a maximum and even removing all the options, that means with 100% CPU at 8 MHz with interrupts ON, you can only reach about 10 KHz for STF and STE stereo and 12.5 KHz for STE mono.
The "composing" frequency, which means with at least data scrolling, will be about 6 KHz which is not very quality friendly.
Of course, you can use STeem at 32 MHz and always use what you want at 50 KHz stereo !

There is an example included, a Fast Tracker 1 mod, in the FST_MOD folder.
To try it, load it, disable all options and select 9.83 KHz on STF/YM output or 12.5 KHz for STE mono or 10 KHz for STE stereo and set V/Q to $4A in mono or $39 in stereo and select PLAY !

About Fast Tracker 1 mods, for the moment Hextracker can not handle the non standard octaves, so if you load a mod that uses a non standard octave you will get some off-key sounds or even crash the application (i hope not). That is explained in the reference text => just have a read at it !

Any voice number from 1 to 16 is suported, like 3 or 5 or 10 or 2 (there is another Fast Tracker example for 2).

Bye,
Paulo.

[Rhino]

Hey Paulo,

I was at Megaleif in 1992! It's so long ago that I can't remember whether I saw your tracker prototype or not, but it's great news that you've finally released it! Better late than never. I look forward to using it.

David Newman (Rhino)

[ljbk]

Hey Paulo,

I was at Megaleif in 1992! It's so long ago that I can't remember whether I saw your tracker prototype or not, but it's great news that you've finally released it! Better late than never. I look forward to using it.

David Newman (Rhino)

Hi !

The preview was not spread, just shown to a bunch of guys at the school otherwise on the same day the tricks used would have been ripped. That is not what was intended when you think about going commercial with a product. So i took my 820K floppy, inserted it in some ST or even STE i think an run the prg. At the time the guys i was showing that to were already only thinkg about the STE having a prog with 8 voices at 25 KHz so ok ...
Their first reaction was: it only a 6 voices tune. I said, ok but that is the best one of these Amiga Oktalyser demo tunes.
Anyway, i remeber i was not too annoyed, because i just went to the party to meet a few people i just knew from the Demo scene or from writing to like Daryl (TEX) who invited me, Gunstick and An Cool. It was nice to see the atmosphere there. But i did no coding at all or bring my machine. I just flew from Berlin were i was on an "On the Job" training to spend Easter there, meet some guys and do a bit of tourism as it was my first and only time in Sweden but with 1 or 2 floppies on the pocket.
I remember also watching the competion show in the main room: one of the prods was a 512 colours animation (anispec like) that was booed and the main prize was a TT030 handed by An Cool to the winner(s). For the music competition Protracker (the green one) was used to play the musics.
That was a long time ago i don't remember that much anymore but some things just come to my mind like the van that took us from Arlanda with Daryl, a lot of sleeping bags at the gimnasium of the school, and the HUGE bike park in from of the Uppsala train station when i went to Stockholm for 1 day.

But that is past.

Let's see if there will be any prods using these digi-routs at least for STF.

We'll see,
Paulo.

[Rhino]

But that is past.

Let's see if there will be any prods using these digi-routs at least for STF.

We'll see,
Paulo.

Haha! Yes, all these names I haven't heard in a long while...TEX, An Cool. I also remember that Future Crew (PC) were upstairs in their own, seperate room. Brings back some great memories.

A couple of us came from England for our first and only time in Sweden, and we were picked up by a couple of guys from Elektra I think it was in their extremely run-down car, which only just made the journey back to Uppsala from Stockholm! It was a fun time, and even though we bought computers, we didn't release any demos or music there, just showed off a few things to nearby people - Electronic Images (Griff and Master) being some of them.

I think it's very unfair for someone to criticise your tracker for having only 6 channels - I think I would have been blown away by it. I think you must have shown it to the wrong people

I like what I've seen of the tracker so far. It will be great for it to be used finally!

Dave (Rhino / Torment)

[dma]

I tried it under STeem, but couldn't get any proper sounding instrument when composing.
All instruments sound like signed samples played as unsigned when entering notes (and i tried setting SIG/UNS in the options menu, no difference).

To test this, I just loaded a track from the "FST_MOD" folder.
It plays fine but any note entered in composing mode sounds trashed.

By the way, the cursors keys lack the usual key repeat.

[ljbk]

I tried it under STeem, but couldn't get any proper sounding instrument when composing.
All instruments sound like signed samples played as unsigned when entering notes (and i tried setting SIG/UNS in the options menu, no difference).

To test this, I just loaded a track from the "FST_MOD" folder.
It plays fine but any note entered in composing mode sounds trashed.

By the way, the cursors keys lack the usual key repeat.

Hi !

The repeat key feature is not included: it was one of the improvements that was planned an never included especially for the arrow keys. If i have time one day ...
The SIG/UNS key is ONLY to select the kind of samples you are loading/saving !
Internally the samples are kept and played in the same way.
Now regarding the trashy sound it seams to be a problem with that 16 voices mod only because if you try on the same folder the brainless.mod everything works fine.
I have to check it. Sorry about that.
Thanks for reporting that bug. Please do as i am not a composer, the full testing has only started now and so the program can only improve with the negative feedback received.


Bye,
Paulo.

PS:
The detected bug has to do with the default volume level of the sample (in that case $30).
If you set it to the maximum ($40) then you get the correct sound.
I will try to find out why this occurs and correct the fault.
Some other faults have been detected meanwhile by me:
1- Any odd number of voices in stereo leads to a crash ... (i did prepare the code for that case but somehow something went wrong );
2- Clicking on the crotchets to enable/disable a voice when having more than 8 and working on the top 8 leads to enabling/disabling the wrong voice;

I will try to solve these faults asap.

[ljbk]

Hi !


Here is the bug corrective version 0.826 with the 3 problems above solved !

From now on, the image file will not be included as the possible composing users will all use an HD.


Bye,
Paulo.

[dma]

Nice!

It would be great to be able to select samples with numeric keyboard, the way most trackers work. Also tab key in edit mode to jump over each channels directly.
It would also be interesting to have a button to wipe all patterns out (leaving samples for a new composition).

The copy/paste system is totally different from usual trackers as well (and not intuitive at all, sorry to say ).
See "FastTracker edit mode / Clipboard operations" section here : http://www.milkytracker.org/docs/MilkyT ... ml#ft2mode

[ljbk]

Nice!

It would be great to be able to select samples with numeric keyboard, the way most trackers work. Also tab key in edit mode to jump over each channels directly.
It would also be interesting to have a button to wipe all patterns out (leaving samples for a new composition).

The copy/paste system is totally different from usual trackers as well (and not intuitive at all, sorry to say ).
See "FastTracker edit mode / Clipboard operations" section here : http://www.milkytracker.org/docs/MilkyT ... ml#ft2mode

Hi !

First of all we should not forget that this tracker was built for an ST back in the 90s which means a ST keyboard.
So first i had to reserve 36 keys to play each possible tone if possible in 3 rows of 12 keys.
So if you look at the ST keyboard pic included with the 0.826 release, you will see that:
- including the tab and excluding the triple size return key, i have 13 keys in that row;
- including the control and excluding the triple size return key, i have 12 keys in that row;
- the row with the shift keys has 13 keys;
So:
=> the middle octave keys are imediatly defined which leaves out the control key;
=> for the bottom keyboard octave, i can use either the left shift or the right shift;
=> for the top keyboard octave, i can use the left shift or the right shift;
=> without the control key, the editing features are then only available via Fx or shift Fx;

So before any change one has to think about which 36 keys to leave to tone playing;

Your tab suggestion is possible if one moves the top octave one key to the right but it will be then a bit of-place unless one also moves the bottom octave to the right shift instead of the left shift;
Another possibility would be to use shift+arrows left/right.

Regarding the numberic keyboard, please don't forget that the program was built to handle up to 255 samples so one would have to give in 3 digits. Instead with the mouse select the hex multiplier you ca reach any sample in a few clicks and you can confirm that it is the correct one as you click.
Example, you want to move from sample $01 to sample $xy:
- you click on the x column and then x times on the small up arrow to reach $x1;
- you click on the y column and then y - 1 times on the small up arrow to reach $xy;


Wiping everything with one key is a very dangerous command especially for a program that was not built with an "UNDO"/"CONFIRM" mechanism .
Anyway, the shift F4 command allows you to do that for 1 voice.
So either you repeat that for a number of voices or you reduce the number of voices to 1 do one time a shift F4 and expand again the number of voices.

One thing i can also offer is to rearrange the Fx and shift Fx functions as you like: you take the 20 types of actions and you redistribute them as you like with the several Fx and shift Fx positions.

As a general info, i would say that this program was not intended to be a copy of any other and it would have requirements of its own.
For example, handling 8 voices on screen means working in medium resolution or high resolution which means handling much more data than the 1 bitplane info of the common low resolution trackers.

Paulo.

[nativ]

kool!

a couple of features I like in Flextracker are

the ability to select note or control data and cut and paste, under or over.

I like the track tab too! shift arrow sounds good

Have you implimented Mod to sample? like octalyser?

How is Midi support? or Midi file support! please see E.P.S.S.
Also how about actual YM sounds?

Cheers

[ljbk]

Hello again !

MOD to sample is not implemented and has never been planned.
I assume the idea is to produce a sample from the mixing buffer(s) and inject it into a(some) file(s), right ?

MIDI and YM are not a part of this project. There are already very good applications for this, i guess.

Ok, i have read a bit the Milky Tracker document.
I will try to have the basic keyboard map a bit more compatible with that but you will have to be patient not because that is dificult but because it takes time and i will be away this weekend.

Anyway, if you detect basic bugs like crashes, please report them. Improvements will always be possible.

Have a nice weekend,
Paulo.

[RA_pdx]

Great work, Paulo!
Never thought that this would be possible on an STF.

Can´t wait to see further versions of your tracker!!!

[jace_stknights]

Great work!

I think if all "oldtimers" show their unrealesed prods nowdays, we should discover some brillant things.

Thank you!

[yerzmyey]

Hello.
I hope You had a nice weekend.

Well, at the moment I can't say anything about composing as I'd like rather to force it to play anything.

Most of songs stop playing instantly or after some time, which - I presume - is related with too high kHz (frequency).
As it has to be set up manually, and program can't chose it by itself, I'd strongly propose to make repeat mode of mouse click on "F /\ \/ " frequency corner.

Or it could recognize number of channels and give max. possible frequency instantly after the MOD is loaded (with assumption that all the 'additional features' like analysers and others) are switched-off. I mean of course ST with YM now.

PS: One thing - I'm talkin'bout emulator because I didn't tested it on ST yet, maybe today I manage, we will see if there are the problems known from another thread.





______________EDIT_________
Although this is not very intuitive after all, as the 2-channels song ("Brainless") cannot go even with 12Khz.




_________EDIT_2______________
On the other hand - "Octopush" 8ch with 17554 hz from 1Mb and YM - pretty shocking, nuff' said. I have to check it out from real machine as soon as possible.



.

[ljbk]

Hello !

The Brainless.mod stops because on pattern 2 there is a reference to sample 6 which does not exist.
Ok, a specific message should appear ...
In the doc there are several indications on why the player can stop.
Now, if you enter edit mode and remove the 06 on voice 2 at position $10 and $30 (present in the original and unchanged Faster Tracker 1 mod), the player works fine.
By the way, if you look at the top right corner, you will see a 4 digit hex number. Only when that number closes to 0 and the player stops will the problem be too high frequency.

The doc also contains the minimum frequency possible for some combinations of number of voices and add-ons.
For 4 voices, the worst case in the one the program starts with.

Bye,
Paulo.

[yerzmyey]

Hi.

Atari 520 STfm 1Mb - 2 bombs. Editor not loaded.

[ljbk]

Hello !


Here comes version 0.830 !

What's new ?

1- STE "on-top" mix is now up to 25% faster;
2- Dist calculation is also faster: it could be even faster as well as the digiscope but i would be wasting memory and i don't want that;
3- Main application file(s) consistency check added to try to avoid corrupt code and data;
4- Editor improvements:
a) function keys rearranged;
b) arrows + shift allow faster movement;
c) clr home goes to first pattern row;
d) # goes to first voice;
e) Insert to create pattern;
f) Shift delete in edit mode deletes tone + sample + effect data;
g) Numpad selects sample: '0' and '.' select a group of 16 and others keys elect one while enter selects number 1;
h) Tone keys are now the same as Standard Protracker with HELP/UNDO to select base octave (EDIT0 or EDIT1 will be displayed on the menu);
i) Information messages added in case of error or markings added (space accepts);


Bye,
Paulo.

PS: Don't forget to read the doc if you want to use this ...

[dma]

Nice! Note entering, moving on pattern, and instrument selection are indeed much better now.

But i can't create a new pattern, pressing insert does nothing (and that key is working fine in the emulator).

[ljbk]

Nice! Note entering, moving on pattern, and instrument selection are indeed much better now.

But i can't create a new pattern, pressing insert does nothing (and that key is working fine in the emulator).

Hi !

Well, sorry that was a copy/paste error where a label was not renamed .
I also added the "shift #" to go directly to the last voice.
The zip only contains the PRG file. Just replace the HEXT830.PRG file.

Bye,
Paulo.