Adlib Tracker II - the official website

If you have experience with the id game Doom, all the game data is stored in what is known as an IWAD file. In there is a 'lump' called 'GENMIDI', Doom's custom instrument set. Those instruments are awesome in-game, so I'm after a way of dumping them as individual SBI files. I've already dumped the GENMIDI lump from the IWAD, that's easy, it's just getting to the instrument data that's difficult.

By looking at information about it, the data is somewhat similar to SBI, except all instruments are in this single file, and there are no obvious tools out there that can dump the instrument data from this file.

There are new versions in download section. For SDL-version users, I recommend to try setting "sdl_screen_mode=1". I hope this will be a nice surprise ;)

@agargara: I think that if you use ZF0 (Release sustaining sound) before new note trigger of such "problematic" instruments, it should solve your problem with popping sound ;)

@Samuel: I'm not sure if I understood correctly what you want, but there are General MIDI ($gmopl.ibk) and Roland MT-32 ($mt32.ibk) iunstrument banks in the "INSTR" directory of the AT2 package.

Is there an easy(-ish) way to dump all of the instruments for Doom's GENMIDI lump in SBI format? I tried a couple of scripts I found within the source of Chocolate Doom and FreeDoom (Ruby and Python respectively) and neither of them worked.

Or, is there already an SBI pack containing said instruments floating around on the internet that I am missing?

Alternatively, is there a way I can convert the GENMIDI lump into Creative IBK format?

Some of my instruments tend to "pop" at the beginning, it sounds like a zero crossing problem. It doesn't happen all the time. Is there a way I can prevent this? I tried lengthening the attack but that doesn't help. Is this just a problem with my soundcard? (I'm using dosbox on a mac.)

@ToriHino: Yes, you are right. Length of block is 4bytes (DWORD) since FFver=9. I have to correct this in new releases. Anyway, I put a link to updated File Format Information in HTML format here on homepage (see in "Contact & Links" section)

I'm trying to read the binary format A2M file (version 11) but i get the impression that the file format description is not correct. It states that the 'length of data block x' are all 2 bytes (1 word) but looks to me like they are 4 bytes long. So the actual first data block should start at 000054 right?

@Diode mA: Really nice FM arrangements! Also lot of AdT2 motives on your page, so nice advertisement - so thanks, hehe... keep up the good work!

Thanks for the amazing software! I made an EP with AT2 and you can order it on 3.5" floppy from my bandcamp :)

https://diodemilliampere.bandcamp.com/merch/ymf262-35-floppy-diskette-handwritten-label

I used container in a loose sense. But you are right.
I'll send you an email now.

@Maan: A2W is so not container "per se", but it's rather a copy of all your instruments-related data from the tracker. As the source code -> please send me your email to subz3ro@hotmail.com. I will send you back the recent sources for AT2 (these on Google Code page are rather obsolete now, as I already mentioned).

I think I got it now with the Macros.
I am going to play with it.

As for the bug I mentioned to you before, this may be a false alarm. I am going to install the latest revision from your page. I was using now 2.4 for DOS from the google code page. But now it appears that this is not the latest (as per the versioning thing you explained).

My question about saving instruments has a reason. So, this is good - an a2i can be saved to a2f even if we do not have any macro data. And a2i can also be saved directly to a2w.
So, A2W is the container to use.

It looks like you have created the ultimate format here :)))))

I am building a second DOS system with a sound card with a true OPL3 chip.
I wanted to use Turtle Beach Tropez Plus. But they are so hard to find.
So I am settling for an Aztech Pro 16 II.

In my free time (which does not exist now) I need to look at the source code of ADT2.
I have enough code here to try and add writing to Rdos RAW, WAV, and multiple waves (one wave per track) via starting adt2 with switches for writing into these formats.
That should be cool.

@Maan: And as your second part of question - yes, it should be always (from principle) possible to put less information to somewhere where is space for more information :)

@Maan: Oh my. I think this macro system in AT2 definitely need some manual... once such skilled person like you is not getting it right, I believe that 95% of users (who try it) won't do either :(
As the table rows - that's the concept of macros in AT2 as figuerd out some longer time ago by Malfunction/Altair. It's very similar to the concept of pattern data - you can on one hand use only one row, or on the other hand use all of the rows. It depends on what you want to do with it. Here, in the macro, you can change all the FM-registers of the instrument in each row (together with the pitch so you can control the note playing), and you can do it sequentially in more rows of the table for the period of playing time between two rows of the pattern being played. Sure, it depends on the tempo/speed of the song and on the tempo/speed of the macro how many rows of the macro table can be played in the "meantime". You have this macro speedup parameter in the Song Variables too, by which you can setup the multiplication factor for the tempo of macro compared to the tempo of song. If factor is 1, it's equal and you can play only one row of the macro table per one row of the pattern (if both speeds are equal), resp. more if the speed parameter of song is higher than the speed paramater of macro. If you setup macro speedup to let's say 10, this means you gain 10 times more space for processing rows in the macro for time playing of one row in a pattern. But again it depends on how you use both speed parameters, then this space can be even higher, or can be lower.

I hope you at least got rough idea how this macro concept works now. If not, please check the use of FM-macro in song "ben\fm63b_rv.a2m" as shown with ins#04 and ins#0E, resp. use of vibrato macro-table as shown with tables 01, 02 and 03 in the pattern data... You can refer to @01, @02 and @03 effect commands and compare the sound result with what is in the macro table data (Ctrl+E pops up the editor window)

Ok. I am talking about one instrument. We go to Instrument Panel. Then to Macro Editor.
As you know there are numbered rows. These rows are all the same.
So, ****for one instrument**** all these numbered rows that are the same are related to this one instrument? I still do not understand :))))

As for saving instrument I understand what you said.
All I need to know:
-- if it is OK to save a2i as a2f then save it to a2w? AND if it is ok:
-- to save a2i directly to a2w.

Answers from tech.support :D
1. They refer to nowhere... but opposite, they are referred by values in the table properties -> directly (as positions) with "loop begin" and "key-off" and indirectly (as constraints) with "length" and "loop length".
2. A2I file carries only FM-register (i.e. carrier and modulator) data, while A2F contains also FM-register macro data (the ones you see as FM-Register Definition Macro Table data in the editor window), which are related to each single instrument. However, what you have to be aware that it doesn't contain, these are the Arpeggio/Vibrato Macro Table data, because they are not related to the single instrument. For these to get archive, you have to save to A2W file (with all instruments together).

Thanks subz3ro for the explanation.
I have some questions for your technical support dept. hehehehehe
1. In Macro Editor, what are the numbered rows of lines? What do they refer to?
2. Just to be sure, if I save an a2i instrument as a2f instrument, does it add to it the macro info which are empty?

@Maan: Yes, there is. And unfortunately this version numbering on AT2 Google Project is quite chaotic now. Cecille probably not intentionally changed my versioning from 2.3.xx to 2.4.xx. He didn't realize there is convention in version numbering: 2.0.xx for the former OPL2-compliant program and 2.3.xx for the G3 (OPL3) line.... I would like to keep the DOS line development in 2.3.xx and the new SDL line development in 2.4.xx instead, therefore I changed it back to 2.3.xx and started from revision 40 (my last release was 38, so this "39" is alternative name which is reserved in revision history for this only one DOS open-source release). I'm not the owner of page at Google Code so can't make proper administration of new sources / release packages, and I didn't have time to contact Cecille (ijsf) so far. Sorry for that. Please simply stick to the download section here at official AT2 page for any new release so far.

subz3ro -
The DOS distribution on
https://code.google.com/p/at2/downloads/list
has DOS Version 2.4 from January 2012 (at2-2.4.00.zip)
Is there a disconnect between their version numbers and yours?

New releases available in download section. Many bugfixes (especially 4OP related) and more. Great improvement with OPL3 emulation in SDL Win32 version ;)

@frtvykbuhni: Good question. Answer is not so simple, though. So far it's not planned to have separate SDL Linux port. Main focus on SDL-version line is to improve the OPL3 emulation. However, the DOS-version line will be continued with all (possible) new features, too... so the DOSBox package will be udpated accordingly with the new versions.

will the sdl version ever be ported to linux? the dosbox version works of course but i like up to date software!

@subz3ro - Actually all you need to do from within adt2 is to load an instrument from a bank which you already do.
And maybe insert an instrument to a bank. Which is not that important.
But creating a bank from within ADT2 is definitely not necessary.
This whole A2B Bank thing should be a separate application. where you can manage a bank and create new banks.
Anyway, I'll write you an email. About this and about your instrument definitions.

@Maan: Yes, I know what you mean - to implement the bank browser like for other bank formats ;) This is hell of work, but I agree with you it would be great improvement. I hope I'll be able to make this stuff, it depends on how much spare time I can get over this spring ;)

@subz3ro - you really should have done all the work while booting to pure DOS 6.2 :)))
Such errors do not bother me. I can always start it from Win95.
However, I am going to send you an email about A2B and A2W banks.
We really need a better and consistent A2B and A2W bank management in the tracker.
Which could be a great improvement.
There is a reason for this that I'll let you know about.

@Maan: I'll try to manage program setup with DOS-VirtualMachine and see if this error can be reproduced at mine. Right now I don't have any idea what can be the problem. But people are reporting many pure DOS-related issues with the tracker, which they do not encounter under Windows DOS console. The fact is, that I didn't use booting to DOS since arrival of Windows 98, I think. So maybe I didn't catch all the related bugs when they raised up.
As the SB AWE series and YMF262 chip emulation -- this is quite funny, because I always liked their OPL sound most :)

Take a look at this subz3ro (and all of course):
midibox.org
jeeeeeeeeeeeeeeeez!
(also check out this list on their web site):
http://midibox.org/dokuwiki/doku.php?id=soundcards_with_ymf262_and_or_yac512_chips

This external midi box uses scavenged Yamaha 262 (OPL3) chips and the Yamah YAC 512 from sound cards.
The YAC 512 is included with sound cards that use OPL3 chip. It is Yamaha's 2-channel DAC.
I did not know that the OPL3 uses 4 channels of audio.
The midibox supports two YAC 512 chips for 4 audio channels from the YMF 262 chip.

The ONLY DOS tracker I know of that can use the OPL3 on the sound card in its tracks and also assign MIDI OUT tracks is the All Sound Tracker (AST) for AWE32/AWE64 cards.
This means with the midi box, up to SIX channels of OPL3 audio.

subz3ro says: Yes, the DOS version is (and always will be) having pure hardware OPL3 output :)

YES! I love the sound of this. PURE hardware OPL3 output hehehehehe.
I just love the sound of it. We opt for purity here :))))

subz3ro says: My intention was to make the least possible A2M module size for, let's say, some intros or small demos which would need it.

Yes of course subz3ro. These older A2M modules running at 5 kb in size can take up space hehehhehehehe. Just kidding :)))

subz3ro -
Thanks for the replies. Now it all came to me :))) Reason for 4-op channel binding on two tracks via connecting 2 2-op instruments but with added opl3 wave forms. My original suggestion from years and years ago was to import Symphonix banks and instruments which are true 4-op instruments and to create a virtual track begind the scenes. But I see now the difficulty and where misuse can happen.
However, your instruments and banks are quite nice in their support for opl3 and their added functionality.

I have two notes for the users:
-- Daniel's plugin in_adt2 supports BOTH newer a2m and a2t. Not a2t only.
-- The other bigger plugin adPlug unfortunately does not support newer AT2 files formats as you know. Nice thing about adPlug is that it can write RAW files that can be converted to MIDI.
But Daniel's work is nice with all the souce code also.

As for the bug:
It happens only when I ran it in DOS (I have DOS 6.2). Meaning I boot PC to DOS.
-- Go to CNTL-F (Song info and where you bind channels). Bind some instruments on two tracks.
-- Go to CNTL-I (Instrument Panel).
PC freezes if there are connected instruments on two channels/tracks. and have to reboot.
To avoid this, one has to UNTIE any connected op4 tracks via CNTL-F before going to Instrument Panel. Go to CNTL-I and do any instrument editing, Then go back to CNTL-F and bind channels again.

In other words, going to Instrument Panel **when you have connected 4-op tracks**, freezes program.

This does NOT happen when program is run in a DOS window from Win 95.

One thing: I an running EMM in DOS. Maybe this is the reason. I do not know.

I want to turn your attention to something you may already know.
Creative AWE32 and AWE64 cards do not have a true Yamaha OPL3 on them. But another chip that does a good job at emulating the Yamaha YMF 262 chip.
A true YMF 262 can be found on many models of Sound Blaster 16 and Sound Blaster Vibra 16.
I am going to get one of those an

@Maan: Here are the A's to your Q's in order of giving:
1. Yes, the DOS version is (and always will be) having pure hardware OPL3 output :)
2. You found yourself an answer for this question already -> it's because of my intention of binding the two 4-OP channels, thus having more user comfort and protection from some user misuse, while giving additional capabilities (autonomous use of effects, etc.) to it. Hope you will like it once you get used to it :)
3. Yes, A2M is featuring the same data compression as the A2T file. However, there is some "garbage" information included (names of instruments, patterns, etc.), which is on other hand not present in A2T files. This is in principle not huge difference, except some really minimalistic modules. My intention was to make the least possible module size for, let's say, some intros or small demos which would need it.

Btw, just curious - what exactly is the bug you mean with "It is already freezing my PC when I bind tracks then go to Inst control panel."? There is one known bug -> marking block section in pattern making endless loop with 4-OP channel pair (which is already fixed in SDL version, and soon it will be fixed in DOS version too)... but from your description this doesn't really look like that one, does it? :)

hmmmm. I figured out the 4op inst connection. It is via binding pairs of tracks from CNTL-F
I remembered a very old discussion and I actually tested the very first version of this. My opinion then was to come up with a new adt2 inst that was 4op opl-3 inst and use it on the track. But I forgot why you opted for binding pairs of tracks instead leaving your inst a 2op inst behind the scenes.

Your approach allows applying effects on each pair. hmmmmmmm
I am going to test with this. It is already freezing my PC when I bind tracks then go to Inst control panel. I'll let you know. This is much fun hehehehe

hehehehe yep. Still alive. I really thought the project was dropped. So I was surprised to see a new version and OPL3 support and a DOS protected mode exe.
I got so busy through the years and I thought of you from time to time. But I really never got any messages from you. So I thought you decided to hide :)) I am glad we are here again. And I'll keep in touch of course.
As to the good old days: I know something is good when I see it and I always thought adt2 was a very good project.

Couple of comments to you:
1. Just wanted to make sure. The protected mode DOS exe does not use emulation, right? It works with the OPL3? Because I am using the program with a sound card that has an opl3 on it. I see there are many emulators these days.
2. I do not understand how you are making a 4OP instrument. I followed the DOC all the way where you mention connecting two instruments to make a 4OP instrument in the 4 different ways. I understand the 4 ways. Can you explain what you are doing exactly to make a 4op instrument? How are you connecting two 2op instruments? I just could not recreate this. It may be very simple and I am missing a step.
3. I noticed that .a2m and .a2t files of same module are almost same size. Are you applying compression on .a2m also?

Dear Lord, Maan is alive :))) I tried to contact you few times within these 13 years with no luck.
Nice to hear from you my old friend. As your name in the DOC - maybe you don't know that yet, but you were the first person who ever gave me feedback about my FM tracker challenging (early predecessor of AT2... that MPU401 Trakker thing which probably nobody knows, except you) so it might be easily possible that without your interest and small donation (money motivation spent quickly, but I still got that book :) there wouldn't ever be this AT2 babe. Your name in the credits is the very least I can do :) Hope you'll stay tuned with us for some time again!

Subz3ro -
Glad to see adlib tracker well and alive.
I still use it with a PC running Win95 and DOS 6.2 with SB 64AWE gold card.
I just downloaded the latest.
Great work. And I am glad to see my name is still in the DOC :))))
Maan

Version under DropBox link meanwhile updated to 2.4.02 (big thanks to Madbrain for quick bug report ;)
Soon will be updated current OPL3EMU (which lacks especially with some 4OP arrangements) or we will switch over to the one used in recent DosBox versions (which seems quite good in many aspects, I must say).
Have a Very Merry Christmas and see you in New Year 2014 with (hopefully) some good news about AdT2 :o)

After long long time I found myself some time to fix the stuff and finally brought the first official SDL version. Sources on Google Code will be updated soon.
So far download the binary package here....
https://dl.dropboxusercontent.com/u/91679779/AdlibTracker2-SDL-Win32.zip

Here maybe some interesting information from VMware development:
https://communities.vmware.com/message/2249754
Found it only today and didn't manage to get it work yet with the VMware player 6.x (it's probably only available in VMware Workstation 10.x) - but it looks promising they tested it with our stuff at first :))))

http://noplanet.bandcamp.com/album/l-operator

oh btw, here's an album I composed directly with AT2, download for free :)

Curious if this is compatible with the ESS FM chips.

@Aaron: w-i-c-k-e-d? :P

Thanks for making this wicked program! I am running Adlib Tracker II on my Macbook pro with Dosbox and it's sure bringing back memories of when i owned a Pentium computer.

@subz3ro: i was trying the sdl version, but the dosbox version works just fine.

@Jason, Yogi: Thanks guys! Welcome to the community and have fun ;)

@ellen_page: What is, actually, the platform / version of AdT2 you encounter problem with peripherals on? Real DOS version @Win9x DOS prompt? DOSBox version? SDL version?

whats a fix for the keyboard and mouse not being recognized problem? i'd like to start working with this but if its not consistently responding it's a bit hard. do i need certain drivers? if so can anyone link them or help me find them? thank's in advance.

ahh fixed, it appears the new MPU-401 was acting a fool.
That and OPL latency did the trick. Sounds beautiful.
Amazing work!

having trouble
Installed with DOS 7.1 and FreeDOS both with same trouble.
I downloaded and installed drivers for SB16 I have a CT1740
Normal music and sound all work perfectly, but when I load Adlib Tracker 2 modules
the instrumentation changes everytime I restart the song. It's noever loaded correctly even on the first load. So when playing back it turns into garbled off key noise.
Not playing back properly.

Hi Ya'll, Just read like the last 50 pages, soaking in as much as possible. I got FM801 setup with a YMF724 card under XP; so I'm trying to figure out a setup with AT2/DOSBox (megaBuild ver) with the OPL pass thru patch/YMF724.
Anyone running this sort of setup? Just started digging into DOSBox so looking for pointers. I know that a few have mentioned this PCI card so I'm hoping this is a 'can do'.
Thanks subz3ro, it's soo cool that you have kept AT2 alive an kick'n for the world to 'discover'