Wesnoth music and the future

[ancestral]

I'd love to see more leitmotif. Having some continuity between pieces would be outstanding (if, perhaps a little hard to coordinate ).


Edit: I missed this, apparently. Yes!

[gmlion]

I should be one of the ONES new composers that West mentioned at the beginning of the post.

We have suggested replacing northerners/loyalists on several occasions when new people have shown up. So far, no takers.

Actually, The Dangerous Symphony was intended to be a replacement for northeners, but since it hasn't been, I suppose it hasn't been considered fitting that role. The Deep Path was also composed with the game needs in mind, since there was need for more underground music.

As for Aleksi, the GPL issue is not secondary for me. Actually, being part of SIAE I haven't still clear how contributing to an open source project could work. This PRO (the only legit in Italy) is considering only at the present time the possibility of using CC for licensing music, with the GPL probably out of question. Actually, I'm not even sure that licensing music with GPL makes ANY sense at all. Has anyone ever given the "source code" of its music, whatever it is supposed to be?
I'm not 100% sure, but everywhere I read the GPL is a license to be used for software only - not music, or art. (This is the main reason CC was invented, I suppose).

Regarding the missing of a clear direction in music development, I agree only partially, and I have a suggestion about this: try, during the revision process, to state clearly whether a critique is a SUGGESTION to improve the music proposed, or a REQUIREMENT for the inclusion.

Anyway, I think Wesnoth has a great number of good music, probably much more of commercial games have.
Morrowind has about four or five music tracks, and it's HUGE.
This perhaps makes people think there isn't so much need of new musics in the game. Probably, as someone did, this also means that a composer prefers to move on a new project.

[ancestral]

As for Aleksi, the GPL issue is not secondary for me. Actually, being part of SIAE I haven't still clear how contributing to an open source project could work. This PRO (the only legit in Italy) is considering only at the present time the possibility of using CC for licensing music, with the GPL probably out of question. Actually, I'm not even sure that licensing music with GPL makes ANY sense at all. Has anyone ever given the "source code" of its music, whatever it is supposed to be?
I'm not 100% sure, but everywhere I read the GPL is a license to be used for software only - not music, or art. (This is the main reason CC was invented, I suppose).

If any works I create get used or submitted, I will whole-heartedly release my Logic Pro files, what I'd consider the music's source.

Since I do not belong to any guild, association, or organization that has any rules or standards on how I release my music, there's no reason for me not to release my project files. So they will absolutely become available for anyone interested.

You are right about Creative Commons. It does fit art, music, and writing much better.

[Dragonchampion]

I know that I'm just an Era manufacturer, but here's what I think: I think people just don't know HOW to contribute. I mean, what will all the amazing music you make, West, people might turn around, listen to it, and say to themselves "There's no way I am ever going to get my music to that quality. I'm going to go be a coder instead."

Another reason is that they don't know what to use and how to make the pieces. They may have an awesome hand-made piece, but have no idea or way to transfer it to computer.

Just a random thought by me.

EDIT: The new location in the Forum Index was a good Idea, it draws more attemtion to this forum.

[Ken_Oh]

scienceguy8 already mentioned something along this line, but to put a personal twist on this...

I actually wanted to make some tracks, if not for mainline then just for my own stuff, but it's not as easy as downloading Gimp. I did look into some software but when I was looking into the voices I wanted to use (e.g. crumhorn, hurdy gurdy) I couldn't find much. So, that's as far as I got.

[Tyler Johnson]

My 2 cents:

If it's not broken, don't fix it.

In my opinion, the music is great. Why add a piece that is not up to the current standards. We wouldn't want to take a step back in that sense, would we? I think it's great how selective we are, it also encourages composers to get better at their craft in order to get something put into the game. Gives them something to shoot for. I don't think it matters how often music is updated to the game, as long as the quality remains the same.

[Death]

My 2 cents:

If it's not broken, don't fix it.

In my opinion, the music is great. Why add a piece that is not up to the current standards. We wouldn't want to take a step back in that sense, would we? I think it's great how selective we are, it also encourages composers to get better at their craft in order to get something put into the game. Gives them something to shoot for. I don't think it matters how often music is updated to the game, as long as the quality remains the same.

Thumbs up.

I could listen to Hell March from C&C, or some of the Warcraft tunes forever. Sure, if I played those games for 10 years, I'd probably switch them off for other music, but that doesn't change the fact that it kicks ass, and I'd still go back to it from time to time.


Incidentally, the Wesnoth music is OK, but it doesn't really grab me and I get sick of it after 1 or 2 go-throughs. Frantic.ogg was actually the only tune to which I'd kinda jive while playing Wes.

[Spi]

I've read this thread through and through and I think Rocket Slug's idea is, simply put, genius (not joking!). Having a certain experience in the field of finding apps and sound for composing music entirely made from the public domain or under GPL license, I know that it is possible to make music (albeit probably at a lesser quality than that of wesnoth), with just a connection to the internet and a minimum ammount of processor power. I've gone through tutorials showing me how to wire up a regular kerboard as a midi piano using software, and how to read outputs from multiple programs to be mixed and fed into the recording program. I've found an entirely free app that allows a moderately experienced user to make every imaginable sound come out of a computer's speakers (given time to configure the sound engine).
My point being that it is possible at least to make midi compositions, to be listened to by the 'pros' as you call the owners of professional-quality sound samples and apps here, so that they can be commented on, critiqued, resent to be worked on with hints and suggestions as to how to make it better. Midi tracks then recorded by professional sounding instruments can sound positively good. And even the low-qual soundfonts and other public-domain samples can give the midi composer an idea of what he's contributing, and can listen to a compressed version of his music with the pro instruments here on the site to see what he could change.
Midi format is much less bandwidth-consuming than any pure or compressed audio format ---> hosting on this server is made easy and lots of midi contributions can be made without overloading the server.
That way we can make efficient use of the professional software certain users already have, and make a unified, cooperating music-contributing community.

I'm not saying there's no drawbacks: losing an important member of said community (real-life situation or other) could cripple the system. I'd be hard sifting through the potential contributors to know which ones could really put up with the job of slowly composing a 4to5 minute long piece. To be quite frank the most I've done so far is one minute, multiple instruments using a synth. I know I could do more given time though, and using soundfonts (or VSTs as West suggested ).
Yet I think it is worth giving it a try (and updating frantic!).
Giving a list of suggested contributions as a sticky in the music/sound development forum could be useful, but I know us normal amateur composers can turn on the tap of creativity just like that so I'd expect some time before someone showed up saying "ok I've got this tune flowing in my head that could fit THIS suggestion...). But eventually sure it would happen.

From what I've learned so far, and seen from screenshots from early versions of wesnoth, I an probably be sure of what I'm typing: Wesnoth has come a long way. And players feel it, they appreciate all the work that's been put in graphics, strategy, AI, and I'm definitely not forgetting Music. In the long list of GPL games Wesnoth's music if first, and it beats a v ery long list of commerceial games I know of, so there's absolutely no reason why I couldn't stand up to be one of the best---> there's always something to do.

Custom tracks for campaigns is a great idea. How about special tracks for skirmish, or even (pushing it) for certain maps? The list could go on...

YES contributions to Wesnoth's music are few and far between. But compare the number of art items in the game, or code items in the game, to the number of sounds in the game, and I'm sure you'll find that the ration of the first two over music is overwhelming. Which justifies the high standards, truly. More ambient variations of music could be made sure (and this would make the great orchestral music seem even greater since people would pay attention to it more ---> I'm not saying ambient music can't be great, but I think it is made to blend in the environment (more bluntly the background) and give an atmosphere, compared to orchestral music which not only gives an atmosphere but adds much more life to the game itself.
Which brings me back to this point: as aleksi said, we shouldn't bring down the standards for wesnoth music ---> new low-quality contributions would clash horribly with the great pieces already there.

As to gamers whih switch off the music after two months playing: what do they want? originality, new melodies, new theme(songs), a new feel to the gamer every time they play it. Making more short tracks 3-5 min would help this I believe.

So that's about it for what I wanted to say (forgot the rest). Even though I haven't contributed music yet I seriously intend on doing so, moreso only with GPL elements which can be found in the open (which in my mind respects the game's GPL license and philosophy; not that I'm saying we should all compose from the public domain, but for me using paid tools to compose for a GPL game deosn't fit the frame, it's all part of the spirit of GPL to be able to make great stuff out of what you have for free).
Cheers, and sorry for the long post,
Spi

[West]

My point being that it is possible at least to make midi compositions, to be listened to by the 'pros' as you call the owners of professional-quality sound samples and apps here, so that they can be commented on, critiqued, resent to be worked on with hints and suggestions as to how to make it better. Midi tracks then recorded by professional sounding instruments can sound positively good. And even the low-qual soundfonts and other public-domain samples can give the midi composer an idea of what he's contributing, and can listen to a compressed version of his music with the pro instruments here on the site to see what he could change.
Midi format is much less bandwidth-consuming than any pure or compressed audio format ---> hosting on this server is made easy and lots of midi contributions can be made without overloading the server.
That way we can make efficient use of the professional software certain users already have, and make a unified, cooperating music-contributing community.

I hate to rain on your parade, but the process you're describing is a lot more complex than you think, and the main workload would end up resting on the people with 'professional sounding instruments'.

First of all, orchestral libraries are not General MIDI compatible so it would be impossible to just load a midi file and play it back with high quality instruments. It doesn't work that way. The person with the good library would need to load up his sampler/sample player and assign suitable samples for each of the 16 tracks of the midi file. Since no two synths or instrument samples are identical he would also need to tweak the tracks by hand, as things like volume, velocity, ranges, expression, etc might be off.

Secondly, GM has only 16 channels. It is simply not possible to make a convincing simulation of a full orchestra with so few instruments. You would need to use catchall "strings" and "brass" patches for playing the majority of the instruments, which a) is not how orchestral arrangements work, and b) doesn't sound even remotely as good as separate instrument sections playing together. Also, with regards to the previous paragraph, what is the guy in question supposed to do when he discovers all string parts in the song is played with the same patch? Manually split all parts up according to ranges and map them to violins, violas, celli, basses?

I realize I'm sounding kind of negative here, sorry about that. It's great that you're enthusiastic about this thing Spi, but I just want you to know that your idea isn't entirely realistic unless you can find someone who's willing to sit down and do all that work for every revision of a midi file (yes, you would need to repeat the process every time).

Personally I think that if we're going to do some collab thing, having decent samples and a way to render these to individual wave tracks is a requirement. Importing a bunch of wave files into a DAW is not much trouble, and you can get to work with mixing and tweaking right away. Midi files is just a hassle.

GM is a useless and outdated format and I honestly wish people would just let it die and rest in peace.

[caslav.ilic]

As for the licensing of music (or any other "non-code" data), I would make two observations based on such discussions so far. Firstly, making it very precise is probably not going to help contributions, as many non-code contributors seem to have own impression of what can be done with the work they contributed (i.e. it's not an issue when they consider to contribute). Secondly, making it very precise is necessary, to avoid the fallout when people do realize the requirements and consequences (which not infrequently results in bad emotions and orphaned data). These two observations, if true, are unfortunate for the OP, but I feel the situation should be cleared out nonetheless, as anything else may be considered "cheating" (even if of "white-lie" kind) the contributors.

So then, I'm neither a lawyer nor a musician, but I do like to bend my mind over the issues of information dissemination :), and therefore I'll propose a framework which people trully involved in a given field (music, graphics,...) may use to thoroughly clear up the matters.

Number one fact is: speaking only about "licensing" is plain not enough. That's the actual tripping stone when one says "GPL is not appropriate for non-code" (implied is: technically not appropriate, rather than not appropriate in spirit). The problem of GPL and most other code-related license is that they establish a link between two as distributed independent pieces of data, the binary and the code. Generalizing, I'd say that the link is made between initial data and data derived by obfuscation from the initial. As discussions so far have shown, especially so for music (but I can think of code-contexts too), it is entirely up to interpretation what the initial and obfuscated data are; so since the two are not even well-defined in general, it is downright impossible to apply blanket source code-like license to their coupling.

As a consequence of number one fact, I propose the following. For each non-code data type, one should determine, as a matter of project policy and not licensing, what constitutes an acceptable contribution. Skeleton of such policy would be:

To make foo-type contribution, yee the contributor:

• must submit A-obfuscated data (e.g. OGG track of such-and-such quality);
• must submit B-initial data (e.g. ... no idea for music);
• are encouraged to submit C-initial data (e.g. samples, sheets -- yes, I'm talking rubbish, I haven't a clue about music);
• must not submit Z-initial data (e.g. the particular pet animal whose purr or roar was the source of inspiration while composing B and C data).

When it is clear what a contribution is (or rather what it can be), then the number two fact is that each of defined pieces of contribution may be covered by a different license, or by the same license if one can be applied to all. So, the final contribution framework is:

To make foo-type contribution, yee the contributor:

• must submit A-obfuscated data under P-license;
• must submit B-initial data under Q-license;
• are encouraged to submit C-initial data under R-license;
• must not submit Z-initial data.

Now the interested parties in foo-contributions, with consideration of how it affects the overal project, determine what A, B, C, Z, P, Q, R are. A, B, C, Z determine the possibilities of evolving contributions. P, Q, R also determine evolution, as well as distribution and use of contributions. Out of all P's, Q's and R's for different contribution types, the one giving least freedom of use may influence the distribution and use of project in total.

* * *

Going into particular licenses (P, Q, R), in particular all CC licenses can be applied to any piece of contribution (A, B, C), as neither makes link between initial and obfuscated data. But, as others have been pointed out earlier, CC licenses are a world of things, with the only unifying aspect that each allows non-commercial distribution and consumption of data.

Thus, on one end of spectrum there is CC-nc-nd: CC license that does not allow earning off of the data, nor making any derivative works. I suspect this is the kind of license e.g. Aleksi would consider appropriate, and to which SACEM and similar institutions would have no objections with (for them it would fall under "promotion"). Of course, if A-obfuscated data is covered by CC-nc-nd, then there is no point in having any B or C-initial data. Furthermore, free-as in-beer distro would have to rip the CC-nc-nd data (e.g. OGG tracks) from their main repositories, and put them somewhere else, to be installed only by direct user's action. It would e.g. not be legal to distribute full Windows or MacOS binaries or Wesnoth on any commercial data carrier, such as on a disk bundled with a for-pay computer magazine.

On the other end of spectrum there is the CC-by-sa: CC license that allows the data to be hacked and slashed in any way, and money to be earned off of it without paying the original author a dime; the only requirement is that original authorship is mentioned, and that anyone else can do the same with derived work. This is the non-coupling (between initial and obfuscated data) counterpart to GPL. Free-as-in-speech distro, e.g. Debian, will accept only data under CC-by-sa to put it into the main repository, and anything else will have to be ripped out and made available only at user's direct request. If all data is under CC-by-sa, Windows/MacOS binaries can be distributed in any fashion imaginable.

If CC licenses are perused, I suspect it would be reasonable to use one and the same CC license for all parts of contribution. If not, then the one on A-obfuscated data should allow more freedom of use; B/C-initial data does not even have to be distributed (an example of non-distributed B/C-initial data are layered Photoshop and Gimp images in branches/resources/cartography).

I also believe that, under clearly defined policies for contributions by parts and no license-coupling different data types, there is no need to consider any but CC licenses (heavens forbid rolling one's own). They were made for precisely the contexts discussed here, and there is a wide range to choose from.

[Kirk]

I think what makes it so hard as a musician, especially for computer games, is that there is a sequence of initial attractiveness towards a game. It is usually prioritized like this:
1. Graphics
2. Game Play
3. Sound FX; Ambiance
4. Soundtrack

Most of the gamers I hang out with rarely say, "Wow! What a soundtrack!" But usually say, "Wow! What awesome graphics!" To be honest with you, version 0.4 turned me off due to the graphics. I'm not saying that this is right. I still have Total Annihilation's soundtracks, Jeremy Soule, on my iPod!

And I agree with previous posts that it's easier and less time consuming using GIMP or Inkscape for artwork, than using ProTools for music.

However, I think you're shooting yourself in the foot by generally not allowing loops. Many programs allow you to adjust loops, edit tempo and key. It just takes a good musical ear and some music theory to make it happen. We need to know what your idea of a loop is. Do you feel Pachelbel's Canon, or the timpani in Strauss' "Also Sprach Zarathustra" loops?

[JLS]

* Get experienced musicians on board : Wesnoth has something to offer to experienced non-pro musicians, an audience... I don't know if there are big amateur composer communities on which we could advertise, you probably know that better than I do, but if such a thing does exist, maybe that's the way to go.

This is a good idea. Unfortunately I'm not a member of any such communities. Suggestions anyone?

Well, I haven't had a chance to read all of the posts in this thread, but I'll take a stab at making a suggestion: OverClocked ReMix, "dedicated to the appreciation and promotion of video game music as an art form." A neat site, with a very large community, and more than a few quite talented artists—some of which do classical/orchestral compositions. For example, just by quickly searching "classical", here are three pieces that I found (click "Dowload" and select a mirror, you should be able to stream the file):

• 'Prelude to Darkness' by Russell Cox
• 'Ice Mountain Symphony' by Mike Vafeas
• 'Cantata for Dancing: I. Mors ego sum mortis' by Derek Oren and Jeremy Robson

Not all of this may be Wesnoth quality, but the potential is there. If you wanted to, you could selectively appeal to a few artists and see if you could get them to join the project. Or, you could probably post a request in the OCR forum—but you'd want to be careful, and clearly state that the music has to be up to a certain standard to be accepted. No sense bringing folks over here only to have to reject their contributions.

[Boucman]

sounds interesting... could we have a music guy attempt to make contact ?

[gmlion]

You could start from this promising guy:D :
http://www.ocremix.org/artist/51/jeremy-soule/remixes

[Jetrel]

Yes, i can do what i want with my work.

So we need to understand, what license(s) can you/are you willing to release your work under and what are the terms of these licenses?

David

I'm honestly a little offended that any artist on this project feels they need special rights on their work when everyone else is willing to be mature and freely give it away. We've got a number of doctoral-level professional coders kicking butt on this project, and not one of them feels that they have to attach special conditions to their work. A gift isn't a gift if it "has strings attached". If you're not interested in freely giving your work away, like everyone else, then please stop teasing us with the carrot.


You should be able to do the work off the record, and not have it protected by SACEM. You don't need the protection if it's a gift; it's contrary to the entire nature of gifts. If you're not allowed to do that - if SACEM owns all of your music and prevents you from working on free hobby stuff, you've been had.

I don't need my work protected here, none of the coders need their work protected, and none of the other musicians we've had here did either. All our work, and this entire project, is a gift to the world at large; we do it because we enjoy the act of creating things, and we want people to freely enjoy it.



All of this said, I wouldn't have anything against dual-licensing our music under something like CC:Attribution in addition to the GPL.