Hundreds of entries.
4 categories.
2 languages.
1 compo.
After much anticipation and delay, Famicompo is finally back!
Famicompo is a Famicom/NES music competition that has gone through various incarnations, with many historic entries and participants. You, too, can become a part of NES music history!
Due to the sudden unexpected timing, Famicompo Pico 3 will last for two months! FCP 3 starts on the 15th of June and ends on the 7th of August before midnight UTC.
The four categories are original, cover, freestyle original and freestyle cover.
Original and cover entries must be submitted in NSF format. All expansions are permitted.
Freestyle entries must be submitted in MP3. The 2A03 must be noticable in the song. Simulation and signal processing are allowed.
Winning entrants in the originals category under 64KB (2A03-only) will be put on a special FCP3 cartridge playable on any NES, or a Famicom with an adapter. One cart will be awarded to the winner of each category! More details will be available by August 1st!
Come join us on
#fcpico3 on EsperNet IRC or on our
Discord server for more info, and visit us on our website at
http://famitracker.org/fcp3/!
We'll see you there!
Why MP3? There are much better codecs, such as Opus.
I like the Famicompo though (with NSF) so I am glad it can return!
retrodpc wrote:
Come join us on #fcpico3 on EsperNet IRC or on our Discord server for more info
Through that server I learned 0CC FamiTracker is better than vanilla FamiTracker in all cases save
one.
zzo38 wrote:
Why MP3? There are much better codecs
Because
MP3's patent recently expired, and "better codecs" that aren't still patented won't play on an iPod touch, iPhone, or iPad.
Opus doesn't play,
Vorbis doesn't play, and
FLAC doesn't play. As far as I'm aware, the
<audio> element in Apple WebKit plays MP3, M4A, and Apple Lossless. That's it.
Also Edge plays Opus but not Vorbis. Go figure.
tepples wrote:
Because
MP3's patent recently expired, and "better codecs" that aren't still patented won't play on an iPod touch, iPhone, or iPad.
Opus doesn't play,
Vorbis doesn't play, and
FLAC doesn't play. As far as I'm aware, the
<audio> element in Apple WebKit plays MP3, M4A, and Apple Lossless. That's it.
Oh for heaven's sake... MP3 needs to die already. Opus blows it right out of the water at a quarter of the bitrate.
What happens when a user of Safari (or any other browser wrapping Apple WebKit) follows a link to an audio file encoded with Opus?
I don't get why anybody is talking about Opus? Everyone knows and uses MP3s and they do the job well. Whether Opus provides a minor improvement to the quality vs size ratio is unimportant.
rainwarrior wrote:
I don't get why anybody is talking about Opus? Everyone knows and uses MP3s and they do the job well. Whether Opus provides a minor improvement to the quality vs size ratio is unimportant.
Because we disagree with the parts I highlighted. I would take 64 kbps Opus over 320 kbps MP3 any day.
Rahsennor wrote:
Because we disagree with the parts I highlighted. I would take 64 kbps Opus over 320 kbps MP3 any day.
I may disagree that the quality : size ratio is
that much better, but even if I accepted your premise that it's 5x smaller, I'd still be of the opinion that MP3s are a much more practical compromise. Size isn't the most important thing, and confusing 99.9% of your users with a file format they haven't heard of is just not a good idea.
I love how you people are acting like MP3 sounds like total shit. I can barely tell the difference between it and uncompressed audio, if I can at all.
rainwarrior wrote:
Rahsennor wrote:
Because we disagree with the parts I highlighted. I would take 64 kbps Opus over 320 kbps MP3 any day.
I may disagree that the quality : size ratio is
that much better, but even if I accepted your premise that it's 5x smaller, I'd still be of the opinion that MP3s are a much more practical compromise. Size isn't the most important thing, and confusing 99.9% of your users with a file format they haven't heard of is just not a good idea.
Exactly. I've never of Opus once before this.
If you haven't heard of Opus, that's probably because it's relatively new. If you care about that kind of stuff, go check it out; if MP3 sounds fine to you then we'll just have to agree to disagree. Lossy compression is always going to be a subjective issue.
I think Opus is cool, I just don't think this competition is the right place to promote it.
I think it depends on who are the target listeners and what kinds of devices those people may use.
As you already put it:
Rahsennor wrote:
that's probably because it's relatively new.
This is an already good enough reason for using MP3 over Opus. MP3 is simply more accessible and nearly all devices nowadays would support, such as playing from a USB stick inserted into a cheap/old Hi-fi and into one's digital TV set, and (obviously) MP3 players. Newer formats may not be supported and these devices may not even have an option to have their firmware upgraded (either not upgradeable or that there is no longer any
new upgrades).
If the files are supposed to be listened by say, someone behind a PC, then Opus is an okay choice.
wow, you guys are really making a deal out of this.
I'm pretty certain no one really gave a thought about the mp3 codec for submissions, it's just the default that most people know and uses. If it bothers, just use a higher bitrate. It's a music competition, not a compression competition
Funny how almost everyone is off-topic; instead of considering the technical details of their potential entries.
From
rules:
[DARK CREDO] After posting, keep traces of you making the song concealed until results are posted. We encourage you not to advertise publicly that you made so-and-so song, especially in a manner that shouts "OH HAY I MADE ENTRY NO. XX". This rule will be more strictly enforced this year.
Is it also off-topic to discuss the intent and practicality of the "[DARK CREDO]" rule requiring keeping the composer's identity secret? Use a less common engine in the NSF, particularly one you developed, and someone might be able to identify you as composer of a particular NSF by
identifying your engine.
I see no issue with discussion that's not completely pertinent but still relevant.
I've already discussed this on the IRC but for the people who weren't there, I will continue to use MP3 simply because of its greater reach, as much as I love Opus' quality.
@tepples: the primary purpose of withholding the author's identity is to prevent listeners from biased voting based on the entrant. As long as the entry is not advertised and the name is withheld, in most cases this is enough to serve its purpose. If you go out of your way to identify instruments and engines to guess the author and base your score on it, that's on you and you should feel bad for it. I'm just trying to protect the innocent from their own unconscious bias.
Rahsennor wrote:
rainwarrior wrote:
I don't get why anybody is talking about Opus? Everyone knows and uses MP3s and they do the job well. Whether Opus provides a minor improvement to the quality vs size ratio is unimportant.
Because we disagree with the parts I highlighted. I would take 64 kbps Opus over 320 kbps MP3 any day.
Actually, 320 kbps MP3 with the right encoder/settings can produce better monophonic fidelity than Opus given the limitations of Opus's full-band implementation, providing no room for frequencies above 20 KHz. Opus is more suited for voice quality than music.
But the point of using MP3 isn't quality or size, it's accessibility… and a bit of a lossy preview to the actual 2A03 executable music. If you want the full fidelity and best size you'd just release only the NSF
As for biased voting, it's inevitable in the sense that different people have different tastes. But withholding names and titles should make it clear that it's about the music standing on its own… before it's put in a compilation, anyway.
ap9 wrote:
[But the point of using MP3 isn't quality or size, it's accessibility… and a bit of a lossy preview to the actual 2A03 executable music. If you want the full fidelity and best size you'd just release only the NSF
The MP3 submissions are just for the "freestyle" category where it is NES + whatever you like. This isn't an NSF category.
I stand corrected. Reading the first post again, I see MP3s are not required for the non-freestyle categories. (Although it would make it technically difficult to tell what NSF engine was used if the voters could only download MP3 recordings.
)
ap9 wrote:
Actually, 320 kbps MP3 with the right encoder/settings can produce better monophonic fidelity than Opus given the limitations of Opus's full-band implementation, providing no room for frequencies above 20 KHz.
How does wasting bits coding something that nobody can hear make it better? Wouldn't it be better to spend those bits on the audible portion of the spectrum?
This whole MP3 vs Opus argument is missing the point, anyway. Why are the freestyle submissions not accepted in a lossless format?
Joe wrote:
Why are the freestyle submissions not accepted in a lossless format?
Because generic audio players don't universally support lossless compression. One may support one format but not another. iOS supports ALAC but not FLAC. (However, it's been reported that
iOS 11 may support FLAC.)
It would be something to allow uncompressed WAV file submissions for freestyle as devices generally support that.
Unfortunately it would be impractical to accept WAV files, due to their immense size.
B00daW wrote:
Funny how almost everyone is off-topic; instead of considering the technical details of their potential entries.
Welcome to Nesdev.com
So, is anyone else on here going to enter something? I think I will, last time I entered something in Famicompo I think it was the very first one. I was the only one who submitted an NSF with more than one track.. whoops.
I know I can't win, but if I'm going to have my ass handed to me, it'll be by the best NES/FC composers in the world, and that's pretty cool!
I'd like to enter. I'd also like to know if this will be preserved and maintained better than the last one that was on bitpuritans.eu (defunct).
I, for one, would like to request the original tagged NSFs be made available once the competition ends. That way I won't still be tagging them myself two years later only to find out about the next Famicompo Pico announcement once I'm finished.
I will make the original NSFs for this compo available; hopefully the ones for the previous ones will be available as well, though since I did not host them I cannot guarantee this.
At any rate, entry period starts in less than 24 hours. Good luck, everyone!
Questions:
1) Multi song NSF?
2) Famicompo Classical vol.2, when? Expansions suck much.
3) Freestyle. "The 2A03 must be noticable in the song." WTF???
Quote:
Multi song NSF?
Last year had a few multi-song NSFs.
Quote:
3) Freestyle. "The 2A03 must be noticable in the song." WTF???
Seems pretty straightforward to me.
Famicom-po.
I'm sure they would accept 'simulated' NES sounds, made with Square or Triangle waves, not just output from Famitracker or actual hardware.
Quote:
3) Freestyle. "The 2A03 must be noticable in the song." WTF???
Quote:
Seems pretty straightforward to me. Famicom-po.
I'm sure they would accept 'simulated' NES sounds, made with Square or Triangle waves, not just output from Famitracker or actual hardware.
Is GameBoy pulse Famicom enough. I have GameBoy through Korg MS-20 and GameBoy+Korg DS-10 songs already. Are they fine for freestyle.
edit: no idea how this quote thing works. sorry.
So long as you don't use any channel 3 waveform other than FEDCBA98765432100123456789ABCDEF, and you don't get fancy with short-period noise, a Game Boy should sound enough like an NES to qualify.
nin-kuuku wrote:
Is GameBoy pulse Famicom enough. I have GameBoy through Korg MS-20 and GameBoy+Korg DS-10 songs already. Are they fine for freestyle.
Sorry for the late reply.
As tepples pointed out, Game Boy is similar enough to Famicom audio for freestyle. I think that some voters would take off points but the restriction is anything that clearly sounds like Famicom + anything else you want.
The program you list by recommended player is Windows program, and some people use Linux (such as myself) or Macintosh computer, too.
If you use GNU/Linux (as I do), then any player based on Blargg's Game_Music_Emu, such as the SDL-based player included with GME's source code distribution, will work. I've made a wrapper for GME and DUMB that'll convert formats they support to a WAV file that'll play in just about anything. It renders NSF, for example, at 1 minute per second on an Atom N450. If you want, I can post its source code.
Game_Music_Emu doesn't support the MMC5, FDS or VRC7 expansion audio chips or the 5B noise or envelope channels. FCEUX (and by extension Mednafen) doesn't support alternative PLAY rates or the 5B noise/envelope. Nestopia doesn't support NSFs at all. Which means the last two editions of Famicompo contain entries that will not play correctly on any Linux emulator I'm aware of.
<plug>Except my own, which is neither stable nor open-source, but if that doesn't deter you you can find it in this thread. I just noticed it's had 37 downloads in the last 6 months so I should get a move on and update it.</plug>
Assuming your Linux computer has either an x86 processor or an x86-64 processor, as opposed to the ARM processor in a Raspberry Pi, how well do they play in an NSF player for Windows in Wine in Linux?
Who maintains Game_Music_Emu in 2017? I should try uploading some MCVEs of NSFs that don't play to its issue tracker.
tepples wrote:
Assuming your Linux computer has either an x86 processor or an x86-64 processor, as opposed to the ARM processor in a Raspberry Pi, how well do they play in an NSF player for Windows in Wine in Linux?
It works, but Wine's audio handling is awful. Dropouts, noise and lag are the order of the day. YMMV of course.
Lag yes, dropouts no. FamiTracker 0.4.6 with the buffer length cranked up to 80 ms is usable even on my Atom potatobook. I imagine that for playing an NSF (as opposed to composing), lag would be even less of an issue.
tepples wrote:
Lag yes, dropouts no. FamiTracker 0.4.6 with the buffer length cranked up to 80 ms is usable even on my Atom potatobook. I imagine that for playing an NSF (as opposed to composing), lag would be even less of an issue.
Not all programs have buffer size options, sadly. Famitracker and NSFplay more or less work, but they still click/stutter slightly every second or two, and there's a faint hiss that doesn't show up when I dump to WAV and play with a native app.
Rahsennor wrote:
faint hiss
.... cheap sample rate conversion? maybe pulseaudio?
Rahsennor wrote:
Game_Music_Emu doesn't support the MMC5, FDS or VRC7 expansion audio chips or the 5B noise or envelope channels. FCEUX (and by extension Mednafen) doesn't support alternative PLAY rates or the 5B noise/envelope. Nestopia doesn't support NSFs at all. Which means the last two editions of Famicompo contain entries that will not play correctly on any Linux emulator I'm aware of.
Nestopia supports NSFs. The Undead Edition has more accurate sound (from what I know). The Linux versions are supposed to work. Do you mean the given entries won't play?
lidnariq wrote:
.... cheap sample rate conversion? maybe pulseaudio?
Whatever it was, it's gone. I have no idea what I changed or when, but right now the Windows version of repeat sounds identical to the Linux version, occasional crackle notwithstanding.
ap9 wrote:
Nestopia supports NSFs. The Undead Edition has more accurate sound (from what I know). The Linux versions are supposed to work. Do you mean the given entries won't play?
Loading an NSF file in the Linux version of Nestopia produces a blank window and no sound. Nestopia UE 1.46.2, from the i386 Debian Jessie repository.
Rahsennor wrote:
lidnariq wrote:
.... cheap sample rate conversion? maybe pulseaudio?
Whatever it was, it's gone. I have no idea what I changed or when, but right now the Windows version of repeat sounds identical to the Linux version, occasional crackle notwithstanding.
I find this commonly a problem when a program outputs at 44100 Hz instead of 48000, and the OS/driver does a poor conversion job in the background.
Nowadays, in
kode54's fork, I'm seeing files with names like "Nes_Fds_Apu.h" and "Nes_Mmc5_Apu.h" and "Nes_Vrc7_Apu.cpp". I haven't been able to verify their accuracy for myself, nor whether the GME in Debian's and Ubuntu's repository includes them.
rainwarrior wrote:
I find this commonly a problem when a program outputs at 44100 Hz instead of 48000, and the OS/driver does a poor conversion job in the background.
Or worse, the hardware itself claiming it supports 44100 natively. But that should have affected everything, not just Wine - there was an audible difference between the two otherwise identical versions of my NSF player.
Anyway, it's working now.
tepples wrote:
Nowadays, in
kode54's fork, I'm seeing files with names like "Nes_Fds_Apu.h" and "Nes_Mmc5_Apu.h" and "Nes_Vrc7_Apu.cpp". I haven't been able to verify their accuracy for myself, nor whether the GME in Debian's and Ubuntu's repository includes them.
Since I know you like data points: the version of Audacious in the Debian Jessie i386 repository credits Game_Music_Emu 0.52 by Shay Green and William Pitcock in its plugin list, and demonstrates the deficiencies I mentioned above. No website is given.
It doesn't depend on the libgme0 package, which is marked as version 0.5.5 and appears to be based on the fork by Michael Pyne that can be found at
https://bitbucket.org/mpyne/game-music-emu/wiki/Home.
retrodpc wrote:
FCP 3 starts on the 15th of June and ends on the 31st of July before midnight. ..... More details will be available by August 1st!
What timezone? You should specify these things
zzo38 wrote:
retrodpc wrote:
FCP 3 starts on the 15th of June and ends on the 31st of July before midnight. ..... More details will be available by August 1st!
What timezone? You should specify these things
I think it said GMT on the website. So, just 2 hours left!
UPDATE: The MySQL package auto-updated and caused all entries submitted from 21 July to not record description and song length. The affected people can be seen on this list:
https://cdn.discordapp.com/attachments/318506559313412097/342042020812161024/fcp3_borked_entrants.txtPlease submit your entries again; I am willing to answer any questions.
Due to this, entry has been extended to 7 August 2017 to 23:59:59 UTC. Good luck everyone!
I resubmitted three of my tracks that were borked. The rest I'll be making changes to.
I'm Arc-Demon, by the way.
Edit: I hope Airwe resubmits all those originals. I really liked their previous entries.