foo_input_usf by PokeParadox at 1:24 PM EDT on August 9, 2007
I'm really glad there's a plugin for foobar finally!

Josh.. I'm just wondering how that's coming along? Any chance of mapping the USF field details to fooabr TAGZ fields? It would make them display much more nicely.
by hcs at 2:35 AM EDT on August 12, 2007
Uh, what? Was this an IRC discussion I missed?
[edit]
Oh, right, that other thread I'd forgotten.

edited 2:37 AM EDT August 12, 2007
by Josh W at 8:13 AM EDT on August 13, 2007
Heh, it was a thread i'd forgotten also.

OK, i got an update here

It has a few things fixed, such as a few memory leaks, and support for playing inside archives.

Ive added reading of the titles to use foobars tagz fields, but tag writing support...well, i wouldn't trust it.

I couldn't however get it to stutter, either because neither of my computers would make it, or most likely fixed it and forgotten about it.

Also the configuration dialog and configuration saving / loading is implemented.
by PokeParadox at 10:42 AM EDT on August 13, 2007
Yay for updates! That's much more usable now!

EDIT: I wouldn't mind access to the source, just to see what's going on there... There are a few plugins that are overdue foobarisation...

But this really is great now! Just settings saving/loading to go I guess.

edited 12:14 PM EDT August 13, 2007
MUCH better!! Great job! by Yoshinkeru at 10:57 AM EDT on August 13, 2007
I knew you wouldn't let us down! :D Thanks for fixing the "memory leak", and I LOVE how it reads the tags; looks much nicer than it did when I had to use the Winamp wrapper for hcs's.

Many props, Josh, many props! You ARE da man!
by mudlord at 9:22 PM EDT on August 13, 2007
Wow! This is impressive! I know some folks on Hydrogen Audio will be very pleased ^^
by mudlord at 7:04 PM EDT on August 14, 2007
I've noticed one teensy problem: The setting for Audio HLE won't save.

Its currently a pain to keep unchecking Audio HLE, and so it would be great if this can be fixed.
by Josh W at 6:00 AM EDT on August 15, 2007
oops, sorry. I forgot to change it when copying/pasting the saving settings code and it wrote the cpu priority instead.

Ok, i fixed it and re-uploaded it :)
by hcs at 11:07 AM EDT on August 15, 2007
Should this be mentioned/linked on USF Central?
by mudlord at 9:55 PM EDT on August 15, 2007
oops, sorry. I forgot to change it when copying/pasting the saving settings code and it wrote the cpu priority instead.

Ok, i fixed it and re-uploaded it :)


Yep, its fixed :). Works like a charm...I noticed a bug with reading large tags also. Look at the Pokemon Stadium rip tags, in the default foobar2000 interface to see what I mean. :)
by Josh W at 9:49 AM EDT on August 16, 2007
okay, that was one of the things i never thought would happen, so i didn't bother with it.

BTW, it wasn't the large tags that caused it, it was that blasted é. So i changed everything to proper converting to UTF8 strings.

Also i added an option to select between Tagz values and custom title formatting, default file names on no title, and added tooltips for help on the config dialog.

Get it here

HCS, it would be good if this were on USF Central, then more people would know about it. Not everyone checks the forum.

edited 9:53 AM EDT August 16, 2007
Let\'s do it! by Yoshinkeru at 1:31 AM EDT on August 17, 2007
hcs said: Should this be mentioned/linked on USF Central?

JoshW said: HCS, it would be good if this were on USF Central, then more people would know about it. Not everyone checks the forum.


There's your answer, hcs! Let's get this party started!!
by hcs at 3:30 AM EDT on August 17, 2007
Do you really consider writing a line of HTML a party?
In any case, up now.
by PokeParadox at 7:46 AM EDT on August 17, 2007
It could be an HTML party... if such things exist...
by hcs at 9:32 AM EDT on August 17, 2007
<party>
w00t
</party>
by Mouser X at 3:00 AM EDT on August 18, 2007
I'm using my brother's Wii to type this. I just wanted to say awesome work on the fb2k v. of 64th Note. Hopefully with it being mentioned on the main page, more people will find, and use it. Of course that probably means more bug reports and feature requests as well... Anyway, good news. Mouser X over and out.
Auto HLE by PokeParadox at 7:55 AM EDT on August 18, 2007
Just wondering if it's possible to add that in somehow? Maybe move the settings for this into a file called auto.ini or something?
Auto AHLE by Yoshinkeru at 3:54 PM EDT on August 20, 2007
Hopefully he's working on it in the next build anyway, but I also second this just in case.

Thing is, I've discovered either the Mario Kart 64 or Mario Party (1) set doesn't play correctly. One or the other, never both, nor neither. It's very strange.

Also, a few of the Donkey Kong 64 tracks start of weird, but after some seconds play normally.
Oh no, another problem... by Yoshinkeru at 9:30 PM EDT on August 22, 2007
Dangit. When I have foobar set on random play, and it finishes a USF track, it seems to hang sometimes.

(Also: should have been "start off weird")

It always seems these plugins always need tinkering...
by Lunar at 5:34 AM EST on November 17, 2014
Bump from hell!

At some point after kode54 took over development of foo_input_usf, it became slow as hell for me, taking up to 7 seconds for songs to initialise. This also has the effect of interrupting the ends of songs as the next USF in a playlist is loaded. foo_input_2sf is similarly sluggish, but at least that one always was.

Just wondering if anyone else has this problem and if there's anything that can be done about it. I've tried the version that's available on Zophar's Domain, which doesn't have this problem, but sadly has others (like crashes with some songs ie. Mario Tennis.) It is possible that an old but more recent version may work better, if anybody could direct me to one, but any suggestions in lieu of that?

edited 11:49 AM EST November 17, 2014
by punk7890-2 at 8:52 AM EST on November 17, 2014
Yeah. I get the same thing now. I think it has something to do with him adding SSE2 support. Though mine isn't much of a problem as yours. Mine takes about 1.5 secs to start now when it was instantaneous.
by kode54 at 11:20 PM EST on November 23, 2014
It has to do with eliminating the dynamic recompilation processor cores, as they were non portable and quite a horrid mess.

Enable HLE in the configuration page if your computer is on par with my iPod Touch, which starts playing comfortably with the HLE mode.

The new, or should I say old, HLE R4300i routines which I dug up from a JoshW josh-use.rar from 2008, add an additional 2 to 5 percent speed boost.

As I said in the other lazyusf topic, if anyone wants dynarec back in, they're going to need to supply a re compilation engine which supports a state machine structure for reentrancy, and they're going to need to target at least x86, x86-64, ARM, and ARM64. Nothing more really, just that for starters.
by Lunar at 2:05 AM EST on November 25, 2014
That's too bad. I'm using 3.1GHz Core i3, not the beefiest processor on the one hand, but less than three years old. I've enabled HLE, it hasn't made a significant difference ~ 5+ seconds on track initialisation (almost 10 seconds for Turok Seeds of Evil.)

I'm guessing the best I can hope for is to revert to the version prior to when the change you mentioned happened, if it is still available.
by djDarkX at 6:01 AM EST on November 25, 2014
First time poster, but I've noticed that 64th Note on Winamp doesn't have this issue. I agree, when I updated the plugin on foobar, aside from slow initialization, certain music from certain games also seems to lag and skip causing audio gaps.

Using 64th Note, it didn't happen. I guess the change was necessary on foo_input_usf, but now I've reverted to using the last version of Winamp and the latest of 64th Note just for USF playback.
by punk7890-2 at 12:03 PM EST on November 25, 2014
Enabling HLE seems to get rid of the time it takes to load a song now. Odd that it takes so long for some other people.
by kode54 at 7:35 PM EST on November 25, 2014
Initialization involves skipping up to the configured 10 second default worth of silence, then buffers up to that amount for end of track detection.

You can turn off both features in the preferences, but know that if a track is untimed and ends with silence, it will continue to play until the default timer runs out.

Yes, the interpreter cores are about four times as slow as the recompilers, but they're portable.

Again, if anyone wants to contribute portable recompilers, they're welcome to.

Or stick to an old version and don't ask foobar2000 for updates. The recompilers are also prone to crashing, while these cores are not, at least in their current state.
by Lunar at 9:02 AM EST on November 26, 2014
Or stick to an old version and don't ask foobar2000 for updates.

I'd be satisfied with this, just need somewhere to download the last version prior to the removal of the recompilers.
by kode54 at 10:09 PM EST on November 26, 2014
Unless joshw updated his copy, he should have a version that predates my fork of lazyusf.
by kode54 at 10:20 PM EST on November 28, 2014
Try the latest update now. If you managed to play full speed before, this will start up quite fast. Robustness of the end of track silence detection depends on machine speed, though.
by djDarkX at 11:24 AM EST on November 29, 2014
New version seems to have fixed the issues I was having. I also was using HLE and was getting slow startups and audio cut-outs from the lag. As I said, however, this seems to have fixed the issue.

Thanks for the update kode54! You do really good work. Your plugins made the transition from Winamp to Foobar so painless. lol
by nothingtosay at 11:28 AM EST on November 29, 2014
"Thanks for the update kode54! You do really good work. Your plugins made the transition from Winamp to Foobar so painless."
Word.
by DrO at 11:48 AM EST on November 29, 2014
who'd use still use Winamp? you'd have be a fool or clinically insane to do something like that..
by Mouser X at 6:05 PM EST on November 29, 2014
I must be insane then. Though, I don't actually use the Dec. 2013 version (I don't remember what version it was, or what version I use). I *did* download the Dec. 2013 version/installer, but I don't use the Winamp.exe it came with. I used the installer to manually update all the plugins (I opened the installer as an archive, and pulled out individual plugins, and updated those that were newer versions than the ones I had installed. Obviously I tried each one to make sure they didn't crash Winamp). So my version of Winamp is older than the "final" release, but I'm using all the newest plugins (that don't cause it to crash).

Yes, actually, I can see that being an acceptable reason to be considered insane. It was a lot pf work, actually. I did it, though, because the newest version of Winamp caused "Next on Stop" to cease functioning. So, I'm using whatever version of Winamp that "Next on Stop" still worked in. I haven't noticed any problems with my "hybrid" Winamp, so I'm happy. :P. Mouser X over and out.
by Tanookirby at 3:15 PM EST on December 5, 2014
@kode54: If you're looking for a challenge, here's an idea: try basing your plugin on the latest version of Project64. If you can also create a new version of 64th Note or equivalent, that would be awesome.
by kode54 at 6:21 PM EST on December 12, 2014
Here's the thing. Does the latest Project64 load save states from 1.4? Because every single USF set in existence depends on loading Project64 1.4 save states.

Also, again, I'd only be looking into the interpreter cores, because I want this to remain portable. I could look into recompiling cores for different platforms, but again, there's the problem of portability.

Oh, and I could update the Winamp plugin, but the container aspect looks like it would be simpler to retrofit it into Neill Corlett's existing ESP framework. It just needs a PSF loader, plugin frontend, tag editor, etc.

And a will to still maintain software for Winamp.
by punk7890-2 at 2:13 PM EST on December 16, 2014
Noticed there's now some odd shakiness in Jet Force Gemini's cymbals and some tracks have a bit of a weird starting. Here's an example (notice the weird start). The first 16 seconds are from the latest Lazy USF and the second is from a recording in PJ64.
Not sure how to explain it. The second version just sounds... more stable?
by kode54 at 8:09 PM EST on December 16, 2014
Probably some weird timing involving the vertical blanking interrupt/timer, which was just changed. It's probably not the best now, but it's better than what it was before. Other than that, I have no clue.
by punk7890-2 at 11:15 AM EST on December 17, 2014
Well that sucks. Is there anyway to revert to older versions that doesn't have this shakiness? It's quite noticeable on Perfect Dark.
by punk7890-2 at 2:16 PM EST on December 19, 2014
Welp. I guess it was the multi resampler I was using in Linear interpolation. All good now. Also a song in Jet Force Gemini no longer works:
Unable to open item for playback (Unknown audio command 0: 000004e0 1bee000e
Unknown audio command 0: 00000650 1bee000e
Unknown audio command 0: 000004e0 1bee000e
Unknown audio command 0: 00000650 1bee000e
Unknown audio command 0: 000004e0 1bee000e
Unknown audio command 0: 00000650 1bee000e
Unknown audio command 0: 000004e0 1bee000e
Unknown audio command 0: 00000650 1bee000e):
"D:\Music\Jet\58 Asteroid.miniusf"

edited 7:50 PM EST December 19, 2014
by kode54 at 6:58 PM EST on December 20, 2014
Looks like naudio_mp3 has an actual command in slot 0, doing something I can't really decipher, since I kind of suck at identifying RSP assembly. For now, use the LLE audio for that file.

Looks like this one uses resonant filtering as well, and something else that a cleverer hacker than I should get to work on converting to C code. :)
by Lunar at 4:57 AM EST on January 15, 2015
Pokemon Stadium - "Magikarp Splash" is crashing fb2k on the current version.

Maybe it could be renamed Magikarp Crash, considering.
by kode54 at 6:31 PM EST on January 15, 2015
Please verify if I've fixed it. There was a bug with start of track silence removal and the buffering that followed it.
by Lunar at 6:05 AM EST on January 16, 2015
Yup, fixed! Thank you kindly.
by punk7890-2 at 12:16 PM EST on January 16, 2015
Is there anyway to apply a sorta hack to a USF set to resample the audio internally? I've tinkered around via hex editing with no results. Kinda like Konami does with their games. I'd love to hear Jet Force Gemini, or another Rare game in this quality. I know it's probably not easy but if it's possible, what would I need to learn?
by kode54 at 10:28 PM EST on January 17, 2015
That would require understanding how each game's sound engine works, since the engine setup routines are probably where the mixing rate is configured. Which also may or may not be called before the save state was made, in which case the entire sound engine setup is already completed and part of the USF save state.
by punk7890-2 at 9:20 AM EST on January 18, 2015
Hm ok. So I guess it's not as simple as searching for the sample rate (like searching for a 22000Hz being 55F0). Would the mixing rate be track specific (a individual miniusf) or as part of the usflib?

EDIT: I have no idea what I'm doing. I think I'll just stick to using your mulitresampler. Thanks for the reply.

edited 4:17 PM EST January 18, 2015


Go to Page 0

Search this thread

Show all threads

Reply to this thread:

User Name Tags:

bold: [b]bold[/b]
italics: [i]italics[/i]
emphasis: [em]emphasis[/em]
underline: [u]underline[/u]
small: [small]small[/small]
Link: [url=http://www.google.com]Link[/url]

[img=https://www.hcs64.com/images/mm1.png]
Password
Subject
Message

HCS Forum Index
Halley's Comet Software
forum source