Fixing all of the too-loud audio packs for good

Thanks for the fixed version, emuandco

qwertymodo wrote:Sweet, that's one less for me to do.  They definitely needed to be redone from scratch, but I could never find the original source files.

You can find em (or better the d/l location) in the archive from now on, but they were here: http://www.dkc-atlas.com/forum/viewtopic.php?t=2073&p=42348 too ^^ Well at least the ones still available after a few years... I had to reuse ONE I did not find anymore. The beginning fanfare, but this one was not clipping afaik. Was a while ago I made the pack. ^^

Hey,

If anyone is interested, I just found an awesome new Genesis style Mega Man X cover by Gecko Yamori/Savage Regime!

https://www.youtube.com/watch?v=9CzkswHHxQI

I downloaded the .VGM files and converted them to WAV and I'm hosting it here: https://www.dropbox.com/s/wbs8qaoiatbgtp9/Mega%20Man%20X%20%20Genesis%20style%20by%20Savaged%20Regime.zip?dl=0

Also, I noticed that there isn't an updated version of Krzysztof Słowikowski's guitar covers of Mega Man X3. It would be great to have an updated version of X3 Zero Project. I hosted the WAV files from his site here if anyone is interested: https://drive.google.com/open?id=0B2lYWBs1GuFaOF9MRHBEMElkRkU

Can't thank you all enough for doing this stuff!

smokemonster wrote:Thanks for the fixed version, emuandco

I tried DKC2 with your newest pack and the big boss music didn't play I am the only one with this problem ? Tested on a SD2SNES Rev.F with borti op-amp board

I had started on Krzysztof Słowikowski's MMX1/2/3 packs awhile back but stopped when I saw they were available for purchase on his Bandcamp (I did post MMX1 because he had a free download of that linked from his Youtube channel). What I didn't realize was that he has the "no minimum" option enabled so they can be downloaded for free, so I'll go ahead and finish those up and get them uploaded. I'll probably take a shot at the Genesis version too if I get the chance, but it'll take awhile (at that point we're talking ~150 tracks to convert between the 4 packs). We'll see how far I get before I burn out again.

I should problaby do MMX2, I think I still have the original files before editing them. Right now I'm back on editing files for Rockman & Forte, should be done this weekend when I go back on it.

I can't wait for any person to do the MMX Savage Rein pack, his Genesis remixes are really good!

Hey DS...no plans on finishing the other one ..that is almost like chrono tirgger..
THE Akutoreiza.........Act Raiser..hmm maybe not!!!

I shall return to ActRaiser someday, it's a unfinished business!

I managed to increase loudness on the mp3 i got from youtube page https://m.youtube.com/watch?v=svWGSOT17XA.. remember DS..he never did better quality songs for you...so,i got it ahain there ..but. I increased the decibels +3 recomended..sounds louder now.. im goin to test it with the file u gave in the past

DarkShock wrote:

smokemonster wrote:Thanks for the fixed version, emuandco

I tried DKC2 with your newest pack and the big boss music didn't play I am the only one with this problem ? Tested on a SD2SNES Rev.F with borti op-amp board

Wooops... seems like I forgot one pcm file... will check what happened later.

New one with the missing file is uploading right now. @Con in 30 mins you have something to fix on your mirror.... sry ^^

@Con in 30 mins you have something to fix on your mirror.... sry ^^

No worries, already updated my mirror with dkc-21.pcm

I'd wish somebody will create DKC 1 anytime in the furure. I really love the life in the mines sound. I think it is one of the best spc ever made (next to zelda ending themes ).

Unrelated, there are so many mega man msu1 versions already that I thought it would be better to split the jum'n run database and make a separate Mega Man database
https://www.zeldix.net/f52-msu-1-database-mega-man

Oh, and RedScorpion managed to make me Secret of Evermore MSU asm. I think he already prepares the pcm:s, so stay tuned

kurrono, if that pack is the one being used here: https://www.youtube.com/watch?v=M52OFyX8cu0 then please fix the volume levels.  Also, if you're speeding up the "last track" music yourself, you've sped it up way too much, only about 15-20% would be correct.

Yes i was using Audacity for almost everything qwerty...but somebody uoloaded 3 different versions of mario kart pcms already...
1.One Ups ukule style guitar and jazz music...
2.Poulygon remixes...
3 .Remastered mixes...
4 will be my unique combined style mix..
Qwerty let me know if u still want mines ..im goin to fix it .if u want me to..
I like mines. More but. I love so much vanilla lake song..so lovely.

My songs only work on,sd2nes and ur snes9x ..and it doesnt work on bsnes..it plays the regular snes music..that got me sad..bsnes dont like me

bsnes requires a manifest. So does higan, unless you rename your files in a very specific manner, then they work without the manifest. Try this, saved to a file named the same as your rom, but .xml (e.g. if your ROM is smk-msu.sfc, then your manifest should be smk-msu.xml)

Code:

<!--?xml version="1.0" encoding="UTF-8"?--><cartridge region="NTSC">
   <rom>
      <map mode="shadow" address="00-3f:8000-ffff">
      <map mode="linear" address="40-7f:0000-ffff">
      <map mode="shadow" address="80-bf:8000-ffff">
      <map mode="linear" address="c0-ff:0000-ffff">
   </map></map></map></map></rom>
   <ram size="0x800">
      <map mode="linear" address="20-3f:6000-7fff">
      <map mode="linear" address="a0-bf:6000-7fff">
      <map mode="linear" address="70-7f:0000-ffff">
   </map></map></map></ram>
   <necdsp revision="uPD7725" frequency="8000000" program="dsp1b.bin" sha256="4d42db0f36faef263d6b93f508e8c1c4ae8fc2605fd35e3390ecc02905cd420c">
        <dr mask="001000" test="000000">
        <sr mask="001000" test="001000">
        <map address="00-1f:6000-7fff">
        <map address="80-9f:6000-7fff">
    </map></map></sr></dr></necdsp>
   <msu1>
        <mmio>
            <map address="00-3f:2000-2007">
            <map address="80-bf:2000-2007">
        </map></map></mmio>
   </msu1>
</cartridge>

It works now...awesome

I already talked with qwerty via pm, that the bszelda and little affected also AST pcm files were a bit too silent. The settings were made to ALTTP volume where the music is in the background and the sfx in the foreground. This made it hard in bszelda map1/2 to understand what is said, when e.g. throwing the sword. The sfx covered the msu voice.

I amplified now the files by qwerty for map1/2 by 7.5 db and AST by 3.5 db. It should be much better now to understand the voices:

map1:
https://drive.google.com/open?id=0B68bmgtrAXQARVEyMUdzZ0ZNeG8

map2:
https://drive.google.com/open?id=0B68bmgtrAXQAQ2VoTVdCdkJOamM

AST:
https://drive.google.com/open?id=0B68bmgtrAXQAeTZGUXRObmNsdUU

I'm finally done with my patch to substitute the spc with pcm files for Ancient Stone Tablets:
https://drive.google.com/open?id=0B3YAcoWoOm9VdTJOSXFXN2FsM00

In contrast to the English Dub, in this version you have alike ALTTP music changes in dependence of your location.
I uploaded in my drive the "normal" ALTTP version amplified to value FF, but I'd actually prefer having the ALBW pcm set

Qwertymode already adjusted it (first post in this thread):
http://www.mediafire.com/?kggx1xt9pp23y3k
But I think some themes are still way too loud (e.g., overworld, 1st AST dungeon), while the cave theme appears to be at the same level (I directly compared by having 2 snes9x open, one with the ALTTP (FF adjusted) and one with the ALBW soundtrack.

Can you confirm, Qwerty, and if yes, is there a chance that you adjust them a bit lower?

Hello! I'm Insidious611, the main developer for FF6 Dancing Mad MSU-1.

A contributor to said project linked me to this thread, as we were already in the process of normalizing many of the tracks and adding the fixed packs as part of our installer. (See this github issue)

qwertymodo, I hope you don't mind if I add your normalized tracks for FF6 to the download path on my server/mirrors? You'll be properly credited.

Hello Insidious611, glad to see you're still working on the project  Absolutely, feel free to use the tracks I posted here.  I actually posted them to your project thread awhile back, but I never saw a reply there, so I was worried you'd disappeared.  They are shorter loops than the OCR ones you're currently using, so if you like the loops you already had, you might want to treat them as completely separate options.

When you're doing your normalization, I highly recommend using nongnu normalize because it uses RMS normalization instead of peak, which is what you want for this kind of application.  I actually wrote up a batch script awhile back that did exactly that, but I'm not sure where it ended up.  Basically, you want to read the loop point from the MSU-1 header, then remove the header and replace it with a standard PCM WAV header.  Then you have a standard .wav file which you can feed into normalize.  Then just pull off the WAV header and copy the MSU header back on.  It's pretty simple.  From my pack, I found a value of -18dB seemed to be correct for FFVI.

Thankfully, now that ikari has updated the hardware on the SD2SNES (and has a firmware workaround for older hardware), the properly normalized audio files should work everywhere.  If you have people complaining that the audio is too SOFT on the SD2SNES after you normalize the files, tell them they need to enable the MSU-1 audio boost in the SD2SNES menu.

If you have any other questions, feel free to ask.

Edit: Here's my pcm2wav utility in C++.  Should be simple enough to port to your language of choice:  https://pastebin.com/PqEg2bdq

As to my disappearance, I've been checking back from time to time. I was a bit discouraged by my inability to work out fading or fixing the opera, plus life, work, and a litany of other projects got in the way. I've just recently gotten back into the swing of things, partly due to moving the project to GitHub in an (apparently successful) attempt to solicit more testing and help from like-minded developers.

It has been brought to my attention that apparently I made a typo in my ALBW OST config and normalized the whole thing at -12dB, not -21dB.  I must have fixed it locally, because I couldn't figure out why people were still complaining about it being too loud

In any case, after giving it a little bit closer listen, I actually think -24dB works better for most of the aLttP track packs, at least ones with really full instrumentation, because at -21dB, even though the level actually matches the original, it still seems to muddle the SFX a little.  At -24dB, everything comes through nice and clear while still sounding balanced.  So, if you've been using my ALBW OST pack... my apologies, go ahead and re-download it now.

Edit: I take it back, -21 really is the right level. I might go back and lower one or two tracks individually, but -24 is too low for a lot of the tracks. Reverted Drac9's pack to -21, I'll give ALBW a closer listen tomorrow.

Thanks, could you please give a list of the ones that you've fixed?

Have you updated any others lately? I'm trying to keep my sets up-to-date.

above post refers to albw set for alttp as the only having issues. You can re-download it from the first post in this thread or, I also listed them here
https://www.zeldix.net/t791-the-legend-of-zelda-a-link-to-the-past

Thanks, Conn. I wasn't sure if that was the only one. Thanks again for keeping up on these, qwertymodo

A while ago I also fixed map1
https://www.zeldix.net/t793-bs-zelda-map1-english-dub
And map2
https://www.zeldix.net/t794-bs-zelda-map2-english-dub

The audio was too silent in comparison to the sfx. If you redownload from these links they're fixed, but dunno whether they're updated in the first post of this thread already