MSU1 Volume Fix, The Second Wave

qwertymodo wrote:Pretty sure that was just a bug in Snes9x, nothing to do with the hack.  I had a similar skip occurring in Chrono Trigger.  Should be fixed now.

Yep, i know. Hence why i pointed Snes9x. It goes away if you load the game a second time in a row, it's the first load that screws up (and pops like hell).

Woaw it is really cool to continue to work on patch to make them better

And it is sooooo cool to fix Volume, so many games have music to hight and we don't listen the game. Thanks

I've been testing out TMNT IV on my sd2snes. Maybe it's just me, but for my money it still seems like it is too loud. I can barely hear any sound effects over the music.

Oh, I remember playtesting TMNT, it was difficult to set a good volume because by default, the songs are in fact very loud.

Having set it too low made my ears to believe I "killed" the guitar in those songs, so yeah, I kinda adjusted those songs following a loud pattern to preserve their frantic ^^''

But I haven't considered the SD2SNES edge case, so I'm gonna experiment a little bit with them once again

However, could you tell which hardware revision of the SD2SNES you are using? And if some of the boosting options are being used?

I have cpu revision 2, ppu1 revision 1, ppu2 revision 3. Firmware version 0.1.7e. I am not using any audio boosting options for the msu tracks.

I think there are maybe also other sets needing revision.

I never played it to certify with experience but DKC2 is also sounding partly a bit too loud in comparison to the SFX when watching the video:

https://youtu.be/7gXHtXn6bGA

TheGameFoxx, he is asking for your SD2SNES revision--it should be a letter, like F or G, printed on the PCB. Without this info, you can't diagnose volume issues with MSU-1 hacks, since volume levels have changed with each SD2SNES revision.

I went for DKC2 the easy way for now and added a dk2_msuMattrizzle_lowVolume.ips
that adjusts $2006 to #$60 instead of former #$FF
Dunno whether anybody is going to normalize the tracks better some time in the future, but until then it fixes the issue.

https://www.zeldix.net/t1445-donkey-kong-country-2#21218

For turtles, this is also possible: open your ~patched~ rom in a hex editor, go to address 0x0788b (Press Ctrl+G, insert 0x0788b) you find the hex value FF there, change it to 60 and you have lower volume when you load it on your sd card afterwards. If nobody wants to normalize the pcm:s this is not the cleanest, but the best way to fix this issue fast.
If wished, I could create also a second patch or whatever for turtles, similar like I did for DKC2 - but it would be better if the pcm:s are normalized instead of providing 2 patches with one being the outdated 60 volume level, hope Colines is going for it

My apologies, i didnt realise i had to open my cart for that info. Looks like its revision H mk II, whatever that means. As for what conn was saying, as patch would be preferred. I don't know the first thing about hex editing, but beggars cant be choosers i guess.

Done!! Sorry for taking so long, had to set up an entire machine from the ground up for the task.

I though emuandco already had normalized the DKC2 tracks? Gonna take a look at them, long ago I was planning to normalize those tracks too, but somehow I truly ended believing emuandco had fixed them X|

Anyway, here it is TMNT4! Same link as in the first page ;-)

Code:

https://drive.google.com/drive/folders/0B-G62GDNCoK0WUM4ZFIydWRkN3c

Please, tell me if it's still loud! This one is tough to get right, the SPC in-game songs are also loud, but maybe for some psychological effect, the MSU-1 songs simply don't mix very well with the SFX...

TheGameFoxx, he is asking for your SD2SNES revision--it should be a letter, like F or G, printed on the PCB. Without this info, you can't diagnose volume issues with MSU-1 hacks, since volume levels have changed with each SD2SNES revision.

Haha! Thanks for pointing that out! Boy, I totally missed he hasn't said his SD2SNES revision, when it was mentioned no boosting option were used and the songs were still loud, I just assumed it was the H rev

I could create also a second patch or whatever for turtles, similar like I did for DKC2 - but it would be better if the pcm:s are normalized instead of providing 2 patches with one being the outdated 60 volume level, hope Colines is going for it

https://www.zeldix.net/t1339-msu-1-volume-fix-the-second-wave#19832

Awesome, I updated:
https://www.zeldix.net/t1443-tmnt-iv-turtle-in-time#21216

https://www.zeldix.net/t1339-msu-1-volume-fix-the-second-wave#19832

Deja vu?

Also thanks for looking into DKC2
Maybe he fixed them, but I tried yesterday (with my mirror, but EmuandCo's are the same) and I still considered them too loud. But this is also a matter of taste I guess. When I tried with 60, they appeared perfect to me.

Colines wrote:Done!! Sorry for taking so long, had to set up an entire machine from the ground up for the task.

I though emuandco already had normalized the DKC2 tracks? Gonna take a look at them, long ago I was planning to normalize those tracks too, but somehow I truly ended believing emuandco had fixed them X|

Anyway, here it is TMNT4! Same link as in the first page ;-)

Code:

https://drive.google.com/drive/folders/0B-G62GDNCoK0WUM4ZFIydWRkN3c

Please, tell me if it's still loud! This one is tough to get right, the SPC in-game songs are also loud, but maybe for some psychological effect, the MSU-1 songs simply don't mix very well with the SFX...

TheGameFoxx, he is asking for your SD2SNES revision--it should be a letter, like F or G, printed on the PCB. Without this info, you can't diagnose volume issues with MSU-1 hacks, since volume levels have changed with each SD2SNES revision.

Haha! Thanks for pointing that out! Boy, I totally missed he hasn't said his SD2SNES revision, when it was mentioned no boosting option were used and the songs were still loud, I just assumed it was the H rev

I could create also a second patch or whatever for turtles, similar like I did for DKC2 - but it would be better if the pcm:s are normalized instead of providing 2 patches with one being the outdated 60 volume level, hope Colines is going for it

https://www.zeldix.net/t1339-msu-1-volume-fix-the-second-wave#19832

It still seems like the sound drowns out the sfx, its weird but it doesnt seem like it made any difference? i repatched my rom and redownloaded the music files and everything.

I think the wrong link might of been posted. the google drive still says it was last updated in 2017? and it only links to the patch, not the music files? im confused...

when i go to download the super mario world fix, the OW folder is empty? and the SFX folder too

The link you used is outdated, try this one:
https://1drv.ms/f/s!AtMUCmAL6x14hVmyYk0PYUmcguvk

Source:
https://www.zeldix.net/t1437-super-mario-world-msu
(for development purposes)

The Google Drive folder will eventually replace Onedrive once I update the Super Mario World plus assets.

Tthe reason the SFX and OW Music folders are empty, is because I haven't normalized those tracks yet. The Onedrive folder is the one outdated actually

For SMW+, please go for Onedrive, use the Google Drive for the native patch only ;D

I made you the author of the first post of
https://www.zeldix.net/t1437-super-mario-world-msu
so you can edit yourself if anything needs to be updated (but I strongly advise to keep the way the pcm are presented (preselected and developer purpose), was quite confusing already ^^ )

...and let us know if links changes there, I try to keep track:
https://www.zeldix.net/t1470-general-msu-1-discussion
(list in the first post).

Greetings!
My name is Chris and as my username suggests i'm from Switzerland. I bought a SD2SNES some months ago and i've become very fond of the capabilities of the device, namely MSU-1 audio. I already have some games enhanced with MSU-1 audio up and running and i really like this addition. I figured that it would just be blatantly rude to just reply to this topic as a new user, so that's why i tried to introduce myself.

First of all thanks for fixing the volume for some of the games i had issues with.

Let's get to the topic: The original MSU-1 hack of Turtles in Time was very very loud, now the current fix (thanks for your effort colines) is just too quiet for my taste. Because of my self-teaching nature i spent some hours learning about the tools you guys used to fix the volume. Long story short:

I used qwertymodos msupcm program to fix the volume to my liking. The shell shocked OC remix FLAC files where used as my source, just like in the original hack. I wrote the corresponding .json file.

I'd love to have some feedback from you guys and i hopefully haven't offended you colines.

How should i proceed with the .json file? Sent it to someone or upload it publicly?

Thanks for you effort so far!

DatSwissGuy wrote:How should i proceed with the .json file?

From qwerty's thread here: https://www.zeldix.net/t1400-msupcm

The two simplest ways to use the application are to name your config file "tracks.json" and place it in the same directory as the .exe (along with the audio files), and just double-click on the .exe. Or, you can drag-and-drop your .json file onto the .exe, in which case it can be named anything (as long as it still has a .json file extension).

Hope that helps.

Thanks Relikk!

I probably have expressed myself incorrectly, i already generated the PCM's. It's just a matter of playtesting .



I noticed something: The Files don't loop even though i set the loop parameter at the imo correct sampling points. Guess i'll have to study further...

Edit:
My .json file is available here (nope! completely wrong!): https://drive.google.com/drive/folders/1EdJS_ZPBBk_amnzgyyACJXtKiAiUk_uZ?usp=sharing

You'll need the Shell Shocked OC Remix FLAC Files and qwertymodo's msupcm in order to generate the PCM's

Ah. Well, if you figure out the looping issue and are happy with the overall result, including your own volume fixes, you can upload them somewhere and link them here so other people can test them if they're willing.

I'm still working on it until i nail it. Since i used to play in 3 different bands (metalcore, blues and rock'n'roll) my personal standards are... let's say harsh.

Thanks anyway!

Edit:
Alright, i figured it out. DarkShocks files have some trimming which i missed, ergo the loop at these specific sample-points ( https://github.com/mlarouche/TurtleInTime-MSU1/blob/master/create_pcm.bat ) won't work. As i looked at his source,  the necessary information to copy/reproduce the base files for looping are missing: https://github.com/mlarouche/TurtleInTime-MSU1/blob/master/edit_audio.bat <- does not look like TMNT.

Oh well...but i already have an idea as how to get what i want: Use the old and very loud pcm files, convert back to .flac or .wav and then recode back with msupcm  

Another edit:
Of course trying to convert a .pcm file with foobar2000 which has looping points will result in an endless file...  

At least i learned a lot... Here are my normalisation values:
for all tracks of TMNT: -16 dbfs
track 232 (Hot for April): -20 dbfs

that's all...

I'll keep working on it.

Very welcome to these forums! Much glad for you joining us =D

DatSwissGuy wrote:The original MSU-1 hack of Turtles in Time was very very loud

They certainly were loud, indeed! But thankfully, DarkShock made sure to not clip the tracks, so in any case, working on his tracks and taking advantage of his edits/loop points won't give you any headaches ;-)

For such, use this tool:
https://github.com/qwertymodo/msu2wav

now the current fix (thanks for your effort colines) is just too quiet for my taste.

Yeah, I lowered them down after receiving a complaint about the tracks being too loud, and even then, they supposedly still kept sounding a bit loud!

I even tried to match them exactly to the same volume level the in-game SPC tracks have, but I still could feel a huge difference between SFX and other not replaced SPC songs. Of course, all of that was subjective, my rationale is that these tracks simply clash too much with the overall game environment, and since the only feedback I had for an entire year was a single report, I couldn't make my mind on this one ='(

Hopefully, with you joining the quest, this will be resolved once and for all!

haven't offended you colines.

Don't worry ^_^

I'm actually very happy, it was always qwertymodo's intention to bring more people to the game, pioneers MSU-1 rom hacks suffer from old and deprecated standards like IPS patching, broken volume levels and emulator versions; every opportunity to fix them is welcome =)

Use the old and very loud pcm files, convert back to .flac or .wav and then recode back with msupcm

Oh no, they are never recoded! The custom PCM format the MSU-1 uses only strips the WAV header. Turning it back would be only a matter of hex editing ;D

Also, the FLAC source volume isn't much different of DarkShock's one. Again, in case you just want to reuse what he has done: ^^
https://github.com/qwertymodo/msu2wav

metalcore, blues and rock'n'roll

Two out of three, not a bad taste XD

Colines wrote: For such, use this tool:
https://github.com/qwertymodo/msu2wav

Awesome! Source included! Although it's been quite a while i dealt with programming. Still have to set up my Dev-Enviroment for two machines (OSX/BSD and Windows). ASM is going to be a little bit tricky for starters, but i know at least some basics. Words like cycles, memory addresses, flags etc. are not unknown to me. It'll be just practice and some trial and error till i get the hang of it

Colines wrote:
Yeah, I lowered them down after receiving a complaint about the tracks being too loud, and even then, they supposedly still kept sounding a bit loud!

I've looked at the audio source...the whole album is just terribly mastered and it's dynamic range is too high for people that don't have the necessary devices like studio monitors. I use a STAX SRM-252S amp + SR 207 electrostatic earspeakers (that's what STAX calls earphones). You can't get closer to reference than these bad boys.

Colines wrote:
I even tried to match them exactly to the same volume level the in-game SPC tracks have, but I still could feel a huge difference between SFX and other not replaced SPC songs.

That's why in some studios they master with a loop-back: audio out looped back into "audio in" to see the real waveform. Yet still it's a tedious process of listening the mastered tracks over and over again till you get insane

Just looking at the waveform and checking it's amplification/gain it might look like they are normalized, but then in practice everything falls apart (worst case scenario)  

Colines wrote:
Oh no, they are never recoded! The custom PCM format the MSU-1 uses only strips the WAV header. Turning it back would be only a matter of hex editing ;D

Big big hug from me! Can't express myself correctly, but reading this i did a little dance of joy!

Colines wrote:
Also, the FLAC source volume isn't much different of DarkShock's one. Again, in case you just want to reuse what he has done: ^^
https://github.com/qwertymodo/msu2wav

I might just do that for the republished/redistributed version. Yet still it's a good practice for me to do the whole process of looping, normalizing etc.

Looking forward to present my work. I'll estimate it'll take me between 3-4 weeks to do so.

If you wish i can have a look at all the other fixed volume packs and give some feedback.

I'ts done! Well...the volume adjustment is at least

Based on DarkShocks .pcm's and with help of qwertymodo's tools msu2wav and msupcm++ i was able to adjust the volume of the whole pack!

Yet one edit has been done: The intro song as been re-trimmt slightly (source is FLAC album). It's just a couple of milliseconds at the beginning and the end where they shout "...turtle power!".

Where can i upload or who shall i contact for playtesting/checking/redistribution?

P.S. i will try to edit the whole pack for myself and maybe provide some new loop-points which include bridges/interludes/solos etc. I will also "fiddle" around with ASM.