A tool used to insert custom music into Super Mario World. This tool is completely hardware compatible and is designed with ease-of-use in mind. This tool supports both songs made for Addmusic 4.05 as well as AddmusicM with minimal tweaking.
Please read the readme for more information.
Note for Mac/Linux users: The binaries inside are for Windows. However, a makefile is present that allows the code to be compiled once src.zip is unzipped.
Novamente não funcionou para mim, eu coloco a música no jogo, porém tudo fica um silêncio, nada toca, apenas o pulo do Mario. Vocês poderiam consertar esse bug?
Yup, I saw your comment the first go round on 1.0.10... and after 1.0.11 released, I ended up recording the issue here and conjured up a possible patch for it here, since I determined that a missed signal could have been the cause.
What I did was remove a few hardcoded NOPs and replaced them with checking for a zero in $2141, since I determined the behavior there is more consistent.
Tested with various songs on both normal and SA-1 SMW roms, everything works as advertised and the changelog is accurate.
Tested on Windows 10 22H2 (OS Build 19045.4842)
O mesmo acontece comigo, eu coloco a música e nem mais o logo da Nintendo faz barulho, porém quando se ativa um p-switch quando se esta na fase com a música personalizada tudo funciona normalmente, mas quando volta para o mapa volta a ficar sem som.
Hello there! Welcome to the commentary section of AddmusicK 1.0.11 (And subsequently, the start of the prerelease update feed for this version)! I initially submitted this on 6/7/24, but I ended up updating the release candidate on 7/2/24 to fix a regression reported by LadiesMan217 regarding an accidental inclusion of an if statement after a && marker, causing it to be misrecognized as a unknown label "if". I also decided to fix something that I suspect could have been a regression, but I ultimately decided to do something else: namely, in the stat .txt files, the tick counters were supposed to be in hex, but were coming out as non-hex numbers due to some code not working. I simply decided to get rid of the hex conversion altogether due to some potential complications in doing so given the C++-side data type the tick counters were using, and also to avoid integer overflowing problems.
For those of you that downloaded this tool prior to 7/2/24, please redownload the tool, as it has been updated.
The main question on my mind is... is there a hijack conflict of some sort going on? I'm wondering because it would be either in the $008179-$0081AA area or at $02A763 (in the latter case, I NOPed out an area where the routine originally was, though I ended up restoring the PLB/RTL opcodes because of a different patch conflict that was reported to me). Whatever the case, it likely happened when the Yoshi Drums tried to activate.
This is very weird and should not happen at all unless you loaded from a savestate and then entered the level with custom music. Please don’t do that, instead boot the game from the very beginning (title screen, intro screen…) and then enter the level from there. Hope I helped!
When re-adding the echo, I edited the code, then patched in the uberASM patch, and initially that caused my ROM to crash. When I disabled all the patch's hijacks, my ROM worked fine, but there was still no echo. Am I missing anything?
What echo do you speak of? SFX echo? Or was it something else?
A couple of things running through my mind with regards to the crash:
Insert order may be a factor here, as I already dealt with a situation regarding the Yoshi Drums hijack when AddmusicK is patched first, then UberASM.
If one of your hijacks relies on a PLB and an RTL opcode at $02A771, then that is a known regression of AddmusicK 1.0.9. Try the AddmusicK 1.0.10 release candidate, as it modifies the way the Yoshi Drums hijack is handled so that those two opcodes are in their original places. It also has a few other regression fixes.
There are some ports that have BRR, if I insert them without BRR they work but when I insert them with BBR it no longer accepts them, I tried in every way but it doesn't accept them why I don't know? I don't know if you know people are Dominator and ◆oL7G5poF4c are Japanese smw hack music creators. Please help me with the songs they are from 2009 2011
Which echo do you speak of? I ask because there is indeed a crashing bug that is fixed when using a global song with echo and then going back to a local song with an EDL of zero through the $FA $04 command.
What song did you try to insert? I'm curious, because the readme does have some extra steps for implementing AddmusicM songs (among them using the #amm flag). And if the special pulse wave is involved, well... it's not in the official AddmusicK to avoid memory constraints, but I did implement that in my fork, AddmusicKFF, in the bleeding-edge branch. And I'm still working on some other historical adaptations.
Continuing off of this previous comment (I'll keep posting these update notices in the comment until the tool exits the prerelease state when it is moderated)...
Anas and Mellonpizza both reported Yoshi-related crashes with this release candidate, and the latter forwarded a fix to Kevin who then forwarded it to me. It turns out that I accidentally introduced a regression due to a stack fault from forgetting to remove a PLB opcode while fixing a patch conflict with the Yoshi Drum hijack, so that was repaired on 12/12/23.
I also implemented a suggestion by Fernap to use padbyte $EA (a bunch of NOPs for padding) and pad on 12/12/23. I did this on the Yoshi Drums hijack and the Start/Select SFX hijack so that they no longer had to be counted manually. I also added warnpc calls prior to the padding to avoid overwriting anything else, since the hijacks are now assumed to be variable size.
On 12/19/23, I added warnings for using non-standard FIR coefficient table IDs with using the $F1 hex command after having been notified of it from Argonfunk (I had accidentally broken them in 1.0.9 without knowing because of ASM modifications). I also, for the non-Windows users, updated make clean in the makefile because I forgot to make it remove AM4Batch and AMMBatch.
For those of you that downloaded this tool prior to 12/19/23, please redownload the tool, as it has been updated.
Heya I'm having troubles with the retry system, aparently, with the update of Addmusick of V1.0.8 (or V1.0.9 XD) to V1.0.10, for thoses that are using the patch, it's will don't work anymore if you decide to run the program, and then, insert the main patch (retry.asm)
I don't know why makes impossible use the patch with this version of Addmusick
Hello there! Welcome to the commentary section of AddmusicK 1.0.10 (And subsequently, the start of the prerelease update feed for this version)! I initially submitted this on 10/23/23, but I ended up updating the release candidate on 10/29/23 to upgrade the song memory overflow error messages. They now distinguish between overflows caused by the samples and overflows caused by the echo buffer, and they no longer round to the nearest $100 bytes.
For those of you that downloaded this tool prior to 10/29/23, please redownload the tool, as it has been updated.
That specific one was contributed towards the beginning of the fork's life by Kevin shortly after I debuted my AddmusicKFF fork on the SMWCentral side during Winter C3 2021. It also came packaged with the first of two rounds worth of global song optimizations.
Continuing off of this previous comment (I'll keep posting these update notices in the comment until the tool exits the prerelease state when it is moderated)...
I have updated the release candidate on 8/28/23 after notifying the moderators about a bug report I had received from Ahrion.
Specifically, the bug in question involved an accidentally duplicated sample causing memory overflows whereas it wasn't doing do in 1.0.8. It turns out that the important markered samples were correctly caught as duplicates, but were then loaded twice. A longer term fix will involve both the C++ and SNES sides because I want to use duplicate pointers in these cases, but for now, I have made it so that duplicate samples are replaced with an empty sample slot like 1.0.8. This also affects sample groups, though, because 1.0.8 still ran into the duplicate situation in those cases.
For those of you that downloaded this tool prior to 9/9/23, please redownload the tool, as it has been updated.
The tool was initially been claimed for moderation on 7/25/23 (it was temporarily unclaimed and reclaimed on 8/28/23 once I notified them, and again on 9/9/23), so unless there's another update during the moderation process, then 9/9/23's update is the last release candidate prior to finalization.
Either way, you're a little late on that one, since it was already claimed for moderation (and if I were to do that, I'd also have to change 1.0.7's method because there's a different problem that causes long notes/legato to not come back).
I'll check and see what Medic did with AddmusicK 1.0.7 on the pause side of things (perhaps the SNES-side modifications may be of note anyways). I may have my own ideas as well...
Can the pause fix be implemented back? Somehow it got reverted in 1.0.8. And sometimes when I pause it makes the game go silent. At least on 1.0.8. Has this been fixed in this version?
I don’t think so sadly. I can still encounter that occasional bug in 1.0.9 as well. At least it is temporary so if you pause and unpause again, all the audio should come back normally. Perhaps this bug fix could be optional?
Either way, you're a little late on that one, since it was already claimed for moderation (and if I were to do that, I'd also have to change 1.0.7's method because there's a different problem that causes long notes/legato to not come back).
Thanks for the update. I don't know if it was intentional, but the zip file doesn't have a top-level version folder. v1.0.8 has all files in a subdirectory of AddmusicK_1.0.8, and v1.0.9 has all files in the top level.
I only noticed because I have some software that automatically extracts the zip files and used that structure. Not a problem for me to adapt to it, but thought I'd report in case it was not on purpose.
It was completely accidental... and it happened as part of the auto-compilation process. The build script forgot to make an outermost layer folder and I never caught it due to the way I extract things on the Mac, as it automatically creates the outermost layer whenever I extract a .zip file. Good thing I have a hex editor for this particular situation. On the flip side, I checked what would happen if I simply re-zipped the entire thing. I ended up getting an outermost layer folder by default.
Continuing off of this previous comment (I'll keep posting these update notices in the comment until the tool exits the prerelease state when it is moderated)...
Kevin provided me with a few contributions that I merged in on 6/20/23. They are:
A series of global song optimizations, shrinking down their memory usage. In addition, the Boss Clear song had a bugfix where the note was not keyed off.
A fastrom-related speedup for explicit JML/JSLs when not using the SA-1.
A bugfix where setting the P-Switch and/or Starman song IDs to zero would result in the song ID being zeroed out on every frame, preventing user code from modifying the song ID while a P-Switch and/or Star was going off.
On 7/11/23, I received a contribution from Atari2.0 that updated all of the DLL bindings to the version that asar 1.81 was using. It also fixes a bug where the detection was being used too late, as asar had already been called beforehand.
For those of you that downloaded this tool prior to 7/11/23, please redownload the tool, as it has been updated.
The tool has been claimed for moderation on 7/25/23, so unless there's an update during the moderation process, then 7/11/23's update is the last release candidate prior to finalization.
did you guys know that you can type ".." (means a privious folder) when you're typing the rom name, for example: "../../SMWedit.smc" (actually this is my rom address)
Continuing off of this previous comment (I'll keep posting these update notices in the comment until the tool exits the prerelease state when it is moderated)...
No it isn't: it's distributed the same way as the original AddmusicK: in a src.zip file. Just extract the .zip to get the C++-side source code. The rest of the source code is still distributed normally.
What do you mean? Are you talking about the site’s SPC player? If so, clicking the pause/play button again should pause it again.
Not the site's SPC player, the option in the AddMusicK Tool itself as shown here. When you play a song, there seems to be no way to play it or pause it. Am I missing something?
Playing the song in AMKGUI means playing the SPC in a separate SPC player that you set as default. Your SPC Player should have the ability of pausing or stopping the song.
Wasn't sure where to post, but just wanted to submit a suggestion, a way to stop or pause the playing music. There is a button for "Play this song", but when playing, there is no discernible way to actually stop the music. Perhaps another button to stop it, or even the same button could function as a stop/pause button when playing a song? This could be very helpful when you want to quickly see what song is what, and there are a lot to go through. Once you find it, you may want to stop listening to it.
Continuing off of this previous comment (I'll keep posting these update notices in the comment until the tool exits the prerelease state when it is moderated)...
On 2/14/23, I updated the sound driver both for a minor optimization and a regression fix to the hot patch system where the second byte was failing to be properly processed due to not being stored to a scratch RAM location.
Is this song in the same folder as your AddmusicK? BTW, it's recommended that you have a song list for your hack in a special folder. Just copy all the AMK components you need to your hack's folder.
Technically one can use the $06 1DFA command to do so, but naturally that requires some 65816 ASM programming and it has to be triggered after the song starts (this is because SFX echo gets cleared under normal circumstances on song initialization). In hindsight, perhaps I could do something so that the SFX echo enable switch only resets upon loading a new local song (or if the user sends a $05 to 1DFA), and have that be a user define.
In both cases, they're not included by default to avoid increasing the SPC700 ASM filesize to beyond 1.0.8's (I also count SFX ASM, but not the SFX data: this is because the pause SFX ASM got relocated outside the SFX instance because I added a !noSFX user define and wanted to keep pause support in that case) (plus, I ended doing other things, with bugfixes being a priority of mine)... and I've already done a massive amount of optimizations just to get the other stuff in there. In particular, the hot patch system (which was designed both to address playback inconsistency between older Addmusics and AddmusicK and also has a few other quirk fixes), one of the hot patches itself for enabling enchanted readahead (it now looks inside the $E6 and $E9 hex commands) and handling noise conflicts between music and SFX (yes, I did this so that they would play nice together... kind of... what really happens is that the SFX's noise mutes the music's noise while it is in use) took up a good chunk of ASM filesize just to implement. There's only so much I can do optimization-wise, and space is extremely tight: right now, I'm only a few bytes away from exceeding that threshold. Yes, I did optimize the SFX data, but I'm taking into account those that don't use the default SFX pack (or the default global instrument data).
A bit complex, but you get the idea. Thank you very much for your attention . (^^)
I think I know what you're talking about... is it, for example, the echo in the SFX that normally is in place in Vanilla Super Mario World in the cave levels?
Yes, exactly. It's because with so many good songs being released and even more so with the echo effect, it's much cooler with the movement sounds that Mario produces.
Technically one can use the $06 1DFA command to do so, but naturally that requires some 65816 ASM programming and it has to be triggered after the song starts (this is because SFX echo gets cleared under normal circumstances on song initialization). In hindsight, perhaps I could do something so that the SFX echo enable switch only resets upon loading a new local song (or if the user sends a $05 to 1DFA), and have that be a user define.
In both cases, they're not included by default to avoid increasing the SPC700 ASM filesize to beyond 1.0.8's (I also count SFX ASM, but not the SFX data: this is because the pause SFX ASM got relocated outside the SFX instance because I added a !noSFX user define and wanted to keep pause support in that case) (plus, I ended doing other things, with bugfixes being a priority of mine)... and I've already done a massive amount of optimizations just to get the other stuff in there. In particular, the hot patch system (which was designed both to address playback inconsistency between older Addmusics and AddmusicK and also has a few other quirk fixes), one of the hot patches itself for enabling enchanted readahead (it now looks inside the $E6 and $E9 hex commands) and handling noise conflicts between music and SFX (yes, I did this so that they would play nice together... kind of... what really happens is that the SFX's noise mutes the music's noise while it is in use) took up a good chunk of ASM filesize just to implement. There's only so much I can do optimization-wise, and space is extremely tight: right now, I'm only a few bytes away from exceeding that threshold. Yes, I did optimize the SFX data, but I'm taking into account those that don't use the default SFX pack (or the default global instrument data).
I think I know what you're talking about... is it, for example, the echo in the SFX that normally is in place in Vanilla Super Mario World in the cave levels?
When I try to insert music into the game it breaks other music I have used values that arent being used by other songs and when i boot the game up the intro music is the yoshis island music and the world map is the castle complete theme
My AddmusicKFF fork has support for the special pulse wave and has some enhanced compatibility with AddmusicM, and I have a prerelease 1.0.9 build that has some compatibility bugfixes as well (though it does not have the special pulse wave: that's only in my fork for now).
Thank you very much! I'll gladly check out your AdmusicKFF fork when I have the time, sounds awesome! I always thought the S-SMP could benefit from some sort of wavetable channel.
Continuing off of this previous comment (I'll keep posting these update notices in the comment until the tool exits the prerelease state when it is moderated)...
In order to incorporate both of these without increasing the filesize to beyond the SPC ASM filesize of 1.0.8's (counting SFX ASM but not counting SFX and instrument data), I performed a series of optimizations. One of these optimizations resulted in the retirement of the GetSampleTableAddress routine in favor of having the SNES side use the standard loading routine to write directly to the DSP registers.
On 10/19/22, I replaced the Bowser Defeated and Bowser Zoom Out ports for better accuracy (including replicating a bug that caused pitch fades to be interrupted). In addition, for Bowser Defeated, it had a 39th rest and an 18th rest, both resulting in fractional tick length warnings. Bowser Zoom In was also analyzed and was verified good by me (it was also a single phrase song, meaning it was not subject to the bug that occurred between phrases).
Then Green notified me of a playback inconsistency with an Addmusic405 song. I identified the cause as a rest being forcibly keyed off upon being encountered, whereas in Addmusic405 it is only keyed off if encountered during readahead. Thus, on 12/13/22, I created a new hot patch (which required me to add a second byte to process, made easier by my variable byte implementation, and also required me to streamline things to avoid adding too many bytes for each subsequent byte I need to add on) and a new preset (vanilla SMW). This also comes with multiple optimizations in order to avoid increasing the ASM filesize past AddmusicK 1.0.8. Please note that this fix only works with songs that use #am4 (I've noted that post-remoderation songs use #amk 1, which makes the fix non-functional on those), as the hot patch I created addresses a playback inconsistency between AddmusicK and vanilla SMW, and this is the first one where the AddmusicK's playback did not match vanilla SMW's.
I corrected a regression on fetching bits for hot patch presets on 1/16/23 due to a missing shift operation.
I additionally updated the manual note alert on 1/16/23 because of a mislabeling case causing me to realize that the warning system had a problem because it wasn't specific enough: what I did was that I upgraded it so that it would separately identify duration/quantization and phrase end markers, the latter one being potentially fatal as it can prematurely terminate a song. I credit Green for notifying me of the songs that caused me to identify them as being mislabeled (and also identifying the cause of why they were fatal to AddmusicK).
Green then notified me of the wrong song being played on a new game (the default level music was playing instead of the Welcome song). I ended up catching conflicting patches, so I cut out two of them (writing !Title to $9724 and skipping Bowser fight music at $9734) due to both of them being overwritten later on, and I merged the other ones together on 1/16/23. Yes, this does affect tweaks.asm as a result of this merger, so be on the lookout for a changed patch location that covers $9723-$973F.
For those of you that downloaded this tool prior to 1/16/23, please redownload the tool, as it has been updated.
An off by one error? Hmm... not sure where that showed up...
At least the issue doesn't happen when using a custom song that's in slot #29 on any level while using the latest build of AddmusicK with Lunar Magic 3.3.1.
Actually, it's not AMK's fault. I'm using a build of AMKFF from 2021, also made by KungFuFurby, and I experience the same problem in LM 3.32. This will need to be reported to FuSoYa.
Oh. OK, that could be problematic... and I'm blind to Lunar Magic for the most part because of a platform incompatibility. The MSC files generated by AddmusicK do not seem to have been adversely affected.
The off-by-one error was reported on the Lunar Magic topic. Thus, it is a regression on Lunar Magic's end. I think it was fixed in Lunar Magic 3.33, going by the notes saying that a bugfix was done with the music dialog in the level editor.
Want to point out an issue while using the 9/27/22 version of AddmusicK 1.0.9 with Lunar Magic 3.3.2.
When using a custom song that's in slot #29 on any level, it plays the song that's in slot 28 instead. For now, use slot 2A and it will play the custom song that's in slot 29.
Continuing off of this previous comment (I'll keep posting these update notices in the comment until the tool exits the prerelease state when it is moderated)...
brickblock369 informed me that the basic use section of the readme was outdated, mostly because it assumed the .txt files were not yet adapted for AddmusicK. I moved and/or replaced the text in question on 9/17/22, as it's still somewhat relevant, but the first section should assume the adaptation has already happened, which also means updating the link... since that was also out of date.
I think the "basic uses" part of the AddMusicK's readme needs an update too, because it's still telling the user to put #am4 on top and telling people to use .bnks, when we don't even use #am4 or .bnks to this day.
The information in there is from a time when SMWCentral was still in the conversion phase from Addmusic405/AddmusicM to AddmusicK which required these kind of updates. I'll post a non-reply comment once I update the documentation. Also, I'll keep the .bnk portion for the Addmusic405 section: to me, .bnk files were part of Addmusic405 in the first place, and I fixed up support for that even though they were all replaced on the web site with a set of .brr files instead.
I'm not sure why this happens, but in the case of some songs using custom samples,
ARAM was not exceeded in version 1.0.8, but in version 1.0.9, ARAM exceeds more than 0x1000 bytes. example: https://www.smwcentral.net/?p=section&a=details&id=11018
It doesn't seem to be happening on my end with UserDefines.asm in its default state on a freshly downloaded copy with your example case. Here are my thoughts on what's going on...
All of my testing was done with asar 1.81, which is the version I distributed here.
I used UserDefines.asm in its default state. Changing them from their default settings can affect the filesize of your SPC-side build, which in turn can affect which songs you can use.
The .asm files were changed quite a bit for AddmusicK 1.0.9, particularly for optimization reasons, and also for a bunch of bugfixes and a few new features here and there.
The SFX data was also changed for cleanup, at least one correction, and optimization reasons.
Although I did not modify #optimized, I did modify #default to get rid of BRR blocks that were redundant.
The global songs were also optimized, along with a correction here and there.
Simply trying to implant the files from 1.0.8 into 1.0.9 may not work too well. I'm not sure what's causing the memory overflows either, but there's plenty of factors that could have come into play that could have caused that.
My AddmusicKFF fork has support for the special pulse wave and has some enhanced compatibility with AddmusicM, and I have a prerelease 1.0.9 build that has some compatibility bugfixes as well (though it does not have the special pulse wave: that's only in my fork for now).
This tool is subject to being revised due to being newly submitted as of 7/18/22, so consider this a prerelease build for now.
Plus, I got a bug report from Selicre involving patching AddmusicK on first, then UberASM. It is a confirmed regression, because it causes a crash on boot, and was fixed on 7/20/22.
A second regression was reported by Anorakun regarding the $F3 command doing bad pitch base fractional multiplier writes due to an accidental corruption from an incorrect label location for the hot patch VCMD. That was fixed on 7/27/22.
I also added a new option directive on 8/14/22, #option amk109hotpatch to streamline the new hot patches that were created for this version, since they are optional (mostly because #amk 4 itself is not intended to be too big of a leap forward too soon by accidentally breaking songs that try to immediately jump to the latest version for those that decide to do so).
For those of you that downloaded this tool prior to 8/14/22, please redownload the tool, as it has been updated.
It's fine on a vanilla ROM but I made a patch to change the music at 99 seconds and it crashes when Mario dies ik it's not the patch because it works fine with vanilla music how do you fix it?
Your patch probably hijacks the same spot as one of AddmusicK's edits. You can check in the hijack map.
It's fine on a vanilla ROM but I made a patch to change the music at 99 seconds and it crashes when Mario dies ik it's not the patch because it works fine with vanilla music how do you fix it?
Sorry about posting too many comments, but I wanted to report that if you use #am4 and the #pad command, you cannot pad the song out by more than 2 digits, everything after the first 2 digits of the value give an error in AMK and only the 2 first digits will be read. Just wanted to report that issue!
I need help! I am new to this page and I am having problems With AddmusicK, when I try to use my Rom to insert Custom Music, it tell me the following:
ERROR: Couldn't not find "Mainloop_pos "
Press "ENTER" to continue...
I have no idea what is happening, I have checked and verified that everything is fine, and everything is fine and is keep coming out.
I have used the AMKGUI, but it did not insert anything... so I concluded that some AddmusicK files had been damaged... I downloaded otros again, I did everything like this in the README but I still gen this error, any help would be very appreciable, and I really don't know what to do With this problem...
Just wanted to post a bug I found, occasionally when you're in a situation that has heavy sfx usage such as grinders, saws, some custom sprites (I'm assuming they are all using $1DF9 sounds) and pause the game then unpause, the game will be muted. You can correct this in-game by pausing/unpausing again or dying/exiting the level. Small issue but I just wanted to make note of this!
I noticed that the UberASM code provided by AMK for enabling echo on SFX doesn't fully work; if a global songs plays and ends afterwards, then the echo may not affect SFX at all or after a few seconds in the song in the level with echo. Try this instead.
I used this tool to insert a sampled track into my ROM to test it out and it worked, though I realized all SFX didn't sound as great anymore.
Afterwards I noticed when Castle 1 is completed, during the castle destruction scene the game totally freezes for about 1 sec when the castle is halfway destroyed.
I played on to Castle 2 that has a slightly different destruction scene, yet the exact same 1 second freeze occurs at the exact destruction point of the castle, when it is halfway going down into the ground.(I didn't test any others)
It appears that it is linked to me using this tool so I reverted back until I learn more of why this is happening.
I don't need the castle cutscenes in my ROM but I am concerned in case there are other problems as well I don't know about that may affect the ROM down the line.
Does anyone know about this issue?
This is due to the fact that it takes the SNES (or emulator) some time to locate and load up the correct set of samples to use, especially if the sampled port you've inserted omits some of the vanilla samples. Or, if you mean that you're transitioning between two different sampled songs, it will definitely take some time to switch samples. This can also be observed at the end of levels in Casio Mario World, where Mario freezes for a second before grabbing the goal tape (that's when the right samples are loaded for the ending theme). It's not really an issue that's feasible to fix.
I used this tool to insert a sampled track into my ROM to test it out and it worked, though I realized all SFX didn't sound as great anymore.
Afterwards I noticed when Castle 1 is completed, during the castle destruction scene the game totally freezes for about 1 sec when the castle is halfway destroyed.
I played on to Castle 2 that has a slightly different destruction scene, yet the exact same 1 second freeze occurs at the exact destruction point of the castle, when it is halfway going down into the ground.(I didn't test any others)
It appears that it is linked to me using this tool so I reverted back until I learn more of why this is happening.
I don't need the castle cutscenes in my ROM but I am concerned in case there are other problems as well I don't know about that may affect the ROM down the line.
Does anyone know about this issue?
Hello there! First of all, awesome job at updating this to fix the slowdown for songs that have high tempos, it's been great to have more songs that do not suffer from this issue anymore!
I did want to report a bug. Some of the songs on the site that are usually used as global songs have echo in them.
What happens for example is if you're playing a level that has music that has no echo, or has a lower echo buffer than the global song with said echo. Once the global song starts playing, there is a small pause when the global song plays which is sometimes accompanied with staticky/crackling noises.
Either that or if you are using a custom starman song that has echo in it, once you lose your star power and the regular level's music plays, the music will either majorly glitch or the music engine will crash, therefore causing the game to crash there or by the next time you exit a level. This doesn't happen when the level' echo buffer matches the same value as the one's with the global song.
When I press start+select quickly (almost at the same time), to leave the level, the song still plays.
A good idea is to use the noise clock mute flag to pause the music instead of backup the echo volume and lowering the master volume.
You're right, that way is much better so I'll update the sfx (although it doesn't solve the issue when doing start+select quickly).
Edit: found out how to fix it, so thanks for reporting the issue.
When I press start+select quickly (almost at the same time), to leave the level, the song still plays.
A good idea is to use the noise clock mute flag to pause the music instead of backup the echo volume and lowering the master volume.
After some people requested it, and in agreement with the music mods we decided to integrate the pause fix in AMK so porters don't have to worry about pausing breaking their songs. Also the fix has been slightly improved from the standalone patch version (the song won't play for a brief second when exiting a level with start+select).
This version fixes the bug where the music would break sometimes when pausing and unpausing **IMPORTANT NOTE: This adds a new freeRAM !FREERAM_FIX_PAUSE used in the asm\SNES\patch.asm which may need to be checked for conflicts**
I've recompiled Addmusick with a handful of reqested fixes. Tested on SA-1, Snes9X, BSNES, Mesen and console verified on NT+SD2SNES. Do note the new FreeRAM usage.
For Linux/Mac users: The included version of Asar is broken in the latest version of Wine. To fix this, drag in a newer Asar.exe, and then you'll need to change the deprecated read(...) to read1(...) in line 56 of asm/SNES/AMUndo.asm.
As someone who often finds myself stressing amk when making wild ports, this is a great improvement! I've noticed slowdown being reduced considerably or completely eliminated on all of my busy ports. Once these changes are approved I will be able to finish and release many of said ports. Yay!
Why does adding music have to be so god damned difficult for noobies like me? Even when following The read.me provided, it seems that wether or not the song will work is random sice every text file for every song is different. I think it's mostly because the songs are not sampled? that or there is'nt a sign of any bnk file anywhere in any of the maybe 30 songs I have downloaded today. help
I'm certain about issues specific to newer versions which weren't even there in the beginning. Stuttering and original music/sound effect oddities still haven't been fixed. The latter is apparently nitpicky.
Pretty sure lag issues with boss fights and castle cutscenes are a thing since earliest versions (if not since the very first), not specific of this one. Either way, annoying as fuck.
AMK is more and more broken once a new update gets submitted. Not only does it change pitch of certain sound effects like jump or rope/chainsaw/grinder, but also makes the game stutter during a boss fight or castle cutscenes and play faster music after beating a level where the hurry up alarm had played. But still, it gets 5 stars because there's no better SMW music insertion tool yet.
For some reason, whenever I add a song or change the slot of an existing song using this, it causes the game to freeze when Mario dies or touches the goal. Can this somehow be explained?
I just used this to replace a song and now for some reason all my music is like... set to the wrong values? Seems offset but I'm unsure how. My level itself now has no music at all. Frustrating because I do not understand how to revert the music and just start over.
i uploaded a song, it said it was fine, i inserted it into my level,
i run the rom in an emulator, and when i go into the level. the musics gone!, how do i fix it?
Olá, eu gostaria de saber como fazer as músicas do jogo ficarem com o efeito do level de caverna, com eco?
Depois que eu coloquei as músicas usando esse programa, os leveis que deveriam ter eco do mario pulando e os efeitos dos sprits também não tem mais eco, os leveis que deveriam ter um efeito de caverna não tem mais.
Como eu faço para resolver esse poblema?
Como colocar o level com efeito de caverna?
Hello, I would like to know how to make the songs in the game have the effect of cave level, with echo?
After I put the songs using this program, the levels that should have echo of Mario jumping and the effects of the sprits also no longer echo, the levels that should have a cave effect has no more.
How do I solve this problem?
How to put the level with cave effect?
To the people who developed this tool, thank you so much!
Hopefully I can get this to work with Lunar Magic 2.53...
(...I never got AddmusicM or K to work, as it always just adds blank music. But, I've always been
the absolute worst at troubleshooting! Ha ha ha. And yes, I'll try first, then look in the forums.
If all else fails, I'll just look it up on YT. I'd do that first but I'm to proud and stubb-
wait wait what am I blathering about on here lol who tf cares about this, this is the download section,
not the noob section, sorry guys )
Ironically, that Readme actually seems to be even less up-to-date than the one included in the ZIP file. The one in the ZIP file actually includes a full changelog up to version 1.0.6. The one on the GitHub only goes up to 1.0.2. Can't really explain that. I assume HertzDevil might have updated the Readme locally and not commited their changes to GitHub?
If you use the tutorial for this tool and you keep getting stuck with it not being able to find a sample. The fix is to put the ENTIRE FOLDER OF YOUR NEW MUSIC'S SAMPLES into addmusicK's "samples" folder, and NOT to put the CONTENTS of your new music's sample's folder into addmusicKs "samples" folder.
That being said, this tool is fantastic!
Grinders SFX was move to the same place they play other important SFX's (and even correlated sfx, such as the spinjump contact sfx), causing them to get drowned out completely or only play for a couple frames.
hey, i have been encountering a problem lately.
every time i add music, it adds an ~ file type and obviously having a .smc~ is causing problems. there any known fixes?
@Vamperumbra thy majestic issue happens due to your magnificentiam collection of samples textona "Addmusic_sample groups.txt" not having the required sample group recalled in the port .txt. Make sure that your "Addmusic_sample groups.txt" includes the samples group recalled by your in-use musically art "port".
If thou still have final technical boss issues, feel obrigado "free" to PM me with the pit port you're having problems with. Excellentiam magister Wakana from the peaceful lands called "e" will try help you out gladly!
Been buggishly problematic exasperation of desultory manifestation. I've never encountered this aberrant computerized contradistinction with this magnificum musicalis program before. I implement euphonious songs desired into canticum enumeration, " add music listing" thus emanates vacuus(blank) box.
Daily reminder that AMK and cloud-syncing software (i.e. Dropbox) do not mesh well together. If you're having bugs (other than music order being out of whack), remove your AMK folder from Dropbox and try again. Same thing with SpriteTool, although that one often corrupts sprites rather than plain crashing.
This tool still needs an option to keep using #4 for the jump SFX, though.
For some unknown reason the music files "Here We Go" and "Welcome" were switched in addmusic_list for me. So if you have any issues with the music being switched around just go to the addmusic_list and switch them with the appropriate song. Also this should be fixed in the next update Kipernal! Thank You.
Late, but in case anyone else needs to know, this fixes a bug where using code insertion in 1DFC sound effects would cause an spc crash. Now it doesn't do that anymore.
Comments (219)
What I did was remove a few hardcoded NOPs and replaced them with checking for a zero in $2141, since I determined the behavior there is more consistent.
Tested on Windows 10 22H2 (OS Build 19045.4842)
For those of you that downloaded this tool prior to 7/2/24, please redownload the tool, as it has been updated.
Tested on Windows 10 22H2 (OS Build 19045.4046)
A couple of things running through my mind with regards to the crash:
Echo buffer exceeds total space in ARAM by 0x2900 bytes.
This is what it says when I add a song from Dominator with AddmusicK
https://www.smwcentral.net/?p=files&sort=date&order=desc
Anas and Mellonpizza both reported Yoshi-related crashes with this release candidate, and the latter forwarded a fix to Kevin who then forwarded it to me. It turns out that I accidentally introduced a regression due to a stack fault from forgetting to remove a PLB opcode while fixing a patch conflict with the Yoshi Drum hijack, so that was repaired on 12/12/23.
I also implemented a suggestion by Fernap to use padbyte $EA (a bunch of NOPs for padding) and pad on 12/12/23. I did this on the Yoshi Drums hijack and the Start/Select SFX hijack so that they no longer had to be counted manually. I also added warnpc calls prior to the padding to avoid overwriting anything else, since the hijacks are now assumed to be variable size.
On 12/19/23, I added warnings for using non-standard FIR coefficient table IDs with using the $F1 hex command after having been notified of it from Argonfunk (I had accidentally broken them in 1.0.9 without knowing because of ASM modifications). I also, for the non-Windows users, updated make clean in the makefile because I forgot to make it remove AM4Batch and AMMBatch.
For those of you that downloaded this tool prior to 12/19/23, please redownload the tool, as it has been updated.
I don't know why makes impossible use the patch with this version of Addmusick
For those of you that downloaded this tool prior to 10/29/23, please redownload the tool, as it has been updated.
unironically the best change, can't believe it took years to build the fix in
For GUI requires .NET 4
OS: Windows XP SP3 or Windows 7 x86
I have updated the release candidate on 8/28/23 after notifying the moderators about a bug report I had received from Ahrion.
Specifically, the bug in question involved an accidentally duplicated sample causing memory overflows whereas it wasn't doing do in 1.0.8. It turns out that the important markered samples were correctly caught as duplicates, but were then loaded twice. A longer term fix will involve both the C++ and SNES sides because I want to use duplicate pointers in these cases, but for now, I have made it so that duplicate samples are replaced with an empty sample slot like 1.0.8. This also affects sample groups, though, because 1.0.8 still ran into the duplicate situation in those cases.
I also took the time to incorporate a fix to werneta's bug report where the outermost folder was missing from the .zip distribution by updating the zip script, also on 8/28/23.
I updated the release candidate on request from Atari2.0 on 9/9/23. Specifically, I added the .dll file from Asar in order to prevent warnings (and also because it speeds up execution), and I also added some documentation for AMMBatch and AM4Batch. Plus, I added a minor bug fix when handling hardcoded 1DFA SFX as they could get prematurely terminated, resulting in a stuck pitch bend on the jump SFX, if multiple SFX instances get triggered in a row that aren't on the same channel as hardcoded 1DFA SFX after getting a report about it from Anas.
For those of you that downloaded this tool prior to 9/9/23, please redownload the tool, as it has been updated.
The tool was initially been claimed for moderation on 7/25/23 (it was temporarily unclaimed and reclaimed on 8/28/23 once I notified them, and again on 9/9/23), so unless there's another update during the moderation process, then 9/9/23's update is the last release candidate prior to finalization.
...and this version was finalized on 10/5/23.
Either way, you're a little late on that one, since it was already claimed for moderation (and if I were to do that, I'd also have to change 1.0.7's method because there's a different problem that causes long notes/legato to not come back).
It only happens in-game, so maybe SNES-side?
Either way, you're a little late on that one, since it was already claimed for moderation (and if I were to do that, I'd also have to change 1.0.7's method because there's a different problem that causes long notes/legato to not come back).
I only noticed because I have some software that automatically extracts the zip files and used that structure. Not a problem for me to adapt to it, but thought I'd report in case it was not on purpose.
Your issue has been acknowledged.
Kevin provided me with a few contributions that I merged in on 6/20/23. They are:
Note that as part of the contributions, the user defines have been modified to not use the # character at all (this also affects the ASM files, since they do that instead) in order to simplify the logic for the comparisons.
On 7/7/23, I merged in a contribution from lx5 that allows a local song to replay after playing a global song without reloading while playing a level. The contribution was modified so that the loading still occurs if the song ID requested doesn't actually match the local song ID and it is not a global song, mostly because I felt the original code was too rigid in that regard... but it's in.
On 7/11/23, I received a contribution from Atari2.0 that updated all of the DLL bindings to the version that asar 1.81 was using. It also fixes a bug where the detection was being used too late, as asar had already been called beforehand.
For those of you that downloaded this tool prior to 7/11/23, please redownload the tool, as it has been updated.
The tool has been claimed for moderation on 7/25/23, so unless there's an update during the moderation process, then 7/11/23's update is the last release candidate prior to finalization.
What antivirus software would you recommend?
Ahrion reported a bug where multiple low note warnings were still causing an error when they weren't supposed to. I corrected this bug on 6/3/23.
For those of you that downloaded this tool prior to 6/3/23, please redownload the tool, as it has been updated.
Not the site's SPC player, the option in the AddMusicK Tool itself as shown here. When you play a song, there seems to be no way to play it or pause it. Am I missing something?
On 2/14/23, I updated the sound driver both for a minor optimization and a regression fix to the hot patch system where the second byte was failing to be properly processed due to not being stored to a scratch RAM location.
Then I perform another optimization and repaired a bug reported by Kevin with regards to the surround settings not being properly cleared out during song initialization when no panning commands are used in the song on 2/17/23.
Then I got a report from LadiesMan217 about a crash that occurred while switching songs. I determined that the cause was the loader going so quickly that the echo buffer didn't have time to finish its writing cycle, and I determined that I had accidentally introduced a regression against this situation by using the DSP register itself instead of !EchoDelay. Thus, I repaired this problem on 2/23/23.
For those of you that downloaded this tool prior to 2/23/23, please redownload the tool, as it has been updated.
They're not officially included at the moment, but Anas had requested a couple of options for replicating a couple of vanilla SMW cases related to the echo (which I did through extra $1DFA commands: namely, matching the music's echo channels to the SFX's and allowing music echo channels to carry between global songs and the currently playing local song). They're partially relevant to your request, and they both happen in the cave as well in vanilla Super Mario World.
In both cases, they're not included by default to avoid increasing the SPC700 ASM filesize to beyond 1.0.8's (I also count SFX ASM, but not the SFX data: this is because the pause SFX ASM got relocated outside the SFX instance because I added a !noSFX user define and wanted to keep pause support in that case) (plus, I ended doing other things, with bugfixes being a priority of mine)... and I've already done a massive amount of optimizations just to get the other stuff in there. In particular, the hot patch system (which was designed both to address playback inconsistency between older Addmusics and AddmusicK and also has a few other quirk fixes), one of the hot patches itself for enabling enchanted readahead (it now looks inside the $E6 and $E9 hex commands) and handling noise conflicts between music and SFX (yes, I did this so that they would play nice together... kind of... what really happens is that the SFX's noise mutes the music's noise while it is in use) took up a good chunk of ASM filesize just to implement. There's only so much I can do optimization-wise, and space is extremely tight: right now, I'm only a few bytes away from exceeding that threshold. Yes, I did optimize the SFX data, but I'm taking into account those that don't use the default SFX pack (or the default global instrument data).
They're not officially included at the moment, but Anas had requested a couple of options for replicating a couple of vanilla SMW cases related to the echo (which I did through extra $1DFA commands: namely, matching the music's echo channels to the SFX's and allowing music echo channels to carry between global songs and the currently playing local song). They're partially relevant to your request, and they both happen in the cave as well in vanilla Super Mario World.
In both cases, they're not included by default to avoid increasing the SPC700 ASM filesize to beyond 1.0.8's (I also count SFX ASM, but not the SFX data: this is because the pause SFX ASM got relocated outside the SFX instance because I added a !noSFX user define and wanted to keep pause support in that case) (plus, I ended doing other things, with bugfixes being a priority of mine)... and I've already done a massive amount of optimizations just to get the other stuff in there. In particular, the hot patch system (which was designed both to address playback inconsistency between older Addmusics and AddmusicK and also has a few other quirk fixes), one of the hot patches itself for enabling enchanted readahead (it now looks inside the $E6 and $E9 hex commands) and handling noise conflicts between music and SFX (yes, I did this so that they would play nice together... kind of... what really happens is that the SFX's noise mutes the music's noise while it is in use) took up a good chunk of ASM filesize just to implement. There's only so much I can do optimization-wise, and space is extremely tight: right now, I'm only a few bytes away from exceeding that threshold. Yes, I did optimize the SFX data, but I'm taking into account those that don't use the default SFX pack (or the default global instrument data).
My AddmusicKFF fork has support for the special pulse wave and has some enhanced compatibility with AddmusicM, and I have a prerelease 1.0.9 build that has some compatibility bugfixes as well (though it does not have the special pulse wave: that's only in my fork for now).
Thank you very much! I'll gladly check out your AdmusicKFF fork when I have the time, sounds awesome! I always thought the S-SMP could benefit from some sort of wavetable channel.
On 9/29/22, I did two updates to the SPC side: I updated the remote gain restoration in order for it to play nice with the SFX, and I added a bugfix regarding playing a global song while a pan fade was running without initializing the panning.
In order to incorporate both of these without increasing the filesize to beyond the SPC ASM filesize of 1.0.8's (counting SFX ASM but not counting SFX and instrument data), I performed a series of optimizations. One of these optimizations resulted in the retirement of the GetSampleTableAddress routine in favor of having the SNES side use the standard loading routine to write directly to the DSP registers.
On 10/19/22, I replaced the Bowser Defeated and Bowser Zoom Out ports for better accuracy (including replicating a bug that caused pitch fades to be interrupted). In addition, for Bowser Defeated, it had a 39th rest and an 18th rest, both resulting in fractional tick length warnings. Bowser Zoom In was also analyzed and was verified good by me (it was also a single phrase song, meaning it was not subject to the bug that occurred between phrases).
On 10/27/22, Masterlink notified me of one of the older AMK songs triggering the too low pitch error, even though it worked on 1.0.8. Upon investigation, I determined that it output a note byte because of a loophole involving quantization being put before it, so it is now a bug exploitation warning for #amk 3 and earlier for 1.0.9, since it's not guaranteed to be beneficial in the first place.
Then Green notified me of a playback inconsistency with an Addmusic405 song. I identified the cause as a rest being forcibly keyed off upon being encountered, whereas in Addmusic405 it is only keyed off if encountered during readahead. Thus, on 12/13/22, I created a new hot patch (which required me to add a second byte to process, made easier by my variable byte implementation, and also required me to streamline things to avoid adding too many bytes for each subsequent byte I need to add on) and a new preset (vanilla SMW). This also comes with multiple optimizations in order to avoid increasing the ASM filesize past AddmusicK 1.0.8. Please note that this fix only works with songs that use #am4 (I've noted that post-remoderation songs use #amk 1, which makes the fix non-functional on those), as the hot patch I created addresses a playback inconsistency between AddmusicK and vanilla SMW, and this is the first one where the AddmusicK's playback did not match vanilla SMW's.
I corrected a regression on fetching bits for hot patch presets on 1/16/23 due to a missing shift operation.
I additionally updated the manual note alert on 1/16/23 because of a mislabeling case causing me to realize that the warning system had a problem because it wasn't specific enough: what I did was that I upgraded it so that it would separately identify duration/quantization and phrase end markers, the latter one being potentially fatal as it can prematurely terminate a song. I credit Green for notifying me of the songs that caused me to identify them as being mislabeled (and also identifying the cause of why they were fatal to AddmusicK).
Green then notified me of the wrong song being played on a new game (the default level music was playing instead of the Welcome song). I ended up catching conflicting patches, so I cut out two of them (writing !Title to $9724 and skipping Bowser fight music at $9734) due to both of them being overwritten later on, and I merged the other ones together on 1/16/23. Yes, this does affect tweaks.asm as a result of this merger, so be on the lookout for a changed patch location that covers $9723-$973F.
For those of you that downloaded this tool prior to 1/16/23, please redownload the tool, as it has been updated.
At least the issue doesn't happen when using a custom song that's in slot #29 on any level while using the latest build of AddmusicK with Lunar Magic 3.3.1.
The off-by-one error was reported on the Lunar Magic topic. Thus, it is a regression on Lunar Magic's end. I think it was fixed in Lunar Magic 3.33, going by the notes saying that a bugfix was done with the music dialog in the level editor.
When using a custom song that's in slot #29 on any level, it plays the song that's in slot 28 instead. For now, use slot 2A and it will play the custom song that's in slot 29.
brickblock369 informed me that the basic use section of the readme was outdated, mostly because it assumed the .txt files were not yet adapted for AddmusicK. I moved and/or replaced the text in question on 9/17/22, as it's still somewhat relevant, but the first section should assume the adaptation has already happened, which also means updating the link... since that was also out of date.
Then LadiesMan217 notified me of a bug where if the noise SFX frequency was not being checked for conflicts, then any SFX instance terminating would set it back to the music noise's frequency, even if there were noise SFX still playing. I fixed that on 9/27/22.
For those of you that downloaded this tool prior to 9/27/22, please redownload the tool, as it has been updated.
The information in there is from a time when SMWCentral was still in the conversion phase from Addmusic405/AddmusicM to AddmusicK which required these kind of updates. I'll post a non-reply comment once I update the documentation. Also, I'll keep the .bnk portion for the Addmusic405 section: to me, .bnk files were part of Addmusic405 in the first place, and I fixed up support for that even though they were all replaced on the web site with a set of .brr files instead.
ARAM was not exceeded in version 1.0.8, but in version 1.0.9, ARAM exceeds more than 0x1000 bytes. example: https://www.smwcentral.net/?p=section&a=details&id=11018
Simply trying to implant the files from 1.0.8 into 1.0.9 may not work too well. I'm not sure what's causing the memory overflows either, but there's plenty of factors that could have come into play that could have caused that.
My AddmusicKFF fork has support for the special pulse wave and has some enhanced compatibility with AddmusicM, and I have a prerelease 1.0.9 build that has some compatibility bugfixes as well (though it does not have the special pulse wave: that's only in my fork for now).
Plus, I got a bug report from Selicre involving patching AddmusicK on first, then UberASM. It is a confirmed regression, because it causes a crash on boot, and was fixed on 7/20/22.
A second regression was reported by Anorakun regarding the $F3 command doing bad pitch base fractional multiplier writes due to an accidental corruption from an incorrect label location for the hot patch VCMD. That was fixed on 7/27/22.
I later got contacted by IronFoxGaming about an uncaught C++ exception relating to some sort of malformed path while looking for samples after locating a sample relative to the music. This appears to also be a 1.0.8 bug. I ended up determining that there was code present that was supposed to make putting samples in the music folder legal, but failed to function when a sample was located due to a fault in the way the path was constructed upon discovery of the sample, thus the bug was repaired on 8/14/22.
I also added a new option directive on 8/14/22, #option amk109hotpatch to streamline the new hot patches that were created for this version, since they are optional (mostly because #amk 4 itself is not intended to be too big of a leap forward too soon by accidentally breaking songs that try to immediately jump to the latest version for those that decide to do so).
For those of you that downloaded this tool prior to 8/14/22, please redownload the tool, as it has been updated.
It's supposed to be in the same folder as your ROM, yes. Paste all contents from inside the folder.
Would it work if I just put a shortcut to the app in the folder rather than all the files in the app's folder?
It's supposed to be in the same folder as your ROM, yes. Paste all contents from inside the folder.
It's probably on your end, not the tool's fault. Try another folder.
Your patch probably hijacks the same spot as one of AddmusicK's edits. You can check in the hijack map.
I never even finished 1 hack because I gave up too easily
ERROR: Couldn't not find "Mainloop_pos "
Press "ENTER" to continue...
I have no idea what is happening, I have checked and verified that everything is fine, and everything is fine and is keep coming out.
I have used the AMKGUI, but it did not insert anything... so I concluded that some AddmusicK files had been damaged... I downloaded otros again, I did everything like this in the README but I still gen this error, any help would be very appreciable, and I really don't know what to do With this problem...
Afterwards I noticed when Castle 1 is completed, during the castle destruction scene the game totally freezes for about 1 sec when the castle is halfway destroyed.
I played on to Castle 2 that has a slightly different destruction scene, yet the exact same 1 second freeze occurs at the exact destruction point of the castle, when it is halfway going down into the ground.(I didn't test any others)
It appears that it is linked to me using this tool so I reverted back until I learn more of why this is happening.
I don't need the castle cutscenes in my ROM but I am concerned in case there are other problems as well I don't know about that may affect the ROM down the line.
Does anyone know about this issue?
This is due to the fact that it takes the SNES (or emulator) some time to locate and load up the correct set of samples to use, especially if the sampled port you've inserted omits some of the vanilla samples. Or, if you mean that you're transitioning between two different sampled songs, it will definitely take some time to switch samples. This can also be observed at the end of levels in Casio Mario World, where Mario freezes for a second before grabbing the goal tape (that's when the right samples are loaded for the ending theme). It's not really an issue that's feasible to fix.
Afterwards I noticed when Castle 1 is completed, during the castle destruction scene the game totally freezes for about 1 sec when the castle is halfway destroyed.
I played on to Castle 2 that has a slightly different destruction scene, yet the exact same 1 second freeze occurs at the exact destruction point of the castle, when it is halfway going down into the ground.(I didn't test any others)
It appears that it is linked to me using this tool so I reverted back until I learn more of why this is happening.
I don't need the castle cutscenes in my ROM but I am concerned in case there are other problems as well I don't know about that may affect the ROM down the line.
Does anyone know about this issue?
There are 2 simple solutions to that issue.
1. Drag and drop your rom into the same folder as the executable you're trying to use and then type your rom name in the cmd window.
2. Drag and drop your rom into the cmd window. (Much simpler imo)
I did want to report a bug. Some of the songs on the site that are usually used as global songs have echo in them.
What happens for example is if you're playing a level that has music that has no echo, or has a lower echo buffer than the global song with said echo. Once the global song starts playing, there is a small pause when the global song plays which is sometimes accompanied with staticky/crackling noises.
Either that or if you are using a custom starman song that has echo in it, once you lose your star power and the regular level's music plays, the music will either majorly glitch or the music engine will crash, therefore causing the game to crash there or by the next time you exit a level. This doesn't happen when the level' echo buffer matches the same value as the one's with the global song.
Just wanted to inform you about this issue!
A good idea is to use the noise clock mute flag to pause the music instead of backup the echo volume and lowering the master volume.
You're right, that way is much better so I'll update the sfx (although it doesn't solve the issue when doing start+select quickly).
Edit: found out how to fix it, so thanks for reporting the issue.
A good idea is to use the noise clock mute flag to pause the music instead of backup the echo volume and lowering the master volume.
Now, that's perfect!
No, several programs are needed, PetiteMM, AddmusicK 1.0.7, Notepad and a MIDI are needed. For more information see the wakana tutorial.
is it the the main software or no
would help so much the problem with songs farting when time goes 099
due the use of vibrato
Edit: with farting i mean, it doesnt go fast when the tempo hike happen
I assume the solution is to just use the program via the command line?
(edit) Doing that only inserts the global songs and doesn't generate any SPCs, which I find exceedingly strange
i run the rom in an emulator, and when i go into the level. the musics gone!, how do i fix it?
Depois que eu coloquei as músicas usando esse programa, os leveis que deveriam ter eco do mario pulando e os efeitos dos sprits também não tem mais eco, os leveis que deveriam ter um efeito de caverna não tem mais.
Como eu faço para resolver esse poblema?
Como colocar o level com efeito de caverna?
Hello, I would like to know how to make the songs in the game have the effect of cave level, with echo?
After I put the songs using this program, the levels that should have echo of Mario jumping and the effects of the sprits also no longer echo, the levels that should have a cave effect has no more.
How do I solve this problem?
How to put the level with cave effect?
Hopefully I can get this to work with Lunar Magic 2.53...
(...I never got AddmusicM or K to work, as it always just adds blank music. But, I've always been
the absolute worst at troubleshooting! Ha ha ha. And yes, I'll try first, then look in the forums.
If all else fails, I'll just look it up on YT. I'd do that first but I'm to proud and stubb-
wait wait what am I blathering about on here lol who tf cares about this, this is the download section,
not the noob section, sorry guys
https://www.smwcentral.net/?p=viewthread&t=94957
I have a fork at https://github.com/jimbo1qaz/AddmusicK/tree/readme/doc with warnings about various bugs I encountered.
That being said, this tool is fantastic!
they are like cut off
every time i add music, it adds an ~ file type and obviously having a .smc~ is causing problems. there any known fixes?
If thou still have final technical boss issues, feel obrigado "free" to PM me with the pit port you're having problems with. Excellentiam magister Wakana from the peaceful lands called "e" will try help you out gladly!
Said options does exists, but you can't toggle it via AMKGUI or Addmusick, you need to change a define inside of asm/SNES/tweaks.asm
Just change it to !false and you'll be fine.
This tool still needs an option to keep using #4 for the jump SFX, though.
AMK2 when?
http://i.imgur.com/lUHetqG.png
I really, REALLY don't think this should be happening.
When you insert the music it inserts the sound effects as music.
error: 'class Music' has no member named 'spaceInfo'
error: 'class Music' has no member named 'pathlessSongName'