Quote from: raekuul on 2024-09-17 15:22:17

The current version seems to be ignoring libADLMIDI and libOPNMIDI if a soundfont path is set - regardless of what player I choose the console reports the following:

Code: [Select]

Opening track for playback: "D:\Synth\rips\1991 - Jimmy White's Whirlwind Snooker\01 - Snooker.mid"
foo_midi failed to initialize FluidSynth: Invalid FluidSynth directory: The system cannot find the path specified (0x00000003)
foo_midi uses SoundFont "D:\soundfonts\HiDef.sf2" with bank offset 0.

clearing the soundfont path allows libADLMIDI and libOPNMIDI to operate as normal, it's only when a soundfont path is set that they're getting overridden.

Thx. for reporting. The player override logic has been updated to support layered soundfonts but it looks like it's a bit too assertive.

I have tried using the Roland Sound Canvas VST and it seems like it doesnt work properly. The midis only play with the default boot settings of the VST and changing them via the "configure" button in preferences has no effect. Maybe also because I cannot access foobar anymore until the VST window is closed.

Not sure if I do something wrong or if its just Foobars VST support. The VST is generally not easy to use though and so far I could only make it fully work in FL Studio with mid files and loopmidi. Sadly FL Studio only supports mid and no deviations like hmp.

Try using the flavor option in foo_midi itself instead of the configuration in the VSTi, and set it to the unit you want to emulate, GS SC-55, GS SC-88Pro, etc. It responds to that in my experience.

Quote from: Slender on 2025-01-23 16:40:23

Try using the flavor option in foo_midi itself instead of the configuration in the VSTi, and set it to the unit you want to emulate, GS SC-55, GS SC-88Pro, etc. It responds to that in my experience.

That actually seems to have done it! I tried comparing some midi recordings between foobar and fl studio now and they look almost identical.

However, I am unsure about the Descent and Descent II exports. The thing is, I compared them to other peoples hardware SC-55 exports and they are different. The playback speed is actually different and I tracked it down to them being based on some hmp to midi conversions but I am not sure if these are accurate or faulty and dont mimic the sound of the real original hmp composition. I even noticed a few additional samples that do not play with the hmp file but now I have no idea if this is due to foo_midi not playing the hmp format properly or if the midi conversions made by other people are faulty.

However, I did rule out the VST plugin as the reason. I also played it with with libADLmidi and descent 2 bank and the speed was the same. So it is foo_midi who is playing the hmp file like that and perhaps its correct.

No idea if anyone can help me with this since this is really specific territory. I uploaded the renders for comparison together with both the original mid and hmp files. Would be interesting to know if the hmp playback with foo_midi is actually the accurate one and the fan conversion midis not but hard to tell for me.

The upload comes with the original hmp, a midi conversion sourced from mirsoft and mp3 renders of both done through foo_midi with the exact same settings played through the Roland Sound Canvas VST and SC-55 mapping. Notice the speed/pitch difference between both and also the midi has a additional sample at 1:07 but its not present with the hmp render at 1:03.

Download: https://www.mediafire.com/file/bqdaqmjbwh3pt1d/Descent_II_foo_midi_HMP_vs_MIDI.7z/file

Quote from: marcoxD95 on 2025-01-24 19:09:24

Quote from: Slender on 2025-01-23 16:40:23

Try using the flavor option in foo_midi itself instead of the configuration in the VSTi, and set it to the unit you want to emulate, GS SC-55, GS SC-88Pro, etc. It responds to that in my experience.

That actually seems to have done it! I tried comparing some midi recordings between foobar and fl studio now and they look almost identical.

However, I am unsure about the Descent and Descent II exports. The thing is, I compared them to other peoples hardware SC-55 exports and they are different. The playback speed is actually different and I tracked it down to them being based on some hmp to midi conversions but I am not sure if these are accurate or faulty and dont mimic the sound of the real original hmp composition. I even noticed a few additional samples that do not play with the hmp file but now I have no idea if this is due to foo_midi not playing the hmp format properly or if the midi conversions made by other people are faulty.

However, I did rule out the VST plugin as the reason. I also played it with with libADLmidi and descent 2 bank and the speed was the same. So it is foo_midi who is playing the hmp file like that and perhaps its correct.

No idea if anyone can help me with this since this is really specific territory. I uploaded the renders for comparison together with both the original mid and hmp files. Would be interesting to know if the hmp playback with foo_midi is actually the accurate one and the fan conversion midis not but hard to tell for me.

The upload comes with the original hmp, a midi conversion sourced from mirsoft and mp3 renders of both done through foo_midi with the exact same settings played through the Roland Sound Canvas VST and SC-55 mapping. Notice the speed/pitch difference between both and also the midi has a additional sample at 1:07 but its not present with the hmp render at 1:03.

Download: https://www.mediafire.com/file/bqdaqmjbwh3pt1d/Descent_II_foo_midi_HMP_vs_MIDI.7z/file

I'll have a look. The HMP conversion code is fairly simple. Maybe the conversion of the variable length values is of.

Just to be clear: is the HMP file the original?

Just to be clear: is the HMP file the original?

Some HMP dialects specify the playback speed. foo_midi uses 160 BPM unless the file contains metadata that specifies a different master tempo.

You can tweak the tempo on the Processing sub page in the Preferences dialog.

Automatic update from within foobar2000 does not work.

v2.16.1.0-alpha2, 2025-03-12

* Fixed: A pending SysEx message would get skipped when the next event used the running status.
* Fixed: More Multi Port MIDI files play correctly now.
  * The first MIDI Port message of a track is now added at the start of a track to make sure it occurs before any Program Change events.

Download only available from GitHub during testing.

Quote from: pqyt on 2025-03-12 17:09:24

v2.16.1.0-alpha2, 2025-03-12

* Fixed: A pending SysEx message would get skipped when the next event used the running status.
* Fixed: More Multi Port MIDI files play correctly now.
  * The first MIDI Port message of a track is now added at the start of a track to make sure it occurs before any Program Change events.

Download only available from GitHub during testing.

The two latest versions cannot open this file

It opens and plays just fine. Which error message do you get? Are there any messages in the fb2k console?

Quote

It opens and plays just fine. Which error message do you get? Are there any messages in the fb2k console?

It's really strange.
The mid file in the attachment I sent is named:
至少还有你 - pai31415926.mid
in my file directory.
When using this file name,
the new version of foo_midi plugin will report an error:
Opening track for playback: "D:\music\MIDI\至少还有你 - pai31415926.mid"
Unable to open item for playback (

If i change to another file name,
such as 01.mid used in my attachment,
it can be played normally.
This has never happened in the old version two versions ago.

This text is translated by Google.
Please forgive me if there are any unclear expressions.

Probably due to the Unicode file name containing Chinese/Japanese(?) glyphs even though nothing was changed in that area of the code.

Can you please attach the file that causes a problem? I'll see if I can replicate the issue.

the new version of foo_midi plugin will report an error:
Opening track for playback: "D:\music\MIDI\至少还有你 - pai31415926.mid"
Unable to open item for playback (

I hope pgyt uses foobar2000 I/O to open and read the files regardless of chosen midi player or other settings. Bypassing the player I/O prevents all fancy features like transparent archive support and network access from working.

v2.16.1.0-alpha4, 2025-03-15