09-05-2023, 04:40 PM
The thing with "old Win95 games" was way back when the toys created by a certain company from Japan were the most desireable...
But today not everyone cares for that General MIDI soundset. From MusicTech Magazine based in the United Kingdom, I picked up a c. 256MiB Soundfont from Pinknoise Studio from Hungary (I hope I'm correct) which had better sounds IMHO. However some of the drum sounds sucked. At one point, that magazine was providing high-quality multisamples from Pinknoise Studio.
https://pinknoisestudio.com/
(Very shabby when there's a different company which web address differs by one letter.)
This cross-platform thing, more than likely is Fluidsynth. I had lost faith in Qsynth a long time ago because it cannot even detect multiple parts calling on the same Soundfont, to prevent it from loading the same Soundfont multiple times and therefore wasting RAM. In the past I used the Fluidsynth Soundfont player as native machine in Psycle, and as DSSI plug-in for 32-bit Ubuntu Studio. It was a bit better than loading Soundfonts into instrument plug-ins created with Synthedit. For a long time, Synthedit didn't support layered Soundfont patches. After stepping up from 32-bit and VST2 to 64-bit and VST3, even loading a plug-in for anything is bugged. The GUI elements have to come with the modules and everything else the plug-in needs.
I have an instrument created with Synthedit in 2009 which has four "oscillators", each one bearing a Soundfont as large as 170MiB! If the same Soundfont is loaded into all four "oscillators", the system cannot detect it. So it was better to use the smallest Soundfont available.
I was supposed to check out Fluidsynth in Csound but the CSD format "scripting" could be a pain. Have to import a MIDI file, or trigger it with a MIDI keyboard, or pretend to trigger MIDI by creating a MIDI note event. This is after the Soundfont is loaded and the bank and program are selected for it. This is not familiar to me or most people.
I also have OpenMPT, free software music tracker for Windows. It supports SF2 and DLS. It was the only way I could make use of DLS extra soundsets I got from buying Sony (earlier Sonic Foundry) ACID Pro. Sadly it could load one sound at a time into a sample slot, so it's a pain using it with layered patches.
These days I use Soundfonts only in LMMS music creation application.
But today not everyone cares for that General MIDI soundset. From MusicTech Magazine based in the United Kingdom, I picked up a c. 256MiB Soundfont from Pinknoise Studio from Hungary (I hope I'm correct) which had better sounds IMHO. However some of the drum sounds sucked. At one point, that magazine was providing high-quality multisamples from Pinknoise Studio.
https://pinknoisestudio.com/
(Very shabby when there's a different company which web address differs by one letter.)
This cross-platform thing, more than likely is Fluidsynth. I had lost faith in Qsynth a long time ago because it cannot even detect multiple parts calling on the same Soundfont, to prevent it from loading the same Soundfont multiple times and therefore wasting RAM. In the past I used the Fluidsynth Soundfont player as native machine in Psycle, and as DSSI plug-in for 32-bit Ubuntu Studio. It was a bit better than loading Soundfonts into instrument plug-ins created with Synthedit. For a long time, Synthedit didn't support layered Soundfont patches. After stepping up from 32-bit and VST2 to 64-bit and VST3, even loading a plug-in for anything is bugged. The GUI elements have to come with the modules and everything else the plug-in needs.
I have an instrument created with Synthedit in 2009 which has four "oscillators", each one bearing a Soundfont as large as 170MiB! If the same Soundfont is loaded into all four "oscillators", the system cannot detect it. So it was better to use the smallest Soundfont available.
I was supposed to check out Fluidsynth in Csound but the CSD format "scripting" could be a pain. Have to import a MIDI file, or trigger it with a MIDI keyboard, or pretend to trigger MIDI by creating a MIDI note event. This is after the Soundfont is loaded and the bank and program are selected for it. This is not familiar to me or most people.
I also have OpenMPT, free software music tracker for Windows. It supports SF2 and DLS. It was the only way I could make use of DLS extra soundsets I got from buying Sony (earlier Sonic Foundry) ACID Pro. Sadly it could load one sound at a time into a sample slot, so it's a pain using it with layered patches.
These days I use Soundfonts only in LMMS music creation application.