PSA: Since the OP's links are broken, I added a note to it with a link to a reply with a still-working link.

Recently I found that someone has made a SoundFont of the ESFM synthesizer and it sounds quite similar although not exactly the same.

Has anyone tried this ESFM SoundFont? https://dropbox.com/s/s7muqffctbty4ju/ESFM.SF2?dl=1

It's a 703 MB file though made by some Japanese guy, I use it in Virtual MIDI Synth and then load the SF2 file in it. Give it a try, the MIDI sounds really nostalgic with it, quite similar to the original hardware ESFM.

autoexecdotbat wrote on 2016-02-26, 16:04:

realnc wrote on 2016-02-17, 05:36:

You know what would be funny? A soundfont where all patches are someone trying to mimic instruments with their mouth. We could call it the Bobby McFerrin SoundFont, and it would sound like this:

https://www.youtube.com/watch?v=ih20Tx-fNJs

I agree with you. I'd love a soundfont like that, too!

Holy spaghetti, somebody actually made such a soundfont 😁 "Acapella GM/GS" by Juicestain:

https://musical-artifacts.com/artifacts/712

"This is a soundfont where I sing everything. It is GM Compatible, so you can play it with MIDIs."

Does anyone know what the heck this is:

https://archive.org/details/sc-55-sf2

"This contains the General MIDI and drum kit presets of the Roland SC-55. Each preset was deepsampled"

🤔

My favorite SC-55 SoundFont is this.

It's the closest I can hear to my real SC-55MKII, especially when playing games like Heretic and DOOM.

Malik wrote on 2022-08-01, 23:15:

My favorite SC-55 SoundFont is this.

It's the closest I can hear to my real SC-55MKII, especially when playing games like Heretic and DOOM.

Yeah, I've used it too before I got the Sound Canvas VA to work in Linux. But I'm curious as to what is meant with "deepsampling". And it's several GBs. Is it machine learning related? A web search turns up nothing.

realnc wrote on 2022-08-02, 05:56:

Malik wrote on 2022-08-01, 23:15:

My favorite SC-55 SoundFont is this.

It's the closest I can hear to my real SC-55MKII, especially when playing games like Heretic and DOOM.

Yeah, I've used it too before I got the Sound Canvas VA to work in Linux. But I'm curious as to what is meant with "deepsampling". And it's several GBs. Is it machine learning related? A web search turns up nothing.

I checked it out.
It's just a standard set of SC-55 instruments, but with a lot of samples.
And when I say "a lot", I mean "A WHOLE BLOODY LOT".

The problem with an "XG soundfont" is that most of what makes XG special is the effects and filters not the sounds themselves. Without that pretty much any XG MIDI that does more than just use the base sounds with no effects or parameter modifications will sound wrong.

xpclient wrote on 2021-04-11, 11:59:

Recently I found that someone has made a SoundFont of the ESFM synthesizer and it sounds quite similar although not exactly the same.

Has anyone tried this ESFM SoundFont? https://dropbox.com/s/s7muqffctbty4ju/ESFM.SF2?dl=1

It's a 703 MB file though made by some Japanese guy, I use it in Virtual MIDI Synth and then load the SF2 file in it. Give it a try, the MIDI sounds really nostalgic with it, quite similar to the original hardware ESFM.

You might want to try this

https://stgiga.itch.io/jummboxsoundfont which is like an ESFM on steroids.

Also, via something called "Silicon SoundFonts" in the SoundFont spec (section 11), I was able to make that bank into an actual synth ROM. It is CC-BY-SA4(+GPL3) like the SoundFont itself.
https://sourceforge.net/projects/stgigasoundf … SF.DAT/download

I want the Dreamblaster to support Silicon SoundFonts (https://github.com/stgiga/siliconsfe/blob/mai … on-1.0-draft.md derived from Section 11 of the SoundFont standard), which is documented here: https://github.com/stgiga/siliconsfe/blob/mai … on-1.0-draft.md and here: https://github.com/stgiga/siliconsfe/tree/main

Basically, it is possible to make a ROMpler/soundchip out of an SF2. And so I did, and now we've beat ESFM at quality high-quality FM. AND it's libre!