This thing is so awesome! Six effects at once, audio/midi over usb. Connected to Renoise, played and recorded a sequence and sampled it right back to Renoise in just minutes.
For the FM-1 (FW version 15), are there any known limitations regarding long SysEx messages, such as loading a voice or an entire bank, over the TRS MIDI connection? I have a DX7 bank manager for the Atari ST that I wanted to try, however nothing happens on the FM-1 when I send a single voice or an entire bank.
Simple MIDI messages such as ānote onā or short SysEx messages such a single-parameter change do work from the same computer and software; so Iām positive that the whole MIDI setup including cable is okay.
I wonder if there is, e.g., a known length limitation for messages specifically over the TRS MIDI interface? (Loading banks from the PC with Dexed via USB MIDI of course works.)
Fwiw, I got it to work by modifying the bank editor to add delays every 32 bytes it sends over MIDI. Now, even long SysEx messages (loading voices or complete banks) work.
I can only assume that the FM-1 has a limited size UART HW buffer for TRS MIDI data and the firmware needs some time to empty it, and otherwise it overruns. Interesting that the original DX7 (running at 1980s CPU speeds) is fast enough to receive this MIDI data without delays in the transmission. ![]()
Does anyone knows a program, other then dexed to upload presets? I also would love a download option from the Fm-1z
Im using macos.
Ps. Still love the device. The amount of stuff you can create with FM is amazing. ![]()
I hope M-Vave will do more synths in the future.
Does anyone knows a program, other then dexed to upload presets?
@KrisG I came across https://fm1-editor.com/ (GitHub - benny-sparra/fm1-dx7-patch-importer: M-VAVE FM1 editor and librarian Ā· GitHub) which seems to do the job for me.
Given that its Browser-based, it should also not be OS-specific.
Wow! Many thanks ![]()
there is another web-based solution introduced earlier in this thread.
https://grandsummoner.github.io/fm-1-helper/FM-1%20Helper.html
(havenāt tried it)
Fwiw, I got it to work by modifying the bank editor to add delays every 32 bytes it sends over MIDI. Now, even long SysEx messages (loading voices or complete banks) work.
I can only assume that the FM-1 has a limited size UART HW buffer for TRS MIDI data and the firmware needs some time to empty it, and otherwise it overruns.
By the way, does anyone here have experience with M-VAVE support? Are they interested in feature requests? Maybe the TRS MIDI buffer handling could be improved in a future firmware update. It would not only benefit my arguably niche application, but also other use-cases sending data over the TRS MIDI port.
They really listen to support questions. Ive filed with other some features they added and/or solved. In short; just send them an email.
In the ā arguably unlikely case ā someone else is using the 3.5mm/TRS MIDI port of the FM-1 like I do, i.e., for sending SysEx data such as voices or voice banks, I wanted to document my findings. By trial and error, I found that the largest chunk of MIDI data that you can send at a time over the TRS MIDI port is exactly 60 bytes (pointing towards a 64-byte buffer somewhere in the firmware). After that, you need to pause for about 10 ms, before sending the next chunk. Without that delay, the FM-1 does not receive the SysEx command correctly. Sometimes, it interprets the partially received data as note commands and plays weird notes.
Whatās more, during my testing, when I sent too much data at once, the FM-1 sometimes locked up hard, with the LEDs turning off and no response to any keys. Luckily, it recovered by turning it off and on again. Imho, you should not be able to crash the FM-1 via MIDI.
Still no reply from M-VAVE support, btw.
EDIT: To make it clear, the USB-MIDI port (that most people probably use) is not affected.
Has anyone noticed that the algorithms 10 and 11 on the FM-1 are different from the standard DX7 algorithms?
When loading some of my carts from Dexed to the FM-1; I noticed some presets sounded completely off. Then I made the connection that it was any preset that was using 10/11. That is when I noticed the algorithms modifiers / loop are slightly different. Please be forgiving as Iām new to FM synthesis but to me something seems different.
Which difference exactly did you notice?
At least whatās shown on screen (FM-1, FW 15):
⦠to me looks identical to Yamahaās algorithm drawings:
When it sounds complete off to you, do you mean in reference to Dexed or to a real Yamaha DX7?
Thank you for providing reference photos. I guess the confusion stems from my lack of understanding FM. I can see that 10 looks the same, but what had me confused was 11 with the modifiers 5 and 6ās order the way it was connected to the carrier 4-1. I just spent about an hour learning more about FM Synthesis and I guess the left/ Right / A/B being in reverse to the DX7 does not matter. A couple of the presets in Dexed that I love, sound different when loaded into the FM-1. I tried to change the LFO and some of the envelopes in the FM-1 to make it sound comparable but sadly not. So I was thinking it was the layout of the algorithms that might have been different. Thanks for taking the time to reply.
At the risk of asking the obvious: How are you comparing the sound? The built-in small speaker of the FM-1 makes things sound different than, e.g., large speakers on your PC with Dexed.
Do you have a preset where this is particularly noticeable that you like to share?
PS: The layout in which algorithms are displayed, e.g., which is displayed left and which is displayed right, does not matter. What matters is the connections between the six operators.
Iām comparing the sound using my studio monitors, then running the FM-1 into my Antelope pre straight into Ableton ext instrument rack. I also A/Bāed using my DT 770 cans.
I created a test cart for you featuring two presets; the rest of the cart is blank INT. Str 1 and Str 2 are great examples of what Iām experiencing. Both sound great in Dexed, and then just off and not comparable when loaded into the FM-1.
Regarding FM algorithms I now understand how the connections to the operators work. Your visuals helped me recognize that more clearly.
Thank you again for taking the time to reply.
You might be up to something here.
Looking at the spectrograms, top one is my FM-1, bottom one is Dexed. You can see that the oscillators are beating against each other differently with your STR-1 patch. That beating typically comes from the ādetuneā parameter. So, I wonder if thatās somehow implemented differently in the FM-1?
You could reach out to their support, I guessā¦
Youāre a real one for taking the time to check this for me. Visually I can see exactly what your findings are.
On preset STR 1; I changed the freq coarse tune from 8 (original) stepped down to 4 on OP 3, in the FM-1 and lowered the out level. Seems to be better, but still something is weird. I tested it with the filter settings completely open , and still itās slightly off to my ears.
Realistically I donāt expect the FM-1 to sound flawless⦠Yet itās weird that some presets from Dexed sound perfect in the FM-1, while some like my example above just sound weird.
Thanks again for your help.
To change the beat frequency, you probably need to vary the ādetuneā parameter (called āTuneā in Dexed), not the coarse frequency ratio.
Realistically I donāt expect the FM-1 to sound flawless⦠Yet itās weird that some presets from Dexed sound perfect in the FM-1, while some like my example above just sound weird.
I mean M-VAVE afaik does not claim it sounds exactly like Dexed. But still weird, yes. If you decide to contact them, let me know what they reply.
Thatās a fair point indeed. On a positive note this is helping me learn FM synthesis a bit more. Iām still really pleased with the FM as is. Iām planning on using it with various Fx pedals, so with a better quality filter and some Eq maybe I can get things sounding better.
Iāll let you know if support gets back to me.
Thanks again for all your help, truly!
I saw this cool 3d printed cover posted at Reddit, link to original post below




