preenfm Forum
PreenFM => preenfm2 and preenfm3 => Topic started by: Mmarsh on December 12, 2013, 01:54:45 AM
-
Newly built and fully working (now!) PreenFM2 playing a DX7 patch: TubBells. Note that I have the overclocked 0.9J software flashed.
https://soundcloud.com/mmarsh/preenfm2tubbells
-
Thanks for the demo.
How does it compares at the low notes mantaining whole sound?
How about other sounds like some strings, octilate preset or bass 2?
Do you think are there power in the low frequencies?
I just sell my yamaha v50 to buy preenFM so I can't compare ever their presets, like lately bass or the general sound.
-
Nice.... :D
Thanks for sharing
-
>> How does it compares at the low notes mantaining whole sound?
Not sure exactly what you mean, but the FM2 has a terrific bass response.
>> How about other sounds like some strings, octilate preset or bass 2?
I'll put some more up. Note that not every DX7 preset translates exactly...
>>Do you think are there power in the low frequencies?
Yes!
>>Nice.... :D
Thanks, Xavier! You've built a cool machine :)
-
Thanks :)
>> Note that not every DX7 preset translates exactly...
You'll hear that some are very far...
This is something i can try to improve in a futur firmware.
-
Xavier it would be great if the preset parity with the DX7 could be improved, just because many have used one before and might think they know what to expect sound wise when choosing a preset from it.
-
The DX7 has 32 algos, but Preen FM2 has 27, not sure how much of a problem that is.
The DX7 feature is useful, but to be fair if you want the perfect DX7 sound there are plenty of ways of getting it. I'd much prefer to move forward than just keep emulating the past.
-
I had to implement all DX7 algos to be able to read the preset. PreenFM algo from 10 to 27 are the same as the 32 DX7 algos.
DX7 algos can be different only because of the place of the feedback (See attached picture).
As the PreenFM2 does'nt have feedback it needs less algo. High feedback value can be simulated by selecting a different waveform of the operator.
What i'll try to do is to have the presets sounding better (not closer the DX7), at least by removing the digital noise that occure in too many presets...
Bad sounding DX7 preset can often be solved by manually decreasing modulation indexes and frequency of the oscilators in the highest positioned operator. Press ENG + MENU to have a quick look at the algo. Operator 5 & 6 are often guilty.
Reseting "fine tune" in the oscillator can also help.
Keep in mind that many presets use corner case or even bug of the DX7 to sound as they do. I learnt that by studying Hexter code (http://dssi.sourceforge.net/hexter.html)
The PreenFM2 will never be able to emulate that
Xavier
-
I also read that people used odd settings (but inaudible ones) as a sort of watermark so they knew if their patches were being copied.
-
I am not actually looking into emulating a dx7 faithfully with my PreenFM2, I agree there might be easier ways.
I was just reminiscing some friend that had prior experience in one and in that sense would have made things easier for people like him, although I would not sacrifice further development in other areas to improve the Preen´s ability to emulate a dx7.
It was just that I thought that Xavier had some more tweaks up his sleeve on this subject. No biggie anyway!
-
One thing I would worry about if the DX7 translation was too perfect is Yamaha taking notice. Not sure how likely that would be, but you know how lawsuit mad companies are these days.
-
Native Instrument's FM8 reads DX-7 patches too - so I wouldn't worry too much about Yamaha...
I'd be happy if DX-7 patch conversion would be improved. A lot of the patches were rather strange so far. It would be a huge library to use with the Preen and I like to load some patches to learn about FM synthesis. It's not intuitive at all for me to program so far...
-
Heter project goes rather far in the DX7 emulation :
http://dssi.sourceforge.net/hexter.html
Not sure yamaha cares about all those project ;-)
I also read that people used odd settings (but inaudible ones) as a sort of watermark so they knew if their patches were being copied.
Interesting... that can explain the strange values i see with some high level modulation oscillator.
The problem is that some of them creates VERY audible aliasing/noise on the PreenFM.
-
Actually after writing that I realised that maybe the FM patents have expired. They're for 25 years?
Although this seems to suggest that FM on hardware is covered by the patent, software isn't covered, hence FM8. Although PreenFM is running software on a computer, it's not an FM IC.
"Fourth, while Yamaha owned the patent on hardware-based FM, which meant that until recently no other manufacturer could offer competing FM instruments, nobody owns FM when it's running on a computer. Each developer can put a fresh spin on the FM concept."
http://oreilly.com/digitalmedia/2006/04/12/fm-synthesis-tutorial.html
CSound apparently can render DX7 sounds, or is it converting sounds to DX7 format?
"CSound also has opcodes for producing DX7 sounds."
I read about the quirky settings thing here:
http://en.wikipedia.org/wiki/Yamaha_DX7#Program_piracy