[YamahaDX] DX7II D - strange problems with fractional scaling

[charles copp [e-mail] [YamahaDX]]

are you changing these with a hex editor
then sending to synth?

[[e-mail] [YamahaDX]]

Just one point: if really there’s such bug, no wonder nobody mentioned it in the past 30 years, because nobody ever understood well what fractional scaling is and how to use it :-)

I'm afraid you could be right, there are thousands of DX patches around, but only very few using fractional scaling - while it's not really complicated, just operator levels up and down for blocks of notes.. hm.


but companies like Sound Source Unlimited released whole soundsets that use this feature (fractional pianos..), strange, that they didn't notice any problem - if it is a problem of all DX7IIs and not only on mine.


That's what I'm trying to find out.

[[e-mail] [YamahaDX]]

no , i did not change anything. First i noticed the problem while transferring dumps with sysex-librarian on MAC, later i tested with SoundDiver on Mac and on PC - the same result. In SoundDiver i could see exactly which data was coming in and out - everything was as expected, except that the DX stored wrong data for the OP6 scaling offset for patch#1.

[[e-mail] [YamahaDX]]

Confirmed with v1.9 ...


1. When receiving fractional scaling cartridge data, the first byte is always set to 0x3C.
... Voice#1 op6 offset -6 (3036) gets offset +70 (3C36)


2. Performance Dual Voice B standard offset values -127 +127 are actually 0 ~ 255.
... Fractional scaling operator output gain +48dB


3. When receiving a master tune parameter change, the value is always set to -64.
... Tuning via midi works fine with v1.3


Enjoy testing !

[[e-mail] [YamahaDX]]

hey - thank you - this is really helpful !
Unbelievable how buggy the good old DX7ii is.
These are exactly the value changes i noticed on mine. Plus - you found an extra bug, the master tuning issue in V 1.9.
The good thing is: i don't need to worry about whether my DX is defective or not. By knowing these problems they can be avoided - not storing fractional patches on #1, not layering such patches with itself etc...

It's really strange, that these bugs remained unknown for 30 years. I have the SSU fractional pianos floppy here, the Mac version with a sysex dump loader on the disk. It was not easy to read this disk and extract the data because of the old exotic 800k floppy format - but that's a different story.
After successfully loading the sounds I noticed that the first patch sounded more like a harpsichord than a piano, while all the others were OK. That's were i stumbled across these bugs.
How could it be that SSU made and sold these discs without taking notice that the first patch was destroyed after transferring it ?
but OK, now that's history - and I'm happy that my DX is OK and working (or not working) like any other DX...

thanks for your help - it saved me a lot of time.

[Daniel Forró [e-mail] [YamahaDX]]

Is it possible that a certain OS version is OK? That would be strange, but
 who knows.

Daniel Forro

Post by ***@yahoo.de [YamahaDX]
How could it be that SSU made and sold these discs without taking notice that the first patch was destroyed after transferring it ?

[[e-mail] [YamahaDX]]

nothing is impossible...
i tested with 1.3 and 1.9, it was the same for the offset bug. Paul wrote, that the master tuning bug is present in 1.9 but not in 1.3.
There were some versions between and one earlier, 1.1.
Maybe the offset bug was introduced in 1.3 and 1.1 was OK ?
I dont have access to any of these Eprom images at the moment. Anyone here with a DX7ii 1.1, or 1,6,1.8 etc... ?

[[e-mail] [YamahaDX]]

I'm not sure of the differences of the II, the IID and the IIFD. I'm guessing the last two just differ by the addition of the floppy.

I have DX7IIV18_48AA.OBJ that I captured in 1989. I will try and upload it.

[[e-mail] [YamahaDX]]

thank you very much - i will test it as soon as i come home - l'm on vacation right now.
as far as i know there are no differences in firmware of dx7ii D and FD, and there was no synth called only dx7ii.
there was the dx7s , but this model was missing a lot of the new features, i think it does not really count as ii - and it had indeed a different firmware.

thanks, Grant - i will test the 1.8 asap !

[[e-mail] [YamahaDX]]

hi - finally i found the time to burn the EPROM and to test firmware 1.8 - with the same results as on 1.9 :-( the fractional offset bug is still present, and master tuning via sysex is stuck at -64.
at the moment so far version 1.3 seems to be the least buggy version :-) - regarding just these two issues.
i wonder if there were versions before 1.3 ?

[Martin Tarenskeen [e-mail] [YamahaDX]]

Post by ***@yahoo.de [YamahaDX]
the fractional offset bug is still present, and master tuning via sysex is stuck at -64.

How are negative numbers coded in DX7II sysex? I have seen different
methods to handle negative and positive values in SysEx messages, and more
than once I have seen editor software sending the wrong values to achieve
this. Is it possible to set mastertuning on the DX7II itself, without
using external software? In that case the external software might be the
problem, not the DX7II or the ROM version. Just guessing.

[[e-mail] [YamahaDX]]

i tested with sound diver, the sysex command for master tune is "F0 43 10 04 40 dd F7" (dd=data).When set on the DX7II itself, master tuning works and the DX is sending the correct data - when sending back the same data, the DX7 ignores all values and sets master tune to -64. Seems to be a bug
I didn't test the master tune on V.1.3, but as Grant reported, master tune via sysex is working with this firmware.