A1Four effects not working on G3n #66
Comments
|
When returning to Linux, I found that the So I had to re-install them... |
|
ToneLib-Zoom is not much help either, though might work for OP. It is possible to navigate and build patches via the Midi interface, however it's quiet 'brutal'... :-( |
|
Firstly you need to know the effect ID in 7-bit values (as will be sent over Midi) Gives The navigate pedal to a preset you don't care about, we'll install the effect in the first (zero-th) slot by sending a SysEx The screen is bust on my G1Four (it's the first one I use for doing 'weird' stuff), but reading the 'screen' through SysEx it looks like effect is installed and working However using ToneLib, it shows as 'Bypass'. |
|
Doing the same on the GCE-3_as_G3n was not as successful. It appears that the 'screen' does NOT show the effect name, this is behavior we have seen before... I then confirmed that the ZD2 are (still) installed. So it appears that these ZD2 are somehow incompatible with the G3n... time to look into the Code segment I guess. |
|
To get at the code segment... EQ_AVN.ZD2.code.readelf.txt There doesn't seem to be anything (too obviously) calling external SONAMEs, which is the case with the modules of the V3 and V6 voice processors. There could be some data value (minimum supported revision etc) which stops the effect loading. These symbols look like they are defining the parameters for each of the EQ sliders, so should be possible to duplicate their values into the other EQ effect, if that's all the OP actually needs. A patched version of Note: that last table is obviously something other than an array of float32. |
|
For reference here's what the normal G3n EQ has for it's settings... |
|
change to |
|
Also noted is that the ASCII of the 'screen' is stored in another symbol It might be necessary to hand edit this symbol/values in a Hex-editor. :-( There are also the 'Text' and 'XML' blocks of the ZD2, but we can automatically copy these with the 'donor' system. |
Can't publish the whole ZD2 for copyright reasons, but can upload a binary patch |
|
Wow, I wasn't expecting it, but happily surprised to see that GuitarLab reports the adjusted values for the sliders...
|
|
Figured out why the 'screen' values are duplicated, this effect is 'wide' and uses the Right button as a A/B toggle. Also noted that although the modified effect has different adjustment range for 270Hz (+/-18dB), this is not shown in Guitar-Lab. Makes me suspicious that there is a bug in effect, and that maybe the Q values aren't different either... |
|
Nice work! Indeed, all of the A1 FOUR EQ effects are 1U, that is "wide" yet made for 1-screen units. These should be compatible with G1 FOUR and similar. G3n has 3 screens, so these EQ modules won't be compatible as-is. ToneLib does not support A1 FOUR in general and its specific effects either. Thus you saw these as BYPASS. So pulling the freqs from the A1 FOUR and plugging them into G3n/G5n "native" There are several "wide" effects for A1 FOUR:
|
It seems that Zoom thinks its worth making multiple EQ for different instruments, but maybe they are just padding the list of effects. Really it's not that much effort given what we know now; and now it's done for the 'violin.patch' all you need is to apply the binary patch file over the top of 'GT_GEQ_7.ZD2'. I even found a Python module to do this... so platform independant, yay! |
Maybe this is the answer to the incompatibility; in that the single screen effects implement a mechanism where the 4th dial is used to change page to get to the extended settings (ie when then are more than 4 settings to be changed). |
|
IIRIC, the paging also involves injection of Dummy effect ( Technically, it's possible to generate a patch of 5 "wide" effects, thus occupying a total of 10 slots (or 15, counting the injected page-slots); however, the firmware caps the G1 FOUR patches to two wide effect per patch. |
|
Back to the question on what effects actually work, and why the others don't... I wrote a script to install each effect in turn into a patch and see what the 'Display' says (ie does it actually accept the effect as usable). Run like this... gce_as_G3n_success_for_A1Four.txt The ones which work are |
|
There are 18 "wide" effects specific to A1 FOUR listed above. Generally, neither should be compatible with G3n/G5n/B3n. As for the "wide" effects from G1/B1 FOUR, there are already non-1U equivalents for the G3n/G5n/B3n. So those should be used instead. |
|
@nomadbyte I'm not quite following what you are saying.... by 'wide' I mean effects which are designed to show across two LCD screens (ie on G3n/G5n). None of the G1-Four/A1-Four will be 'wide'. As you note the 1U equivalents will use the 4th knob for paging on a single screen. I double checked my GCE-3_as_G3n and manually removed ALL effects, before installing all that GL7 lists for the A1-Four. That gives me: Looks like only 59 (out of the 176 possible) effects actually install, add to patch and report parameters.... this can't be just down to the amount of parameters. A couple of the effects (AG_PUSL.ZD2 and ANTI_FB.ZD2) only have 1 parameter. |
|
We know the highest byte of the Effect ID is the 'group', do we know what the next byte is (D23..16)?? Seems that most of the A1-Four effects have bits set in this, whereas most of the G3n do not... Is this a bit field for 'requirements'? |
|
I mean "wide" as designed to occupy 2 slots (rather 2x4 parameter screens). These effects usually have 8 parameters. 1U models page these parameters; 3U and more models spread these 8 parameters over 2 screens. Just as I mentioned earlier, 1U models have their own version of such "wide" modules (and blacklist the non-1U modules in |
|
I think of the highest byte of the Effect ID as category id. The 7th byte is used more like a mnemonic to further classify the effects: |
|
|
So just running the numbers... and removing the (obvious?) know reasons for not installing - ie '3S' and '1U' effects, and 0x14 group/categories. We still have the following failing to install. So of these have more than 4 params, but quiet a few have 4 or less. |
|
So, if the Effect ID is a determining factor... can we just change that? But that does NOT work, the modified Effect still does not install. :-( Maybe 'Effect ID' is also contained within the ELF binary... |
Reported by user on Reddit, the A1Four effects (
EQ_AVN.ZD2andEQ_EVN.ZD2) are not working on his G3n.Using my GCE-3 I was able to confirm the failure. I installed the effects (as the only ones on the device)
There are actually more ZD2 files on the device, but not in the
FLST_SEQ.ZT2. Not sure whether safe to delete them...I also checked to see if these were listed in the Guard file, they are not.
Then I jumped into Windows and looked with Guitar-Lab. It seemed 'confused', front panel only noted the presence of other effects, but when editing a patch NO effects were present (other than the pseudo Bypass/Rythm/Loopers).
The text was updated successfully, but these errors were encountered: