Plex shift

edo

Power User
Vendor
Spent the last couple hours on my new fm3, I had never questioned / doubted about the sound quality, which is top notch like expected. I also had to reconsider the 3 switches (I was afraid it wasn’t enough, but the way they work is just pure genius, so that’s a big positive surprise!). However, I’ve been trying to copy a preset from the af3 that uses plex shift, and with all parameters set identical, I can’t make it sound the same. So I was curious whether the plex shift algorithm in the fm3 is different from the one in the af3 (because of cpu limitations), or if I’m missing something or I’m doing something wrong..
 
Can you post both presets (III + FM3)?

Yeah, well, actually it's a preset from one of my commercial preset packs, but I can make an edited / stripped down version to focus on the Plex Shift block only! Give me a minute! (P.S. How do I upload a preset here??)
 
Just upload the presets with only the Plex Delay.
You can drag the presets in the text box of your answer.
 
here you go
 

Attachments

  • PLEX SHIFT AF3.syx
    48.2 KB · Views: 14
  • PLEX SHIFT FM3.syx
    24.1 KB · Views: 7
The Plex Delay algo is the same. But the FM3 trails the III a little. The latest improvements in the III version have not yet been implemented in the firmware of the FM3. For example, the Crossfade parameter has been replaced in the III.

Also, some of the parameter values in the preset are different, such as the Detune values.

But I noticed that the stereo imaging seems "off" in the FM3 version. Changing "Stereo Spread" emphasises this. Can you confirm this?
 
Detune values should be irrelevant as I'm only using the first 4 delays (the other ones are off with volume set to 0%). I find the behavior of the repeats is substantially different, the af3 Plex has a different rhythmic pattern and each repeat sounds 1 octave higher than the previous one, while in the fm3 the rhythmic pattern sounds off and each Plex stays within the same octave. I don't know if this makes sense, maybe it's easier if I just post a clip with the part I'm playing to give an actual sample of what I'm describing.
 
Yes, definitely different! Seems odd, I have no doubt Fractal will address this in next firmware.

i don’t necessarily see it as a “problem” to address in the next fw, I would just like to know whether the algorithm works differently to figure out a workaround!
 
i don’t necessarily see it as a “problem” to address in the next fw, I would just like to know whether the algorithm works differently to figure out a workaround!
Yeah, but you would expect it to be similar... it would be a real pain if like for like blocks needed wildly different parameters to get the same result.
 
Yeah, but you would expect it to be similar... it would be a real pain if like for like blocks needed wildly different parameters to get the same result.

yeah, well, this is the only block where I encountered such difference, everything else is pretty much spot on identical. That’s why I asked to see if there’s a technical reason for this and in that case learn to live with that!
 
yeah, well, this is the only block where I encountered such difference, everything else is pretty much spot on identical. That’s why I asked to see if there’s a technical reason for this and in that case learn to live with that!
Edo, I would think you might want to enter this as a potential bug, unless someone has heard that there is supposed to be a significant difference in this area.
 
FM3 has the pitch shift before the feedback loop rather than within it. It would be nice to have the option for either routing. Page 64 of the blocks guide has a description of the two routings as used in the pitch block.
 
FM3 has the pitch shift before the feedback loop rather than within it. It would be nice to have the option for either routing. Page 64 of the blocks guide has a description of the two routings as used in the pitch block.
So you mean the implementation IS different?
 
anyone else notice that any of the Plex Delays using any kind of pitch shifting sound and react different ever since the 12.08 update?
I guess it's to be expected since the entirely reworked pitch algo, which works great for all the virtual capo stuff etc., but it hasn't done anything good for the Delays, imho. especially the Plex Detune has become next to unusable for me.
this may sound funny, but it's almost like the new algorithm is too good (too fast and clean). I don't think I'm imagining things here as I had just tweaked Plex blocks shortly before the update and the sound was still pretty fresh in my ears.
probably not gonna happen, but it would be cool to have the old algorithm back for everything that's not 100% pitch shift related...
 
Back
Top Bottom