Tripleplay Connect with all high E strings - .012 gauge

[kimyo]

When I use it in "Trigger" mode there is no issue whatsoever, but as soon as I try the other modes ( Auto and Smooth ) I cannot control the Sax PitchBend, and worse, some notes are out of tune.  So finally I gave up trying different PitchBend combinations

you've probably already checked this, but this behavior could be the result of a mis-match between the bend range settings on the tripleplay and swam.

[kimyo]

1.  Ability to use all high E or B strings or best choice/optimal string gauge for fastest, most accurate tracking.

i replaced my tuners this weekend and while i was messing around decided to do some experiments in order to try to determine why our experience differs.

what i didn't realize previously is that with a 'b' string in place of the low 'e', notes are not transmitted beyond 8 or 9 frets (the c or c#). 

with a high 'e' string in place of the low 'e', you were likely not getting notes beyond the 3rd or 4th fret.

i left the upper strings alone and swapped the low 'e' for a 'b' again, with the same experience as before, ~13ms latency.  there were no detectable differences glitch-wise between the 'a'/'d' strings and the 'b' taking the place of the 'e'.  hand position was off.  finger style was on.

if i were fishman i'd probably do exactly what they're doing.  there's a significant minority who aren't going to use zones or hand position sensing, or who might prefer the lowest latency performance by re-stringing, but fishman has to focus on their primary audience.

it is extremely nice, though, to have exactly the same latency for every note you play.  and that is achievable today, with the current software, as long as one is willing to sacrifice the upper range of the low 'e' string and the upper upper range of the 'a'. 

to me, the benefit of equal latency is as or perhaps even more important than low latency.  keyboard players would lose their minds if the lower part of the keyboard responded less quickly than the upper.

[kimyo]

i had read that finger style takes two cycles instead of one, but haven't seen specific times in msec before, thanks.

and i've only upgraded recently, so the latency tests i did were with the previous version.  which likely explains the consistency across the fretboard.

where i see an issue is when you have latency differences of 20msec or more between the low and the high strings.  i agree that 7msec is not a problem. or, certainly not for me.

do you recall which version of the software was 14ms across all the strings?  i would like to keep a copy handy.  if the choice for me is uniform latency vs fast latency i will generally choose uniform.

also, aside from zones and the faster processing on high strings, does the monitor hand position function serve other purposes?

[kimyo]

thank you for sharing your thoughts.  you make a strong argument.

i still would like more info on the earlier version with the fixed latency.  this is the firmware, correct? 

i don't mind if companies change things, however, i had it working just the way i like it.  i prefer dealing with companies which provide access to earlier versions of software/firmware for this exact reason.

for my main guitar, i have no choice but to use standard strings and so none of this matters there.  however, i am repurposing an old one as a midi-control only and i would like the option of using the setup i have already developed. 

another issue is that i don't use a pick, and so it's fingerstyle for me regardless. 

[philjynx]

....

another issue is that i don't use a pick, and so it's fingerstyle for me regardless. 

1970 something, the last time I tried to use a plectrum, never got on with them... 

[bonanote]

you've probably already checked this, but this behavior could be the result of a mis-match between the bend range settings on the tripleplay and swam.
[/quoteat

you've probably already checked this, but this behavior could be the result of a mis-match between the bend range settings on the tripleplay and swam.

Thanks for the answer. Yes , indeed , it is one of the first things I tried. Also, bear in mind that on the iPad Connect app there are no options to modify the Pitch Bend Range , so I could only play around with the SWAM Sax settings. I monitored the MIDI Data output from the Fishman connect and what seems to happen is that when "Smooth" or "Auto" settings , there is some PB data that flows together with the Note ON data , and this is what results initially in  out of tune  notes randomly. Also, afterwards when I actually bend the string , the PB data does not flow in an orderly fashion, it sort of jumps around. WHen I set the PB preferences to "Trigger" or "Stepped" the notes are perfectly in tune , because there is no PB data together with the NOte On .

[BigJim]

i replaced my tuners this weekend and while i was messing around decided to do some experiments in order to try to determine why our experience differs.

what i didn't realize previously is that with a 'b' string in place of the low 'e', notes are not transmitted beyond 8 or 9 frets (the c or c#). 

with a high 'e' string in place of the low 'e', you were likely not getting notes beyond the 3rd or 4th fret.

i left the upper strings alone and swapped the low 'e' for a 'b' again, with the same experience as before, ~13ms latency.  there were no detectable differences glitch-wise between the 'a'/'d' strings and the 'b' taking the place of the 'e'.  hand position was off.  finger style was on.

if i were fishman i'd probably do exactly what they're doing.  there's a significant minority who aren't going to use zones or hand position sensing, or who might prefer the lowest latency performance by re-stringing, but fishman has to focus on their primary audience.

it is extremely nice, though, to have exactly the same latency for every note you play.  and that is achievable today, with the current software, as long as one is willing to sacrifice the upper range of the low 'e' string and the upper upper range of the 'a'. 

to me, the benefit of equal latency is as or perhaps even more important than low latency.  keyboard players would lose their minds if the lower part of the keyboard responded less quickly than the upper.

Thanks again for your clear and detailed suggestions.
Keep in mind that it is usually a lot easier to make changes to the software and/or controller firmware than to (for example) modify the mechanical design of the pickup. Also, as you are likely aware, the string sensitivity settings can compensate for quite a wide range of string height adjustment, especially on the wireless version.
Finally, FYI, the tuning settings in preferences are not "custom tuning settings", instead they are telling the software how the guitar is actually tuned, which allows the fret markers to work properly, and allows the feature "Monitor hand position" to work properly. Adding a custom per-string tuning setting for each patch makes a lot of sense as a feature request.

OK, around 20 years ago, I tried doing what the OP suggested (just equip a spare guitar with a Roland midi pickup and string it with all high-E strings) in an attempt to overcome latency problems, and ran into the same issues. In the end I guessed the problem was in the converter software making assumptions about what each hex pickup "should" receive, and getting confused when it received "impossible" frequencies. I contacted someone at Roland at the time (no, I don't remember who) and suggested they add a feature in their software that enabled people to bypass the "assumptions" function.

So, here I am ~20 years later, wondering if manufacturers had addressed this, and... it appears not. So I've signed up to the forum specifically to plead the case.

OK, business case: Although there are many guitarists who would like to be able to use midi on live playing (and thus need it on a guitar with standard tuning) most guitarists are actually sitting in their bedrooms noodling and recording. What's more, they have at least one "spare" axe that they would love to use solely as a midi controller, and would buy your system in the blink of an eye if they could mount the midi pickup in one of their spare axes, strung with all high-Es, and play lovely responsive synth/piano/drum/whatever samples... if they only could.

So please please PLEASE make this a priority in your R&D. I'm a software developer myself, and can't imagine this would require much effort to make happen. It would MASSIVELY expand the market for this product.