[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Date Index][Thread Index][Author Index]

Re: guitarists: triple play?





On 05/01/2015 20:09, In Mobile wrote:
Hi to all

Maybe we can elaborate..

glad to ;-)

Andy's assumptions are based on the fact the Axon
processing of the signal has already peaked near
the physical barrier (or so to say, adds no time lag to the response).

No, it's not my assumption, it's my reasoning.
Were I to follow the DSP experts then the limit is firmly set
at one period.

Basically you can't measure something before it's happened.



This is probably a bit far fetched,
and there is probably a lot of room for algorithm improvement,
not even to mention alternative means.

Not really, it's physics.
Unless by alt means you intend to use something other than the regular hex 
p/u.

I have owned and still use a Roland GR50, a Yamaha G50 ( a rebadged  Axon 
NGC 77) and an Axon AX100Mk2.

Provided I use the internal sounds, the old GR50 yields better results 
than the latters on tapping, hammer-on and pull-off techniques.
My guess was always that it used internally pitch to voltage tech
while using pitch to midi for the midi out hence explaining why the
built in sound did not match the midi out at the same time (I checked).

More likely the internal sounds are derived from the audio from the 
pickups.

I'm not really familiar with the various Roland editions but if you
wanna believe what you hear on Looper's Delight the situation
seems to be.

1) Roland stuff is a bit slower on Midi conversion
2) Some Roland models use the p/u audio directly, so 'tracking' is 
excellent.

I don't think any Roland gear does pitch to voltage,
but if it did the physics is still the same in terms of
how fast you can identify the pitch.



If the Axon AX100MkII provides pick position recognition,

Yes, I didn't say, but it does.

...but pick position information is always one note delayed.

The *second* time you play at any given position the Axon algorithm can 
look at the waveform
and have a pretty good guess at what the pitch is.
I figure it could actually do a bit better than the one period limit
that is general accepted in the DSP world.

 it doesn't track any better nor quicker than its rebadged grand father 
Yamaha G50.

No, but it has the most unappealing computer soundcard wavetable synth on 
board that
you ever heard!
They both have the same tech, but neither machine is built by Axon.
Terratec used the Axon name however.

The G50 even provides functions not available on the latter like octave 
transposing switches using S1/s2 switches.

..and the mono input is handy.


There is probably room for improvement.

Sure, but tracking speed isn't going to get better using the same p/u 
config.

We now find software that can extract poly note information from a guitar 
output without any divided pick up.

Yeah, that's genius, but more directed towards autogenerating a score than 
real time playing.


Gareth here proposed at a time some very interesting software projects.

Indeed.

I've discussed this with Bill Walker and I think he'd agree that the future
lies in using the pitch info to control effects that work on the
audio from the string.
e.g. filter frequency tracking note


The question is more : this old race of guitar to midi or to synth can 
sustain innovation with this niche market.

It's an interesting field, but really it never seems to catch on.

I don't think the guys who used it to play piano wavetables helped the 
case much :-)


andy



Regards

Olivier