For a horn (lip-vibrated aero-On Sunday, January 16, 2022 at 1:09:31 AM UTC, Don Y wrote:
On 1/15/2022 6:01 PM, Rich S wrote:
On Saturday, January 8, 2022 at 10:13:46 PM UTC, Don Y wrote:
On 1/8/2022 12:57 PM, John Walliker wrote:
On Friday, 7 January 2022 at 00:15:40 UTC, Don Y wrote:
On 1/6/2022 12:50 PM, bitrex wrote:
I\'ve also seen a product or two that look at the *audio* and try to reconstruct
the fingering from that. Theoretically possible but a fair bit of work to
accomplish in real-time. Not sure how that latency stacks up with the
rest of the processing chain...
I haven\'t tried one of those in some time but the last I did I wasn\'t
impressed. The G50 is reported to have an average latency of about 10 ms on the
low strings and feels pretty acceptable to me (haven\'t measured it myself) for
guitar playing that\'s not speed metal riffs at least.
For playing USB keyboards though, like with virtual instruments, it has to be
lower than that, I find around 2ms is about what I can tolerate, I can tell the
difference between a 2ms response and a 5 ms, and 10 ms starts to feel like an
eternity.
Well, at least there is an explicit (industry-wide) acknowledgement that there
is a significant issue, there. And, attempts to put a number on it. Too many
other interfaces just leave that unspoken -- /caveat emptor/.
I question whether the physical-virtual instruments (\"funky UI\'s that try to
LOOSELY resemble a real instrument\") end up altering *how* you play. The
\"physical presence\" of an instrument seems to be a part of the performance
experience; would a \"sousaphone emulator\" (that you can hold in one hand)
*play* like a real sousaphone? Even if the emulator was a 100% faithful
reproduction...
How about a pneumatically controlled pipe organ in a large church where
the latency could be around 200ms?
Most wind instruments have high latency -- esp if excited by
human breath.
My above comment was wrt the *feel* of the instrument-wannabe vs. the
real instrument. What\'s it like to play a \"little box with a mouthpiece\"
that emulates a tuba vs. having to carry the tuba on your shoulders
while trying to summon up enough air to excite it? Does the little
box represent the same \"impedance\" to your windpipe as the genuine
article? Do you have to move the same amount of air to produce a
particular sound?
[I.e., playing trills on a trumpet is a piece of cake compared to
a trombone, baritone horn, etc. Does a trombone-emulator present
the same PHYSICAL playing challenges as a real trombone?]
Latency is hard to avoid unless youre putting a pressure sensor right
at the stimulus point. For a guitar, that would be under each fret
position, for each string. i.e. ~15 x 6 = 90 sensors.
Not use how to do *that*, reversably & non destructively.
and while not interefering with normal acoustics, ergonomics,
etc. = RS
But *players* of particular instruments already have a \"feel\" for
the inherent characteristics of an instrument and adjust their
play, accordingly. Change that \"feel\" (because you\'re using an
approximation of the instrument) and your \"play\" changes, as well.
E.g., The basics of playing any horn are similar. Yet, the actual
results vary considerably as the physics of the instrument vary (e.g.,
with size, required air volume, etc).
I\'d imagine playing a ukulele to be significantly different than
a double bass.
Sure, most acoustical instruments have an inherent acousto-
mechanical \"latency\" and ergonomic factors that set a upper-
limit on fluent note change. You could retrofit one
of these types with electronics that make it outperform
the original instrument.
phone), we could trigger on finger- or other actions, and
not rely on the lip-buzz.
cheers, RS
On 1/15/2022 6:49 PM, Rich S wrote:
On Sunday, January 16, 2022 at 1:09:31 AM UTC, Don Y wrote:
On 1/15/2022 6:01 PM, Rich S wrote:
On Saturday, January 8, 2022 at 10:13:46 PM UTC, Don Y wrote:
On 1/8/2022 12:57 PM, John Walliker wrote:
On Friday, 7 January 2022 at 00:15:40 UTC, Don Y wrote:
On 1/6/2022 12:50 PM, bitrex wrote:
I\'ve also seen a product or two that look at the *audio* and
try to reconstruct
the fingering from that. Theoretically possible but a fair bit
of work to
accomplish in real-time. Not sure how that latency stacks up
with the
rest of the processing chain...
I haven\'t tried one of those in some time but the last I did I
wasn\'t
impressed. The G50 is reported to have an average latency of
about 10 ms on the
low strings and feels pretty acceptable to me (haven\'t measured
it myself) for
guitar playing that\'s not speed metal riffs at least.
For playing USB keyboards though, like with virtual instruments,
it has to be
lower than that, I find around 2ms is about what I can tolerate,
I can tell the
difference between a 2ms response and a 5 ms, and 10 ms starts
to feel like an
eternity.
Well, at least there is an explicit (industry-wide)
acknowledgement that there
is a significant issue, there. And, attempts to put a number on
it. Too many
other interfaces just leave that unspoken -- /caveat emptor/.
I question whether the physical-virtual instruments (\"funky UI\'s
that try to
LOOSELY resemble a real instrument\") end up altering *how* you
play. The
\"physical presence\" of an instrument seems to be a part of the
performance
experience; would a \"sousaphone emulator\" (that you can hold in
one hand)
*play* like a real sousaphone? Even if the emulator was a 100%
faithful
reproduction...
How about a pneumatically controlled pipe organ in a large church
where
the latency could be around 200ms?
Most wind instruments have high latency -- esp if excited by
human breath.
My above comment was wrt the *feel* of the instrument-wannabe vs. the
real instrument. What\'s it like to play a \"little box with a
mouthpiece\"
that emulates a tuba vs. having to carry the tuba on your shoulders
while trying to summon up enough air to excite it? Does the little
box represent the same \"impedance\" to your windpipe as the genuine
article? Do you have to move the same amount of air to produce a
particular sound?
[I.e., playing trills on a trumpet is a piece of cake compared to
a trombone, baritone horn, etc. Does a trombone-emulator present
the same PHYSICAL playing challenges as a real trombone?]
Latency is hard to avoid unless youre putting a pressure sensor right
at the stimulus point. For a guitar, that would be under each fret
position, for each string. i.e. ~15 x 6 = 90 sensors.
Not use how to do *that*, reversably & non destructively.
and while not interefering with normal acoustics, ergonomics,
etc. = RS
But *players* of particular instruments already have a \"feel\" for
the inherent characteristics of an instrument and adjust their
play, accordingly. Change that \"feel\" (because you\'re using an
approximation of the instrument) and your \"play\" changes, as well.
E.g., The basics of playing any horn are similar. Yet, the actual
results vary considerably as the physics of the instrument vary (e.g.,
with size, required air volume, etc).
I\'d imagine playing a ukulele to be significantly different than
a double bass.
Sure, most acoustical instruments have an inherent acousto-
mechanical \"latency\" and ergonomic factors that set a upper-
limit on fluent note change. You could retrofit one
of these types with electronics that make it outperform
the original instrument.
phone), we could trigger on finger- or other actions, and
not rely on the lip-buzz.
cheers, RS
Yes, but the point of \"instrumenting\" an instrument is to
record your *current* playing proficiency. Change the way the
instrument plays/\"feels\" and do you end up turning a prodigy
into a neophyte? The instrument is an integral part of the
performance, alongside the performer.
\"Rich S\" wrote in message
news:fa1c2bfe-a14c-419e-8b0e-d2a9e0d4c74an@googlegroups.com...
On Saturday, January 8, 2022 at 10:13:46 PM UTC, Don Y wrote:
On 1/8/2022 12:57 PM, John Walliker wrote:
On Friday, 7 January 2022 at 00:15:40 UTC, Don Y wrote:
On 1/6/2022 12:50 PM, bitrex wrote:
I\'ve also seen a product or two that look at the *audio* and try to
reconstruct
the fingering from that. Theoretically possible but a fair bit of
work to
accomplish in real-time. Not sure how that latency stacks up with the
rest of the processing chain...
I haven\'t tried one of those in some time but the last I did I wasn\'t
impressed. The G50 is reported to have an average latency of about 10
ms on the
low strings and feels pretty acceptable to me (haven\'t measured it
myself) for
guitar playing that\'s not speed metal riffs at least.
For playing USB keyboards though, like with virtual instruments, it
has to be
lower than that, I find around 2ms is about what I can tolerate, I can
tell the
difference between a 2ms response and a 5 ms, and 10 ms starts to feel
like an
eternity.
Well, at least there is an explicit (industry-wide) acknowledgement
that there
is a significant issue, there. And, attempts to put a number on it. Too
many
other interfaces just leave that unspoken -- /caveat emptor/.
I question whether the physical-virtual instruments (\"funky UI\'s that
try to
LOOSELY resemble a real instrument\") end up altering *how* you play. >> The
\"physical presence\" of an instrument seems to be a part of the >> performance
experience; would a \"sousaphone emulator\" (that you can hold in one >> hand)
*play* like a real sousaphone? Even if the emulator was a 100% faithful
reproduction...
How about a pneumatically controlled pipe organ in a large church where
the latency could be around 200ms?
Most wind instruments have high latency -- esp if excited by
human breath.
My above comment was wrt the *feel* of the instrument-wannabe vs. the
real instrument. What\'s it like to play a \"little box with a mouthpiece\"
that emulates a tuba vs. having to carry the tuba on your shoulders
while trying to summon up enough air to excite it? Does the little
box represent the same \"impedance\" to your windpipe as the genuine
article? Do you have to move the same amount of air to produce a
particular sound?
[I.e., playing trills on a trumpet is a piece of cake compared to
a trombone, baritone horn, etc. Does a trombone-emulator present
the same PHYSICAL playing challenges as a real trombone?]
Latency is hard to avoid unless youre putting a pressure sensor right
at the stimulus point. For a guitar, that would be under each fret
position, for each string. i.e. ~15 x 6 = 90 sensors.
Oh dear....er.... that will be 24 x 6 mate.....