4 Mar
2010
4 Mar
'10
4:48 a.m.
Morning LAU,
As I cycled to work this morning, I had a crazy idea. It's just
daydreaming and will probably never happen, but I wondered if anyone on
the last has any useful/interesting thoughts.
I want to measure how fast I'm turning the pedals on my bike (the
"cadence") and synthesize/sample the sound of an internal combustion engine.
As far as I can work out, there are three major parts to this.
1. A sensor that can measure my cadence. A simple magnet switch that
triggers once a revolution won't be enough to measure the cadence with
sufficient resolution, since my cadence is usually between 50 and 80
rpm. I would probably need to mount multiple magnets spaced equally
around the chainwheel and have a single sensor on the frame. Then I have
to get it to supply this information to my control program.
2. I need a control program that can read in the input from my cadence
sensor and convert a cadence reading of "66 rpm" into a frequency that
should be sampled/synthesised, e.g. "500 Hz" (I'm making these numbers
up). It will also need to be able to somehow smooth out the readings,
perhaps by interpolation, so when I accelerate, the sound of the revs
climbing doesn't increase in obvious steps. It could also have other
logic, e.g. when my cadence is 0 rpm, the sound of the engine is idling
rather than off.
3. I need a synthesiser or sampler that can take an input from my
control program and make the sound of an engine (or more likely, a sine
wave to start with). I've never sampled or synthesised on a computer
before but this engine-specific sampling technology already exists in
video games, such as torcs[1].
I have absolutely no idea why I would want such a device - just for the
fun of building it, I guess. I would like it to work in realtime (rather
than later generating the soundtrack from recorded cadence data). The
thought of sitting at the traffic lights with my earphones in and then
hearing the mighty roar of a V8 as I pull away would be really satisfying...
[1] http://torcs.sourceforge.net/
Any thoughts - useful, interesting, humorous, or otherwise - are welcome!
Cheers,
Jonathan
----------------------------
Jonathan Gazeley
Systems Support Specialist
ResNet | Wireless& VPN Team
Information Services
University of Bristol
----------------------------
4 Mar
4 Mar
5:09 a.m.
Hi Jonathan!
I like the idea. :-)
Well I don't have exact answers for you, but at least tendential answers. I
hope this will satisfy you, until the real thing comes tyour way. :-)
I know that you can get several controllers (to connect via USB or similar)
and if you are good at real simple hardware work, then you can modify your own
controllers. I've seen (on lac), that Marije used a set of such controllers in
a realtime performance (I think using sc3 or another synth of that category -
csound, CLM, PD...).
I think either we here on LAU or the chaps on csound mailinglist had a
discussion about synthesizing the sound of an engine.
So ingeneral it should be possible to make your idea into reality. Csound -
or its competitors - certainly can do a lot of stuff for you (ramping, maybe
interpolating,...). So I'm not too sure, if you will need a control program in
the end. The hardest part here would be the modification of the controllers (I
suspect).
If you decide on csound or CLM I know, that you have to write some code. If
I'm correct about the synthesis of an engine, you might be lucky and only need
to implement your specific realtime control to plug into the engine-sound
itself.
Sorry, I can't be more specific. Perhaps this post is pointless, but I was
really intrigued by the idea. :-)
Kindest regards
Julien
--------
Music was my first love and it will be my last (John Miles)
======== FIND MY WEB-PROJECT AT: ========
http://ltsb.sourceforge.net
the Linux TextBased Studio guide
======= AND MY PERSONAL PAGES AT: =======
http://www.juliencoder.de
5:20 a.m.
On Thu, March 4, 2010 09:48, Jonathan Gazeley wrote:
1. A sensor that can measure my cadence. A simple
magnet switch that
triggers once a revolution won't be enough to measure the cadence with
sufficient resolution, since my cadence is usually between 50 and 80
rpm. I would probably need to mount multiple magnets spaced equally
around the chainwheel and have a single sensor on the frame. Then I have
to get it to supply this information to my control program.
Cycle computers which measure cadence work with a single magnet on the
crank and can measure cadence in that range and above. I think the fastest
RPM I ever got was (only a vague memory) around 120 (maybe?) when I was
being silly trying to see how fast I could pedal.
I have absolutely no idea why I would want such a
device - just for the
fun of building it, I guess. I would like it to work in realtime (rather
than later generating the soundtrack from recorded cadence data). The
thought of sitting at the traffic lights with my earphones in and then
hearing the mighty roar of a V8 as I pull away would be really
satisfying...
Would it not be better to have a pair of mighty speakers in the shape of
twin exhaust pipes sticking out the back so every one else can hear the
roar :-D
5:24 a.m.
The mighty exhaust pipe speakers sound even better. But while we're at it.
There are those game controls for your chair, that can vibrate and probably
siomluate some other shocking experiences. Why not integrate that in the
settle? :-) Wait a second: How powerful do you need to be to drive all these
gadgets with the dynamo? :-)
Kindest regards
Julien
--------
Music was my first love and it will be my last (John Miles)
======== FIND MY WEB-PROJECT AT: ========
http://ltsb.sourceforge.net
the Linux TextBased Studio guide
======= AND MY PERSONAL PAGES AT: =======
http://www.juliencoder.de
5:31 a.m.
On 04/03/10 10:24, Julien Claassen wrote:
The mighty exhaust pipe speakers sound even better.
But while we're at
it. There are those game controls for your chair, that can vibrate and
probably siomluate some other shocking experiences. Why not integrate
that in the settle?
Not sure that's necessary - the roads around Bristol are bumpy enough as
it is!
Wait a second: How powerful do you need to be to drive
all these
gadgets with the dynamo?
I once calculated approximately that when I'm cycling hard up a gentle
hill, I can probably develop a peak power of around 300W. Maybe this is
more like 100-150W sustained power.
Assuming that most of that power goes into forward kinetic energy, I
don't know how much I could tap off to run my on-board synthesiser,
gaming chair, amp, speakers, headlights, and champagne cooler.
Maybe I should fit a generator?
5:40 a.m.
Oh, the fine gentlleman needs his champagne cooller! :-) But you know, that
the real biker implement is either a small fridge for beer or the microwave
(for those bikers gone soft and family :-).
Kindly yours
Julien
--------
Music was my first love and it will be my last (John Miles)
======== FIND MY WEB-PROJECT AT: ========
http://ltsb.sourceforge.net
the Linux TextBased Studio guide
======= AND MY PERSONAL PAGES AT: =======
http://www.juliencoder.de
5:28 a.m.
On 04/03/10 10:20, James Morris wrote:
On Thu, March 4, 2010 09:48, Jonathan Gazeley wrote:
Yes - I have a cycle computer that can measure cadence. However, I
typically pedal between 60 and 90 rpm which means my reading would only
be available for sampling every 0.7 - 1.0s. It would work, but in the
space of one second my cadence can vary a lot if I'm accelerating so the
sound of the rising revs would be jumpy.
Also, if I suddenly stopped pedalling, it might take up to a second for
the sampler to register and kill the engine noise. I don't think it
would be a very satisfactory experience, which is why I was thinking of
mounting 10 or so magnets around the chainwheel, to be able to sample
the cadence every 0.1s.
1. A sensor that can measure my cadence. A simple
magnet switch that
triggers once a revolution won't be enough to measure the cadence with
sufficient resolution, since my cadence is usually between 50 and 80
rpm. I would probably need to mount multiple magnets spaced equally
around the chainwheel and have a single sensor on the frame. Then I have
to get it to supply this information to my control program.
Cycle computers which measure cadence work with a single magnet on the
crank and can measure cadence in that range and above. I think the fastest
RPM I ever got was (only a vague memory) around 120 (maybe?) when I was
being silly trying to see how fast I could pedal.
6:32 a.m.
On Thu, Mar 04, 2010 at 10:27:43AM +0000, Jonathan Gazeley wrote:
Yes - I have a cycle computer that can measure
cadence. However, I
typically pedal between 60 and 90 rpm which means my reading would only
be available for sampling every 0.7 - 1.0s. It would work, but in the
space of one second my cadence can vary a lot if I'm accelerating so the
sound of the rising revs would be jumpy.
Also, if I suddenly stopped pedalling, it might take up to a second for
the sampler to register and kill the engine noise. I don't think it
would be a very satisfactory experience, which is why I was thinking of
mounting 10 or so magnets around the chainwheel, to be able to sample
the cadence every 0.1s.
That would be a good rate. But for a realistic result you also
need to measure the force you are putting on the pedals, or the
chain tension - the noise of an engine depends very much on the
amount of power it has to deliver.
I've been working on engine noise simulation - it's not very
difficult unless you want to exactly simulate an existing car.
Ciao,
--
FA
O tu, che porte, correndo si ?
E guerra e morte !
6:39 a.m.
On 04/03/10 11:27, fons(a)kokkinizita.net wrote:
On Thu, Mar 04, 2010 at 10:27:43AM +0000, Jonathan
Gazeley wrote:
Nick Copeland suggested simulating the force applied to the pedals by
calculating the rate of change of the cadence. I reckon that might do
the trick.
Yes - I have a cycle computer that can measure
cadence. However, I
typically pedal between 60 and 90 rpm which means my reading would only
be available for sampling every 0.7 - 1.0s. It would work, but in the
space of one second my cadence can vary a lot if I'm accelerating so the
sound of the rising revs would be jumpy.
Also, if I suddenly stopped pedalling, it might take up to a second for
the sampler to register and kill the engine noise. I don't think it
would be a very satisfactory experience, which is why I was thinking of
mounting 10 or so magnets around the chainwheel, to be able to sample
the cadence every 0.1s.
That would be a good rate. But for a realistic result you also
need to measure the force you are putting on the pedals, or the
chain tension - the noise of an engine depends very much on the
amount of power it has to deliver.
I've been working on engine noise simulation -
it's not very
difficult unless you want to exactly simulate an existing car.
That sounds promising :)
6:54 a.m.
On Thu, Mar 04, 2010 at 11:39:40AM +0000, Jonathan Gazeley wrote:
Nick Copeland suggested simulating the force applied
to the pedals by
calculating the rate of change of the cadence. I reckon that might do
the trick.
It doesn't. Just consider going uphill 5% or downhill 5% at
the same and constant cadence. Makes some difference to your
legs and to a motor.
Ciao,
--
FA
O tu, che porte, correndo si ?
E guerra e morte !
8 a.m.
Fons:
On Thu, Mar 04, 2010 at 11:39:40AM +0000, Jonathan
Gazeley wrote:
Well, then you need an accelerometer [1]. [2] is said to be in the
iPhone, but you probably want something easier to solder.
With it you could get the inclination angle and then get the applied
force. If you do it right you could even get a crashing sound when
falling...
Regards,
/Karl
[1] http://en.wikipedia.org/wiki/Accelerometer
[2] http://www.st.com/stonline/products/literature/ds/12726/lis302dl.htm
-----------------------------------------------------------------------
Karl Hammar Aspö Data karl(a)aspodata.se
Lilla Aspö 148 Networks
S-742 94 Östhammar +46 173 140 57 Computers
Sweden +46 70 511 97 84 Consulting
-----------------------------------------------------------------------
Nick Copeland suggested simulating the force
applied to the pedals by
calculating the rate of change of the cadence. I reckon that might do
the trick.
It doesn't. Just consider going uphill 5% or downhill 5% at
the same and constant cadence. Makes some difference to your
legs and to a motor. 
4 p.m.
Nick Copeland suggested simulating the force applied
to the pedals by
calculating the rate of change of the cadence. I reckon that might do
the trick.
It doesn't. Just consider going uphill 5% or downhill 5% at
the same and constant cadence. Makes some difference to your
legs and to a motor.
4:30 p.m.
On Thu, Mar 04, 2010 at 10:00:09PM +0100, Nick Copeland wrote:
Well, then you need an accelerometer [1]. [2] is said to be in the
iPhone, but you probably want something easier to solder.
With it you could get the inclination angle and then get the applied
force. If you do it right you could even get a crashing sound when
falling...
Actually it does do the trick, Fons doesn't understand the british turn of
phrase: Jon is not implying an attempt to win the 2010 Tour de France
with his contraption, he is going to have fun on the way to work so I am sure
he can extract acceleration (from traffic lights, Fons, he doesn't actually want
to be the King of the Mountains or anything, not even give or take 5%) based
on the rate of change of a signal.
With such noise generating things on your bike you'd be
excluded from the TDF anyway :-)
Of course you can find acceleration from the rate of change
of cadence plus gear ratio etc. And according to Newton,
F = ma and we know m. But that assumes that it is the force
on the pedals that causes the acceleration, which is not
always the case. And the roar of a motor is very dependent
on the work it has to do.
For the simulations I've been working on (Italian cars,
fast, usually red, I can't say more), we used RPM and
throttle position as the inputs to the motor model.
RPM defines the fundamental frequency, but the power
delivered by the moter defines the spectrum.
Ciao,
--
FA
O tu, che porte, correndo si ?
E guerra e morte !
Nick Copeland
suggested simulating the force applied to the pedals by
calculating the rate of change of the cadence. I reckon that might do
the trick.
It doesn't. Just consider going uphill 5% or downhill 5% at
the same and constant cadence. Makes some difference to your
legs and to a motor.
5:13 p.m.
With such noise generating things on your bike
you'd be
excluded from the TDF anyway :-)
Yeah, I heard you couldn't enter without a cocaine dependency
either - Jon, do you have a habit?
Of course you can find acceleration from the rate of
change
of cadence plus gear ratio etc. And according to Newton,
F = ma and we know m. But that assumes that it is the force
on the pedals that causes the acceleration, which is not
always the case. And the roar of a motor is very dependent
on the work it has to do.
Chasing away from lights in the city is a pretty close thing. This
was a fun project proposal, now it needs pulse detectors, some
accelerometers, pressure guages, a small computer, gps, 2 separate
lead acid accumultators to take care of the Omron X5MF 24V (one in
each pannier assumably), a speaker set in a rucksack and a wind vane
on a 10 meter pole.
Man, I would start suggesting that a better method would be to strap
a 2-stroke to his back and mic it up - I am sure LAU could advise on
optimal positioning.
Now if he does go for strapping the motor on his back then
the Omron could also be placed in a small tank mounted on the
crossbar (that will keep them dry and could also house the mic
pre-amps), he could chrome up his front forks, get a full face visor
crashhelmet on and splash out on a set of leathers. Anybody on
the list know of a few tattoo artists to boot? Oh, and now he will
need those TDF steroids to even get the whole lot moving in the
first place - is Lance on the list?
Many thanks for the extra little bit of selfpublicity, that will
certainly help me sleep better.
Regards, nick.
_________________________________________________________________
Hotmail: Free, trusted and rich email service.
https://signup.live.com/signup.aspx?id=60969
5 Mar
5 Mar
3:29 a.m.
On Thursday 04 March 2010 23:13:21 Nick Copeland wrote:
I would suggest a small one- or two cylinder engine and some modifications to
the exhaustion pipes. Gives a really realistic engine sound and it saves you
from pedalling :-)
Arnold
Of course you
can find acceleration from the rate of change
of cadence plus gear ratio etc. And according to Newton,
F = ma and we know m. But that assumes that it is the force
on the pedals that causes the acceleration, which is not
always the case. And the roar of a motor is very dependent
on the work it has to do.
Chasing away from lights in the city is a pretty close
thing. This
was a fun project proposal, now it needs pulse detectors, some
accelerometers, pressure guages, a small computer, gps, 2 separate
lead acid accumultators to take care of the Omron X5MF 24V (one in
each pannier assumably), a speaker set in a rucksack and a wind vane
on a 10 meter pole.
Man, I would start suggesting that a better method would be to strap
a 2-stroke to his back and mic it up - I am sure LAU could advise on
optimal positioning. 
4:12 a.m.
On 03/05/2010 09:28 AM, Arnold Krille wrote:
I would suggest a small one- or two cylinder engine
and some modifications to
the exhaustion pipes. Gives a really realistic engine sound and it saves you
from pedalling :-)
wow, cool. simulating the sound of a motorbike with a motorbike. that
will turn numerical physics on its head...
4:10 a.m.
On 03/04/2010 11:13 PM, Nick Copeland wrote:
This
was a fun project proposal, now it needs pulse detectors, some
accelerometers, pressure guages, a small computer, gps, 2 separate
lead acid accumultators to take care of the Omron X5MF 24V (one in
each pannier assumably), a speaker set in a rucksack and a wind vane
on a 10 meter pole.
Man, I would start suggesting that a better method would be to strap
a 2-stroke to his back and mic it up - I am sure LAU could advise on
optimal positioning.
come on, bullshitting is only fun when you take it really seriously :-D
3:54 p.m.
Karl Hammar wrote:
Fons:
I would recommend the 'bowsense' from Notam/Hans Wilmers:
http://lad.linuxaudio.org/events/2009_cdm/Friday/12_Wilmers/index.html
It can measure the acceleration (including gravitation) vector in 3D,
and can transmit it at a rate of 100 or 200 Hz via bluetooth to your PC.
There you can transform it into OSC messages to control your synth
(together with the pedal rate). And the included gyrometer might improve
the the falling detection...
Best regards,
Giso
On Thu, Mar 04, 2010 at 11:39:40AM +0000,
Jonathan Gazeley wrote:
Well, then you need an accelerometer [1]. [2] is said to be in the
iPhone, but you probably want something easier to solder.
With it you could get the inclination angle and then get the applied
force. If you do it right you could even get a crashing sound when
falling... Nick Copeland suggested simulating the force
applied to the pedals by
calculating the rate of change of the cadence. I reckon that might do
the trick.
It doesn't. Just consider going uphill 5% or downhill 5% at
the same and constant cadence. Makes some difference to your
legs and to a motor.
Regards,
/Karl
[1] http://en.wikipedia.org/wiki/Accelerometer
[2] http://www.st.com/stonline/products/literature/ds/12726/lis302dl.htm
-----------------------------------------------------------------------
Karl Hammar Aspö Data karl(a)aspodata.se
Lilla Aspö 148 Networks
S-742 94 Östhammar +46 173 140 57 Computers
Sweden +46 70 511 97 84 Consulting
-----------------------------------------------------------------------
------------------------------------------------------------------------
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
6:11 p.m.
Giso:
Fun project that also, but note, with Bluetoth they need long time to start
the connection. And they don't explain in [3], page 18, why they add
epsilon * a to the gravity estimater.
Karl Hammar wrote:
... [pedalling up/downhill]
...
Well, then you
need an accelerometer [1]. [2] is said to be in the
iPhone, but you probably want something easier to solder.
With it you could get the inclination angle and then get the applied
force. If you do it right you could even get a crashing sound when
falling...
I would recommend the 'bowsense' from Notam/Hans Wilmers:
http://lad.linuxaudio.org/events/2009_cdm/Friday/12_Wilmers/index.html It can measure the acceleration (including
gravitation) vector in 3D,
and can transmit it at a rate of 100 or 200 Hz via bluetooth to your PC.
There you can transform it into OSC messages to control your synth
(together with the pedal rate). And the included gyrometer might improve
the the falling detection...
Isn't gyrometers a little expensive?
> [1] http://en.wikipedia.org/wiki/Accelerometer
> [2] http://www.st.com/stonline/products/literature/ds/12726/lis302dl.htm
[3]
http://lad.linuxaudio.org/events/2009_cdm/Friday/12_Wilmers/bowsense-lac200…
Regards,
/Karl
-----------------------------------------------------------------------
Karl Hammar Aspö Data karl(a)aspodata.se
Lilla Aspö 148 Networks
S-742 94 Östhammar +46 173 140 57 Computers
Sweden +46 70 511 97 84 Consulting
-----------------------------------------------------------------------
17 Mar
17 Mar
9:31 p.m.
Might be handy somehow.
http://www.larsby.com/johan/e-bike/
On Sat, Mar 6, 2010 at 10:10 AM, Karl Hammar <karl(a)aspodata.se> wrote:
Giso:
Fun project that also, but note, with Bluetoth they need long time to start
the connection. And they don't explain in [3], page 18, why they add
epsilon * a to the gravity estimater.
Karl Hammar wrote:
... [pedalling
up/downhill] ...
Well,
then you need an accelerometer [1]. [2] is said to be in the
iPhone, but you probably want something easier to solder.
With it you could get the inclination angle and then get the applied
force. If you do it right you could even get a crashing sound when
falling...
I would recommend the 'bowsense' from Notam/Hans Wilmers:
http://lad.linuxaudio.org/events/2009_cdm/Friday/12_Wilmers/index.html It can measure the acceleration (including
gravitation) vector in 3D,
and can transmit it at a rate of 100 or 200 Hz via bluetooth to your PC.
There you can transform it into OSC messages to control your synth
(together with the pedal rate). And the included gyrometer might improve
the the falling detection...
Isn't gyrometers a little expensive?
> [1]
http://en.wikipedia.org/wiki/Accelerometer
> [2]
http://www.st.com/stonline/products/literature/ds/12726/lis302dl.htm
[3]
http://lad.linuxaudio.org/events/2009_cdm/Friday/12_Wilmers/bowsense-lac200…
Regards,
/Karl
-----------------------------------------------------------------------
Karl Hammar Aspö Data karl(a)aspodata.se
Lilla Aspö 148 Networks
S-742 94 Östhammar +46 173 140 57 Computers
Sweden +46 70 511 97 84 Consulting
-----------------------------------------------------------------------
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
4 Mar
4 Mar
1:14 p.m.
On Thu, 04 Mar 2010 10:27:43 +0000
Jonathan Gazeley <jonathan.gazeley(a)bristol.ac.uk> wrote:
On 04/03/10 10:20, James Morris wrote:
Instead of directly trying to measure RPM, measure milliseconds per
rev. then invert it. Once you have your first reading, then start to
average over 2/3 revs. gives you a pretty fast responding figure with a
simple setup.
Also, use a short range photocell on the spokes (say 50mm). Easy to
mount, reliable results with minimal effort. As these can usually
respond up to 10-30kHz you won't have to worry about pedalling too
fast :)
--
Will J Godfrey
http://www.musically.me.uk
Say you have a poem and I have a tune.
Exchange them and we can both have a poem, a tune, and a song.
On Thu, March 4, 2010 09:48, Jonathan Gazeley
wrote:
Yes - I have a cycle computer that can measure cadence. However, I
typically pedal between 60 and 90 rpm which means my reading would only
be available for sampling every 0.7 - 1.0s. It would work, but in the
space of one second my cadence can vary a lot if I'm accelerating so the
sound of the rising revs would be jumpy.
Also, if I suddenly stopped pedalling, it might take up to a second for
the sampler to register and kill the engine noise. I don't think it
would be a very satisfactory experience, which is why I was thinking of
mounting 10 or so magnets around the chainwheel, to be able to sample
the cadence every 0.1s. 1. A sensor that can measure my cadence. A simple
magnet switch that
triggers once a revolution won't be enough to measure the cadence with
sufficient resolution, since my cadence is usually between 50 and 80
rpm. I would probably need to mount multiple magnets spaced equally
around the chainwheel and have a single sensor on the frame. Then I have
to get it to supply this information to my control program.
Cycle computers which measure cadence work with a single magnet on the
crank and can measure cadence in that range and above. I think the fastest
RPM I ever got was (only a vague memory) around 120 (maybe?) when I was
being silly trying to see how fast I could pedal.
7:30 p.m.
On Thu, 4 Mar 2010, Folderol wrote:
Also, use a short range photocell on the spokes (say
50mm). Easy to
mount, reliable results with minimal effort. As these can usually
respond up to 10-30kHz you won't have to worry about pedalling too
fast :)
Sensing the spokes measures vehicle speed, not pedaling speed, you'd want
your synth to drop to idle while you're coasting. I'm for a photocell
watching the crank sprocket teeth in the rear interval where there's no
chain.
You also need to measure torque, or chain tension, to drive your synth's
volume. An idler with a load cell on the top of the chain between the
crank and the wheel can be used for a single-speed or three-speed in-hub
transmission, but if you have a rear derailleur, you'll need another fixed
idler between the sensor and the rear wheel. If you have a front
derailleur as well, it just gets too complicated...
Maybe the accelerometer is simpler after all, even if you have to
compensate for incline. That's just math, not hardware.
--
Rick Green
"Those who would give up essential Liberty, to purchase a little
temporary Safety, deserve neither Liberty nor Safety."
-Benjamin Franklin
"As for our common defense, we reject as false the choice between our
safety and our ideals."
-President Barack Obama 20 Jan 2009
11:52 p.m.
I don't mean to sound like a wowser or a troll, as this sounds like
kind of a cool project, but ahh, I'd say stick with the headphones.
Background sound levels in urban areas continue to increase relatively
unabated, and given the connections between stress levels and noise,
headphones may be more civic-minded. It's bad enough that people want
to start blasting synthesized engine sounds out of (thankfully)
quieter electric and hybrid vehicles.
just a thought
-m
5 Mar
5 Mar
1:41 a.m.
On Fri, 5 Mar 2010, michael noble wrote:
It's bad enough that people want
to start blasting synthesized engine sounds out of (thankfully)
quieter electric and hybrid vehicles.
Many yeaars ago, I read a science fiction story, set in a time well past
peak oil, when conservation of hydrocarbons was at the forefront of
everyone's mind, and society would heap much shame upon anyone who wasted
even a drop. The story was about a father and son who owned an antique
mid-20th century 'muscle car', and spent their every spare night or
weekend in the garage lavishing attention upon it. Once a month or so,
much to the consternation of the neighbors, they would open up the
garage, start it up, and very noisily parade it around the block. After
recounting the disgust of each neighbor they pass, the narrator follows
the pair back into their garage. As soon as they close the door
securely behind them, they open a locked cupboard, remove a large crank,
and insert it into a socket concealed behind the gas filler cap. As the
Dad begins to turn the crank, the son checks a counter concealed behind
the front grill. "I think we may have gone a bit far today. We're each
going to have to spend a full hour a day at the crank if we're to have
any hope of taking it out again next month".
--
Rick Green
"Those who would give up essential Liberty, to purchase a little
temporary Safety, deserve neither Liberty nor Safety."
-Benjamin Franklin
"As for our common defense, we reject as false the choice between our
safety and our ideals."
-President Barack Obama 20 Jan 2009
2:40 a.m.
On Fri, 2010-03-05 at 13:52 +0900, michael noble wrote:
noise,
headphones may be more civic-minded. It's bad enough that people want
to start blasting synthesized engine sounds out of (thankfully)
quieter electric and hybrid vehicles.
Loud Pipes Save Lives.
Hybrid and electric vehicles should be banned entirely from the roads,
until they have some sort of audible warning device fitted to warn
pedestrians (especially partially-sighted ones) of their approach.
There's a reason why the electric guided vehicles and forklifts in
factorys beep when they're moving.
Gordon MM0YEQ
5:13 a.m.
michael noble wrote:
I don't mean to sound like a wowser or a troll, as
this sounds like
kind of a cool project, but ahh, I'd say stick with the headphones.
Background sound levels in urban areas continue to increase relatively
unabated, and given the connections between stress levels and noise,
headphones may be more civic-minded. It's bad enough that people want
to start blasting synthesized engine sounds out of (thankfully)
quieter electric and hybrid vehicles.
just a thought
That's a fair point, and one of the reasons that I cycle to work anyway.
If I do ever get rounding to building this, I don't think I'd be likely
to use it all the time anyway - just to "impress my mates" on rare
occasions...
4 Mar
4 Mar
7:37 a.m.
Jonathan Gazeley:
...
I want to measure how fast I'm turning the pedals
on my bike (the
"cadence") and synthesize/sample the sound of an internal combustion engine.
...
Sounds
like fun.
1. A sensor that can measure my cadence. A simple
magnet switch that
triggers once a revolution won't be enough to measure the cadence with
sufficient resolution, since my cadence is usually between 50 and 80
rpm. I would probably need to mount multiple magnets spaced equally
around the chainwheel and have a single sensor on the frame. Then I have
to get it to supply this information to my control program.
...
Why don't you put an inductive sensor [1] at the pedal gear teeth?
You could probably mount it at the chain protective cover if the
tolerances are the right. Have a look at Omron "E2E-X5MF1 2M" [2,3]
(you might have to paint it to protect it from water). For that one
you need 12-24V DC.
Regards,
/Karl
[1] http://en.wikipedia.org/wiki/Inductive_sensor
[2] http://www.ia.omron.com/product/family/449/index_fea.html
[3] http://www.ia.omron.com/product/item/e2e_7258d/index.html
-----------------------------------------------------------------------
Karl Hammar Aspö Data karl(a)aspodata.se
Lilla Aspö 148 Networks
S-742 94 Östhammar +46 173 140 57 Computers
Sweden +46 70 511 97 84 Consulting
-----------------------------------------------------------------------
9:09 a.m.
On Thu, 4 Mar 2010, Karl Hammar wrote:
Why don't you put an inductive sensor [1] at the
pedal gear teeth?
You could probably mount it at the chain protective cover if the
tolerances are the right. Have a look at Omron "E2E-X5MF1 2M" [2,3]
(you might have to paint it to protect it from water). For that one
you need 12-24V DC.
This is a good idea... as long as your sensor and your uC
can keep up with the pulses. I didn't look into the sensors
that you picked, but they usually publish their response
times... so you need to check that out.
How many teeth does a typ. sprocket have? Say about 40T.
Lance Armstrong spins his cranks at 80RPM. That would be
53.3 pulses/sec (or 1 pulse every 18.75 ms).
Sometimes the sprockets have spokes or holes in them to make
them lighter. Like 6 or 8. Perhaps sensing those would
provide a more sensible signal if 18ms is too fast for your
sensor/uC.
-gabriel
8:10 a.m.
On Thu, Mar 4, 2010 at 4:48 AM, Jonathan Gazeley
<jonathan.gazeley(a)bristol.ac.uk> wrote:
http://www.patentstorm.us/patents/6394875/description.html
this is an available product, too. its not based on cadence, just wheel rpm's.
a google for "bicycle spokes rear wheel sound engine simulated" gets
quite a few similar hits.
--p
8:15 a.m.
On 04/03/10 13:10, Paul Davis wrote:
http://www.patentstorm.us/patents/6394875/description.html
this is an available product, too. its not based on cadence, just wheel rpm's.
Ah, that's no good. Makes it sound like I've only got one gear!
I did think about shoving a tin can in my wheel but decided it would be
more entertaining to build a sensor circuit, write a control program and
mess with a synthesiser. Perhaps.
Cheers,
Jonathan
a google for "bicycle spokes rear wheel sound
engine simulated" gets
quite a few similar hits.
8:56 a.m.
On Thu, 4 Mar 2010, Jonathan Gazeley wrote:
I want to measure how fast I'm turning the pedals
on my bike (the
"cadence") and synthesize/sample the sound of an internal combustion engine.
Been there, done that. Forty years ago: Attach a common playing card
to your front fork with a clothespin, so that the end of the card is
struck by the passing spokes. Vrooom!
BTW: The human body is considered capable of producing 1/20hp, or 35
watts on a sustained basis.
--
Rick Green
"Those who would give up essential Liberty, to purchase a little
temporary Safety, deserve neither Liberty nor Safety."
-Benjamin Franklin
"As for our common defense, we reject as false the choice between our
safety and our ideals."
-President Barack Obama 20 Jan 2009
10 a.m.
On Thu, Mar 4, 2010 at 8:45 AM, Rick Green <rtg(a)aapsc.com> wrote:
BTW: The human body is considered capable of
producing 1/20hp, or 35
watts on a sustained basis.
hmm. when i'm trained up, i can generate about 240W for an hour, and
at that point in the year, 100W is basically a "do nothing" state.
world-class Ironman racers and crazy 100 mile time-trial cyclists
generate 250-300W for 3.5-5 hrs. so your number must reflect
an untrained, rather unfit human being or must mean something other
than what i understand it to mean.
--p
6:13 p.m.
On Thu, 4 Mar 2010, Paul Davis wrote:
On Thu, Mar 4, 2010 at 8:45 AM, Rick Green
<rtg(a)aapsc.com> wrote:
I think that figure goes back to the days of farming with draft animals,
as in "One horse can do the work of 20 men". So 'sustained' in this
case
means 'every daylight hour in a six-day work week'.
--
Rick Green
"Those who would give up essential Liberty, to purchase a little
temporary Safety, deserve neither Liberty nor Safety."
-Benjamin Franklin
"As for our common defense, we reject as false the choice between our
safety and our ideals."
-President Barack Obama 20 Jan 2009
BTW: The human body is considered capable of
producing 1/20hp, or 35
watts on a sustained basis.
hmm. when i'm trained up, i can generate about 240W for an hour, and
at that point in the year, 100W is basically a "do nothing" state.
world-class Ironman racers and crazy 100 mile time-trial cyclists
generate 250-300W for 3.5-5 hrs. so your number must reflect
an untrained, rather unfit human being or must mean something other
than what i understand it to mean.
10:04 a.m.
1. A sensor that can measure my cadence. A simple
magnet switch that
triggers once a revolution won't be enough to measure the cadence with
sufficient resolution, since my cadence is usually between 50 and 80
rpm. I would probably need to mount multiple magnets spaced equally
around the chainwheel and have a single sensor on the frame. Then I have
to get it to supply this information to my control program.
Once per revolution is okay. b/c what you've really asked for it not
that the estimation be accurate to some threshold throughout the
cycle, but that the output be smooth and continuous.
An F/V converter takes pulses, irregularly spaced, and creates
identical pulses of current. Those pulses get integrated across an RC
network. It's not accurate during the time between pulses, but it
doesn't have to be. It just has to have smooth response when the
pulses start to get faster or slower. You choose a RC time constant
that is long compared to the slowest cycle you want to use and is
small compared to the time over which the pulses
accelerate/decelerate.
Chuck
10:34 a.m.
On Thu, Mar 4, 2010 at 9:04 AM, Charles Henry <czhenry(a)gmail.com> wrote:
Let me also add that multiple magnets would definitely *help* the
process, if an F/V method were used. My point was simply that any
rate of discrete events can be smoothly counted, if you don't care
about the accuracy between those events. If you really need
resolution (like with control systems), go with a rotary encoder.
Chuck
1. A sensor
that can measure my cadence. A simple magnet switch that
triggers once a revolution won't be enough to measure the cadence with
sufficient resolution, since my cadence is usually between 50 and 80
rpm. I would probably need to mount multiple magnets spaced equally
around the chainwheel and have a single sensor on the frame. Then I have
to get it to supply this information to my control program.
Once per revolution is okay. b/c what you've really asked for it not
that the estimation be accurate to some threshold throughout the
cycle, but that the output be smooth and continuous. 
1:35 p.m.
Jonathan Gazeley wrote:
Morning LAU,
As I cycled to work this morning, I had a crazy idea. It's just
daydreaming and will probably never happen, but I wondered if anyone on
the last has any useful/interesting thoughts.
I want to measure how fast I'm turning the pedals on my bike (the
"cadence") and synthesize/sample the sound of an internal combustion engine.
What you really need is a bike powered sound system on a trailer to make
the sound:
http://web.ukonline.co.uk/baka/rinky/index2.htm
And while you're at it, why does the sound have to sound like a normal
engine? A spaceship sound effect from a sci fi film would be cool.
andy
5 Mar
5 Mar
2:15 a.m.
You should check out the S.F. Bicycle Music Festival
http://bicyclemusicfestival.com/
They cycled from San Francisco all the way to Vancouver, B.C. with
bands and a bicycle powered PA system. I caught them here in Seattle
and helped broadcast a singer in a cargo trike during a parade over my
bike sound system via FM transmitter. They set up 3 stationary bikes
attached to generators with ultracaps to power a PA system for shows
in parks and wherever.
I didn't have my 12" JBL subwoofer bike system with me at the time,
but it requires a lot more power than my legs can put out :)
--Joseph
On Thu, Mar 4, 2010 at 1:48 AM, Jonathan Gazeley
<jonathan.gazeley(a)bristol.ac.uk> wrote:
Morning LAU,
As I cycled to work this morning, I had a crazy idea. It's just
daydreaming and will probably never happen, but I wondered if anyone on
the last has any useful/interesting thoughts.
I want to measure how fast I'm turning the pedals on my bike (the
"cadence") and synthesize/sample the sound of an internal combustion engine.
As far as I can work out, there are three major parts to this.
1. A sensor that can measure my cadence. A simple magnet switch that
triggers once a revolution won't be enough to measure the cadence with
sufficient resolution, since my cadence is usually between 50 and 80
rpm. I would probably need to mount multiple magnets spaced equally
around the chainwheel and have a single sensor on the frame. Then I have
to get it to supply this information to my control program.
2. I need a control program that can read in the input from my cadence
sensor and convert a cadence reading of "66 rpm" into a frequency that
should be sampled/synthesised, e.g. "500 Hz" (I'm making these numbers
up). It will also need to be able to somehow smooth out the readings,
perhaps by interpolation, so when I accelerate, the sound of the revs
climbing doesn't increase in obvious steps. It could also have other
logic, e.g. when my cadence is 0 rpm, the sound of the engine is idling
rather than off.
3. I need a synthesiser or sampler that can take an input from my
control program and make the sound of an engine (or more likely, a sine
wave to start with). I've never sampled or synthesised on a computer
before but this engine-specific sampling technology already exists in
video games, such as torcs[1].
I have absolutely no idea why I would want such a device - just for the
fun of building it, I guess. I would like it to work in realtime (rather
than later generating the soundtrack from recorded cadence data). The
thought of sitting at the traffic lights with my earphones in and then
hearing the mighty roar of a V8 as I pull away would be really satisfying...
[1] http://torcs.sourceforge.net/
Any thoughts - useful, interesting, humorous, or otherwise - are welcome!
Cheers,
Jonathan
----------------------------
Jonathan Gazeley
Systems Support Specialist
ResNet | Wireless& VPN Team
Information Services
University of Bristol
----------------------------
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user(a)lists.linuxaudio.org
http://lists.linuxaudio.org/listinfo/linux-audio-user
5185
days inactive
5198
days old
linux-audio-user@lists.linuxaudio.org
36 comments
19 participants
participants (19)
-
andy baxter -
Arnold Krille -
Charles Henry -
Folderol -
fons＠kokkinizita.net -
Gabriel M. Beddingfield -
Giso Grimm -
Gordon JC Pearce -
James Morris -
Jonathan Gazeley -
Joseph Sheedy -
Julien Claassen -
Jörn Nettingsmeier -
karl＠aspodata.se -
michael noble -
Nick Copeland -
Paul Davis -
Rick Green -
tdrecordable