Forums

Synthesizing Engine Sounds

Started by Tim Wescott July 30, 2013
I don't think engine sounds are purely periodic. If you synthesize a series of sine waves locked to the tach rate it just sounds like a tinny little buzzing sound. 
radams2000@gmail.com writes:

> I don't think engine sounds are purely periodic. If you synthesize a > series of sine waves locked to the tach rate it just sounds like a > tinny little buzzing sound.
Wouldn't that be a function of the lowest-frequency sine wave in the series? Surely if you just keep an engine at a constant RPM its sound is periodic. -- Randy Yates Digital Signal Labs http://www.digitalsignallabs.com
On Tue, 30 Jul 2013 23:44:33 -0400, Randy Yates wrote:

> radams2000@gmail.com writes: > >> I don't think engine sounds are purely periodic. If you synthesize a >> series of sine waves locked to the tach rate it just sounds like a >> tinny little buzzing sound. > > Wouldn't that be a function of the lowest-frequency sine wave in the > series? > > Surely if you just keep an engine at a constant RPM its sound is > periodic.
I doubt that. There's randomness in the gasses whooshing out the exhaust valves and into the intakes, and if your want to get really accurate there'll be various clatters and clanks from the valve train. But I sincerely hope that the whooshing can be simulated by periodically changing the magnitude of a random wave; I most surely do! -- Tim Wescott Control system and signal processing consulting www.wescottdesign.com
On Tue, 30 Jul 2013 22:53:20 -0500, Tim Wescott
<tim@seemywebsite.please> wrote:

>On Tue, 30 Jul 2013 23:44:33 -0400, Randy Yates wrote: > >> radams2000@gmail.com writes: >> >>> I don't think engine sounds are purely periodic. If you synthesize a >>> series of sine waves locked to the tach rate it just sounds like a >>> tinny little buzzing sound. >> >> Wouldn't that be a function of the lowest-frequency sine wave in the >> series? >> >> Surely if you just keep an engine at a constant RPM its sound is >> periodic. > >I doubt that. There's randomness in the gasses whooshing out the exhaust >valves and into the intakes, and if your want to get really accurate >there'll be various clatters and clanks from the valve train. > >But I sincerely hope that the whooshing can be simulated by periodically >changing the magnitude of a random wave; I most surely do! > >-- >Tim Wescott >Control system and signal processing consulting >www.wescottdesign.com
I think what makes this hard is that you don't want to use a speaker or other audio transducer. Not sure what exactly you want to modulate to get the desired sound unless you really do use something like an updated card-in-the-spokes, like this: http://www.turbospoke.com/ It actually works pretty well: http://www.youtube.com/watch?v=iWl-aKMRq40 Not sure how you'd fit that on an electric model airplane, though, especially without loading it down in undesirable ways. Eric Jacobsen Anchor Hill Communications http://www.anchorhill.com
Even if you ignore clatters and random components, each individual cylinder does not produce the same magnitude or spectrum as any other cylinder, so there are components at all multiples of the single-cylinder firing rate as well as all multiples the fundamental rpm rate. This has been well-studied recently because electric cars need to make fake external engine noises so that innocent pedestrians are not run over. So you want your Prius to sound like a Porshe. Also the video game industry has gotten pretty sophisticated about making engine noises. I have seen a few papers on this and they all start with "periodic waveforms are not nearly good enough".  There was an AES some years back with an entire track devoted to audio for games. You can probably dig up the preprints.
I think the computer games industry has done a lot on the topic. It seems
to be a surprisingly difficult problem. 

PS I remember one job advertisement I saw in the university's electronics
department, from Harley Davidson, on exhaust research. At first glance, it
looked more like they were into musical instruments...
	 

_____________________________		
Posted through www.DSPRelated.com
Tim Wescott <tim@seemywebsite.please> writes:

> On Tue, 30 Jul 2013 23:44:33 -0400, Randy Yates wrote: > >> radams2000@gmail.com writes: >> >>> I don't think engine sounds are purely periodic. If you synthesize a >>> series of sine waves locked to the tach rate it just sounds like a >>> tinny little buzzing sound. >> >> Wouldn't that be a function of the lowest-frequency sine wave in the >> series? >> >> Surely if you just keep an engine at a constant RPM its sound is >> periodic. > > I doubt that. There's randomness in the gasses whooshing out the exhaust > valves and into the intakes, and if your want to get really accurate > there'll be various clatters and clanks from the valve train.
(Note to self: go sample some YouTube audio of an engine, loop it, and see if it doesn't damn well sound like an engine). -- Randy Yates Digital Signal Labs http://www.digitalsignallabs.com
Tim Wescott <tim@seemywebsite.please> writes:

> On Tue, 30 Jul 2013 23:44:33 -0400, Randy Yates wrote: > >> radams2000@gmail.com writes: >> >>> I don't think engine sounds are purely periodic. If you synthesize a >>> series of sine waves locked to the tach rate it just sounds like a >>> tinny little buzzing sound. >> >> Wouldn't that be a function of the lowest-frequency sine wave in the >> series? >> >> Surely if you just keep an engine at a constant RPM its sound is >> periodic. > > I doubt that. There's randomness in the gasses whooshing out the exhaust > valves and into the intakes, and if your want to get really accurate > there'll be various clatters and clanks from the valve train.
Yes, I don't disagree that in reality there will be minor variations, and of course those variations will not be representable by a periodic waveform. But I bet you can get damn close - at least for a constant RPM. -- Randy Yates Digital Signal Labs http://www.digitalsignallabs.com
Agreed that constant rpm will be closer to periodic, not sure how close. It also occurs to me that anything driven by a belt from the main shaft will have a frequency ratio given by the relative pulley sizes which is probably non-integer, so there will be some non-harmonic components related to the alternator, AC compressor, power steering pump, etc. 

Bob
On Wed, 31 Jul 2013 04:32:29 +0000, Eric Jacobsen wrote:

> On Tue, 30 Jul 2013 22:53:20 -0500, Tim Wescott > <tim@seemywebsite.please> wrote: > >>On Tue, 30 Jul 2013 23:44:33 -0400, Randy Yates wrote: >> >>> radams2000@gmail.com writes: >>> >>>> I don't think engine sounds are purely periodic. If you synthesize a >>>> series of sine waves locked to the tach rate it just sounds like a >>>> tinny little buzzing sound. >>> >>> Wouldn't that be a function of the lowest-frequency sine wave in the >>> series? >>> >>> Surely if you just keep an engine at a constant RPM its sound is >>> periodic. >> >>I doubt that. There's randomness in the gasses whooshing out the >>exhaust valves and into the intakes, and if your want to get really >>accurate there'll be various clatters and clanks from the valve train. >> >>But I sincerely hope that the whooshing can be simulated by periodically >>changing the magnitude of a random wave; I most surely do! >> >>-- >>Tim Wescott Control system and signal processing consulting >>www.wescottdesign.com > > I think what makes this hard is that you don't want to use a speaker or > other audio transducer. Not sure what exactly you want to modulate to > get the desired sound unless you really do use something like an updated > card-in-the-spokes, like this: > > http://www.turbospoke.com/ > > It actually works pretty well: > > http://www.youtube.com/watch?v=iWl-aKMRq40 > > Not sure how you'd fit that on an electric model airplane, though, > especially without loading it down in undesirable ways.
Current electronic motor drives for model airplanes beep, by driving a square wave to the motor, and using the propeller and the frame of the plane to couple the sound to the air. My plan is to see what kind of quality I can get out of that. I also plan on getting an engine starting sound, then have it fade out as the speed goes up. And, as I said at the head of the thread, there may just be too many barriers. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com