GPL: A "musical" note.

Hi! When I tried Patrick's replacement engine sounds for GPL, I noticed, that they have different pitch from the originals. So I did some experimenting and found that neither Patrick's nor the original sounds have pitch, which matches with the rpm in the game. If you want to verify this, switch to chase view so you can see rpm, switch to neutral and rev up to 6600. 6600 rpm is 110 Hz, which is the lowest "A" you can sing or the second lowest string in guitar. If the engine sound was correctly in tune, you would now hear 110 Hz and/or some of it's multipliers. Another good rpm to try is 8810, which is 147 Hz = D = the third lowest string in guitar. I found that some Papyrus sounds have 220 Hz at 6600 rpm, but Patrick's Lotus/Eagle sound was closest to what I would expect - only a bit too high. Actually I am not quite sure about that multiplier thing. Engines can produce both harmonic and inharmonic overtones. However, I would wxpect some "musical" relationship between the rpm and the engine sound. Sorry about a bit pointless post. I have nothing better to do today. Jari Jokinen

Also bear in mind that the number of cylinders will affect the note at any given RPM... You can't just divide RPM by 60 and derive the pitch of the exhaust sound. Trips

Quoted text (28 lines)

> Hi! > When I tried Patrick's replacement engine sounds for GPL, > I noticed, that they have different pitch from the originals. > So I did some experimenting and found that neither Patrick's > nor the original sounds have pitch, which matches with the > rpm in the game. > If you want to verify this, switch to chase view so you can > see rpm, switch to neutral and rev up to 6600. 6600 rpm is > 110 Hz, which is the lowest "A" you can sing or the second > lowest string in guitar. If the engine sound was correctly in > tune, you would now hear 110 Hz and/or some of it's > multipliers. Another good rpm to try is 8810, which is > 147 Hz = D = the third lowest string in guitar. > I found that some Papyrus sounds have 220 Hz at 6600 rpm, > but Patrick's Lotus/Eagle sound was closest to what I would > expect - only a bit too high. > Actually I am not quite sure about that multiplier thing. Engines > can produce both harmonic and inharmonic overtones. However, > I would wxpect some "musical" relationship between the rpm and > the engine sound. > Sorry about a bit pointless post. I have nothing better to do > today. > Jari Jokinen

>Also bear in mind that the number of cylinders will affect the note at >any given RPM... As would the length of the exhaust system (manifold and headers) itself, as in any wind instrument with a tone generator at one end? Trey Behan

You even further must remember that you can change the sampling RPM from inside the app text file so if you REALLY want the pitch to be perfect, just change that number until it fits right and you have a perfectly tuned, musically sound, engine. hehehehe Jesse | |>Also bear in mind that the number of cylinders will affect the note at |>any given RPM... | |As would the length of the exhaust system (manifold and headers) itself, as in |any wind instrument with a tone generator at one end? | |Trey Behan

I didn't know it was that easy! Anyway the tuned WAV files are now posted at alt.binaries.autos.simulators for musically sensitive drivers to try. >You even further must remember that you can change the sampling RPM from >inside the app text file so if you REALLY want the pitch to be perfect, just >change that number until it fits right and you have a perfectly tuned, >musically sound, engine. >hehehehe >Jesse

I have listened two Ferrari samples and there seems to be a very strong component at frequency rpm x 4.5. That's 9th harmonic for rpm divided by two and makes two octaves plus a major second as a musical interval.

Quoted text (8 lines)

>Not true. These are 4 stroke engines. Each cylinder fires >only once every two revolutions. Also, the pitch is definitely >affected by numbers of cylinders and other factors. Compare >the sound of a real-life Ferrari F1 and almost anything else >at the same RPM....... :) >-- >// rrevved posts from mindspring dot com

As an example of what is being discussed here, I set up a web page with a chart showing the "Power Spectral Density" of a Ducati motorcycle sound recording. The chart has the frequency scaled by the RPM of the sample so you can easily see just how much of the sound's power is concentrated in the frequency ranges of half per rev, one per rev, one and a half per rev, two per rev, etc. http://www.motorsims.com/users/mjessick/sample_psd.htm - Matt

Quoted text (16 lines)

> I have listened two Ferrari samples and there seems to be a very > strong component at frequency rpm x 4.5. That's 9th harmonic > for rpm divided by two and makes two octaves plus a major second > as a musical interval. > >Not true. These are 4 stroke engines. Each cylinder fires > >only once every two revolutions. Also, the pitch is definitely > >affected by numbers of cylinders and other factors. Compare > >the sound of a real-life Ferrari F1 and almost anything else > >at the same RPM....... :) > >-- > >// rrevved posts from mindspring dot com

-- Matthew V. Jessick Motorsims Vehicle Dynamics Engineer (972)910-8866, Fax: (972)910-8216

That was interesting. The base frequency of engine tone should thus be rpm divided by two. My ears tell me, that of Papyrus' original sounds only Lotus and Eagle match this criteria. Brabham, Murasama adn Coventry are a minor third too high. BRM is a perfect fourth too low and Ferrari is a major third too low. The Papyrus Ferrari has a *** harmonic at 10 x base frequency, which can confuse the ear. It also sounds very bright, if transposed upwards to the correct (?) pitch. A bit of filtration propably helps. I will verify the frequencies, when I find a program, that can do frequency analysis, but I am pretty sure of what I hear. Jari Jokinen

Quoted text (14 lines)

>As an example of what is being discussed here, >I set up a web page with a chart showing the >"Power Spectral Density" of a Ducati motorcycle >sound recording. >The chart has the frequency scaled >by the RPM of the sample so you can easily see >just how much of the sound's power is concentrated >in the frequency ranges of half per rev, one per rev, >one and a half per rev, two per rev, etc. >[link lost in import] >- Matt

As a user of your 'musical' engine sounds, I am looking forward to hearing any developments from you guys. Mikey

I did just post some "new" engine sounds to alt.binaries.simulators.autos and alt.binaries.misc. I think the engines play nicely together now :) Jari Jokinen >As a user of your 'musical' engine sounds, I am looking forward to hearing >any developments from you guys. >Mikey