Engine Sounds - Page 2 - ED Forums
 


Notices

Reply
 
Thread Tools Display Modes
Old 10-10-2019, 08:20 AM   #11
philstyle
Senior Member
 
philstyle's Avatar
 
Join Date: Jan 2017
Location: The Aeotearoan Republic of Niederbayern
Posts: 1,142
Default

Quote:
Originally Posted by hazzer View Post
Cleary you haven't. None of them sound anything like the real life counterparts I'm afraid.
I have seen many warbirds and piston aircraft and none sound like they do in DCS.

DCS (and many sims) problem with audio is not always that the raw sound files are bad, it's often due to the way the totality of sound effects end up getting mixed together.
In many cases, the -0DB limit is used for all audio files (I am not saying this is the case with DCS), so everything is effectively normalised. Äunder such a scenario, a 1 second wav file sound of a car door closing has the same peak as a 1 second explosion sound. And once a file hits the limit, there is nowhere to go when it comes to increasing the gain - becasue digital distortion will occur.

In real life, this digital max dB limit for sound doesn't exist. If one noise source has more sound-power than another it just does . . . there'as no scaling required.

Furthermore, games have real trouble with attenuation and refelction of sound. Most don't even bother to model reflective sound at all - it would take as much computational power as ray tracing for light (more potentially if you take stereophonics into account).
Attenuation is (generally) proportional to the square of sound frequency, so different tones attentuate differently . . . some stuff "sounds louder" at distance, even though it comes from a source with the same sound power.
Games alsmot never replicate this - the simply just drop the overall gain of a file's playback with distance from the source. The result is a muchless dynamic sound profile over distance - unlike in real life.

In addition, even if a game did have a high-fidelity recording of an engine from real life to use, the software would still need to do the following
1. Adjust the frequeny of the sound file to account for all the miniscule RPM variations
2. Find a way to distinguish between airborn and structure borne noise and vibreation, and then reproduce the relevant audible tones
3. Attenuate/ interefere alongside all the other sounds in the environment
Each of these modifications will make the sound more synthetic, adding in artifacts and removing fideilty.

My suggestion is that the best way to "fix" audio for a flight sim is to adjust the sliders in game to heavily prejudice the engines (fortunately DCS is quite customisable like this) and maybe to get a decent ASIO sound card and high-quliaty headphones or monitor speakers. Adding an EQ to your audio output can work wonders too.
__________________
On YouTube @:https://www.youtube.com/philstylenz

Current: AMD 3700x, GTX 1080Ti, 32GB 3200 RAM, Auros Pro X570, 2TB M.2 drive
Previous build: i5-6600k, GTX 1080Ti, 16BG 2666 RAM, ASUS Maximus XIII Ranger Mobo
philstyle is offline   Reply With Quote
Old 10-10-2019, 08:23 AM   #12
grafspee
Senior Member
 
Join Date: Nov 2017
Location: Poland
Posts: 1,326
Default

Quote:
Originally Posted by Ala13_ManOWar View Post
Well, that wasn't exactly "copied", that was produced as M-25 under license (paying fees and everything, just with metric stuff) and later on M-25 was developed into M-62/3 and further.


S!
Yes, you are right. But even this dont bring ash82 any closer to bmw801 BMW 801 develops like twice the power

Last edited by grafspee; 10-10-2019 at 08:35 AM.
grafspee is online now   Reply With Quote
Old 10-10-2019, 09:02 AM   #13
OttoPus
Member
 
Join Date: Aug 2012
Location: Sardinia, Italy
Posts: 522
Default

Quote:
Originally Posted by philstyle View Post
DCS (and many sims) problem with audio is not always that the raw sound files are bad, it's often due to the way the totality of sound effects end up getting mixed together.
In many cases, the -0DB limit is used for all audio files (I am not saying this is the case with DCS), so everything is effectively normalised. Äunder such a scenario, a 1 second wav file sound of a car door closing has the same peak as a 1 second explosion sound. And once a file hits the limit, there is nowhere to go when it comes to increasing the gain - becasue digital distortion will occur.

In real life, this digital max dB limit for sound doesn't exist. If one noise source has more sound-power than another it just does . . . there'as no scaling required.

Furthermore, games have real trouble with attenuation and refelction of sound. Most don't even bother to model reflective sound at all - it would take as much computational power as ray tracing for light (more potentially if you take stereophonics into account).
Attenuation is (generally) proportional to the square of sound frequency, so different tones attentuate differently . . . some stuff "sounds louder" at distance, even though it comes from a source with the same sound power.
Games alsmot never replicate this - the simply just drop the overall gain of a file's playback with distance from the source. The result is a muchless dynamic sound profile over distance - unlike in real life.

In addition, even if a game did have a high-fidelity recording of an engine from real life to use, the software would still need to do the following
1. Adjust the frequeny of the sound file to account for all the miniscule RPM variations
2. Find a way to distinguish between airborn and structure borne noise and vibreation, and then reproduce the relevant audible tones
3. Attenuate/ interefere alongside all the other sounds in the environment
Each of these modifications will make the sound more synthetic, adding in artifacts and removing fideilty.

My suggestion is that the best way to "fix" audio for a flight sim is to adjust the sliders in game to heavily prejudice the engines (fortunately DCS is quite customisable like this) and maybe to get a decent ASIO sound card and high-quliaty headphones or monitor speakers. Adding an EQ to your audio output can work wonders too.
Great analysis.

If I may add:
  1. Human hearing follows very specific and mathematically described rules.

    These rules led to the HRTF (Head Related Transfer Function). If a 3D sound environment implements this modeling, and uses binaural recordings, everything comes to life using a pair of headphones.

    You can compare this to "VR for sound" in a certain way.

  2. Every solid object traveling through air generates a typical sound (like a flanger on pink noise) and the "breaking" of the air around the object introduces a modulation of the original sound source (like a phaser that changes speed consonant to the doppler effect).

    This effect summed to the doppler effect and the distant reflections, modeled through HRTF, gives the sounds that we hear in real life when an airplane (or a car, or an helicopter) passes by.
OttoPus is offline   Reply With Quote
Old 10-10-2019, 09:58 AM   #14
hazzer
Member
 
Join Date: Jan 2014
Posts: 440
Default

Quote:
Originally Posted by grafspee View Post
So you saying that you have heard bf190 k4 and fw190 D9 ??
ED claims that sounds are from real recordings when this is possible.
And i will never accept sounds recorded by some crapy things like videos on youtube.
I was talking in general about the warbirds. I have heard many P-51s, Spitfires, P-47s. Neither the Mustang or the Spitfire sound like the real life counterparts; and yes hearing a high quality sound recording from a video with mustang's and a bf109 I would say it is pretty accurate.
__________________
Asus GTX1080ti,I7-7800k, 32gb ram, SSD, Samsung Odyssey VR, MSFFB2, T-50 Throttle, Thrustmaster Rudder Pedals
hazzer is offline   Reply With Quote
Old 10-10-2019, 10:39 AM   #15
Ala13_ManOWar
Senior Member
 
Ala13_ManOWar's Avatar
 
Join Date: Apr 2010
Location: Spain
Posts: 1,960
Default

Quote:
Originally Posted by hazzer View Post
Cleary you haven't. None of them sound anything like the real life counterparts I'm afraid.

I have seen many warbirds and piston aircraft and none sound like they do in DCS.
Maybe my four trips to Duxford didn't happen, but last time I just wondered being there how close everything sounded to the sim, fly bys, aircraft taxiing, take offs, start-ups, everything. If you look for feeling the roar in your chest, the smell and everything that's not gonna happen ever neither in DCS or any other game/audio setting.


S!
__________________
"I went into the British Army believing that if you want peace you must prepare for war. I believe now that if you prepare for war, you get war." -- Major-General Frederick B. Maurice
Ala13_ManOWar is offline   Reply With Quote
Old 10-10-2019, 10:50 AM   #16
Ala13_ManOWar
Senior Member
 
Ala13_ManOWar's Avatar
 
Join Date: Apr 2010
Location: Spain
Posts: 1,960
Default

Quote:
Originally Posted by hazzer View Post
... and yes hearing a high quality sound recording from a video with mustang's and a bf109 I would say it is pretty accurate.
Well, most of the time I wouldn't, videos usually never depict the thing like it was, and YT is plagued with cellphone videos which tend to be really bad in audio.


Anyway as we know indeed (said several times here) DCS is recorded at Duxford place from real aircraft, one of the few things I think we can do is getting better audio cards and audio devices to honour the original sound quality.



S!
__________________
"I went into the British Army believing that if you want peace you must prepare for war. I believe now that if you prepare for war, you get war." -- Major-General Frederick B. Maurice
Ala13_ManOWar is offline   Reply With Quote
Old 10-10-2019, 11:17 AM   #17
DD_Fenrir
Member
 
Join Date: Jun 2011
Posts: 776
Default

Quote:
Originally Posted by hazzer View Post
I was talking in general about the warbirds. I have heard many P-51s, Spitfires, P-47s. Neither the Mustang or the Spitfire sound like the real life counterparts; and yes hearing a high quality sound recording from a video with mustang's and a bf109 I would say it is pretty accurate.
Having flown a two-seater Spit, I can tell you that the internal sounds are almost spot-on, for the Spit at least.

Externally, they could be better but they have improved even in the last year. Just last night sat on dispersal watching a buddy takeoff in a P-51 and it sounded recognisably like a Mustang taking off; not exactly, but the characteristics were damn close.

Flybys are still lacking that certain something but as alluded to earlier by others here, the complex nature of varying attenuation of sound levels depending on their frequency and Doppler effects are always going to be a challenge.

Also bear in mind you are not listening to pre-recorded samples, but a sound engine that attempts to synthesise the sounds - Christ, the sheer number of samples required to try and match every combination of Manifold Pressure and RPM alone would be staggering and unwieldy in terms of file size.

Add to that the following problems:

a) most of the engines these days are rarely run anywhere over max continuous cos they're kinda expensive, you know. So how do you get a sample set of an engine running full bore at WEP if no-one's gonna let you operate their priceless pride and joy at that power regime?

b) so how do you propose to record the external sound of a Mustang/Spitfire/109/*insert warbird name here* doing 400mph? Where are you gonna put the microphone so you don't get 400mph of wind noise? Even 10mph of wind noise can ruin a recording. Recording from the ground gives great flyby samples but how useful is that? How do you tell at which point in the sample represents the 0 Doppler moment? Is that snippet actually long enough to make a meaningful sample that can be looped? How far away is the aircraft at that point? What frequency losses might be incurred? Will it sound right still at a camera viewpoint 5m from the aircraft? What about 5km?

Getting this right takes a huge amount of time, resources and knowledge and is an evolving process.
DD_Fenrir is offline   Reply With Quote
Old 10-10-2019, 11:46 AM   #18
grafspee
Senior Member
 
Join Date: Nov 2017
Location: Poland
Posts: 1,326
Default

Quote:
Originally Posted by DD_Fenrir View Post
Having flown a two-seater Spit, I can tell you that the internal sounds are almost spot-on, for the Spit at least.

Externally, they could be better but they have improved even in the last year. Just last night sat on dispersal watching a buddy takeoff in a P-51 and it sounded recognisably like a Mustang taking off; not exactly, but the characteristics were damn close.

Flybys are still lacking that certain something but as alluded to earlier by others here, the complex nature of varying attenuation of sound levels depending on their frequency and Doppler effects are always going to be a challenge.

Also bear in mind you are not listening to pre-recorded samples, but a sound engine that attempts to synthesise the sounds - Christ, the sheer number of samples required to try and match every combination of Manifold Pressure and RPM alone would be staggering and unwieldy in terms of file size.

Add to that the following problems:

a) most of the engines these days are rarely run anywhere over max continuous cos they're kinda expensive, you know. So how do you get a sample set of an engine running full bore at WEP if no-one's gonna let you operate their priceless pride and joy at that power regime?

b) so how do you propose to record the external sound of a Mustang/Spitfire/109/*insert warbird name here* doing 400mph? Where are you gonna put the microphone so you don't get 400mph of wind noise? Even 10mph of wind noise can ruin a recording. Recording from the ground gives great flyby samples but how useful is that? How do you tell at which point in the sample represents the 0 Doppler moment? Is that snippet actually long enough to make a meaningful sample that can be looped? How far away is the aircraft at that point? What frequency losses might be incurred? Will it sound right still at a camera viewpoint 5m from the aircraft? What about 5km?

Getting this right takes a huge amount of time, resources and knowledge and is an evolving process.
I totally agree with you, i will add that DCS actually has the best flyby sounds, comparing this t other sims DCS sound are much better.
grafspee is online now   Reply With Quote
Old 10-10-2019, 12:44 PM   #19
btd
ED Team
 
btd's Avatar
 
Join Date: Feb 2011
Location: Moscow
Posts: 190
Default

Many thanks DD_Fenrir for the detailed answer. As a sound designer - I fully agree to your every word)
Yes, flybys sound recorded from the ground is useless for us. We should re-create it by many sounds and sound engine and a lot of code. It is not too simple.
But we do not stand still. Each new module released sounds better than the previous one. And I'm sure you will really like the sound of the P47 from a series of WW2-birds. There we were able to improve the sound scheme specifically for flybys.
Some modules were released many years ago. Therefore, their level is slightly behind modern. We will upgrade their sound necessarily as soon as free time appears.

I really don’t know a simulator where flybys would sound better than ours. Especially in jets.
__________________
Best regards,
Kanstantsin Kuzniatsou (btd)
https://soundcloud.com/btd

Last edited by btd; 10-10-2019 at 12:48 PM.
btd is offline   Reply With Quote
Old 10-10-2019, 01:29 PM   #20
grafspee
Senior Member
 
Join Date: Nov 2017
Location: Poland
Posts: 1,326
Default

Quote:
Originally Posted by btd View Post
I really don’t know a simulator where flybys would sound better than ours. Especially in jets.
Agree, flyby sounds are the best of all sims i've played.

Last edited by grafspee; 10-10-2019 at 01:32 PM.
grafspee is online now   Reply With Quote
Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

All times are GMT. The time now is 05:37 PM. vBulletin Skin by ForumMonkeys. Powered by vBulletin®.
Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.