For me personally, sounds (especially engine sounds) in sims are more important than graphics. Just listen...it's the best music in the world (ok, Pink Floyds are still better ) That's why I love DRM (1 & 2) mod for rf 1.
If you do decide to give it a whirl, just remember: I'm purely talking about conveying information regarding engine load. The sounds themselves are very fake, electronic and synth-y. There are no 'bells and whistles,' so-to-speak. It's a complete trade-off, unfortunately. I didn't know that the GT series used this technique (though, it makes total sense from a file-size standpoint, as you said) but I've always found their sounds lacking in impact as well. I was under the impression that RBR was some kind of mix of the two systems, though I've never really delved into it.
I just read paper describing grip and friction design for highway designers, tire sound is because of macro texture, so perfectly glass smooth surface and slick tire should be rather silent, but also grooved tire should be almost silent on such surface. It was really good read in many other reasons too, road noise was really small part of that, I put link to tgm + ttool thread. Netkar pro also had synthetic sound system at some point, can't remember if there still is possibility to choose sample sound engine or synth, it was there at the time when software was not as refined as it is today so many probably forgot that feature. Even that in this video engine sound is rather loudly captured, there are lot of other sounds too, also great reference of how bumpy race tracks really are: This is what I meant with wind sounds:
Yes, I was hoping someone would post up a great GT on-board video. Love this. Thank you. Let's actually think about what we hear in this clip, though. First and foremost, let's think about how we are hearing this clip. The first issue is, unavoidably, Youtube's compression. There's just no way around it. Though for as loud of a source as this is, the actual effect is rather minimal. Second is the source recording. The sound is coming from what I assume is a very tiny microphone that lives inside of the camera the video came from. The quality and SPL handling of this microphone actually has a great deal of bearing on the sound we hear. Most notably is the overdriven microphone and corresponding electronics. The sound of the engine that you hear in the video is literally distorted in the same manner that a hard rock guitar is, by overloading the circuitry or the microphone element itself. The interesting part of that is that our ears perceive this distortion as "loud" and "powerful" and "edgy" which are emotions that we correlate to the sound of the distortion, which we then also associate to the source sound. However, if we were sitting in the passenger seat we wouldn't hear that distortion, or at least to nowhere near the same extent, because our ears do not have electronics to overdrive. This is why I argue that on-board vids like these are not truly representative of real sound and should be viewed with a large grain of salt with regards to sound. The other great bit in here is, what do you actually hear? It's pretty much all engine all the time, with the exceptions of: - Transmission whine on the downshifts - Brake squeal - Wind noise plus cabin vibration from the engine (hard to distinguish what it actually is) - The car that is in front and ONLY when right on its bumper (at 1:14 or so) - Lots of stone chips or other pinging sound by stuff picked up and then ejected by the tyres And really, that's about it. If you really want to factor in the helmet over top, then that's even less detail you'll hear. So when we talk about threads like this that want a simulator to be "more realistic" by having details like more accurate tire states and reverb and etc. the actual reality is, you never hear any of it in a race car! Really loving this thread guys, it's really great to hear what the sim crowd thinks of how the industry is doing. Keep it coming!
Actually, one part that is really interesting to me is at 5:00 where he locks up the brakes. It sounds like you hear the whole car chuddering. That seems to be missing in most if not every racing game (at least that I've seen/heard). Like you've said, so many games focus on what the tires themselves are doing. I think there's a lot more feedback we could be getting from chassis/cockpit sounds so we can bring the tire sounds back into the range of 'somewhat realistic'.
What we hear from engine is quite different what we hear from engine in car, again my street car things, but I prefer them over race cars because I'm slow His clips are probably best quality audio from car engine sounds that you can find from youtube.
I noticed a few notes about improper Doppler effect in sims. I'm curious what is wrong with Doppler effect in rF/rF2. I'm far from saying it is 100% realistic because I have no idea how to prove it, but for sure it is not about volume modulation only. For sure also frequency is changing due to position change relative to camera. If there is something wrong with it i would like to learn about.
It might just be that the effect has been done proper and I'm not noticing it because it's not wrong (likely) but I quite often notice in games that the car just gets louder as it approaches the camera, and quieter after it passes. As you say, there needs to be a frequency change, relative to speed/camera position/volume/etc. If rF/rF2 has this, great. I've just not noticed it myself, yet. And that could be because it is proper, so it doesn't stick out as being wrong.
Yes, there is freq change in rf1 and rf2 (don't know if 100% correct but it is present). check the first seconds of this video. Sound is recorder from rf1
How about fake enviromental effects that some sound card drivers do offer? Here is video from rallying event, there are forest around the place, so just for experiment I did set my audio hq profile to forest, effect was not very huge, nothing silly like sewer or room, but it did made sound to regain some aspects that recording had made it to disappear, it gave me much better impression of sitting right next to those roads. Effect is of course very fake, but if we think about sitting inside of car, without poor or no sound proofing, it is always quite confined space that reflect sounds quite well, so maybe some kind of reflection of sound (is that reverb?) could be implemented to rFactor engine and it could be adjusted by sfx file to give more illusion about that confined space that reflects sounds well? Especially outside views lack a lot of such effect, imo. Is there someone who drives with such fake effects on?
an ultra realistic graphics is useless during replay if u haf a laughable external sound. Theres even sound characteristics difference running in the morning and evening.
I noticed that I keep getting views on my website from this thread, so I hope the thread res is ok. Unless the car was so loud that the camera had to adjust mic gain to compensate, the video should be pretty much accurate in terms of reverb. The sound of the car is is hitting the trees and etc. and then coming back to the camera, just as it would your ears were you standing there. If you take the video and then apply another reverb on top of that (and potentially equalization depending on how your soundcard software works) then you're actually putting echoes on echoes, if you want to think of it that way. Functionally you've made a thicker forest. When you're in a car that's as raced-out as possible the things like body metal and windows are as thin as they can be and actually aren't super reflective. Plus the car itself would be so brutally loud that the sounds being reverberated are indistinguishable from the rest of the engine noise. (This assuming a "cockpit" cam position in your game.) I agree that most games/sims suffer from poor external view cameras (especially replay cameras) but not quite in the way you describe - what I think they really lack is oversimplifying "listening distance." The way it works behind the scenes is that the game will say "if a car is x feet away, it is inaudible." Often that distance gets shorter and shorter to make space for other sounds. (Games/sims are limited by their system as to how many sounds can be processed/playing at one time) While recently at Road America for the ALMS race I can say from experience that you can hear a car pretty much anywhere on the track if you are in the infield. In fact, I even made a recording of this: http://snd.sc/OQFUoT (listen to #1/2) - I recorded this in the middle of the woods on a trail that goes from Turn 6 to Canada Corner if that helps. However if you listen to #3 or #4 (recorded right on the side of Turn 6) you'll notice that much of the "reverb" is inaudible until the car has driven past the corner (and most of the way down the hurry downs). You'll also notice that the car goes from pretty quiet to LOOUUUDDD as it passes by and then quickly back to medium-quiet again. In a game it's really tough to achieve dynamic contrast (the difference between quiet and loud) so this often results in kinda-loud to loud to kinda-loud for your replay cameras. It's one area that I hope to really see some improvement in my future work!
It is quite possible for the sounds to be different at different times of day. In the morning the air is thicker in a lot of places that have high humidity and thins out a little as the day warms up. This will cause some of the sounds to stay low to the ground and will make ambient sounds change. When the air warms up, the sound is able to travel quite a bit more so everything changes.
Has anyone been out when it is really really dense fog out there so you can't see your arm? Eerie sounding enviroment? Water molecules that have air between them absorb the sound I believe and this makes sounds becoming muffled, same thing with morning / mid day, only smaller scale. If someone don't believe me, then only thing you need is to eyeball this study, Page 6 has graph from effects of temperature and humidity and humidity part information, first pages cover temperature effects. Speed of sound and volume of sound is different at different temp and humidity and all that would be relatively easily put into sim, who knows maybe they already are at least partly. It is just adjusting sound emitting ranges (one variable) and speed of sound (one variable) formulas are in study. http://duckduckgo.com/k/?u=http%3A%2F%2Fwww.rane.com%2Fpdf%2Feespeed.pdf It has also tables which would be more than sufficient to use as modifier to values to create difference in sound. There was setting in plr file that defines distance from wall that reverb effect has some effect, it was something near 10 meters as default (maybe 13?), not sure if increasing that would have positive effect to overall feel of sound, it is easily hundred meters that we can hear echoing from forests and so on, even when ridingin in a car passing forest and passing field sounds different, even it is very minor difference, very hard to pinpoint what it is. For me, racing cars can be racing cars and I leave them alone, I focus street cars again as always. So inside a car effects coming from outside of car there should be adjustable attenuation parameter, this would then cause outside views to have different nature to sound and some effects would be much more clear, of course another variable to be added to sound engine. And as cars emit different sounds from front of car and rear of car it would need to be made for sound engine too. Then one would not need to really adjust range of sound anymore as attenuation of sound would take care of that, which should result much more realistic sounding enviroment. Greatest challenge would then be to create simple method which would make walls, buildings and terrain reflect and dampen the sound, which would require, imo. fog of war kind of computing for the sound, this would need to be done in 3d of course and it should probably be done to table, with 1m resolution from every drivable area, but it would make sound enviroment to be best there is. I think it would be possible to use some existing fog of war code or basic idea for that, if doing grayscale maps one could use colour as modifier %, top and left + right views at minimum would be needed, maybe resolution could be less than 1m as sound might be quite easy to interpolate? Anyway table might be rather big, like HAT table. Anyway that is how I currently see of near perfect sound simulation in sim enviroment. Also I think we will never see such detail in sims, but simple parts of that would already be huge improvement. Of course even currently track makers probably could make invisible walls that reflect sound to mimic forests, buildings etc. But there is challenge if plr file is set to something like 10m reverb distance, then you make invisible sound reflecting walls 30m from track, those have no effect, but when plr file value is increased, then it can cause too many reverb sounds in some locations as if there is wall in front of wall, I doubt current engine understands that, I think it behaves then like two walls? I'm sure it is area where also modders could do more about it, bit same as damage in rF1 where there were more possibilities than was typically used.
In regards to time of day and affects on sound, yes differences in air pressure and humidity have effects, but our perception of the sound tends to regulate away the actual changes from consciousness. Think of it like white balance for light. In the morning, sunlight is very amber; at mid day bright white to blueish. Yet we as humans will correctly identify white as "white" as our brain knows to tune out differences in white balance. (Unlike your camera, which needs to be told what white is. which is why your indoor pictures look orange and your evening pictures look blue. ) Your brain does a similar thing with sound at times of day. What your brain really cues on is environmental sounds (ie bird chirping)! There might be a subconscious awareness going on both with light and sound but I'm not getting in to that. The effect is so subtle that I wouldn't expect to have it in a sim for a very, very long time.
Changing humidity from 30 to 80 with rest of properties same had something like 5db difference at 140ft away if I don't recall wrong from document I posted. 3db change is what human think is double the volume, I think.
3dB is the effect of doubling amplitude. The average human will identify ~10dB as "doubled" (depending on starting volume and etc). The average human can identify 3dB as a difference in volume. Plus, consider how often in your sim your camera is a full 140' away from the car. Further, the attenuation is not even across all frequencies, so functionally it behaves more like an equalization curve. For a game to feature a humidity parameter to tie into an equalizer DSP in concert with volume attenuation PLUS factor distance, it becomes too much for current game engines for such a subtle effect. (Not trying to make an excuse, just trying to explain.)
