posted
Anyone ever tried anything like this? You could get perfect path length differences to your ears BUT how badly would the speaker on the opposite side of the car f*ck up the sound stage?
Here's a little idea. Probably going to hard to follow so get out the ritilan. To correct for the problem of the far speaker interfering with the sound stage you could 1.take a tap of the far channel. 2.Electronically delay that information by a couple of feet or whatever the distance is from the center channel to the far channel. 3.Add the delayed tap from the far channel back into the center speaker "out of phase" with the far speaker. The sound you tapped off would cancel out the sound comming from the far speaker, making it seem as if the far speaker is not there at all. This would create a perfect sound stage because the driver side would effectively have a left speaker on the left side of the dash and a right speaker in the center of the dash. The passenger would have the reverse, a left speaker in the right corner of the dash and a right speaker in the center of the dash. The interferance from the far speakers from each listener would be cancelled out and not heard.
posted
i followed until you switched channels with the passanger....
?
-------------------- Get loud daily... not burping. From Finals 2005:
quote: It's World Finals. A certified event. A 4x scoring event. It's the "pinnacle" of the "sport".
If I showed up to Game 7 of the World Series and was found to have a Corked Bat in pre game warm-ups, what would happen to me? Even if I said that I 'didnt see it in the rulebook, my bad'.
From Finals 2004:
quote:Originally posted by dBSteve: Everybody in the top 8 in every class bent the rules.
posted
he is using only three speakers, so the passenger's image would be backwards. Where the driver would hear drums on the left side, the passenger would hear the same information but on the right side.
-------------------- life is a dance you learn as you go, sometimes you lead, sometimes you follow. Posts: 1391 | From: southern cali | Registered: Oct 2004
| IP: Logged |
posted
okay, i got that much... buuuut what signal do you feed the centralspeaker? you have headunit's that offer dolby 5.1?
-------------------- Get loud daily... not burping. From Finals 2005:
quote: It's World Finals. A certified event. A 4x scoring event. It's the "pinnacle" of the "sport".
If I showed up to Game 7 of the World Series and was found to have a Corked Bat in pre game warm-ups, what would happen to me? Even if I said that I 'didnt see it in the rulebook, my bad'.
From Finals 2004:
quote:Originally posted by dBSteve: Everybody in the top 8 in every class bent the rules.
posted
the center gets the right channel information.
-------------------- life is a dance you learn as you go, sometimes you lead, sometimes you follow. Posts: 1391 | From: southern cali | Registered: Oct 2004
| IP: Logged |
-------------------- Get loud daily... not burping. From Finals 2005:
quote: It's World Finals. A certified event. A 4x scoring event. It's the "pinnacle" of the "sport".
If I showed up to Game 7 of the World Series and was found to have a Corked Bat in pre game warm-ups, what would happen to me? Even if I said that I 'didnt see it in the rulebook, my bad'.
From Finals 2004:
quote:Originally posted by dBSteve: Everybody in the top 8 in every class bent the rules.
You could get perfect path length differences to your ears BUT how badly would the speaker on the opposite side of the car f*ck up the sound stage?
You wouldn't get perfect pathlength differences, but I don't understand how the speaker on the far side of the car is screwing anything up (before your idea) You would also probably royally screw up your stage width. Here's a little idea. Probably going to hard to follow so get out the ritilan. To correct for the problem of the far speaker interfering with the sound stage you could
What problem???
1.take a tap of the far channel. 2.Electronically delay that information by a couple of feet or whatever the distance is from the center channel to the far channel.
You have essentially just created 2 far channel speakers.
3.Add the delayed tap from the far channel back into the center speaker "out of phase" with the far speaker.
Now its out of phase with both speakers and while I've never done it, I would guess that your width on the far side would not extend (convicingly) far beyond your center channel. You also only have one side of your stereo image feeding to your center, right or left, which will screw things up.
The sound you tapped off would cancel out the sound comming from the far speaker, making it seem as if the far speaker is not there at all.
Not really, you will have late and early reflections coming into play as well as 2 speakers playing the same frequency in 2 different places with a messed up stereo image. Also, you don't want to cancel out the far speaker, thats what gives you your width.
This would create a perfect sound stage because the driver side would effectively have a left speaker on the left side of the dash and a right speaker in the center of the dash.
That wouldn't create a perfect soundstage at all.
The passenger would have the reverse, a left speaker in the right corner of the dash and a right speaker in the center of the dash.
If I'm following you right, the passenger would have 2 rights, one on the right and one in the center out of phase with each other, and a left on the left.
The interferance from the far speakers from each listener would be cancelled out and not heard.
I understand what you mean here, but it wouldn't be the case on the passenger side. Aside from that I don't understand what interference your talking about from the far speakers. Your best bet is kicks (or some other form of equalizing pathlengths as much as possible) with proper angling so that the "loudness" of the near speaker is offset by the angle, and the on-axis of the far speaker makes up for its distance, this can be done from both sides. Also, in your example your running a center channel out of phase. Phasing is something that shouldn't be relied up on IMO. In my short experience competing, as soon as I sit in a car I can tell if stuff is out of phase, it has that out of phase tonality, a sort of glow around the instruments and voices that detracts from the focus and tonality, and a sort of backround glow too. If you have a deck with that capability switch the phase on something back and forth while listening to music, you will see what I mean. Its normally done for height, or in 2 seat cars to get a better "center" from both seats, but its at the expense of your tonality, which, If I'm gonna sacrafice something, I'd rather it be height. Some people don't hear that "glow" i'm talking about so they don't care but it drives me nuts. As far as center thats why my kicks take me 50 hours, to make sure center and height are good to begin with.
If it must be done it must be done but I first try to optimize things in phase, so I wouldn't build a car with your design in mind.
posted
I think you need a bit more ritalin. Let me try and explain this again in a different way.
How would it sound in the drivers seat if you could have a let channel on the left side of the dash and a right channel in the center of the dash? If you think that would not sound good, please explain. I don't believe that the width of the soundstage would be comprimised. The closer speakers are to your head, the less significant physical seperation of the speakers is. In this scenario, I believe that speakers will be close enough to the listener that the vectors of eminating sound will not narrow the stage. The sound I get from my cruddy computer spakers in my home (about 2 feet in front of my head and only 3 feet apart) and from my headphones (only half a foot of my fat head seperates them) sounds beautifully seperated and imaged accurately. Having speakers physically farther away from you (unless they are both the exact same distance and angle from you) does nothing but detract from imaging (especially in a car where there are so many reflective surfaces). Anyways, with tonally accurate speakers placed the way I mentioned, I believe it would sound excellent. So the problem is, what about the passenger? Well, if we give him a left channel in the right corner of the dash, he will have a nice stage too, although he'll have a left speaker on the right and vice-versa. So would this sound good? I doubt it because for the driver, the extra left channel in the right corner is going to interfer with the right channel in the center of the dash. So is tehre a way to cancel out the sound coming from the far corners (left channels) and not interfer with the far corners output themselves? I believe there is... What you COULD do is add the far channel (left channel) information to the right channel (in the center of the dash) out of phase with itself. well, you'd have to delay it a foot or two so that it arrives at your ear (the drivers right ear and the passengers left ear) at the same time as the far channel to make sure they cancel. A continuously variable delay would make it easy to tweak. You mention that there would be secondary reflections and that sort of stuff, well, yeah, there would. Maybe they would interfere and maybe they would be insignificant. I don't know. I'll be sure to experiment with is someday and let you know. This is just an idea, no need to come down too hard on it, I don't believe you fully understood the theory, which is to be expected. I'm a technologist, not an instructor. I appreciate your input and I'm glad you took the time to go through it.
Thanks for the feedback, Rrrrolla
Posts: 112 | From: NB, Canada | Registered: Mar 2004
| IP: Logged |
posted
How would it sound in the drivers seat if you could have a let channel on the left side of the dash and a right channel in the center of the dash? If you think that would not sound good, please explain.
Crappy. For the same reason that it probably would not sound as good if you put the speakers withing 8 inches of each other. Stereo image works by left and right channel independent combining with each other to form imaging, so in a car, if you narrow that space between the speakers for stereo image, you are cramping everything. Try it on your home system with some test material, like the iasca disc with the vocal and drum imaging tracks. Put the speakers side by side and see how it sounds, then spread them out farther and farther, you will see that the more space you give them, the wider your stage will get.
I don't believe that the width of the soundstage would be comprimised.
See above.
The closer speakers are to your head, the less significant physical seperation of the speakers is. In this scenario, I believe that speakers will be close enough to the listener that the vectors of eminating sound will not narrow the stage.
I've read that 7 times and I still don't get it.
The sound I get from my cruddy computer spakers in my home (about 2 feet in front of my head and only 3 feet apart) and from my headphones (only half a foot of my fat head seperates them) sounds beautifully seperated and imaged accurately.
Ok, i'm assuming the 3 feet apart is your computer width (or close to it) Now pretend that its supposed to extend 3 more feet to the right, how deos it sound....or move your right computer speaker halfway across the computer and see how it sounds.
Having speakers physically farther away from you (unless they are both the exact same distance and angle from you) does nothing but detract from imaging (especially in a car where there are so many reflective surfaces).
Thats the exact reason you put them farther away from you, to try to equalize everything. The less difference between left and right will only ADD to your imaging. Keep in mind we are talking about a car, having speakers in the kicks is not "miles" away.
Anyways, with tonally accurate speakers placed the way I mentioned, I believe it would sound excellent.
Then do it and let us know.
So the problem is, what about the passenger? Well, if we give him a left channel in the right corner of the dash, he will have a nice stage too, although he'll have a left speaker on the right and vice-versa.
You don't see that as a problem for proper staging?? And now you've got left AND rights on one side of the car....and still a center channel only playing one side....I assure you, it ain't gonna work.
So would this sound good? I doubt it because for the driver, the extra left channel in the right corner is going to interfer with the right channel in the center of the dash.
You've got left and right signal all over the place, it isn't going to work for the driver, the passenger, people in the back, it may sound OK to someone in the trunk if there head is in the port and the sub level is cranked way up. So is tehre a way to cancel out the sound coming from the far corners (left channels) and not interfer with the far corners output themselves? I believe there is... What you COULD do is add the far channel (left channel) information to the right channel (in the center of the dash) out of phase with itself.
Then you can only add the far channel of one side of the car, and the putting it out of phase is not going to do what you think it is, and if it did, you've cut your width by a large margin. Aside from that you've got 2 speakers playing the same information, and what you've (in your theory) succeeded in doing on one side, is going to be the opposite on the other.
well, you'd have to delay it a foot or two so that it arrives at your ear (the drivers right ear and the passengers left ear) at the same time as the far channel to make sure they cancel. A continuously variable delay would make it easy to tweak.
It doesn't matter, as its one sides information, I understand what you are saying, but you've got 2 rights and one left.
You mention that there would be secondary reflections and that sort of stuff, well, yeah, there would. Maybe they would interfere and maybe they would be insignificant. I don't know. I'll be sure to experiment with is someday and let you know. This is just an idea, no need to come down too hard on it, I don't believe you fully understood the theory, which is to be expected. I'm a technologist, not an instructor. I appreciate your input and I'm glad you took the time to go through it.
I completely understand what you are saying, i'm not coming down hard on it. So if I'm coming off that way don't think its intentional, I enjoy conversations like these.
Rrrrolla [/QB][/QUOTE]
Posts: 2276 | From: East | Registered: Jan 2003
| IP: Logged |
posted
How would it sound in the drivers seat if you could have a let channel on the left side of the dash and a right channel in the center of the dash? If you think that would not sound good, please explain.
Crappy. For the same reason that it probably would not sound as good if you put the speakers withing 8 inches of each other. Stereo image works by left and right channel independent combining with each other to form imaging, so in a car, if you narrow that space between the speakers for stereo image, you are cramping everything. Try it on your home system with some test material, like the iasca disc with the vocal and drum imaging tracks. Put the speakers side by side and see how it sounds, then spread them out farther and farther, you will see that the more space you give them, the wider your stage will get.
I agree that if you listen to an Iasca disc in your living room with the speakers side by side and you sitting back a ways, it would sound like crap, but if you're sitting right in front of them (which is how it would be in a car) it will sound just as good as sitting back ten feet and seperating them alot. The closer you are to the speakers, the less they seperation there has to be to achieve proper imaging. In fact, it will sound even better if you sit right up close and have the speakers on each side of your head (but still a bit in front of course) because there would be no interferance from things like walls and coffee tables.
I don't believe that the width of the soundstage would be comprimised.
See above.
The closer speakers are to your head, the less significant physical seperation of the speakers is. In this scenario, I believe that speakers will be close enough to the listener that the vectors of eminating sound will not narrow the stage.
I've read that 7 times and I still don't get it.
The sound I get from my cruddy computer spakers in my home (about 2 feet in front of my head and only 3 feet apart) and from my headphones (only half a foot of my fat head seperates them) sounds beautifully seperated and imaged accurately.
Ok, i'm assuming the 3 feet apart is your computer width (or close to it) Now pretend that its supposed to extend 3 more feet to the right, how deos it sound....or move your right computer speaker halfway across the computer and see how it sounds.
Having a pair of speakers 1 foot in front of you and 1 foot apart is the same as having a pair of speakers 10 feet in front of you and ten feet apart (just like a home theatre setup) and sounds great either way. Personally I don't like having my soundstage skewed to the right. I know this is what the SQ guys are looking for in a car (a stage all the way acress the dash), but I much prefer the stage created by headphones which create a huge stage (no reflections to limit it's size). The mind creates width well beyond the physical limits of the speakers physical seperation as long as there are no significant reflections from other surfaces. I'm sure you've noticed this effect in your headphones.
Having speakers physically farther away from you (unless they are both the exact same distance and angle from you) does nothing but detract from imaging (especially in a car where there are so many reflective surfaces).
Thats the exact reason you put them farther away from you, to try to equalize everything. The less difference between left and right will only ADD to your imaging. Keep in mind we are talking about a car, having speakers in the kicks is not "miles" away.
I agree that kicks sound much better than speakers on the dash. I've had kicks in 2 of my previous cars and it sounded good. It didn't sound as big and full as my headphones though. Unequal pathlengths is the number one problem with stage width and height. You seem to think that a soundstage will sound wider and taller if the passenger kickpanel is in it's standard location whereas I think it will sound better if it's brought closer to the driver so that the pathlengths are exactly equal. I have played around with this stuff in my car in the past and if you get the speakers in the perfect position the stage will not be limited to the dash, it will extend well beyond the limits of the car; an orchestra will sound HUGE and well defined and will not be limited to the dashboard. Kinda like how it sounds on headphones.
Anyways, with tonally accurate speakers placed the way I mentioned, I believe it would sound excellent.
Then do it and let us know.
So the problem is, what about the passenger? Well, if we give him a left channel in the right corner of the dash, he will have a nice stage too, although he'll have a left speaker on the right and vice-versa.
You don't see that as a problem for proper staging?? And now you've got left AND rights on one side of the car....and still a center channel only playing one side....I assure you, it ain't gonna work.
So would this sound good? I doubt it because for the driver, the extra left channel in the right corner is going to interfer with the right channel in the center of the dash.
You've got left and right signal all over the place, it isn't going to work for the driver, the passenger, people in the back, it may sound OK to someone in the trunk if there head is in the port and the sub level is cranked way up. So is tehre a way to cancel out the sound coming from the far corners (left channels) and not interfer with the far corners output themselves? I believe there is... What you COULD do is add the far channel (left channel) information to the right channel (in the center of the dash) out of phase with itself.
Then you can only add the far channel of one side of the car, and the putting it out of phase is not going to do what you think it is, and if it did, you've cut your width by a large margin. Aside from that you've got 2 speakers playing the same information, and what you've (in your theory) succeeded in doing on one side, is going to be the opposite on the other.
well, you'd have to delay it a foot or two so that it arrives at your ear (the drivers right ear and the passengers left ear) at the same time as the far channel to make sure they cancel. A continuously variable delay would make it easy to tweak.
It doesn't matter, as its one sides information, I understand what you are saying, but you've got 2 rights and one left. Actually you've got 2 lefts and one right. The farthest speaker from the listener on each side is the same sound (left channel). So we need to get rid of it by "subtracting" it from the center channel. This is done by adding it out of phase. This seems to be a difficult idea for you to understand, I can't really explain it very well, if I knew how to get pictures on here, I could show you alot easier. Do you know how to do it?
You mention that there would be secondary reflections and that sort of stuff, well, yeah, there would. Maybe they would interfere and maybe they would be insignificant. I don't know. I'll be sure to experiment with is someday and let you know. This is just an idea, no need to come down too hard on it, I don't believe you fully understood the theory, which is to be expected. I'm a technologist, not an instructor. I appreciate your input and I'm glad you took the time to go through it.
I completely understand what you are saying, i'm not coming down hard on it. So if I'm coming off that way don't think its intentional, I enjoy conversations like these.
I actually appreciate your feedback. I'm glad you're taking the time to read it. I posted this idea about 5 years ago on here when I used to post as a different user. People didn't understand it much and didn't even take the time to try really.
quote:Originally posted by Rrrrolla: Anyone ever tried anything like this? You could get perfect path length differences to your ears BUT how badly would the speaker on the opposite side of the car f*ck up the sound stage?
Here's a little idea. Probably going to hard to follow so get out the ritilan. To correct for the problem of the far speaker interfering with the sound stage you could 1.take a tap of the far channel. 2.Electronically delay that information by a couple of feet or whatever the distance is from the center channel to the far channel. 3.Add the delayed tap from the far channel back into the center speaker "out of phase" with the far speaker. The sound you tapped off would cancel out the sound comming from the far speaker, making it seem as if the far speaker is not there at all. This would create a perfect sound stage because the driver side would effectively have a left speaker on the left side of the dash and a right speaker in the center of the dash. The passenger would have the reverse, a left speaker in the right corner of the dash and a right speaker in the center of the dash. The interferance from the far speakers from each listener would be cancelled out and not heard.
LOL, can anyone oven follow this?
Rrrrolla
Think about what would happen when "left only" signal occurs, like a drum on the far left of the soundstage.
Your setup would create a mono image for both sides of the vehicle.
Posts: 144 | From: Midwest | Registered: Oct 2003
| IP: Logged |
posted
I agree that if you listen to an Iasca disc in your living room with the speakers side by side and you sitting back a ways, it would sound like crap, but if you're sitting right in front of them (which is how it would be in a car) it will sound just as good as sitting back ten feet and seperating them alot.
No it won't, and you keep skipping the issue of not having the speakers at your stage width points, you are only running them half-way, with you sitting on one side. Is this how you would run your home setup??
Having a pair of speakers 1 foot in front of you and 1 foot apart is the same as having a pair of speakers 10 feet in front of you and ten feet apart (just like a home theatre setup) and sounds great either way.
Nope, how wide will it be in the first scenario vs the second scenario, there will be a huge difference in perspective as well. You have to keep in mind that there is a set perspective in a car, the stage should seem to extend from side mirror to side mirror.
Personally I don't like having my soundstage skewed to the right. I know this is what the SQ guys are looking for in a car (a stage all the way acress the dash), but I much prefer the stage created by headphones which create a huge stage (no reflections to limit it's size).
It should be all the way to the right but their should also be left stage that extends beyond your pillar, you shouldn't feel as though you are sitting at the far left boundary of the soundstage. In my car vs other cars, you sit very far to the left, you close the door and its almost on your leg, but the stage on that side extends well beyond the pillar and windsheild, closing your eyes and listening you wouldn't think that the car stops where it does. Your whole plan is to get proper staging with your idea, but I don't think you know what proper staging in a car is.
Your headphones have all optimal conditions to function correctly. THey should sound damned good. You are dead center, no reflections, few speakers, equal pathlengths....you can't really compare the 2 as none of that is possible in a car.
I agree that kicks sound much better than speakers on the dash. I've had kicks in 2 of my previous cars and it sounded good. It didn't sound as big and full as my headphones though.
They will probably never sound as good as a good set of HP, but they are 2 different things. You should be able to get close enough that you enjoy listening to your car almost as much as you do your headphones though...
Unequal pathlengths is the number one problem with stage width and height.
What??
Speakers in the door will create the maximum width and that is one of the most unequal pathlengths you could have, now put them high in the doors you have your height
You seem to think that a soundstage will sound wider and taller if the passenger kickpanel is in it's standard location whereas I think it will sound better if it's brought closer to the driver so that the pathlengths are exactly equal.
Your width will suffer, believe me. As for height we haven't been discussing that at all.
I have played around with this stuff in my car in the past and if you get the speakers in the perfect position the stage will not be limited to the dash, it will extend well beyond the limits of the car; an orchestra will sound HUGE and well defined and will not be limited to the dashboard. Kinda like how it sounds on headphones.
Lets say your right, you still have left and right signal all over the place, and a mono center channel that isn't getting "mono" signal, its being fed left or right channel info. The intensity of far left info and far right info needed to create image in the correct spot will no longer be correct as you have another speaker in the center creating the same image, what you'll get is confusion. Take 2 pairs of headphones and put a left and right on each ear and see what you get. I understand your theory about cancelling things out, and that would work if the speaker was in the exact same spot, which just isn't possible. Throwing your center out of phase to cancel out another speaker isn't going to do the job anywhere near convincingly, its in a different location, different reflections, different pathlengths etc.
Posts: 2276 | From: East | Registered: Jan 2003
| IP: Logged |
posted
I was with you right up until you started talking about cancelling sound.
If you can hook up two speakers playing music containing information over 100hz, in ANY sort of phase, put them more than 1 foot apart, put them in a car, and cancel a very subjectively noticeable amount of sound across their bandwidth between the two of them, I will be not only shocked, but bow down.
Interesting idea though, serving as the basis for some more fleshed out designs having engineered center channels and DSP.
ShadowStar
-------------------- You can't build a reputation for what you're GOING to do.. But you can build one for TALKING about it!
It's all about knowledge, love and respect. Posts: 2579 | From: Somewhere In the Northeast | Registered: May 1999
| IP: Logged |
posted
In this case, tyhere is NO center channel. There is a channel that eminates from the center of the dash. It is the "right channel only" to both driver and passenger. Of course the passenger has reversed channels, no big deal. The biggest problem with the whole deal I believe will be what Shadowstar has said. Can you actually get a full (or at least pretty much full) cancelling effect at the higher frequencies. That would be difficult I agree. However, as the frequencies get higher they also get more directional making them of less significance because they would be more directed in the direction which they are intended for. Possibly a low pass filter for the middle channels left side cancelling audio would be needed. It would be something to play around with to get it right. I really wish I could play arond with it, but here in Atlantic Canada, the weather conditions are keeping me out of my car for the next few months.
Here's a bird's eye view, I think I did this right... Posts: 112 | From: NB, Canada | Registered: Mar 2004
| IP: Logged |
quote:Originally posted by Rrrrolla: The biggest problem with the whole deal I believe will be what Shadowstar has said.
WTF?
Have you not been reading my posts?
And that is nowhere near you biggest problem. Your biggest problem is that you have left signal on both sides of the car. and right signal in the center. Think of how stereo image works, then look at your diagram.
Now that I see exactly what you are talking about (had the same idea before, just different L/R scenario) it makes even less sense how you think that is going to work. You are creating 2 stages, one of which is reversed. So while that may seem like a good idea (as you have 2 passengers) there is no way to seperate them, (unless you build a wall down the center) so all you will get is mass confusion. Both passengers will be hearing left signal on both sides of the car, thats not a problem??
And in the example you cite, aside from already mentioning you won't effectively cancel anything, you don't even have the same signal on the channel you want to cancel, it can't cancel something that it isn't playing.
IE, let's say you've got a triangle playing on the left (in a correct setup) you're in the passenger seat, and its on your left AND right (leading to not being able to place it at all) but you want to cancel one side by running right signal, in the center of the car, out of phase.....well your right channel isn't playing a triangle, so your hopes of canceling are 1 million times more slim then they were before.
If you still don't see a problem, lord help you.
Posts: 2276 | From: East | Registered: Jan 2003
| IP: Logged |
posted
You do not understand how this works at all my friend. If a triangle is playing in the left channel, it is playing on the left for driver and right for passenger. It is also playing on the far right for the driver and the far left for the passenger. That is the problem that we have BOTH identified. Now I will present you with a solution. If you add some left channel information to the right channel out out phase with the left channel information, it will eliminate the problem where the far left channel is interfering with BOTH peoples right channel. This is a difficult concept to grasp. I don't know of another way to explain it though. If I add left channel info out of phase into the right channel (like in parallel with the right channel), it will cancel the information being received at both parties problem side ie. the drivers right ear and the passengers left ear. Lets look at it strictly from the drivers point of view. The problem is that his far right has left channel information making it to his ear. So I propose we add information to the right channel (in the center of the dash) to cancel this out at the drivers right ear. The left channel is making it to the drivers left ear without issue, so we will not look at this at all as being a problem. So you will have a left channel pointing directly at the drivers left ear, taht's good. You will have a right channel pointing at the drivers right ear, that's good. Now you will also have a left channel pointing kinda at the drivers right ear from the far right side of the dash. This fuks everything up except that we also have added left channel (out of phase with itself) to the right channel in the middele of the dash to cancel it out at the drivers right ear.
I see the problem you are seeing, and the whole point of this thread is to deal with that issue. The right channel plays exactly as it's recorded in the center of the dash by the way, all we do is add some (delayed and out of phase) left in there with it.
The blue signal is added so it will cancel out the red signal. The sound gets cancelled at the peoples ears, not electrically in the sinal path. There are three sets of sounds making it to the drivers right ear. 2 of them are out of phase and cancel each other (the blue and red).
I'm sorry for being so unclear, but it is a very difficult theory to convey.
posted
I understand it pretty good, my main point is the methods you are trying to combat the problem with aren't going to be even remotely successful.
You've yet to identify what the problem is with normal setup...the "right channel interference" you were talking about.
Posts: 2276 | From: East | Registered: Jan 2003
| IP: Logged |
posted
ALright I know enough to know that what Rrrrolla is trying won't work BUT I have had an idea for wuite some time that I have never had the time to try. What if you did some digital processing and generated the following three channels: Right only Center Left only
The right only channel would contain ONLY the information that was not present at all in the left channel.
THe center channel would contain ONLY the information that was common to both left and right channels.
The left only channel would contain ONLY the information that was not present at all in the right channel.
Actually, if I implemented this I would provide an adjustment that would allow you to remix in a portion of the original right and left channels back into the corresponding outputs (it is just more complicated to explain that scenario).
How do you think that it would affect the imaging?
My guess is that it would take an extreme amount of proessing power to seperate the signals as I have indicated but I'm not extremely proficient with digital processing of audio signals so who knows.
-------------------- 2004 Saturn Vue Red Line Edition Alpine DVA9860 Alpine H701 MB Quart QTD-25 B&C 6MD38 6.5" Mid B&C 8NDL51 8" Midbass IDMAX (3) MTX 81000D (3) MTX 8302 (2) MTX 6304 (2) Posts: 284 | From: Huntsville, AL (go to BFE and take a left) | Registered: Oct 2003
| IP: Logged |
posted
I see what you're saying. And.. you've got a problem. Your center channel speakers are only going to be directive for a certain portion of their frequency response.
Once you go under the "beam" frequency of a driver, it sounds the same pretty much no matter where you are dependant on the axis, it is only dependant on distance
So, your left center will be heard by the right passenger at 99% of its level for 75% or more of its frequency range, and vice versa so for the right center and left passenger.
ShadowStar
-------------------- You can't build a reputation for what you're GOING to do.. But you can build one for TALKING about it!
It's all about knowledge, love and respect. Posts: 2579 | From: Somewhere In the Northeast | Registered: May 1999
| IP: Logged |
Thanks for the response. I'm not sure that I follow what you are trying to say. I understand that below a certain freq sound in omnidirectional but I am really talking about freqs above 250Hz. The processing that I am referring to should create a three channel front stage that has a VERY predominant center channel (that's why I mentioned that you would probably want to mix back in some portion of the original signal into the left and right channels).
I have run center channels before where you sum the left and right channel for your center channel. Those types of center channels typically destroy your stereo seperation (if the volume level of the center channel is not extremely and therefore not very effective). What I am referring to here is a different concept.
The right only channel would contain ONLY the right channel information having no information that was present in the left (this assumes that there will be no remix of the original right channel signal for the sake of simplicity). The left only channel would contain ONLY the left channel information having no information that was present in the right. The center only channel would have ONLY information that was common to both left and right channels.
This seems to be what most surround sound systems employ for the front channels. What I would like to do is deelop a processor that emulates that type of system. I know that there are home components that will emulate this output from a stereo input. Most of those do a poor job and I am wondering what the difficulties in implementation would be for the car environment.
Again thanks for the input and it is very possible that I am missing something here. If I am please explain.
After writing this post I re-read your post for the umpteenth time and picked up on something. YOu mention a "left cetner" channel. There are only three channels right only, center only, and left only. I don't know maybe this clears things up a bit.
-------------------- 2004 Saturn Vue Red Line Edition Alpine DVA9860 Alpine H701 MB Quart QTD-25 B&C 6MD38 6.5" Mid B&C 8NDL51 8" Midbass IDMAX (3) MTX 81000D (3) MTX 8302 (2) MTX 6304 (2) Posts: 284 | From: Huntsville, AL (go to BFE and take a left) | Registered: Oct 2003
| IP: Logged |
posted
Hey Snake, that's a cool idea too. Hell, why stop with just 3 channels, you could even put a channel in between the left and center and one in between the right and center. If you can make it work for 3, I bet you could make it work for as many as you wanted.
BTW, I think Shadowstar's response was aimed at me.
If you look at this drivers response, you will see several things.
Lets say that the driver rolls off at 12db/o at 120hz (and down) and then is ruler flat up to 5khz, where it likewise rolls off at 12db/o.
Draw yourself a picture of this to help, if you like.
Now, lets look at the response of this driver 30 degrees off the axis. (30 degrees away from a line drawn from the center of the dustcap straight out)
What we see is that it rolls off from 120hz and down, and then is ruler flat from 120 to 4000 hz.. and then something changes: now this driver rolls off at 12 db/o or so at 4khz
Now look at the response of the driver 60 degrees off axis: 120hz roll off, flat to 3500hz, and then roll off.
Now, consider the region from 120hz to 3500hz
For all three curves, 0, 30 and 60 degrees off axis, the response was exactly the same. Which means that an ear, placed the same distance away and listening to information from the speaker between 120hz to 3500hz, could go from 0 to 60 degrees (almost sideways) to the driver and hear roughly the same information.
Your center drivers are located very close together and are pointed mostly towards the back of the car: both driver and passenger will get almost the same amount of information through both of them, preventing your center devices from having much distinguishable L-R identity. The passenger will hear mostly a "bunch" more sound coming from the left and the driver will hear a "bunch" more sound coming from the right, however, because there is also information in front of the driver some of this info will be interpreted as being on center: but not most of it. This will serve mostly to muddy up the stereo effect altogther and make you lose L-R placement, which happens in a lot of cars that use center channels: Everything seems to come from a smaller region in the center because more sound energy is coming from that direction. Which isn't necessarily bad.. nor is it necessarily good
ShadowStar
-------------------- You can't build a reputation for what you're GOING to do.. But you can build one for TALKING about it!
It's all about knowledge, love and respect. Posts: 2579 | From: Somewhere In the Northeast | Registered: May 1999
| IP: Logged |
![[Wink]](wink.gif)
Printer-friendly view of this topic