0:00:13 | okay so the i mean that's is that the um the a is about |
---|
0:00:18 | spatial sound reproduction but you see i would lost because |
---|
0:00:22 | so |
---|
0:00:24 | um so that's |
---|
0:00:25 | let me |
---|
0:00:26 | just discrete we go to the um somewhat deviation or some background |
---|
0:00:30 | this you know i mean we could do these bases all reproduction and |
---|
0:00:35 | with in you and side um |
---|
0:00:37 | in a given region of space |
---|
0:00:39 | so the idea is to you were besides sound feel |
---|
0:00:42 | but spatial sound you produce it using a |
---|
0:00:45 | a bunch of laws speak |
---|
0:00:47 | so there are two common techniques meets people use |
---|
0:00:50 | um in the a great is if you want to use the if you synthesis so |
---|
0:00:54 | using using good job of most in table |
---|
0:00:56 | okay cage to uh and in the one is sound |
---|
0:00:59 | you know |
---|
0:01:00 | called seed in the or spherical harmonic decomposition all |
---|
0:01:04 | oh commonly high is one |
---|
0:01:06 | um |
---|
0:01:07 | so |
---|
0:01:09 | one thing |
---|
0:01:10 | we the what of these techniques are uh um |
---|
0:01:14 | with a pure tried to use them |
---|
0:01:16 | to reproduce a sound in a you each and |
---|
0:01:19 | you need actually be a a lot of a large number of |
---|
0:01:21 | also |
---|
0:01:23 | works on but um |
---|
0:01:24 | you you |
---|
0:01:26 | for a frequency say a um a or we have number K with though |
---|
0:01:30 | we have let them down |
---|
0:01:32 | um |
---|
0:01:33 | if the reproduction region is a radius of a a a a a in a number loudspeakers you need is |
---|
0:01:39 | approximately |
---|
0:01:40 | so to um |
---|
0:01:42 | to times they are bus once a proportion to the frequency |
---|
0:01:46 | um |
---|
0:01:47 | is |
---|
0:01:47 | what was it may not bad in the use of the agent so |
---|
0:01:51 | mean you |
---|
0:01:52 | what to have a you |
---|
0:01:54 | not i mean and if you want to increase D frequencies then |
---|
0:01:57 | or be see the number of all goes are going to be a um |
---|
0:02:03 | in a sense on really |
---|
0:02:04 | so for example a um a production or what a point one meter radius at sixteen khz takes quite like |
---|
0:02:10 | sixteen all speak |
---|
0:02:12 | so that's in a a a a a few assume a to be |
---|
0:02:15 | a three |
---|
0:02:16 | but if |
---|
0:02:17 | it's the three D case this and number of speakers are going to be the while K are one |
---|
0:02:23 | approximate |
---|
0:02:24 | so that a seven one O the |
---|
0:02:26 | uh a go back so he |
---|
0:02:28 | i i some um |
---|
0:02:30 | that's |
---|
0:02:32 | a another another issue use um |
---|
0:02:34 | i you try to do this production or when an yeah yeah |
---|
0:02:37 | or B S C |
---|
0:02:40 | you do you know room there would be a real a coming from "'cause" his own was |
---|
0:02:44 | you know outside region as well |
---|
0:02:46 | so there a unit selection it back so that we |
---|
0:02:50 | um |
---|
0:02:51 | change you or or you know so on so |
---|
0:02:53 | to compensate that |
---|
0:02:56 | sometimes you can be black station type need so |
---|
0:02:59 | okay be in these algorithms |
---|
0:03:02 | to the um |
---|
0:03:05 | um |
---|
0:03:05 | compose |
---|
0:03:09 | and and also the it has been shown that if you use |
---|
0:03:12 | rather than just the only directional all speakers by |
---|
0:03:16 | um |
---|
0:03:17 | first the save a type or |
---|
0:03:19 | or fixed directivity loss because you can actually |
---|
0:03:24 | right we use used is on feel going outside the your regional well |
---|
0:03:28 | you you uh decide which region |
---|
0:03:30 | so the mean we all the X |
---|
0:03:32 | in do you have a less if you in the |
---|
0:03:35 | region channel of interest and then to be outside you'd all speaker arrays |
---|
0:03:39 | exterior F feel |
---|
0:03:40 | so |
---|
0:03:41 | you things say takes into the of if it's indices |
---|
0:03:44 | using a a |
---|
0:03:46 | type was you can actually eliminate the outside side uh three |
---|
0:03:50 | well of course uh are you is a problem that |
---|
0:03:55 | the |
---|
0:03:55 | the coming from |
---|
0:03:57 | um was and things like a |
---|
0:04:00 | um |
---|
0:04:01 | a another a a local background um so |
---|
0:04:04 | there is a thing about this frequency bin you do in the boat also we got a sweet he's um |
---|
0:04:10 | which just a simple D |
---|
0:04:12 | if we can corresponding to the um i'll wavelet separation in a lot of speakers you looking at say |
---|
0:04:18 | um |
---|
0:04:19 | my my of that |
---|
0:04:22 | um um so as i say in U K to have a some works which shows that if you use |
---|
0:04:26 | the fixed that to this because you can actually use these |
---|
0:04:29 | some be going outside |
---|
0:04:31 | just codecs C |
---|
0:04:32 | P |
---|
0:04:34 | okay so i |
---|
0:04:35 | so that is a summary of the shows these spatial sound you button use one is the you need a |
---|
0:04:40 | large number of those speakers and do is these |
---|
0:04:42 | reverberation operation and from if you give you |
---|
0:04:46 | so you we nine see if we could use the less number of all speakers |
---|
0:04:50 | and also |
---|
0:04:51 | um |
---|
0:04:52 | you can have |
---|
0:04:53 | you going outside to sixty if you |
---|
0:04:56 | less me my |
---|
0:04:58 | so what i Q can be of this paper is that |
---|
0:05:01 | we use a i would've be able to to be with the loudspeakers speakers was sound reproduction in this particular |
---|
0:05:08 | there but be just using two D case |
---|
0:05:10 | um so that means |
---|
0:05:12 | use to you have used |
---|
0:05:14 | all speakers |
---|
0:05:15 | which one |
---|
0:05:17 | be able to produce a variable a really factor |
---|
0:05:21 | and then you combine the or |
---|
0:05:24 | in a area to produce your |
---|
0:05:27 | and then be show that if you use |
---|
0:05:29 | and in for the variable |
---|
0:05:31 | um |
---|
0:05:32 | also does |
---|
0:05:33 | it actually you can increase that night it's we can still the maximum operating frequency by in time |
---|
0:05:39 | well the on the |
---|
0:05:41 | well |
---|
0:05:42 | i if you keep the frequency constant can increase the regional reproduction but it does it's all alternate view you |
---|
0:05:49 | can |
---|
0:05:50 | the D V A the uh each of size and in the um |
---|
0:05:54 | if we can see then you can use the number of rows because by a one in time |
---|
0:06:00 | okay so |
---|
0:06:01 | oh what is fairly small the about these |
---|
0:06:04 | how to design |
---|
0:06:06 | these are what the lost |
---|
0:06:08 | so this is about you know |
---|
0:06:10 | for some way P can have a was five the because of you can use them to um a is |
---|
0:06:15 | it sounds |
---|
0:06:17 | okay so |
---|
0:06:20 | normally a it decides on a you put really i mean would be written using |
---|
0:06:25 | this see question this is just a |
---|
0:06:28 | does this is an interior feel so we in the last week at right |
---|
0:06:32 | so this is given by a |
---|
0:06:34 | a gain for the two dimensional S which means um you just assume you i mean |
---|
0:06:40 | so that's is given by a |
---|
0:06:41 | that's a function of which you know be a you know |
---|
0:06:44 | and frequency frequency and the yeah |
---|
0:06:46 | and then |
---|
0:06:48 | the |
---|
0:06:48 | i decides on a you you can be given by a bunch of corporations |
---|
0:06:52 | C |
---|
0:06:54 | so because of the do to some of the as a function for this you can get these |
---|
0:06:59 | spanish and do in |
---|
0:07:01 | um say in the as minus in the class |
---|
0:07:05 | so |
---|
0:07:06 | if we use see um a only directional also speak at it in both C |
---|
0:07:12 | it'll it'll all speakers V L need to be able to say when class |
---|
0:07:19 | so i |
---|
0:07:20 | so it's a close we all but was that sound really is given by that |
---|
0:07:24 | and then the idea ugly i good how do you are this this you using a us position at a |
---|
0:07:29 | and for example of this paper or you see that just a circular |
---|
0:07:34 | okay so |
---|
0:07:35 | if |
---|
0:07:36 | just use uh when many directional a speakers does then |
---|
0:07:40 | you put to D case |
---|
0:07:41 | the um |
---|
0:07:44 | each house case of more what which is you one by a proportional to this |
---|
0:07:49 | five the at least the H not use this zero or and and of function |
---|
0:07:53 | and then these are |
---|
0:07:55 | are is a a a way to to what any point in B |
---|
0:08:00 | we a um is and then are in the of a point where |
---|
0:08:05 | oh with that the uh |
---|
0:08:07 | but to were a one of the last |
---|
0:08:09 | so you can expand that a function |
---|
0:08:13 | you seeing his the known i is so |
---|
0:08:16 | for or do you favour yeah the are is that they use of the also that it you have this |
---|
0:08:21 | expansion |
---|
0:08:22 | and a full |
---|
0:08:24 | exterior for you you get is too small and then you have that |
---|
0:08:29 | so yeah in in a is the M board in did not handle a function and the G in he's |
---|
0:08:33 | the import a a a a a as a function |
---|
0:08:36 | so it |
---|
0:08:37 | if you use it go to |
---|
0:08:39 | sound source sort all speakers |
---|
0:08:41 | then |
---|
0:08:43 | the |
---|
0:08:44 | you can expect you would have a |
---|
0:08:47 | um |
---|
0:08:48 | a song you given by in these fact that we use in the the um |
---|
0:08:53 | that it would be better |
---|
0:08:55 | we the a a |
---|
0:08:57 | detection and debate |
---|
0:09:00 | so |
---|
0:09:02 | yeah |
---|
0:09:03 | B that really is a um and the B |
---|
0:09:07 | this uh |
---|
0:09:11 | and the to the vector from whatever the set |
---|
0:09:14 | so you can you can expand this this the |
---|
0:09:17 | you two cases the exterior case and they a so you use and the next year |
---|
0:09:22 | so um again you don't use a a a as a function come here |
---|
0:09:29 | so do |
---|
0:09:30 | you can see that there is a and and and and is in plastic that yeah |
---|
0:09:36 | otherwise the laws |
---|
0:09:37 | the the difference between X T D and the D V is that |
---|
0:09:40 | uh a the change of changing |
---|
0:09:44 | okay so |
---|
0:09:46 | a a dollar to sound so is this i mean if you combine |
---|
0:09:50 | say a a the and mine mine have lead to be that you can pretty much yeah it a five |
---|
0:09:55 | and signed by data |
---|
0:09:57 | be that but |
---|
0:09:59 | so it a lot of each use it general and the order to do the variable T |
---|
0:10:06 | we uh |
---|
0:10:07 | yeah can be additional these still we can say a and and of the other that in order of this |
---|
0:10:13 | will produce a fear |
---|
0:10:15 | given by then |
---|
0:10:17 | you know who the these of the month of what we still for each |
---|
0:10:20 | in the war |
---|
0:10:22 | uh |
---|
0:10:23 | a uh that so be |
---|
0:10:26 | this C B |
---|
0:10:28 | uh |
---|
0:10:29 | yeah i'll speak and in U the |
---|
0:10:32 | what's the waiting for the in the word |
---|
0:10:35 | so |
---|
0:10:37 | you can do it and huh |
---|
0:10:39 | what we could do is you can you quit that should do |
---|
0:10:43 | you you desired sound field so this is you'll some field due to the you use a means in the |
---|
0:10:49 | all all speakers and then |
---|
0:10:51 | you can do a great that do you would decide on we |
---|
0:10:54 | you know to good both sides by a um |
---|
0:10:57 | E J say in five and then um |
---|
0:11:01 | and get a more demanding solutions so called more a solution for |
---|
0:11:06 | will be to get a feel that exterior |
---|
0:11:09 | so you |
---|
0:11:10 | do you have a we have a a bunch of corporations speech use your decide |
---|
0:11:15 | of which i |
---|
0:11:16 | um this as your decide sound and then what we stay at a model |
---|
0:11:20 | a sixty via field um you are required a you would like |
---|
0:11:24 | that is no in it to goes outside so you can and |
---|
0:11:27 | assume be coefficients were you zero |
---|
0:11:31 | so you call these two haitians for you get more than X T V more full in goes from minus |
---|
0:11:37 | in |
---|
0:11:38 | so now you can actually to combine these two iterations to get the |
---|
0:11:42 | for a single |
---|
0:11:44 | system C |
---|
0:11:45 | see |
---|
0:11:46 | so the |
---|
0:11:47 | um at icsi creation |
---|
0:11:50 | so |
---|
0:11:51 | you can combine density H |
---|
0:11:54 | does include a um |
---|
0:11:57 | you know what these five does to get as so |
---|
0:11:59 | yeah i a it is then to try to find you E |
---|
0:12:03 | main peaks |
---|
0:12:04 | we make good she's for each again in each model me |
---|
0:12:09 | each also |
---|
0:12:10 | so |
---|
0:12:11 | a you can so that a and then um |
---|
0:12:13 | you can use it to uh |
---|
0:12:17 | yeah can |
---|
0:12:19 | so |
---|
0:12:21 | a was to to do that you need to have a |
---|
0:12:24 | you know a little available more you to be less that the |
---|
0:12:29 | it put at all |
---|
0:12:31 | yeah um |
---|
0:12:33 | C the in nice the high as a little but speaker and it is the number of lost |
---|
0:12:39 | um and also we know for a given region |
---|
0:12:42 | to have that |
---|
0:12:44 | uh you a number of we produce a little more of are given by frequency time i |
---|
0:12:51 | so uh we do matter that's a relation of the so we choose a circular area and at fifty i |
---|
0:12:57 | of that all speakers |
---|
0:12:59 | and add it is easy and he does |
---|
0:13:02 | so for this particular case if you just use a zero order that the corresponding i |
---|
0:13:07 | frequencies |
---|
0:13:08 | hundred |
---|
0:13:09 | so |
---|
0:13:10 | so this is i |
---|
0:13:12 | really is a so you have a result is due to a |
---|
0:13:17 | a source located at radius of six meters and and uh that is sixty three keys in fact ah ha |
---|
0:13:24 | two loudspeakers |
---|
0:13:27 | so |
---|
0:13:29 | this is the uh |
---|
0:13:31 | he produces on the other |
---|
0:13:34 | will be used |
---|
0:13:35 | the first-order loss does |
---|
0:13:37 | and |
---|
0:13:37 | frequency hundred hundred by uh |
---|
0:13:39 | so as you can see so the the the or see some yeah there are six meters does that they |
---|
0:13:44 | so you can see a feel inside and then |
---|
0:13:48 | um the on side exterior field is um |
---|
0:13:51 | lace |
---|
0:13:55 | and then in this |
---|
0:13:56 | case yeah D |
---|
0:13:58 | the same system |
---|
0:14:00 | so we got you know it's because the the but the data and |
---|
0:14:04 | three hundred fifty the so was you know this is a ball the night this frequency so that |
---|
0:14:09 | um corresponding E |
---|
0:14:13 | the |
---|
0:14:13 | the set little |
---|
0:14:16 | you know who a a a guess is yeah okay K be pretty C |
---|
0:14:20 | some has has been really use |
---|
0:14:22 | from the origin of that |
---|
0:14:27 | and |
---|
0:14:28 | and really got it yeah that's S here is defined by a suppress a |
---|
0:14:33 | difference be in the but is field and that the original you divided by the patient the origin |
---|
0:14:40 | um that |
---|
0:14:41 | for do you feel that well |
---|
0:14:43 | X T because all the of you use you zero so is just |
---|
0:14:48 | uh |
---|
0:14:51 | section |
---|
0:14:53 | so i this is the average production um there uh |
---|
0:14:57 | first order of speech data |
---|
0:15:00 | and we we hundred three fifty and it |
---|
0:15:05 | so as you can see um |
---|
0:15:08 | there is you can do uh side and |
---|
0:15:11 | yes the the inside that region |
---|
0:15:14 | uh you have a |
---|
0:15:18 | and |
---|
0:15:20 | okay and then you look at |
---|
0:15:22 | you see |
---|
0:15:23 | the door the |
---|
0:15:26 | also so on and still we can be not of lost just to fifty |
---|
0:15:29 | and then we increase |
---|
0:15:31 | the uh |
---|
0:15:32 | you you a region also but the three |
---|
0:15:35 | and the at that to you have so in this case of B S you can see |
---|
0:15:42 | original |
---|
0:15:42 | reproduction production has increased |
---|
0:15:45 | again |
---|
0:15:46 | and then all side but it's see a few of them |
---|
0:15:49 | be the listening to |
---|
0:15:53 | and this is the you but i it or is this case of as you can see "'em" was you |
---|
0:15:57 | mean you will was it well as the |
---|
0:15:59 | shah |
---|
0:16:01 | the door to |
---|
0:16:02 | mean |
---|
0:16:02 | got a lot of is but inside and outside this list of this |
---|
0:16:07 | and this C is a case we a seek store the |
---|
0:16:11 | also because again in know speakers and eight hundred uh so this case |
---|
0:16:16 | um no no is the other side |
---|
0:16:17 | use small zero |
---|
0:16:23 | so yes yes a could close and um use these file it allows is just to see |
---|
0:16:29 | you know what C whom and you can have a so what we have concluded that |
---|
0:16:33 | but using in order |
---|
0:16:35 | as you can |
---|
0:16:36 | L of them by in time |
---|
0:16:38 | you know i the frequency or radius of convergence all |
---|
0:16:42 | in this is that you can do not loss because by one or |
---|
0:16:48 | Q |
---|
0:16:51 | i |
---|
0:16:52 | i |
---|
0:16:53 | i |
---|
0:16:56 | oh |
---|
0:17:05 | uh_huh |
---|
0:17:06 | so |
---|
0:17:08 | um |
---|
0:17:09 | i think you can only code |
---|
0:17:13 | yeah i mean for example it that the what allows a cat who am and that order out could be |
---|
0:17:20 | say you produced by |
---|
0:17:21 | so see um |
---|
0:17:23 | number of circular |
---|
0:17:25 | all speakers it's so |
---|
0:17:26 | but it yeah |
---|
0:17:28 | yeah i i and that's that's it i haven't got |
---|
0:17:31 | well i mean of was that would be you know you tried to |
---|
0:17:34 | are you mean have a |
---|
0:17:36 | uh uh we go to that all speaker a problem |
---|
0:17:40 | um |
---|
0:17:42 | omnidirectional all speakers |
---|
0:17:44 | B |
---|
0:17:45 | yeah but but sites going to be increased well |
---|
0:17:48 | yes so at the end of the day are they going to get get we because so |
---|
0:17:52 | of course |
---|
0:17:53 | some |
---|
0:17:55 | a of this kind of things the the would look at at the |
---|
0:17:58 | um can of properties and things like that are are are that just |
---|
0:18:02 | based on uh |
---|
0:18:03 | action uh |
---|
0:18:08 | yeah |
---|
0:18:15 | i |
---|
0:18:22 | oh |
---|
0:18:25 | oh |
---|
0:18:26 | okay so i mean you know as you know there are not that many |
---|
0:18:31 | higher order to |
---|
0:18:32 | do to get all speakers that such |
---|
0:18:34 | so it is a game |
---|
0:18:36 | people have done yeah |
---|
0:18:37 | right |
---|
0:18:38 | so i mean if you just |
---|
0:18:40 | say kind of of what man vision of a higher order last he's using a |
---|
0:18:44 | um one a whole all speakers to get a |
---|
0:18:46 | so you that we at least to save number i |
---|
0:18:49 | well i was then |
---|
0:18:50 | it is more um |
---|
0:18:54 | so so other that have been um more else because are around the circle to use could be a |
---|
0:18:59 | a same number of rows because the around you |
---|
0:19:02 | point |
---|
0:19:02 | or around the set |
---|
0:19:03 | uh |
---|
0:19:06 | but i can i mean as i say you know that's kind of know how well the version of a |
---|
0:19:10 | high would allows speaker right |
---|
0:19:12 | well you know it's a technology that the video low |
---|
0:19:15 | we you know you using we gone a couple days and stuff like that |
---|
0:19:22 | i |
---|
0:19:22 | more |
---|
0:19:26 | i |
---|
0:19:26 | or |
---|
0:19:28 | one |
---|
0:19:29 | oh |
---|
0:19:34 | i |
---|
0:19:36 | and |
---|
0:19:37 | uh_huh or |
---|
0:19:38 | why |
---|
0:19:39 | yeah |
---|
0:19:44 | or |
---|
0:19:47 | so |
---|
0:19:50 | oh |
---|
0:19:52 | am |
---|
0:19:53 | and it |
---|
0:19:56 | no no i have a look at you mean a lower power consumption known no one |
---|
0:20:01 | that way |
---|
0:20:04 | i |
---|
0:20:07 | yeah |
---|
0:20:10 | i |
---|
0:20:10 | well |
---|
0:20:13 | you |
---|
0:20:14 | a |
---|
0:20:14 | oh |
---|
0:20:19 | okay |
---|
0:20:20 | just six |
---|
0:20:22 | i |
---|
0:20:22 | i |
---|
0:20:23 | i |
---|
0:20:26 | man |
---|
0:20:27 | i |
---|
0:20:30 | why |
---|
0:20:34 | all i |
---|
0:20:36 | i |
---|
0:20:37 | hmmm |
---|
0:20:41 | yeah |
---|
0:20:45 | okay |
---|
0:20:45 | okay |
---|
0:20:48 | i |
---|
0:20:48 | i |
---|