JP4427915B2 - Virtual sound image localization processor - Google Patents

Description

【００３４】
この数１の式における各符号の意味するものは、下記の通りである。
Ｓ：モノラル音響信号
ＳＬ：左耳用音響信号
ＳＲ：右耳用音響信号
ＨＬ：左耳用音響伝達関数
ＨＲ：右耳用音響伝達関数
ｎ：サンプル
ｓＬ：ＨＬの次数
ｓＲ：ＨＲの次数
【００３５】
直接音音響信号生成に使用される音響伝達関数ＨＬ、ＨＲは、無響室等の通常響きの少ない空間で測定されたデータに基づいて、作成される。
【００３６】
直接音音響信号生成手段２により生成された音響信号は、聴取者に音像定位させたい位置にスピーカを設置し、そのスピーカからモノラル音響を放声した際に聴取者の耳元に届く音響をシミュレートしたものである。従って、これをヘッドフォンで受聴した聴取者は上述したスピーカの位置に音像定位すると考えられるが、実際にはほとんど頭外定位しない。無響室など響きの少ない、つまり略直接音のみの空間では音像の距離感はほとんど出ないためである。
【００３７】
音像の距離感を得るための要因としてはいくつかの物理的要因や心理的要因が挙げられているが、その中で間接音の存在は重要な要因の一つとされている。そこで、直接音音響信号に加えて更に間接音音響信号成分を生成、付加することによりヘッドフォン受聴における音像の頭外定位感を向上させる手法が行われている。
【００３８】
間接音音響の物理的役割は両耳に受ける音響の相関性（両耳相関）を減らすことであると言われている。相関性を減らすことにより音像が頭外にあるものとして知覚されると考えられている。
【００３９】
ところで、通常自然界での間接音音響は、複数の方向から到来し、時間的には数百ミリ秒から数秒程度まで遅延した音響から構成される。このような間接音音響信号を上述したような音響伝達関数の重畳演算にて生成するとなると膨大な演算量となり、特にリアルタイム演算が要求されるテレビゲーム機などの分野では非現実的なものとなってしまう。
【００４０】
そこで、従来、より少い演算量ながら両耳相関を効果的に低くする手法として、聴取者側方から到達する数少ない本数の間接音を使用するものが提案されていた。しかしながら、この手法においても、テレビゲーム機の分野などにおいては、未だ高負荷となっており、より低負荷な間接音音響信号の生成が求められていた。
【００４１】
ここで、先に述べたように間接音は両耳相関を減らす、という物理的機能に主として注目し、間接音音響成分に使用される音響伝達関数の細部構造の頭外定位感効果に対する寄与は小さいものと推測した。ここで側方間接音音響の構造を極力単純化させるための試聴及び実験を行ったところ、仮想聴取者の左右の耳に対し、左及び右側方の仮想音源からの音響伝達関数は、例えば、時間軸インパルス、その逆側である右及び左側方の仮想音源からの音響伝達関数は、例えば、時間軸矩形波で代用しても有意な頭外定位感効果が得られることが判明した。
【００４２】
図１に戻り、モノラル音響信号源１よりのモノラル音響信号を遅延させる遅延手段３からの音響信号を入力とする左側間接音音響信号生成手段５は、上述した聴取者側方（ここでは左側方）から到達する間接音音響信号を生成するものである。しかしながら、上述したような従来手法が音響伝達関数を使用した高負荷な高次重畳処理であるのに対し、本発明で提案する左側間接音生音響信号成手段５は、仮想聴取者の左耳までの音響伝達特性を時間軸インパルスにて近似代用し、仮想聴取者の右耳までの音響伝達特性を、例えば、時間軸矩形波にて近似代用する演算量的に低負荷なものである。同様に、右側間接音音響信号生成手段６は、仮想聴取者の右耳までの音響伝達特性を時間軸インパルスにて近似代用し、仮想聴取者の左耳までの音響伝達特性を、例えば、時間軸矩形波にて近似代用する演算量的に低負荷なものである。
【００４３】
図２は、図１における遅延手段３、４並びに左及び右間接音音響信号生成手段５、６を更に具体的に示したものである。以下、図２について説明する。ＤＬは１サンプル周期に等しい遅延時間を有する遅延器を示す。モノラル音響信号源１よりのモノラル音響信号が、縦続接続された所定個数の遅延器ＤＬから構成される遅延手段３によって遅延された後、左側間接音音響信号生成手段５に供給される。
【００４４】
左側間接音生成手段５では、遅延手段３よりの遅延モノラル音響信号を、振幅重み付け係数がα１の係数乗算器（減衰手段）を通じて、左耳用間接音音響信号５Ｌを得るようにする。また、遅延手段３よりの遅延モノラル音響信号を、縦続接続された所定個数の遅延器ＤＬから構成される遅延手段５Ｄを通じて、デジタルフィルタ５Ｆに供給する。このフィルタ５Ｆは、遅延手段５Ｄの出力側の１個のタップ及び遅延手段５Ｄに縦続接続された所定個数の遅延器ＤＬから構成される遅延手段における、遅延手段５Ｄの出力側の１個のタップに連続した（ｒ−２）個、計（ｒ−１）個のタップを取り出し、それぞれのタップを振幅重み付け係数がα２、α３、…、α（ｒ−１）、αｒの係数乗算器を通じて、加算器５Ｓに供給して加算して、右耳用間接音音響信号５Ｒとして出力する。即ち、デジタルフィルタ５Ｆは、ＦＩＲ（Finit Impulse Response）（有限インパルス応答）フィルタを構成している。
【００４５】
ここで、左側間接音音響信号生成手段５の左耳用間接音音響信号５Ｌに重畳した特性は時間軸インパルスとなる。フィルタ５Ｆにおける振幅重み付け係数α２、‥‥、α（ｒ−１）、αｒの値を全て等しくすると、左側間接音音響信号生成手段５の右耳用音響信号に重畳した特性は時間軸矩形波となる。更に、次の数２の式の
【００４６】
【数２】

【００４７】
ような関係に設定することにより、左側間接音音響信号生成手段５の左耳用間接音音響信号５Ｌ及び右耳用間接音音響信号５Ｒに重畳した低域振幅特性は略等しくなり、中域から高域にかけての右耳用間接音音響信号５Ｒの振幅特性のエンベロープは、左耳用間接音音響信号５Ｌのそれに対して高域方向に減衰してゆく特性となる。このような相対的な振幅特性はマクロ的には、仮想聴取者左側方から到達する音響伝達関数のそれと類似したものとなる。
【００４８】
また、左側間接音音響信号生成手段５の時間軸上の左右出力波形の相対位置（時間遅れ）を、仮想聴取者の左側方から到達する音響伝達関数のそれと類似したものに設定することにより、位相特性に関してもマクロ的には、仮想聴取者の左側方から到達する音響伝達関数のそれと類似したものにすることができる。即ち、仮想聴取者の左側方から到達する音波は、先ず、左耳に到達した後に、左右の両耳間距離を音波が伝播する時間だけ、やや遅れて右耳に到達する。そのとき、聴取者の頭部による回折効果などにより、右耳へ到達する音波は相対的に中高音域が減衰されて到達する。
【００４９】
このようにして左側間接音音響信号生成手段５の出力音響信号は、仮想聴取者の左側方から到達する音響伝達関数を簡略化したものとして扱うことができる訳である。
【００５０】
次に、図５を参照して、左側間接音音響信号生成手段５の原理を説明する。図５Ａに示すように、聴取者ＬＮの左側方に仮想音源ＩＳを設け、その仮想音源ＩＳから、聴取者ＬＮの左耳Ｅ_L 及び右耳Ｅ_R までの音響伝達関数（インパルス応答）をそれぞれＨ_L 、Ｈ_R とする。
【００５１】
この音響伝達関数（インパルス応答）Ｈ_L 、Ｈ_R を示す時間軸波形の一例を、図５Ｂ、Ｃに示す。尚、図５Ｂ、Ｃにおけるｔは時間を示す。仮想音源ＩＳから聴取者ＬＮの右耳Ｅ_R に到達する音波は、左耳Ｅ_L に到達する音波に比べてΔｔだけ遅れて到達する（図５Ｃ）。また、図５Ｂ、Ｃの時間特性を、周波数軸に変換した特性を、図５Ｄに、それぞれ▲１▼及び▲２▼として示す。図５Ｂ、Ｃ及びＤに明らかなように、仮想音源ＩＳから聴取者ＬＮの右耳Ｅ_R に到達する音波は、仮想音源ＩＳから聴取者ＬＮの左耳Ｅ_L に到達する音波に比べて、聴取者ＬＮの頭部の存在によって、中高音成分が減衰していることが分かる。
【００５２】
図５Ｂ、Ｃに示すインパルス応答を示す時間軸波形を単純化した音響伝達関数（インパルス応答）を示す時間軸波形を、図５Ｅ及びＦにそれぞれＨ_L ′、Ｈ_R ′として示す。図５Ｅの音響伝達関数（インパルス応答）Ｈ_L ′は、振幅がα１のインパルスであり、図５Ｆの音響伝達関数（インパルス応答）Ｈ_R ′は、図５Ｅの音響伝達関数（インパルス応答）Ｈ_L ′よりΔｔだけ遅延した矩形波応答で、各サンプリング時間毎の振幅値の総和が、α１となるように設定されている。図５Ｆの矩形波応答が、図５Ｇに丸印で示すように、例えば、３サンプルから構成されている場合は、各サンプル値は共にα１／３となる。
【００５３】
図５Ｅ及びＦの音響伝達関数（インパルス応答）を示す時間軸波形を、周波数軸波形に変換して示すと、それぞれ図５Ｈの▲３▼、▲４▼のようになり、▲３▼は周波数の変化に対し平坦な特性を呈し、▲４▼はローパスフィルタ特性（中高音域を単純減衰させる特性）を示す。また、図５Ｈの応答▲３▼、▲４▼の積分値、即ち、各サンプル点の振幅値の総和が、両者で一致しているので、低音域での周波数特性は、略一致する。
【００５４】
再び、図２に戻って説明する。モノラル音響信号源１よりのモノラル音響信号が、それぞれ縦続接続された所定個数の遅延器ＤＬから構成される、縦続接続された遅延手段３、４によって遅延された後、右側間接音音響信号生成手段６に供給される。遅延手段４は、上述の遅延手段５Ｄ、フィルタ５Ｆを構成する各遅延器ＤＬ及びその他の各遅延器５Ｄから構成される。
【００５５】
右側間接音音響信号生成手段６では、遅延手段３及び４によって遅延されたモノラル音響信号、即ち、遅延モノラル音響信号を、振幅重み付け係数がβ１の係数乗算器を通過させることによって、右耳用音響信号６Ｒを得るようにする。また、遅延手段４よりの遅延モノラル音響信号を、縦続接続された所定個数の遅延器ＤＬから構成される遅延手段６Ｄを通じて、デジタルフィルタ６Ｆに供給する。このフィルタ６Ｆは、遅延手段６Ｄの出力側の１個のタップ及び遅延手段６Ｄに縦続接続された所定個数の遅延器ＤＬから構成される遅延手段における、遅延手段６Ｄの出力側の１個のタップに連続した（ｋ−２）個、計（ｋ−１）個のタップを取り出し、それぞれのタップを振幅重み付け係数がβ２、β３、…、β（ｋ−１）、βｋの係数乗算器を通じて、加算器６Ｓに供給して加算して、左耳用音響信号６Ｌとして出力する。このデジタルフィルタ６Ｆも、上述のデジタルフィルタ５Ｆと同様に、ＦＩＲフィルタを構成している。
【００５６】
ここで、右側間接音音響信号生成手段６の右耳用音響信号６Ｒに重畳した特性は時間軸インパルスとなる。フィルタ６Ｆにおける振幅重み付け係数β２、‥‥、β（ｋ−１）、βｋの値を全て等しくすると、左側間接音音響信号生成手段５の右耳用音響信号に重畳した特性は時間軸矩形波となる。更に、次の数３の数式
【００５７】
【数３】

【００５８】
のような関係に設定することにより、右側間接音音響信号生成手段６の右耳用音響信号６Ｒ及び左耳用音響信号６Ｌに重畳した低域振幅特性は略等しくなり、中域から高域にかけての左耳用音響信号６Ｌに重畳した振幅特性のエンベロープは、右耳用音響信号６Ｒのそれに対して高域方向に減衰してゆく特性となる。このような相対的な振幅特性はマクロ的には、仮想聴取者の左側方から到達する音響伝達関数のそれと類似したものとなる。
【００５９】
また、右側間接音音響信号生成手段６の時間軸上の左右出力波形の相対位置（時間遅れ）を、仮想聴取者の右側方から到達する音響伝達関数のそれと類似したものに設定することにより、位相特性に関してもマクロ的には、仮想聴取者の右側方から到達する音響伝達関数のそれと類似したものにすることができる。
【００６０】
このようにして右側間接音音響信号生成手段６の出力音響信号は、仮想聴取者の右側方から到達する音響伝達関数を簡略化したものとして扱うことができる訳である。
【００６１】
図２では、仮想聴取者の左側方から到達する間接音音響信号の方が、仮想聴取者の右側方から到達する間接音よりも時間的に早く、仮想聴取者の耳元に到達する関係になっているが、これは逆でも良い。
【００６２】
また、上述の左側及び右側間接音音響信号生成手段５、６のいずれか一方のみを使用しても、音像の頭外定位感や奥行き感が十分に改善されることが、実験的に確かめられている。
【００６３】
左側及び右側間接音生成手段５、６において、仮想音源と逆側の耳に到達する波形を生成するために重畳した伝達関数は時間軸矩形波に限ったものではなく、数２及び数３の式の条件を満たし、かつ、時間軸短時間波形に置き換えることも有効である。すなわち、α₂ 、α₃ 、‥‥‥‥、α_r に関して必ずしも等しい値に設定する必要はない。同様に、β₂ 、β₃ 、‥‥‥‥、β_k に関しても必ずしも等しい値に設定する必要はない。
【００６４】
図５Ｇに示す時間軸波形の変形例を、図６Ａ、Ｂ、Ｃに示す。尚、図６Ａ、Ｂ、Ｃにおける丸印はサンプリング点を示す。以下、これらの変形例について説明する。
【００６５】
図６Ａに示す時間軸波形は、２次または３次多項式を適用した例であり、図６Ｂは４次または５次多項式を適用した例である。図５Ｇに示す矩形波を含めて、いずれも移動加算平均演算に用いられる重み係数列である入力信号に、これらの時間軸波形（伝達関数）を重畳演算すれぱ、いずれも高音域が減衰した出力言号が得られるが、時間軸波形によって高音域の減衰特性が変化する。
【００６６】
尚、図５Ｇに示す矩形波の場合は、例えば、ＤＳＰ（デジタル・シグナル・プロセッサ）を使用した信号処理では、係数データを格納するメモリが１つで済み、更に、サンプリング点の数が２の羃乗、例えは、４であれば、それぞれの重み係数値は、α₁ ／４となるので、ビットシフト演算により求めることができ、演算処理の負荷がより一層軽減されるので、好適である。
【００６７】
図６Ｃに示す時間軸波形は、指数関数を適用して疑似的にＲＣフィルタを模した例である。この場合も、サンプリング点は高々数点でも良いが、後述のＩＩＲフィルタによって実施された場合には、係数データメモリや演処理量が非常に少なくて済む利点がある。
【００６８】
また、伝達関数が時間軸矩形波であるＦＩＲフィルタの代わりとして中高音域の振幅を単純減衰させる特性を有するフィルタやイコライザにて近似代用することも有効である。図３は、この場合における、図１における遅延手段３、４並びに左及び右間接音音響信号生成手段５、６を具体的に示したものである。
【００６９】
以下、図３について説明する。基本構成は図２と同様であり、異なるのは左側間接音音響信号生成手段５の右耳用音響信号生成部及び右側間接音音響信号生成手段６の左耳用音響信号生成部である。即ち、左側間接音音響信号生成手段５では、左耳用音響信号として取り出したタップ位置より数タップから数十タップ後方の位置より１タップを取り出し、振幅重み付け係数α₂ を乗じた後、中高音域の振幅を単純減衰させる特性（図５Ｈの特性曲線▲４▼参照）であるイコライザ５Ｅにより処理した信号を右耳用間接音音響信号５Ｒとして出力する。イコライザ５Ｅとしては、例えば低次ＩＩＲ( Infinite Impulse Response ）フィルタからなるＬＰＦ( Low Pass Filter ) で構成する。
【００７０】
右側間接音音響信号生成手段６では、右耳用間接音音響信号６Ｒとして取り出したタップ位置より数タップから数十タップ後方の位置より１タップを取り出し、振幅重み付け係数β₂ を乗じた後、中高音域の振幅を単純減衰させる特性であるイコライザ６Ｅにより処理した信号を左耳用間接音音響信号６Ｌとして信号出力する。イコライザ６Ｅとしては、例えば低次ＩＩＲフィルタで構成されたＬＰＦで構成する。
【００７１】
尚、本発明の実施の形態では、デジタル信号処理による構成としたが、この振幅重み付け係数α₂ 、β₂ を乗ずる係数乗算器やイコライザ５Ｅ、６Ｅの前にＤ／Ａ変換器を設けてアナログ信号として取り出せば、これらを簡単なＣＲフィルタなどのアナログローパスフィルタで構成するようにしても良い。
【００７２】
左側及び右側間接音生成手段５、６において、右耳用間接音音響信号５Ｒ及び左耳用間接音音響信号６Ｌに重畳した低域振幅特性がそれぞれ左耳用間接音音響信号５Ｌ及び右耳用間接音音響信号６Ｒに重畳した低域振幅特性とが略等しくなるように、イコライザ５Ｅ及び６Ｅのの構成及び係数乗算器の係数α２及びβ２をそれぞれα₁ 及びβ₁ と互いに略等しくなるように設定するのが望ましい。
【００７３】
図４は、図１における加算手段７を更に具体的に示したものである。以下、図４について説明する。
【００７４】
この加算手段７は、左耳用音響信号用加算器７ＳＬ及び右耳用音響信号用加算器７ＳＲを備える。そして、直接音音響信号生成手段２よりの左耳用音響信号２Ｌ｛ＳＬ〔ｎ〕｝、左側間接音音響信号生成手段５よりの左耳用間接音音響信号５Ｌ及び右側間接音音響信号生成手段６よりの左耳用間接音音響信号６Ｌが、左耳用音響信号用加算器７ＳＬに供給されて加算されて、その左耳用加算信号７Ｌがヘッドフォン８の左耳用スピーカに供給される。又、直接音音響信号生成手段２よりの右耳用音響信号２Ｒ｛ＳＲ〔ｎ〕｝、左側間接音音響信号生成手段５よりの右耳用間接音音響信号５Ｒ及び右側間接音音響信号生成手段６よりの右耳用間接音音響信号６Ｒが、右耳用音響信号用加算器７ＳＲに供給されて加算されて、その右耳用加算信号７Ｒがヘッドフォン８の右耳用スピーカに供給される。
【００７５】
このようにして、聴取者はヘッドフォン受聴による音像の頭外定位感を享受することが可能となる。
【００７６】
尚、上述の実施の形態においては、各信号処理ブロックを回路として説明したが、その代わりに、マイクロコンピュータやＤＳＰなどのソフトウェアとして信号処理を行うようにしても良いことは勿論である。
【００７７】
更に、間接音音響信号生成方法、又は、それを含んだ仮想音像定位処理方法をプログラム・ステップとして記録した記録媒体（光ディスク、光磁気ディスク、磁気ディスク、磁気テープ、半導体メモリ等）として提供することも可能である。例えば、数種類の間接音音響信号生成パラメータ（遅延時間や高音域減衰特性など）をこの処理プログラムと共に記録媒体に記憶しておき、聴取者の嗜好に応じて選択させたり、再生信号ソースの制作者側の意図により、適宜切り換えたりすることが可能となる。
【００７８】
これらのプログラムは、テレビゲーム機のために提供されるゲームソフト用記録媒体に記録されても良いし、別体の記録媒体に記録されて供給されても良い。
【００７９】
上述の遅延時間は、上述の図１の遅延手段３、４で与えられるものに相当する。上述の高音域減衰特性は上述の図１の左側及び右側間接音音響信号生成手段５、６で与えられる時間軸伝達特性または周波数特性に相当し、これらによって形成される間接音音響信号のレベルに対して、相対的に変化させたり、これらのパラメータを変化させることにより、音像の当該定位感や奥行き感を変化させることができる。この場合にも、頗るコンパクトな信号処理量を以て、間接音音響信号を得ることができるので、マイクロコンピュータなどの信号処理への負荷は頗る小さく、また、パラメータとしての情報量も頗る小さいので、記録媒体における記録容量を大きく占有することはない。
【００８０】
上述のヘッドフォンは、聴取者の頭部に装着して聴取するものに限らず、聴取者の耳介や耳道に直接装着するイヤーホンでも良く、又、ヘッドフォンの代わりに、聴取者の耳近傍に配置されたスピーカ装置であっても良い。
【００８１】
【発明の効果】
本発明によれば、コンパクトな信号処理量の音像定位処理装置を得ることができる。
【図面の簡単な説明】
【図１】本発明の実施の形態の仮想音像定位処理装置が適用されたヘッドフォン供給用音響信号生成装置（ヘッドフォン装置）の例を示すブロック図である。
【図２】図１の本発明の実施の形態の仮想音像定位処理装置が適用されたヘッドフォン供給用音響信号生成装置（ヘッドフォン装置）の例の左側及び右側間接音音響信号生成手段の具体構成例を示すブロック図である。
【図３】図１の本発明の実施の形態の仮想音像定位処理装置が適用されたヘッドフォン供給用音響信号生成装置（ヘッドフォン装置）の例の左側及び右側間接音音響信号生成手段の他の具体構成例を示すブロック図である。
【図４】図１の本発明の実施の形態の仮想音像定位処理装置が適用されたヘッドフォン供給用音響信号生成装置（ヘッドフォン装置）の例の加算手段の具体構成例を示すブロック図である。
【図５】右側間接音音響信号生成手段の原理説明のための図である。
【図６】時間軸波形の他の例を示す波形図である。
【符号の説明】
１ モノラル音響信号源、２ 直接音音響信号生成手段、３ 遅延手段、４ 遅延手段、５ 左側間接音音響信号生成手段、６ 右側間接音音響信号生成手段、７ 加算手段、７ＳＬ、７ＳＲ 加算器、８ ヘッドフォン。[0001]
  The present inventionVirtual sound localization processing device.
[0002]
[Prior art]
Normally, when a stereo sound signal or a monaural sound signal is directly supplied to, for example, a headphone and reproduced, a sound image perceived by a listener is generally localized in the head. In order to obtain a natural out-of-head localization feeling during headphone playback, an acoustic signal subjected to virtual sound image localization processing may be supplied to the headphones.
[0003]
On the other hand, in recent years, in the field of video game machines and the like, there has been an increasing demand for positioning the monaural sound in various directions when viewed from the listener. Headphones are used as one of the sound reproduction devices, and virtual sound image localization processing is performed. There has also been proposed an apparatus in which a sound signal is supplied to headphones. Furthermore, a speaker is used as the sound reproducing device, and the sound signal subjected to the virtual sound image localization processing is further subjected to crosstalk processing, and then supplied to two speakers arranged on the front left and right of the listener to be placed at an arbitrary position. Some have been proposed to localize the sound image. In some cases, the sound signal subjected to the virtual sound image localization process is supplied to the headphones after being subjected to a crosstalk process.
[0004]
Here, an arbitrary monaural sound signal is signal-processed and supplied to, for example, headphones, and at the time of reproduction, the sound image is localized outside the listener's head, or reproduced by a speaker to localize the sound image at an arbitrary position. In order to give a sense of depth, binaural signal processing that superimposes the acoustic transfer function from the position to be localized to the ear of the listener is performed, and the generated acoustic signal is supplied to the headphones as an input signal, or Furthermore, it is good to add a crosstalk / cancellation process and supply it to the speaker.
[0005]
  In this specification, such binaural signal processing, or crosstalk keyTurbochargerNseLeThe signal processing to which the processing is added is referred to as virtual sound image localization processing.
[0006]
However, even if the monaural sound signal is subjected to the virtual sound localization process, if the indirect sound component of the monaural sound signal itself is small, the sound image is sufficient even if the virtual sound localization process is supplied to the headphones or speakers. It is difficult to obtain a sense of out-of-head localization and a sense of depth.
[0007]
  So, in the past,NoAdd an indirect sound signal such as a reflected sound of a suitable room to the sound signal, or add a sound signal that has been delayed, phase-shifted, or attenuated to the original sound signal as an indirect sound signal. Was done.
[0008]
Here, by adding a few indirect sound signals to the monaural sound signal, it is effective to set the direction of arrival of the indirect sound to about 45 to 90 degrees in the horizontal plane when viewed from the listener, that is, to the side of the listener. It is said that there are (refer to the AA-81-32, 1981 of the Acoustical Society of Japan, Architectural Acoustics Research Committee document, “The effect of two and three reflected sounds on the sense of distance and spread of sound” .
[0009]
Therefore, it is effective to use as an indirect sound signal an acoustic signal that has been subjected to a virtual sound localization process that superimposes the acoustic transfer function from the side to the listener's ear on the monaural sound signal delayed by a predetermined time. It is.
[0010]
[Problems to be solved by the invention]
However, in general, the acoustic transfer function obtained by measurement has, for example, several tens of samples on the time axis when the sampling frequency is 44.1 kHz, and this is superimposed on the monaural acoustic signal for indirect sound generation. The processing to be performed is a problem in fields such as real-time processing games where a large amount of processing is required and a compact signal processing amount is required, and a demand for a more compact virtual sound image localization processing is increasing.
[0011]
  In view of this point, the present invention provides a compact amount of signal processing.A virtual sound localization processing device is proposed..
[0012]
[Means for Solving the Problems]
In order to solve the above problems, a virtual sound localization processing apparatus according to the present invention is a direct sound / acoustic signal generator that receives a monaural sound signal source and generates a sound signal portion that becomes a direct sound / acoustic signal for a virtual listener. And a monophonic sound signal source, a first delay unit to which one or more delay units having a delay time equal to one sample period are connected in series, an output signal of the first delay unit is input, and the delay unit is The second delay unit connected in series, the output signal of the second delay unit is input, the first digital filter of the FIR type that adds all the output signals of the delay devices connected in series with the same tap coefficient, The output signal of the first digital filter is input, and the third delay unit to which one or more delay units are connected in series and the output signal of the third delay unit are input, and the output signals of the plurality of delay units connected in series are all equal. T One of the stereo outputs by adding the FIR type second digital filter added by the coefficient, the output signal of the direct sound and acoustic signal generation unit, the input signal of the second delay unit, and the output signal of the second digital filter And adding the output signal of the direct sound / acoustic signal generation unit, the input signal of the third delay unit, and the output signal of the first digital filter to the other channel of the stereo output It comprises.
[0023]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an example in which a virtual sound image localization processing device and a processing method according to an embodiment of the present invention are applied to a headphone supply acoustic signal generation device (headphone device) will be described with reference to the drawings. FIG. 1 shows the overall configuration of this headphone supply acoustic signal generation device (headphone device) example, which will be described below. Here, it is assumed that a digital monaural sound signal (digital input sound signal) is generated from the monaural sound signal source 1, but in the following description, it is simply referred to as a monaural sound signal or an input sound signal. 1 constitutes a headphone device as a whole, and a portion excluding the headphone 8 constitutes a headphone supplying acoustic signal generating device.
[0024]
It is also possible to add a crosstalk cancellation means to the final stage of the portion excluding the headphone 8 in FIG.
[0025]
The monaural sound signal from the monaural sound signal source 1 is supplied to the direct sound sound signal generation means 2 and subjected to virtual sound image localization processing to generate a direct sound sound signal, and the direct sound sound signal is supplied to the addition means 7. Is done.
[0026]
The monaural sound signal from the monaural sound signal source 1 is delayed for a predetermined time by the delay means 3 and then supplied to the left indirect sound sound signal generating means 5 to generate the left indirect sound sound signal, and the left indirect sound sound is generated. The signal is supplied to the adding means 7.
[0027]
The monaural sound signal from the monaural sound signal source 1 is delayed for a predetermined time by the delay means 3 and 4 and then supplied to the right indirect sound sound signal generation means 6 to generate the right indirect sound sound signal and the right indirect sound signal. A sound-acoustic signal is supplied to the adding means 7. The delay means 4 also includes a delay means in the left indirect sound acoustic signal generation means 5.
[0028]
The delay means 3 and 4 are connected in parallel to the monaural sound signal source 1, and the delayed monaural sound signals from the delay means 3 and 4 are individually connected to the left indirect sound sound signal generating means 5 and the right indirect sound sound signal. You may make it supply to the production | generation means 6. FIG.
[0029]
Then, first and second channel headphone supply sound signals obtained from the adding means 7 are supplied to the headphone 8.
[0030]
The direct sound / acoustic signal generating means 2 performs a process of generating a sound signal portion that becomes a direct sound / acoustic signal for a virtual listener using the monaural sound signal from the monaural sound signal source 1 as an input signal. The direct sound can be generated by superimposing an acoustic transfer function from a position where the virtual listener wants to perceive a sound image to the ear of the virtual listener on a monaural acoustic signal that is an input signal.
[0031]
  In this specification, a convolution integral operation of impulse time response is performed on the input time signal, or a transfer frequency characteristic is applied in order to change the frequency characteristic of the input signal like an equalizer or a filter described later. Are collectively referred to as superposition processing or superposition computation..
[0032]
The acoustic transfer function is an acoustic impulse response from the position where the virtual listener wants to perceive a sound image to the ear of the virtual listener, and the acoustic transfer function HL to the left ear and the acoustic transfer function HR to the right ear for one position. Exists. The direct sound signal for left ear is obtained by superimposing the sound transfer function HL up to the left ear on the monaural sound signal. Similarly, a direct sound signal for right ear is obtained by superimposing a sound transfer function HR up to the right ear on a monaural sound signal. Specifically, the superimposition calculation is expressed by the following equation (1).
[0033]
[Expression 1]

[0034]
The meaning of each symbol in the formula 1 is as follows.
S: Monaural sound signal
SL: Acoustic signal for left ear
SR: Acoustic signal for right ear
HL: Acoustic transfer function for left ear
HR: Sound transfer function for right ear
n: Sample
sL: HL order
sR: HR order
[0035]
The acoustic transfer functions HL and HR used for generating the direct sound acoustic signal are created based on data measured in a space with little normal sound such as an anechoic room.
[0036]
The sound signal generated by the direct sound signal generating means 2 simulates the sound that reaches the listener's ear when a speaker is installed at a position where the listener wants to localize the sound image and monaural sound is emitted from the speaker. Is. Therefore, it is considered that the listener who has listened to this with the headphones is localized at the position of the speaker described above, but in reality, it is hardly localized outside the head. This is because there is almost no sense of distance in the sound image in an anechoic room where there is little reverberation, that is, in an almost direct sound space.
[0037]
There are several physical factors and psychological factors as factors for obtaining a sense of distance in the sound image. Among them, the presence of indirect sound is regarded as one of the important factors. In view of this, a technique for improving the sense of out-of-head localization of a sound image in listening to headphones by generating and adding an indirect sound sound signal component in addition to the direct sound sound signal has been performed.
[0038]
It is said that the physical role of indirect sound acoustics is to reduce the correlation between the sounds received by both ears (binaural correlation). It is considered that the sound image is perceived as being out of the head by reducing the correlation.
[0039]
By the way, the indirect sound sound in the natural world is usually composed of sound that arrives from a plurality of directions and is delayed from several hundred milliseconds to several seconds. If such an indirect sound signal is generated by the superimposition calculation of the acoustic transfer function as described above, the calculation amount becomes enormous, and it becomes unrealistic particularly in the field of video game machines and the like that require real-time calculation. End up.
[0040]
Therefore, conventionally, as a technique for effectively lowering the binaural correlation while using a smaller amount of calculation, a technique using a small number of indirect sounds reaching from the listener side has been proposed. However, even in this method, in the field of video game machines and the like, the load is still high, and the generation of a lower load indirect acoustic signal has been demanded.
[0041]
Here, as mentioned above, the indirect sound mainly focuses on the physical function of reducing the binaural correlation, and the contribution to the out-of-head localization effect of the detailed structure of the acoustic transfer function used for the indirect sound acoustic component is I guess it was small. Here, when the trial listening and the experiment for simplifying the structure of the side indirect sound as much as possible were performed, the acoustic transfer functions from the left and right virtual sound sources for the left and right ears of the virtual listener are, for example, It has been found that the sound transfer function from the right and left virtual sound sources on the opposite side of the time axis impulse can obtain a significant out-of-head localization effect even if the time axis rectangular wave is substituted, for example.
[0042]
Returning to FIG. 1, the left indirect sound sound signal generation means 5 that receives the sound signal from the delay means 3 that delays the monaural sound signal from the monaural sound signal source 1 is connected to the listener side (here, the left side). ) To generate an indirect sound signal. However, while the conventional method as described above is a high-load high-order superimposition process using an acoustic transfer function, the left indirect sound signal generating means 5 proposed in the present invention is a virtual listener's left ear. The acoustic transmission characteristics up to and including the time axis impulse are approximated, and the acoustic transmission characteristics up to the right ear of the virtual listener are approximated using, for example, the time axis rectangular wave and the amount of computation is low. Similarly, the right-side indirect sound acoustic signal generation means 6 approximates the sound transfer characteristic up to the virtual ear's right ear with a time axis impulse, and uses the sound transfer characteristic up to the virtual listener's left ear as, for example, time The amount of calculation is low and the load is approximated by an axial rectangular wave.
[0043]
FIG. 2 shows more specifically the delay means 3 and 4 and the left and right indirect sound signal generation means 5 and 6 in FIG. Hereinafter, FIG. 2 will be described. DL denotes a delay device having a delay time equal to one sample period. The monaural sound signal from the monaural sound signal source 1 is delayed by the delay means 3 composed of a predetermined number of delay devices DL connected in cascade, and then supplied to the left indirect sound sound signal generation means 5.
[0044]
The left indirect sound generating means 5 obtains the left ear indirect sound acoustic signal 5L from the delayed monaural acoustic signal from the delay means 3 through a coefficient multiplier (attenuating means) having an amplitude weighting coefficient α1. Further, the delayed monaural sound signal from the delay means 3 is supplied to the digital filter 5F through the delay means 5D constituted by a predetermined number of delay devices DL connected in cascade. This filter 5F has one tap on the output side of the delay means 5D in the delay means composed of one tap on the output side of the delay means 5D and a predetermined number of delay devices DL cascaded to the delay means 5D. (R-2) and (r-1) total taps are extracted, and each tap is passed through a coefficient multiplier with amplitude weighting coefficients α2, α3,..., Α (r−1), αr, The signal is supplied to the adder 5S, added, and output as an indirect sound signal 5R for the right ear. That is, the digital filter 5F constitutes an FIR (Finit Impulse Response) filter.
[0045]
Here, the characteristic superimposed on the left ear indirect sound acoustic signal 5L of the left indirect sound acoustic signal generating means 5 is a time-axis impulse. When the values of the amplitude weighting coefficients α2,..., Α (r−1), αr in the filter 5F are all equal, the characteristic superimposed on the right ear acoustic signal of the left indirect sound acoustic signal generating means 5 is a time axis rectangular wave. Become. Furthermore, the following equation 2
[0046]
[Expression 2]

[0047]
By setting the relationship as described above, the low frequency amplitude characteristics superimposed on the left ear indirect sound acoustic signal 5L and the right ear indirect sound acoustic signal 5R of the left indirect sound acoustic signal generating means 5 become substantially equal, The envelope of the amplitude characteristic of the indirect sound acoustic signal 5R for the right ear over the high frequency is a characteristic that attenuates in the high frequency direction with respect to that of the indirect sound acoustic signal 5L for the left ear. Such a relative amplitude characteristic is macroscopically similar to that of an acoustic transfer function that arrives from the left side of the virtual listener.
[0048]
Also, by setting the relative position (time delay) of the left and right output waveforms on the time axis of the left indirect sound signal generation means 5 to be similar to that of the acoustic transfer function that arrives from the left side of the virtual listener, The phase characteristic can be made macroscopically similar to that of the acoustic transfer function that arrives from the left side of the virtual listener. That is, the sound wave that arrives from the left side of the virtual listener first reaches the left ear, and then reaches the right ear with a slight delay for the time that the sound wave propagates through the distance between the left and right ears. At that time, due to the diffraction effect by the listener's head, etc., the sound wave reaching the right ear arrives with the mid-high range relatively attenuated.
[0049]
Thus, the output acoustic signal of the left indirect sound acoustic signal generating means 5 can be treated as a simplified acoustic transfer function that arrives from the left side of the virtual listener.
[0050]
Next, with reference to FIG. 5, the principle of the left indirect sound acoustic signal generation means 5 will be described. As shown in FIG. 5A, a virtual sound source IS is provided on the left side of the listener LN, and the left ear E of the listener LN is provided from the virtual sound source IS._LAnd right ear E_RThe acoustic transfer function (impulse response) up to_L, H_RAnd
[0051]
This acoustic transfer function (impulse response) H_L, H_R5B and 5C show examples of the time axis waveform indicating the above. In addition, t in FIG. 5B and C shows time. Right ear E of listener LN from virtual sound source IS_RThe sound wave reaching the left ear E_LCompared to the sound wave that reaches, the wave arrives with a delay of Δt (FIG. 5C). Also, the characteristics obtained by converting the time characteristics of FIGS. 5B and 5C to the frequency axis are shown as (1) and (2) in FIG. 5D, respectively. As apparent from FIGS. 5B, 5C and 5D, the right ear E of the listener LN from the virtual sound source IS._RSound waves that reach the left ear E of the listener LN from the virtual sound source IS_LIt can be seen that the mid-high sound component is attenuated by the presence of the head of the listener LN as compared to the sound wave that reaches.
[0052]
5B and 5C, the time axis waveform showing the acoustic transfer function (impulse response) obtained by simplifying the time axis waveform showing the impulse response shown in FIGS._L', H_RShown as'. 5E acoustic transfer function (impulse response) H_L′ Is an impulse having an amplitude α1, and the acoustic transfer function (impulse response) H in FIG._R′ Is the acoustic transfer function (impulse response) H of FIG._LThe sum of the amplitude values at each sampling time is set to α1 with a rectangular wave response delayed by Δt from ′. When the rectangular wave response of FIG. 5F is composed of, for example, three samples as indicated by a circle in FIG. 5G, each sample value is α1 / 3.
[0053]
When the time-axis waveform showing the acoustic transfer function (impulse response) in FIGS. 5E and 5F is converted into a frequency-axis waveform, the waveforms are as shown in (3) and (4) in FIG. (4) indicates a low-pass filter characteristic (characteristic for simply attenuating the mid-high range). Further, since the integral values of the responses (3) and (4) in FIG. 5H, that is, the sum of the amplitude values of the respective sample points are the same, the frequency characteristics in the bass range are substantially the same.
[0054]
Again, referring back to FIG. After the monaural sound signal from the monaural sound signal source 1 is delayed by cascaded delay means 3 and 4 each comprising a predetermined number of delay devices DL connected in cascade, the right indirect sound and sound signal generating means is generated. 6 is supplied. The delay means 4 includes the delay means 5D and the delay elements DL constituting the filter 5F and the other delay elements 5D.
[0055]
In the right-side indirect sound signal generating means 6, the monaural sound signal delayed by the delay means 3 and 4, that is, the delayed monaural sound signal is passed through a coefficient multiplier having an amplitude weighting coefficient β1, thereby allowing the right-ear sound signal. The signal 6R is obtained. Further, the delayed monaural sound signal from the delay means 4 is supplied to the digital filter 6F through the delay means 6D composed of a predetermined number of delay devices DL connected in cascade. This filter 6F has one tap on the output side of the delay means 6D in the delay means composed of one tap on the output side of the delay means 6D and a predetermined number of delay devices DL cascaded to the delay means 6D. (K-2) and (k-1) total taps are taken out, and each tap is passed through a coefficient multiplier with amplitude weighting coefficients β2, β3,..., Β (k−1), βk, The signal is supplied to the adder 6S, added, and output as the left ear acoustic signal 6L. This digital filter 6F also constitutes an FIR filter in the same manner as the digital filter 5F described above.
[0056]
Here, the characteristic superimposed on the right ear acoustic signal 6R of the right indirect sound acoustic signal generating means 6 is a time axis impulse. When the values of the amplitude weighting coefficients β2,..., Β (k−1), βk in the filter 6F are all equal, the characteristic superimposed on the right ear acoustic signal of the left indirect sound acoustic signal generating means 5 is a time axis rectangular wave. Become. Furthermore, the following mathematical formula 3
[0057]
[Equation 3]

[0058]
By setting the relationship as follows, the low-frequency amplitude characteristics superimposed on the right-ear acoustic signal 6R and the left-ear acoustic signal 6L of the right indirect acoustic signal generation means 6 are substantially equal, and from the middle range to the high range. The envelope of the amplitude characteristic superimposed on the left-ear acoustic signal 6L is a characteristic that attenuates in the high frequency direction relative to that of the right-ear acoustic signal 6R. Such relative amplitude characteristics are macroscopically similar to those of an acoustic transfer function that arrives from the left side of the virtual listener.
[0059]
Also, by setting the relative position (time delay) of the left and right output waveforms on the time axis of the right indirect sound signal generation means 6 to be similar to that of the acoustic transfer function that arrives from the right side of the virtual listener, The phase characteristics can be made macroscopically similar to that of the acoustic transfer function that arrives from the right side of the virtual listener.
[0060]
Thus, the output acoustic signal of the right indirect sound acoustic signal generating means 6 can be handled as a simplified acoustic transfer function that arrives from the right side of the virtual listener.
[0061]
In FIG. 2, the indirect sound acoustic signal that arrives from the left side of the virtual listener reaches the ear of the virtual listener earlier in time than the indirect sound that arrives from the right side of the virtual listener. However, this may be reversed.
[0062]
Further, it has been experimentally confirmed that even if only one of the left and right indirect sound acoustic signal generating means 5 and 6 is used, the out-of-head localization feeling and the depth feeling of the sound image are sufficiently improved. ing.
[0063]
In the left and right indirect sound generating means 5 and 6, the transfer function superimposed to generate the waveform that reaches the ear opposite to the virtual sound source is not limited to the time-axis rectangular wave, It is also effective to replace the waveform with a short time waveform while satisfying the condition of the equation. That is, α₂, Α_Three, .................., α_rNeed not be set to equal values. Similarly, β₂, Β_Three, ………………, β_kIt is not always necessary to set equal values for.
[0064]
6A, 6B, and 6C show modifications of the time axis waveform shown in FIG. 5G. The circles in FIGS. 6A, 6B and 6C indicate sampling points. Hereinafter, these modifications will be described.
[0065]
The time axis waveform shown in FIG. 6A is an example in which a second-order or third-order polynomial is applied, and FIG. 6B is an example in which a fourth-order or fifth-order polynomial is applied. Including the rectangular wave shown in FIG. 5G, these time-axis waveforms (transfer functions) are superimposed on the input signal, which is a weighting coefficient sequence used for the moving average calculation, and the high frequency range is attenuated. Although the output symbol is obtained, the attenuation characteristic in the high frequency range changes depending on the time axis waveform.
[0066]
In the case of the rectangular wave shown in FIG. 5G, for example, in signal processing using a DSP (digital signal processor), only one memory is required to store coefficient data, and the number of sampling points is two. If the power, for example, is 4, then each weighting factor value is α₁/ 4, which is preferable because it can be obtained by a bit shift operation and the load of the arithmetic processing is further reduced.
[0067]
The time axis waveform shown in FIG. 6C is an example of imitating an RC filter by applying an exponential function. In this case, the number of sampling points may be at most several, but when implemented by an IIR filter described later, there is an advantage that the coefficient data memory and the amount of performance processing can be very small.
[0068]
It is also effective to use an approximate substitution with a filter or equalizer having a characteristic of simply attenuating the amplitude of the mid-high range instead of the FIR filter whose transfer function is a time-axis rectangular wave. FIG. 3 specifically shows the delay means 3 and 4 and the left and right indirect sound acoustic signal generation means 5 and 6 in FIG. 1 in this case.
[0069]
Hereinafter, FIG. 3 will be described. The basic configuration is the same as that in FIG. 2, and the difference is the right ear acoustic signal generation unit of the left indirect sound acoustic signal generation unit 5 and the left ear acoustic signal generation unit of the right indirect sound acoustic signal generation unit 6. That is, the left-side indirect sound acoustic signal generation means 5 extracts one tap from a position several taps to several tens of taps behind the tap position extracted as the left ear acoustic signal, and the amplitude weighting coefficient α₂, The signal processed by the equalizer 5E, which is a characteristic that simply attenuates the amplitude of the mid-high range (see characteristic curve (4) in FIG. 5H), is output as an indirect sound signal 5R for the right ear. The equalizer 5E is composed of, for example, an LPF (Low Pass Filter) composed of a low-order IIR (Infinite Impulse Response) filter.
[0070]
The right indirect sound and acoustic signal generation means 6 extracts one tap from a position several taps to several tens of taps behind the tap position extracted as the right ear indirect sound and acoustic signal 6R, and an amplitude weighting coefficient β₂Then, the signal processed by the equalizer 6E, which has the characteristic of simply attenuating the mid-high range amplitude, is output as an indirect sound signal 6L for the left ear. The equalizer 6E is configured by an LPF configured by a low-order IIR filter, for example.
[0071]
In the embodiment of the present invention, the configuration is based on the digital signal processing.₂, Β₂If a D / A converter is provided in front of the coefficient multipliers and equalizers 5E and 6E multiplied by, and these are taken out as analog signals, these may be constituted by a simple analog low-pass filter such as a CR filter.
[0072]
In the left and right indirect sound generation means 5 and 6, the low-frequency amplitude characteristics superimposed on the right ear indirect sound acoustic signal 5R and the left ear indirect sound acoustic signal 6L have the left ear indirect sound acoustic signal 5L and the right ear indirect sound acoustic signal 5L, respectively. The configurations of the equalizers 5E and 6E and the coefficients α2 and β2 of the coefficient multiplier are set to α so that the low-frequency amplitude characteristic superimposed on the indirect sound acoustic signal 6R is substantially equal.₁And β₁It is desirable to set so that they are substantially equal to each other.
[0073]
FIG. 4 shows the adding means 7 in FIG. 1 more specifically. Hereinafter, FIG. 4 will be described.
[0074]
The adding means 7 includes a left ear acoustic signal adder 7SL and a right ear acoustic signal adder 7SR. Then, the left ear acoustic signal 2L {SL [n]} from the direct sound acoustic signal generation means 2, the left ear indirect sound acoustic signal 5L and the right indirect sound acoustic signal generation means from the left indirect sound acoustic signal generation means 5. 6 is supplied to and added to the left ear acoustic signal adder 7SL, and the left ear addition signal 7L is supplied to the left ear speaker of the headphone 8. Also, the right ear sound signal 2R {SR [n]} from the direct sound sound signal generation means 2, the right ear indirect sound sound signal 5R and the right indirect sound sound signal generation means from the left indirect sound sound signal generation means 5. 6 is supplied to and added to the right ear acoustic signal adder 7SR, and the right ear addition signal 7R is supplied to the right ear speaker of the headphone 8.
[0075]
In this way, the listener can enjoy the sense of out-of-head localization of the sound image by listening to headphones.
[0076]
In the above-described embodiment, each signal processing block has been described as a circuit. However, it goes without saying that signal processing may be performed as software such as a microcomputer or DSP instead.
[0077]
Furthermore, it is provided as a recording medium (optical disk, magneto-optical disk, magnetic disk, magnetic tape, semiconductor memory, etc.) in which an indirect sound acoustic signal generation method or a virtual sound image localization processing method including the method is recorded as a program step. Is also possible. For example, several types of indirect sound signal generation parameters (delay time, high-frequency attenuation characteristics, etc.) are stored in the recording medium together with this processing program, and can be selected according to the listener's preference, or the producer of the reproduction signal source It is possible to switch appropriately according to the intention of the side.
[0078]
These programs may be recorded on a game software recording medium provided for a video game machine, or may be recorded and supplied on a separate recording medium.
[0079]
The delay time described above corresponds to that given by the delay means 3 and 4 in FIG. The treble attenuation characteristics described above correspond to the time-axis transmission characteristics or frequency characteristics given by the left and right indirect sound acoustic signal generating means 5 and 6 in FIG. On the other hand, the localization feeling and the depth feeling of the sound image can be changed by changing them relatively or changing these parameters. Even in this case, since the indirect sound and acoustic signal can be obtained with a small amount of signal processing, the load on the signal processing of the microcomputer and the like is small and the amount of information as a parameter is also small. It does not occupy a large recording capacity on the medium.
[0080]
The headphones described above are not limited to those that are attached to the listener's head, but may be earphones that are attached directly to the listener's pinna or ear canal. An arranged speaker device may be used.
[0081]
【The invention's effect】
According to the present invention, it is possible to obtain a sound image localization processing device with a compact signal processing amount..
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating an example of a headphone supply acoustic signal generation apparatus (headphone apparatus) to which a virtual sound image localization processing apparatus according to an embodiment of the present invention is applied.
2 is a specific configuration example of left and right indirect sound acoustic signal generation means of an example of a headphone supply acoustic signal generation device (headphone device) to which the virtual sound localization processing device of the embodiment of the present invention of FIG. 1 is applied; FIG.
3 is another specific example of the left and right indirect sound sound signal generating means of the headphone supply sound signal generating device (headphone device) to which the virtual sound localization processing device of the embodiment of the present invention of FIG. 1 is applied. It is a block diagram which shows a structural example.
4 is a block diagram showing a specific configuration example of an adding unit of an example of a headphone supply acoustic signal generation device (headphone device) to which the virtual sound image localization processing device of the embodiment of the present invention of FIG. 1 is applied. FIG.
FIG. 5 is a diagram for explaining the principle of right-side indirect sound acoustic signal generation means.
FIG. 6 is a waveform diagram showing another example of a time axis waveform.
[Explanation of symbols]
1 monophonic sound signal source, 2 direct sound sound signal generating means, 3 delay means, 4 delay means, 5 left indirect sound sound signal generating means, 6 right indirect sound sound signal generating means, 7 adding means, 7SL, 7SR adder, 8 Headphones.

Claims (1)

  A direct sound signal generation unit that receives a monaural sound signal source and performs a process of generating a sound signal part that is a direct sound signal for a virtual listener;
  A first delay unit to which the monaural acoustic signal source is input and one or more delay units having a delay time equal to one sample period are connected in series;
  An output signal of the first delay unit, and a second delay unit to which the one or more delay units are connected in series;
  An FIR type first digital filter that receives the output signal of the second delay unit and adds a plurality of output signals of the delay devices connected in series with equal tap coefficients;
  A third delay unit to which an output signal of the first digital filter is input, and one or more delay devices are connected in series;
  An FIR type second digital filter that receives the output signal of the third delay section and adds all the output signals of the delay devices connected in series with equal tap coefficients;
  The output signal of the direct sound / acoustic signal generation unit, the input signal of the second delay unit, and the output signal of the second digital filter are added and output to one channel of stereo output, and the direct sound / acoustic signal is output. An adder for adding the output signal of the signal generation unit, the input signal of the third delay unit, and the output signal of the first digital filter to output to the other channel of the stereo output;
A virtual sound localization processing apparatus.

Images