TW201119422A

TW201119422A - Dynamic volume control and multi-spatial processing protection

Info

Publication number: TW201119422A
Application number: TW98138835A
Authority: TW
Inventors: Christopher M Hanna; Gregory Benulis; Scott Skinner
Original assignee: That Corp
Priority date: 2008-11-14
Filing date: 2009-11-16
Publication date: 2011-06-01

Abstract

A disclosed system and method dynamically controls the perceived volume of a stereo audio program including left and right channel signals. The system comprises: a dynamic volume control configured and arranged so as to maintain a perceived constant volume level of the stereo audio program; and an excessive spatial processing protection processor configured and arranged for controlling the level of a difference signal, created as a function of the right channel subtracted from the left channel signal (L-R), relative to the level of a sum signal, created as a function of the right channel signal plus the left channel signal; wherein the excessive spatial processing protection processor processes the audio signals so as to control the difference (L-R) signal relative to the sum (L+R) signal. A system and method are also provided for dynamically controlling the perceived volume of a stereo audio program including left and right channel signals, comprising: a dynamic volume control configured and arranged so as to maintain a perceived constant volume level of the stereo audio program; and a program change detector configured and arranged to provide a program change signal indicating that the volume of the left and right channel signals has dropped below a threshold level for at least a threshold time period so as to anticipate a possible change in the sound level of the left and right channel signals; wherein the dynamic volume control is responsive to the program change signal.

Description

201119422 \、發明說明：【發明所屬之技術領域】本發明係關於並且以Christopher M. Hanna，Gregory Benulis與Scott Skinner為名申請在2008年11月14日提出申請之美國臨時申請案第61/114,684號並且以201119422 \, invention description: [Technical field of the invention] The present invention relates to the US Provisional Application No. 61/114,684 filed on November 14, 2008, in the name of Christopher M. Hanna, Gregory Benulis and Scott Skinner. Number and

Christopher M. Hanna 與 Gregory Benulis 為名申請在 2008 年11月14曰提出申請之第61/114,777號的優先權，兩申請案在此均以引用的方式併入本文。本申請案亦相關於申凊中的美國專利申請案號__(代理人案號 56233-428-THAT-27 )，其係以 Christopher M. Hanna 與Christopher M. Hanna and Gregory Benulis, the priority of which is hereby incorporated by reference in its entirety, the entire disclosure of which is incorporated herein by reference. This application is also related to the U.S. Patent Application Serial No. __ (Attorney Docket No. 56233-428-THAT-27), which is filed by Christopher M. Hanna and

GreforyBenulis為名而與本申請案同時申請，並且受讓給本受讓人。本發明係關於聲音訊號處理，且更特別地關於聲音訊音量控制與多空間的處理保護。【先前技術】在看電視的時候，音量改變令人惱怒，其係並且常常包含觀看者所進行的手動音量調整。其中一個例子是當改變電視頻道時經常產生之感知音量的改變。另一個例子則是在播放電視節目與廣告之間所產生之感知音量（p^ived v〇iume) 彳目収縣本上歸因域生期間所引入之播點上缺乏水平控制。所增加^到響度的稍微已知因素係為多空的聲音會在播音室中處社-即目材枓中里以在一聲道糸統中引入環繞空 201119422 間效應（模擬環繞）。假如隨後在電視處理此種聲音播放以引入二聲道環繞效果的話，如目前在許多電視模式之作法，感知之位準可能會劇烈變化。此額外的空間處理會造成中央影像（一般為對話）幾乎難以理解。在所有情形中，自動音量控制技術會將聽者的不適度最小化並且維持更一致的音量水平。當在播放點上，更注意調整聲音音量水平的同時，該問題似乎難以避免。事實上，由於高動態範圍 DTV播放的出現，電視收視者目前已可察覺較廣之響度差異。。【發明内容】根據所揭露系統與方法的一個態樣，本發明提供一種系統以用來動態控制包括左與右聲道訊號之立體聲音節目之感知音量，包含：動態音量控制器，其係予以架構與排列以維持立體聲音節目之感知固定音量水平；以及過度的空間處理保護處理器，其係予以架構與排列以控制產生當做左聲道訊號減去右聲道訊號（L — R)之函數的差訊號水平，其係相對於產生當做右聲道訊號加左聲道訊號之函數之合訊號的水平；其中，過度的空間處理保護處理器會處理聲音訊號以便控制差（L — R)訊號的增強。根據另一態樣，可提供一種系統，以用來動態控制包括左與右聲道訊號之立體聲音節目的感知音量，包含：動態音量控制器，其係予以架構與排列以便維持立體聲音節目之感知固定音量水平；以及節目改變檢測器，其係予以架構與排列以提供節目改變訊號，其指出左與右聲道訊號的 6 201119422 音量已經掉到臨界水平以下至少一臨界時期，以便預期在左與右聲道訊號之聲音水平的可能改變；其中’動態音量控制器易對節目改變訊號起反應。根據仍另一態樣，可提供一種系坑以用來動態控制包括左與右聲道訊號之立體聲音卽目之感知音量，包含：動態音量控制器，其係予以架構與排列以維持立體聲音節目之感知固定音量水平’該動態音量控制器包括至少易對高與低上升與釋放比率臨界起反應的壓縮器，以便定義安靜、正常與吵雜之感知音量水平。根據仍另一態樣，可提供一種系統以用來動態控制包括左與右聲道訊號之立體聲音節目的感知音量，包含：過度空間處理保護處理器，其係予以架構與排列以控制左聲道訊號減去右聲道訊號（L-R)所產生的差訊號水平，以差訊號成型的輪廓濾器。 & 根據仍另左與右聲道訊狀含：過度空間處理賴處理V":/知音4 °該系统包制左聲道喊減去右聲道^ ㈣與排列叫現在討論說明性實施例。苴地使用。顯而易見或不-要的細節會=另更間或更有效率地顯示。相及名略以郎省反地，在不揭露所有細節之下 7 201119422 可實施一些實施例。動態音量控制器（DVC)系統 DVC系統係予以描述來動態控制聲音訊號的音量。該系統係予以架構與排列，以便當突然發生改變時能夠動態操作與修改聲音音量。在此所說明的實施例係予以架構與排列以維持聲音頻帶應用的感知固定音量水平。DVC系統可完全數位化並可經濟性地在軟體（C、組合器等等）或數位硬體（HDL說明）中實施，雖然該系統很顯然的是完全類比或混合類比/數位系統。市場應用包括電視音效、DVD 播放器音效、機上盒音效、收音機音效與其他高度傳真性與非局度傳真性音效產品。缺乏在此所描述種類的Dvc系統，隨著節目材料在已知播放/來源内改變或隨著音效播放/ 來源改變，感知音量水平可巨幅地改變。這些音量改變令人煩躁，其係並且經常包含聽眾所進行的手動音量調整。其中一個特定實例係為當改變電視頻道時所產生的音量改變。另—實例係為在電視節目與電視商業廣告之間的音量改變。在兩實例中，DVC系統可消除收聽者不舒服的感知並且維持更一致的音量水平。圖1顯示此一 DVC系統1〇〇的一個實施例。系統1〇〇接收兩個輸入訊號，在輸入1〇2上的左訊號L以及在輸入 104上的右訊號。在該些實施例中所描述的DVC系統架構係依據具有僅僅在數位實施過程中具有可能性之彈性與額外修改之古典壓縮器設計（THAT公司Design Notell8)的 8 201119422 數位實施過程。系統100包括用來提供代表左與右訊號L 與R之RMS平均總和之訊號的RMS水平檢測器11〇二對數變換方塊112與訊號平均AVG方塊114。對數變換方塊 Π2將RMS水平檢測器11〇的輸出從線性轉換成對數領域。系統100易對許多控制訊號起反應，每—個皆指出是否特定情況的存在需要來自該系統的反應。系統1〇〇亦包括架構與排列以實施DVC系統1〇〇之操作的主處理器顯示）。所示實施例會對許多控制訊號起反應，包括：目桿讯號產生裝置116所提供的目標水平訊號、上升臨界訊號裝置118所產生的上升臨界訊號、釋放臨界（未顯示）、閘極臨界訊號裝置120所產生的閘極臨界訊號、上升比率^ 界（未顯示）、釋放比率臨界（未顯示）、比率訊號裝置122 所產生的比率訊號以及易對節目改變檢測器（pcD_未顯示）起反應之靜音鎖定裝置124所產生的靜音鎖定訊號。裝置 116、118、120、122僅僅是使用者所易使用的可調整使用者控制。裝置124可排列成當頻道改變時接收來自電視抑制’或者來自檢測輸人102與104兩者是否靜音之靜音檢測器（未顯示）的訊號。相對於全幅輸人，目標訊號水平 116以分貝(dB)表示其位準，其係為目標音量。上升臨界ιΐ8 代表在上升時間由因子N減少以前咖必須超過腦的犯數，在此N縣任何數。在―個_性實_中，n=i〇。釋放臨界訊號代表在上升日相㈣Μ減少以前ref必須低於AVG的dB數，在此M係為任何數，且在一個顯示性貫施射’ 。難臨界m代表麵有左與右增益調 201119422 整停止以前REF在AVG以下的數目，負dB數。上升比率臨界代表在音量控制使輸入訊號開始變弱以前rEF超過目標訊號水平116的絕對數，以dB表示。釋放比率臨界代表在音量控制開始增加增益到輸入訊號以前REF低於目標訊號水平116的絕對數’ dB。比率訊號122會藉由希望的壓縮比率來調整AVG值。藉由訊號加法器126從對數轉換方塊112的輸出減去目標水平訊號116，以便提供ref訊號給訊號平均AVG方塊114、比較器128與第二比較器130。REF訊號代表相對於希望聽力臨界之輸入訊號的音量水平。AVG訊號亦可予以視為同時發生（在上升/釋放處理以前）的理想增益推薦。訊號平均方塊114的輸出係為AVG訊號，該訊號係為REF 訊號之平均的函數。AVG訊號會予以施加到訊號加法器 132 ’在此它會予以添加到上升臨界訊號118。在類似的方式中（未顯示），AVG訊號會與釋放臨界一起相加。AVG 訊號亦會予以施加到訊號加法器134，在此它會予以添加到閘極臨界訊號120。訊號加法器132的輸出會予以施加到上升臨界比較器128，在此將與REF訊號互相比較，同時，訊號加法器134的輸出會予以施加到閘極臨界比較器 U0，在此將與REF訊號互相比較。AVG訊號亦可藉由訊號乘法器136而乘以比率訊號122。比較器128的輸出會施加到上升/釋放選擇方塊138，其係依次提供Att (上升）訊號或Rel (釋放）訊號到訊號平均方塊114，其係依據並且易對靜音鎖定訊號124的狀態起反應。釋放臨界AVG加法 201119422 器（未顯示）亦會與REF訊號相比較並施加到上升/釋放選擇方塊。比較器130提供輸出到訊號平均方塊U4的鎖定輸入。最後，訊號乘法器136提供輸出到對數_至_線性 (l〇g-t〇-linear)訊號轉換器140，其係依次提供施加到各訊號乘法器142與144的輸出，其中它各別地將提供在相應輸入102與1〇4的左與右訊號比率化，以便提供該輸出修改左與右訊號Lo與R〇。參考第1圖，RMS水平檢測器110感應輸入訊號的聲音水平。應該注意的是，在顯示RMS水平檢測器的同時，可使用任何型態的訊號水平檢測器。例如，峰值檢測器、平均檢測器、以感知為主的水平檢測器（譬如ITU177〇響度檢測器或CBS響度檢測器）、或可使用來感應聲音水平的其他檢測器。這些水平檢測器通常具有可動態與獨立調整的時間常數。-種調整這些時間常數的方法係為依據輸入訊號之封包或一般形狀，以致於該時間常數會隨著訊號改變。在其他實施例中，時間常數是固定的。為了簡化資° 料處理’聲音水平可使㈣數轉換方塊112轉換成對數二域，如圖所示。在多頻帶系統中，分開的RMS檢測器可使用於每-頻帶。訊號平均謂114係予以架構與排列以計算腳相對於上升與釋放時間的平均。訊號平均方塊ιι4 的輸出訊號AVG係經由乘法器136、藉由希望壓縮比率來調整，以產生齡加的增益值。最後，該增益可藉由對數_ 至-線性轉換器140而往後轉換成線性領域，以施^到左與右訊號L與R，以便產生修改的左與右訊號Lg*r〇。’、 11 201119422 將目標水平訊號116所代表的目標輸出水平從在對數轉換方塊112之輸出上的感應水平減去，以決定真實與希望聲音水平之間的差。此一差異，代表輸入訊號相對於目標水平訊號116的水平’視為參考（REF)訊號。目標水平訊號係為使用者輸入’譬如簡單的旋紐或其他預設設定，以便控制所希望的聲音水平。此臨界可予以固定或它可以輸入訊號水平的函數來改變，以相對於輸入動態範圍來較佳放置該壓縮。一旦得到REF訊號的話，它可做為一輸入地提供到平均方塊114、上升臨界比較器128與閘極臨界比較器130。上升臨界比較器128的輸出可予以施加到上升/ 釋放選擇方塊138，其係依次從節目改變檢測器接收訊號，靜音鎖定訊號124。當加入到現有平均AVG時，閘極臨界訊號12〇代表在冷凍左與右增益調整（142與144)以前能夠得到最低值 REF»閘極臨界比較器130接收即時訊號水平（REF)訊號並且決定REF所代表的聲音水平是否掉到已知上述的臨界以下。假如即時訊號水平（REF)超過出現在方塊114之輸出之平均訊號水平下㈣極臨界數量的話，箱在訊號路徑上施加到訊號的增益會維持固定，直到錢水平上升到臨界以上。目的係為阻止系％⑽施加增加的增益到非常低水平的輸人訊號，譬如噪音。在無限鎖定系統中，該增益會水遠S1定，直龍號水平上升為止。在㈣的鎖定系統中’該增益可以漸增的步調（比釋放時間更慢）來增加。在-個實施例中，此閘極鎖定臨界可調整，同時在另一實 12 201119422 施例中，閘極臨界134所設定的臨界可予以固定。當輸入是、、無聲"時，節目改變檢測器或靜音鎖定會感應出來。當使用者改變電視（TV)頻道時，二聲道之^ 的聲音水平可改變，可明顯增加或減少。基本上，電視製造商將短暫地使音效靜音，同時改變頻道，以避免觀眾^ 為聲音的暫態而煩躁。節目改變檢測器係設計來藉由決定聲音水平是否掉到預定臨界（MuteLev)以下達一^定數量的時間（MuteTime)而檢查此種靜音。假如即時聲音水平 (REF)在臨界以下達一特定時期或、、靜音時間的^，那麼節目的改變則會予以檢測出來。假如節目改變可予以檢測出來的話，那麼上升與釋放時間的速度（在以下有進一步詳細說明）則會增加。因為此增加，假如將噪音頻道改成無聲頻道的話，那麼所增加的釋放時間則會允許更快的增益增加，以符合目標聲音輸入水平。相反地，假如將無聲頻道改變為噪音頻道的話，那麼所增加的上升時間會^ 許更快的增益減少’以符合此目的。假如在、、靜音時間夕截止以前’聲音水平上升到臨界以上的話，那麼則無法檢測到節目改變。在替代性實施例中，、、靜音時間，，與靜音臨界會予以固定、供使用者調整、改變或別的方式。第2圖顯示用來操作節目改變檢測器之靜音檢測運算法則之狀態圖的一種實施例。操作200包括三個狀態、^ 音關閉（MUTE OFF)狀態202、靜音開啟（mute ON)狀態208與靜音鎖定（MUTE HOLD)狀態212。在靜音關閉狀態202中’可將在訊號加法器126之輪出上的REF訊號 13 201119422 與在204的MuteLev臨界水平週期性地比較，以決定是否 REF > MuteLev 或者 REF < MuteLev。假如 REF > MuteLev，那麼操作會維持在狀態202並且持續在那狀態。在此狀態中，靜音開啟=0，靜音鎖定=〇，且上升與釋放時間則在匕們的正常設定。不過，假如REF<MuteLev，那麼可檢測出靜音，且操作會在206轉變為狀態208靜音開啟。一旦轉移到狀態208，靜音開啟=1，且在狀態2〇8中，節目改變檢測器接著會決定靜音情況是否維持一預定時間。假如靜音情況不會維持夠長且REF>MuteOffLev發生在計時器截止以别的話，那麼檢測器則會往回轉變到狀態202。此可發生在聲音部分無聲處的節目停止處。不過，在計時器決定靜音時間截止的地方，節目改變會發生。在此狀態中，當REF>MuteOffLev恢復時，檢測者將在21〇轉換到靜音鎖疋狀態212。在此狀態中，上升與釋放時間會加速，以便使相對的噪音訊號更柔軟，且在預定時間界限内（靜音時間），相當柔軟的訊號會更吵雜。在第2圖中，在狀態2〇8 中的&十時器设疋會顯示與在狀態212中的相同。它們明顯為不同值。當在狀態212中，假如在靜音時間截止以前’ Ref減少到MuteLev設定以下的話（亦即，Ref< MuteLev )，該狀態會從214往回轉換到狀態2〇8。不過，假如MuteTime ;又有截止的8舌’檢測器將從216往回轉換到狀，雖202。在一種實施例中’可將MuteTime(靜音時間）與MuteLev (靜音水平）調整。靜音時間與靜音水平亦可在已知實施過程中固定。靜音臨界可設定成比閘極臨界還低。靜音檢 201119422 測浪异法可成自動或手動模式地起作用。在自動模式中，系、’’先100可在頻道改變期間内檢測靜音情況。節目改變檢測器亦可呈手動模式地操作，在此，、、靜音〃訊號可從指出頻道正在改變的電視或其他裝置接收。再者，節目改變檢測益亦可接收來自使用者遠距控制的訊號，以解釋使用者 ^否正在改變頻道。系統100亦可使用上升與釋放臨界來操作。在已知時間視窗中，假如聲音水平跳到穿越上升臨界118之範圍内的話，那麼系統1〇〇則可呈、、快速上升" 模式地操作。在一個實施例中，假如REF藉由上升臨界超過AVG的話，那此快速上升模式則會增加上升時間常數，以快速減少此增加聲音水平的增益。同樣地，假如穿過釋放臨界的話，那麼系統則會在快速釋放模式中操作，在此増盈會快速地增加。這些上升與釋放時間常數可在彼此之間並可同樣地在多頻系統中的高與低頻之間獨立地調整。在一些貫施過程中，施加到輸入訊號的最大增益可予以限制。這可限制施加到無聲聲音通過的增益數量。假如噪音通過（在電影中的打雷聲）緊接著無聲聲音通過的話，那麼在上升時間將增益減少以前，不受限制的增益則會造成明顯的聲音過衝。平均方塊114接收REF、上升、釋放與鎖定訊號，並且決定REF訊號的平均值（AVG)，其係依據上升、釋放與鎖定訊號並且當做其函數。隨後可藉由施加到原始訊號的壓縮比率來調整AVG訊號，以用於音量控制。AVG訊號代表以上升/釋放時間常數來處理的REF訊號。一旦REF中 15 201119422 的改變波動經過平均方塊U4以影響AVG訊號的話’那麼它首先則必須藉由希望的壓縮比率來調整。應該理解的是，系統100不會無限地壓縮。一旦AVG訊號值受到壓縮比率所調整的話，那麼AVG訊號則會經由比率設定裝置 122與乘法器136而乘以-(1-比率）。因此，藉由實例’ 4 : 1壓縮以率將使AVG訊號乘以-(1-1/4)或-3/4。所以’假如聲音是在臨界值以上2〇dB的話，那麼AVG訊號將等於 20dB (在上升時間常數消逝以後）。將2〇 dB乘以-3/4會產生-15 dB值。結果，在臨界值以上2〇dB的聲音會在施加 -15dB增益以後減弱到5dB。20/5 = 4，其係為4 : 1的壓縮比率。施加到訊號的壓縮比率係為單一傾斜比率。例如，依據水平臨界’ 4 : 1比率可予以施加到進入訊號。假如AVG 在臨界以上的話，那麼該訊號可減少因子4(以上升比率）。相反地，假如AVG在臨界以下的話，那麼該訊號可放大因子4 (以釋放比率）。在另一實施例中，壓縮比率會不同，其係取決於AVG 號在疋否裝置116所提供的目標水平臨界以上或以下。例如，假如AVG訊號在目標水平臨界以上的話，那麼該訊號可減少因子4，如在先前實例中。不過，相反地，假如 AVG在臨界以下的話，那麼不同比率則可予以施加，以放大輸入訊號，稱為1.5 : 1比率。此排列情形允許在比率臨界以上噪音訊號的壓縮，並可同樣地保存無聲對話的聲音水平，譬如耳語。以上所說明的排列情形可以電影模式^ 201119422 慮，它從吵嘈聲音移去刺耳的邊緣，但卻允許無聲聲音（留下沙沙聲等等），以維持它們最初的水平。這是吵雜音量設定的良好模式。因此則可得到更完整的動態範圍，同時仍 ~τ壓細吵雜討厭的訊號。另一排列情形包含在水平臨界以上與以下之AVG值的重大壓縮。重大濃縮在此視為、、夜晚模式"，因為你可聽到節目中的所有聲音而無需將音量調大 (針對輕柔的聲音）或調小（針對吵嘈的聲音）。夜晚模式對低音量設定是好的，其係經常在深夜時刻受到電視觀眾所偏愛。更進一步，另一實施例詮釋高與低上升與釋放比率臨界的使用。在此一實施例中，兩臨界定義響度空間的三個區域：無聲、正常與吵嘈。在每一個這些視窗中，可施加不同的壓縮比率。例如，1·5 : 1比率可使用來放大無聲訊號，1 . 1比率可使用來保存正常訊號，且4 ··丨比率可使用來減弱聲音訊號。由於此種多視窗系統，最初動態範圍可予以更準確地保存，同時邊緣吵嘈與輕柔訊號可分別地減弱與放大。最後，假如在對數領域中進行加工處理的話，那麼在施加增益到輸入訊號以前，算出的壓縮比率可在140上、、線性化〃。、第3圖顯示單一頻帶系統300 ’其中一個Dvc系統搬可施加相同的增益到予以施加到各別輸入3〇4與篇的母個左（L )與右（R )訊號。特別地，如第3圖所見， DVC系統3G2的輸出（由對數_至'線性信號轉換器14〇所 17 201119422 提供）可各別動態地設定每一放大器308與31〇的增益，其係依次放大施加到系統300之兩輸入的相應左與右訊號，以提供Lout與Rout訊號在輸出316與318 ° DVC系統302可對每一 L與R訊號的整個頻率範園起反應’或者僅僅例如如第3圖所示之每一個的選擇性頻帶’高通量濾器312與314每一個僅僅將各別L與R訊號的高頻率部分通到DVC系統302，以致於後者僅僅可對每一訊號的高頻率内容起反應。或者，可架構多頻帶系統，以致於選擇性頻帶每一個均可各別由其自己的DVC系統所處理，以便能夠獨立控制 L與R訊號。如第4圖所示，例如，雙頻道系統40〇使用兩DVC系統406與408，每一個均用於L與R訊號’以致於施加到輸入402與404的L與R訊號能夠享用獨立的增益控制。如圖所示，L訊號可予以施加到高通量濾器410 與低通量濾器412，同時R訊號可予以施加到高通量濾器 414與低通量濾器416。在具有高與低頻帶之第4圖的雙頻道系統中’藉由施加每一 DVC系統的輸出到高與低通量濾器的各別輸出’ DVC系統（406與408)可施加增益到在高頻帶中的L與R訊號。特別地，可施加DVC系統406的輸出’以控制接收與放大高通量濾器41〇與412之高頻率輸出之每一放大器418與420的增益。同樣地，可施加DVC 系統408的輸出’以控制接收與放大低通量濾器412與416 之低頻率輸出之每一放大器422與424的增益。放大器418 與420的輸出可予以添加在訊號加法器426，以便產生輸出 201119422 訊號L〇Ut於輸出428上，同時放大器422與424的輸出則可藉由δί1旒加法器430來添加’以便產生輸出訊號R0ut於輸出432上。在另一實施例中，假如在多頻帶訊號之每一 L與R訊號的獨立增益控制令人希望的話’那麼可使用各別的DVC 系統於每一 L與R訊號的每一頻帶。再者，代替多頻帶系統，高通量濾器可使用來刪除對低頻率沒反應之系統的低頻率，如第3圖所示。關於使用以多頻帶DVC系統的濾器而言，在每一鄰近頻帶（為低與高通量頻帶的雙頻道系統）之間頻率上的交叉則可予以調整。同樣可能的則是使頻率上的交叉固定。其中一個實例係為依據所得到濾器的數位實施過程來交叉。所得到的濾器係說明於來自Milford，MA之THAT公司的 THAT 公司 Application Note 104 與 B〇hn，D. ( Ed.)，如出0 國立半導體公司，Santa Clara，CA 1976)§ 5.2.4。在所得到遽器實施過程的一個實例中，該交叉使用第二級 Butterworth LPF與所得到的HPF，其係加總為一，如第5 圖所示。在另一實施例中，該交又係為傳統的數位第二級，其具有HPF反轉的Q=〇.5，以致於頻帶會加總為一單元，如第6圖所示。在仍另一實施例中，交叉係依據第4級 Linkwitz-Riley濾器，其係加總為一單元，如第7圖所示。在單一頻帶音量控制中，高通量遽器會控制RMS檢測器的輸入。 201119422 多空間處理保護（Mpp ) SRS立體聲環環繞（模擬環繞）技術（例如，輸出路後。此二聲at 術等等）於二聲道電視聲音安裝在電視外牆的味彳。:3可進行到電視外面的喇叭或提高存在於立“放；的環繞技術會藉由操作與善可在產生聲音節二二：通常’此種型態的空間改吸引繃罘、:匕的』間内進行。這特別適用於為了引觀眾左忍力而增強的電視廣告。當聲音且空間增強時(例如在生產點與在電視聲音處二 2 P麼在聲音品質上則會有明顯的退化。相對於L+R 此重’預先處理的聲音傾向於具有明顯的L_R能量。增強處理的第二、串聯階段傾向於更增加L—R能量數量。最近的研究顯示，過量的L—w_為使聽眾疲勞的其中一個重要因素。同樣地，會有明顯的音量増加。於是，根據本發明的-個態樣’可提供MPP系統。在一個實施例中’在電視的立體增強技術以前，MPP係為雙重處理保護（DPP)系統，其係為-部份的電視聲音訊號接收與錄音再mMPP緖在下城域擬環繞訊號處理器。示範性的DPP系統處理聲音訊號，以便最小化在生產點上引入的差（L —R)增強（亦即，相對於合（l+r) 訊號’最小化差（L-R)訊號的能量水平）。這會使電視的 20 201119422 空間增強技術以心理聲學场悅聽眾的方絲處理聲音訊號。在電視空間増強聲音處理以前的Dpp系統串聯，盆係已經證實在緩和雙重空間處理的祕效果上非常有效:、在 -個實施射，DPP Μ完全触，錢並且可在軟裝配器）或數位硬體（HDL說明）中料轉性實施。麻該理解岐，DPP系統亦可完全為類比，或者類比二元件的混合。在一個實施例中，相對於相應的L+R水平，Dpp系統會減少L - R增強。該實施㈣少多個二聲道空間效果處理的效果。此-系統的-個實施例顯稀第8圖中的_。左況被L與右訊號R各別施加到系統8⑻的輸入搬與购。 L與R訊號可施加到由兩訊號加法器8〇6與8〇8所/代表的矩陣。訊號加法器806與808構成提供合（L+R)與差（l —R)訊號的矩陣。 ^ 、在s (L+R)路徑中，該訊號一般未受影響。合訊號通¥包含不-定需要局部化的聲音内容。不過，在替代性實施例中，鮮輪廓成型行以進行，以增㈣如對話的 2音内容如所示，合訊號會在提供到以訊號加法器812 一 814顯不之矩陣以前乘以在訊號乘法器"ο的中心常數士。中心常數使中心影像（L+R)水平得以調整，假如希望 ^舌以助於理解對話化从與L — R訊號的添加提供在輸各816上的左輸出訊號L〇,同時自L + R減去L —R則會在輪出+818提供右輪出訊號R〇。Grefory Benulis applied for this application at the same time as the application and was assigned to the assignee. The present invention relates to audio signal processing, and more particularly to audio volume control and multi-space processing protection. [Prior Art] When watching TV, the volume change is irritating, and it often includes manual volume adjustments made by the viewer. An example of this is the change in perceived volume that is often produced when changing TV channels. Another example is the perceived volume (p^ived v〇iume) generated between the broadcast of a TV show and an advertisement. The lack of level control on the broadcast point introduced during the attribution period is reported. The slightly known factor for increasing the loudness is that the long and short sounds will introduce a surround space 201119422 effect (analog surround) in the one-channel system in the studio. If such a sound is subsequently processed on the television to introduce a two-channel surround effect, as is currently the case in many television modes, the level of perception may vary drastically. This extra spatial processing can make the central image (usually a conversation) almost incomprehensible. In all cases, the automatic volume control technique minimizes the discomfort of the listener and maintains a more consistent volume level. This problem seems to be difficult to avoid while paying more attention to adjusting the sound volume level at the playback point. In fact, due to the emergence of high dynamic range DTV playback, TV viewers are now aware of the widest differences in loudness. . SUMMARY OF THE INVENTION According to one aspect of the disclosed system and method, the present invention provides a system for dynamically controlling a perceived volume of a stereo sound program including left and right channel signals, including: a dynamic volume controller Architecture and arrangement to maintain a perceived fixed volume level of the stereo sound program; and excessive spatial processing to protect the processor, which is structured and arranged to control the generation of the left channel signal minus the right channel signal (L - R) The level of the difference signal relative to the level at which the signal is generated as a function of the right channel signal plus the left channel signal; wherein the excessive spatial processing protection processor processes the voice signal to control the difference (L - R) signal Enhancement. According to another aspect, a system can be provided for dynamically controlling the perceived volume of a stereo sound program including left and right channel signals, including: a dynamic volume controller that is structured and arranged to maintain the perception of stereo sound programs. Fixed volume level; and program change detector, which is structured and arranged to provide a program change signal indicating that the left and right channel signals of the 6 201119422 volume have fallen below a critical level for at least a critical period in order to be expected to be left The sound level of the right channel signal may change; where 'the dynamic volume controller is easy to react to the program change signal. According to still another aspect, a pit can be provided for dynamically controlling the perceived volume of the stereo sound including the left and right channel signals, including: a dynamic volume controller that is structured and arranged to maintain stereo sound. Perceived fixed volume level of the program' The dynamic volume controller includes at least a compressor that is susceptible to high and low rise and release ratio thresholds in order to define a quiet, normal and noisy perceived volume level. According to still another aspect, a system can be provided for dynamically controlling the perceived volume of a stereo sound program including left and right channel signals, including: an over-space processing protection processor that is structured and arranged to control the left channel The signal is subtracted from the difference signal level generated by the right channel signal (LR), and the contour filter is formed by the difference signal. & According to still left and right channel signals: excessive spatial processing processing V": / 知 4 ° The system package left channel shout minus right channel ^ (four) and arrangement called now discuss illustrative examples . Use it 苴. Obvious or not - the details will be = more or more efficient. The implementation of some embodiments can be implemented in the context of the lang province, without revealing all the details. Dynamic Volume Controller (DVC) System The DVC system is described to dynamically control the volume of the sound signal. The system is structured and arranged to dynamically manipulate and modify the sound volume when sudden changes occur. The embodiments described herein are structured and arranged to maintain a perceived fixed volume level of the sound band application. The DVC system can be fully digitized and economically implemented in software (C, combiner, etc.) or digital hardware (HDL specification), although the system is clearly a fully analog or mixed analog/digital system. Market applications include TV sound, DVD player sound, set-top box sound, radio sound and other highly fax and non-local fax sound effects. In the absence of a Dvc system of the kind described herein, the perceived volume level can vary dramatically as the program material changes within a known play/source or as the sound play/source changes. These volume changes are irritating, and often include manual volume adjustments made by the listener. One of the specific examples is the volume change that occurs when the TV channel is changed. Another example is the volume change between the TV show and the TV commercial. In both instances, the DVC system can eliminate the viewer's uncomfortable perception and maintain a more consistent volume level. Figure 1 shows an embodiment of this DVC system. System 1〇〇 receives two input signals, the left signal L on input 1〇2 and the right signal on input 104. The DVC system architecture described in these embodiments is based on the 8 201119422 digital implementation of a classical compressor design (THAT Design Notell 8) with flexible and additional modifications that are only possible in digital implementation. System 100 includes an RMS level detector 11 〇 two logarithmic transform block 112 and a signal averaging AVG block 114 for providing signals representative of the RMS average sum of the left and right signals L and R. The logarithmic transformation block Π2 converts the output of the RMS level detector 11〇 from linear to logarithmic. System 100 is susceptible to reacting to many control signals, each indicating whether the presence of a particular condition requires a response from the system. The system 1 also includes a main processor display that is structured and arranged to implement the operation of the DVC system. The illustrated embodiment reacts to a number of control signals, including: a target level signal provided by the mast signal generating device 116, a rising threshold signal generated by the rising threshold signal device 118, a release threshold (not shown), and a gate critical signal. The gate critical signal generated by the device 120, the rising ratio limit (not shown), the release ratio threshold (not shown), the ratio signal generated by the ratio signal device 122, and the program change detector (pcD_not shown) The mute lock signal generated by the mute lock device 124 of the reaction. The devices 116, 118, 120, 122 are merely adjustable user controls that are easy for the user to use. The device 124 can be arranged to receive a signal from the television suppression' or from a silence detector (not shown) that detects whether both the input persons 102 and 104 are muted when the channel changes. The target signal level 116 is expressed in decibels (dB) relative to the full-size input, which is the target volume. The rising threshold ιΐ8 represents the number of crimes in the N-county before the rise time is reduced by the factor N. In the _ _ real _, n = i 〇. The release of the critical signal represents that the ref must be lower than the AVG's dB number before the ascending phase (four) Μ reduction, where M is any number and is applied at one display. Difficult critical m represents the left and right gain adjustments 201119422 The number of negative REFs below AVG before the whole stop, negative dB number. The rise ratio threshold represents the absolute number of rEF exceeding the target signal level 116 before the volume control causes the input signal to begin to weaken, expressed in dB. The release ratio threshold represents the absolute number 'dB below the target signal level 116 before the volume control begins to increase the gain to the input signal. The ratio signal 122 adjusts the AVG value by the desired compression ratio. The target level signal 116 is subtracted from the output of the logarithmic conversion block 112 by the signal adder 126 to provide a ref signal to the signal average AVG block 114, the comparator 128 and the second comparator 130. The REF signal represents the volume level relative to the input signal that is expected to be critical to hearing. The AVG signal can also be considered as an ideal gain recommendation for simultaneous occurrence (before the rise/release process). The output of the signal average block 114 is an AVG signal, which is a function of the average of the REF signals. The AVG signal is applied to the signal adder 132' where it is added to the rising threshold signal 118. In a similar manner (not shown), the AVG signal is added to the release threshold. The AVG signal is also applied to the signal adder 134 where it is added to the gate threshold signal 120. The output of the signal adder 132 is applied to the rising threshold comparator 128 where it is compared with the REF signal, and the output of the signal adder 134 is applied to the gate critical comparator U0, where it will be associated with the REF signal. Compare with each other. The AVG signal can also be multiplied by the ratio signal 122 by the signal multiplier 136. The output of comparator 128 is applied to rise/release selection block 138, which in turn provides an Att (rising) signal or a Rel (release) signal to signal average block 114, which is based on and readily reacts to the state of mute lock signal 124. . Release critical AVG addition The 201119422 (not shown) is also compared to the REF signal and applied to the rise/release selection block. Comparator 130 provides a lock input that is output to signal averaging block U4. Finally, signal multiplier 136 provides an output to a log-to-linear converter 140 that in turn provides an output applied to each of signal multipliers 142 and 144, where it will be provided separately The left and right signals of the respective inputs 102 and 1〇4 are scaled to provide the output to modify the left and right signals Lo and R〇. Referring to Figure 1, the RMS level detector 110 senses the sound level of the input signal. It should be noted that any type of signal level detector can be used while displaying the RMS level detector. For example, a peak detector, an average detector, a perceptual level detector (such as the ITU177〇 Loudness Detector or CBS Loudness Detector), or other detectors that can be used to sense the level of sound. These level detectors typically have time constants that can be dynamically and independently adjusted. The method of adjusting these time constants is based on the packet or general shape of the input signal such that the time constant will change with the signal. In other embodiments, the time constant is fixed. In order to simplify the processing of the sound level, the (four) number conversion block 112 can be converted into a logarithmic two field as shown. In a multi-band system, separate RMS detectors can be used for each band. The signal average 114 is structured and arranged to calculate the average of the foot relative to the rise and release times. The output signal AVG of the signal average square ι4 is adjusted by the desired compression ratio via the multiplier 136 to generate an aged gain value. Finally, the gain can be converted back to a linear domain by a log-to-linear converter 140 to apply left and right signals L and R to produce modified left and right signals Lg*r〇. ', 11 201119422 The target output level represented by the target level signal 116 is subtracted from the sense level at the output of the logarithmic conversion block 112 to determine the difference between the true and desired sound levels. This difference, representing the level of the input signal relative to the target level signal 116, is considered a reference (REF) signal. The target level signal is for the user to enter 'such as a simple knob or other preset settings to control the desired sound level. This threshold can be fixed or it can be changed by a function of the input signal level to better place the compression relative to the input dynamic range. Once the REF signal is obtained, it can be provided as an input to the average block 114, the rising threshold comparator 128 and the gate critical comparator 130. The output of the rising threshold comparator 128 can be applied to the rise/release selection block 138, which in turn receives the signal from the program change detector, muting the lock signal 124. When added to the existing average AVG, the gate critical signal 12〇 represents the lowest value REF» before the frozen left and right gain adjustments (142 and 144). The gate critical comparator 130 receives the instantaneous signal level (REF) signal and determines Whether the sound level represented by REF falls below the above known criticality. If the instantaneous signal level (REF) exceeds the critical threshold level (four) at the average signal level of the output at block 114, the gain applied to the signal by the box on the signal path will remain fixed until the money level rises above the critical level. The purpose is to prevent the %(10) from applying an increased gain to a very low level of input signals, such as noise. In the infinite locking system, the gain will be determined by the distance S1 and the level of the straight dragon will rise. In the locking system of (4), the gain can be increased by an increasing step (slower than the release time). In one embodiment, the gate lock threshold is adjustable, while in another embodiment 12 201119422, the threshold set by the gate threshold 134 can be fixed. When the input is yes, no sound, the program change detector or mute lock will be sensed. When the user changes the television (TV) channel, the sound level of the two channels can be changed, which can be significantly increased or decreased. Basically, the TV manufacturer will briefly mute the sound and change the channel to avoid the viewer's irritability for the transient state of the sound. The program change detector is designed to check for such silence by determining if the sound level falls below a predetermined threshold (MuteLev) for a certain amount of time (MuteTime). If the immediate sound level (REF) is below a critical level for a specific period of time, or a silent time, then the program change will be detected. If the program change can be detected, then the speed of the rise and release time (described further below) will increase. Because of this increase, if the noise channel is changed to a silent channel, the increased release time will allow for a faster gain increase to match the target sound input level. Conversely, if the silent channel is changed to a noisy channel, then the increased rise time will result in a faster gain reduction' to suit this purpose. If the sound level rises above the critical level before and after the silence time, the program change cannot be detected. In an alternative embodiment, the silence time, and the silence boundary are fixed for the user to adjust, change or otherwise. Figure 2 shows an embodiment of a state diagram of a silence detection algorithm for operating a program change detector. Operation 200 includes three states, a MUTE OFF state 202, a mute ON state 208, and a MUTE HOLD state 212. In the mute off state 202, the REF signal 13 201119422 on the round of the signal adder 126 can be periodically compared with the MuteLev critical level at 204 to determine whether REF > MuteLev or REF < MuteLev. If REF > MuteLev, then the operation will remain in state 202 and continue in that state. In this state, mute is on = 0, mute is locked = 〇, and the rise and release times are in our normal settings. However, if REF<MuteLev, then mute can be detected and the operation will transition to 206 in state 208 to mute. Once transitioned to state 208, mute is on, and in state 2〇8, the program change detector then determines if the mute condition is maintained for a predetermined time. If the mute condition does not remain long enough and REF>MuteOffLev occurs at the timer cutoff, then the detector will transition back to state 202. This can occur where the program where the sound portion is silent is stopped. However, where the timer determines that the silence time has expired, a program change will occur. In this state, when REF > MuteOffLev is restored, the detector will transition to the silent lock state 212 at 21〇. In this state, the rise and release times are accelerated to make the relative noise signal softer, and within a predetermined time limit (silent time), a fairly soft signal can be more noisy. In Fig. 2, the & chronograph setting in state 2〇8 is displayed the same as in state 212. They are obviously different values. In state 212, if Ref is reduced below the MuteLev setting (i.e., Ref < MuteLev) before the silence time expires, the state transitions from 214 back to state 2〇8. However, if MuteTime; there is a cut-off 8 tongue' detector will be converted back from 216 to shape, although 202. In one embodiment, MuteTime and MuteLev can be adjusted. The mute time and mute level can also be fixed during known implementations. The mute threshold can be set to be lower than the gate threshold. Silent check 201119422 The wave test can work in either automatic or manual mode. In the automatic mode, the system 100 can detect the mute condition during the channel change period. The program change detector can also be operated in a manual mode where the mute signal can be received from a television or other device indicating that the channel is changing. Furthermore, the program change detection benefit can also receive a signal from the user's remote control to explain whether the user is changing the channel. System 100 can also operate using rise and release thresholds. In the known time window, if the sound level jumps to the extent of crossing the rising boundary 118, then the system 1 可 can be operated in a fast rising " mode. In one embodiment, if REF exceeds AVG by a rising threshold, then the fast rising mode increases the rise time constant to quickly reduce the gain of this increased sound level. Similarly, if the release threshold is exceeded, then the system will operate in a fast release mode where the profit will increase rapidly. These rise and release time constants can be adjusted independently of each other and equally between high and low frequencies in a multi-frequency system. The maximum gain applied to the input signal can be limited during some implementations. This limits the amount of gain applied to the silent sound. If the noise passes (the thunder in the movie) followed by the silent sound, then the unrestricted gain will cause a significant sound overshoot before the gain is reduced during the rise time. The average block 114 receives the REF, rise, release, and lock signals and determines the average of the REF signals (AVG) based on the rise, release, and lock signals and as a function of them. The AVG signal can then be adjusted for volume control by the compression ratio applied to the original signal. The AVG signal represents the REF signal processed by the rise/release time constant. Once the change in REF 15 201119422 fluctuates through the average block U4 to affect the AVG signal, then it must first be adjusted by the desired compression ratio. It should be understood that system 100 does not compress indefinitely. Once the AVG signal value is adjusted by the compression ratio, the AVG signal is multiplied by -(1-rate) via ratio setting means 122 and multiplier 136. Therefore, the AVG signal will be multiplied by -(1-1/4) or -3/4 by the example '4:1 compression rate. So if the sound is 2〇dB above the threshold, then the AVG signal will be equal to 20dB (after the rise time constant has elapsed). Multiplying 2〇 dB by -3/4 produces a -15 dB value. As a result, a sound of 2 〇 dB above the critical value is attenuated to 5 dB after the -15 dB gain is applied. 20/5 = 4, which is a compression ratio of 4:1. The compression ratio applied to the signal is a single tilt ratio. For example, a ratio can be applied to the incoming signal based on a horizontal critical '4:1 ratio. If the AVG is above the critical level, then the signal can be reduced by a factor of 4 (in ascending ratio). Conversely, if AVG is below the critical value, then the signal can amplify factor 4 (to release ratio). In another embodiment, the compression ratio will vary depending on whether the AVG number is above or below the target level threshold provided by the device 116. For example, if the AVG signal is above the target level threshold, then the signal can be reduced by a factor of four, as in the previous example. However, conversely, if the AVG is below the critical level, then different ratios can be applied to amplify the input signal, called the 1.5:1 ratio. This arrangement allows compression of the noise signal above the ratio boundary and can likewise preserve the level of sound of a silent conversation, such as a whisper. The arrangement described above can be considered in the movie mode ^ 201119422, which removes the harsh edges from the noisy sound, but allows silent sounds (with rustling, etc.) to maintain their original level. This is a good mode for noisy volume settings. Therefore, a more complete dynamic range can be obtained, while at the same time, the ~τ is awkward and annoying. Another arrangement involves significant compression of AVG values above and below the horizontal threshold. Significant enrichment is treated here as, night mode " because you can hear all the sounds in the show without having to turn up the volume (for soft sounds) or turn down (for noisy sounds). Night mode is good for low volume settings, which are often preferred by TV viewers at late nights. Still further, another embodiment illustrates the use of high and low rise and release ratios. In this embodiment, the two thresholds define three regions of the loudness space: silent, normal, and noisy. In each of these windows, different compression ratios can be applied. For example, a ratio of 1·5:1 can be used to amplify a silent signal, a ratio of 1.1 can be used to store a normal signal, and a ratio of 4·· can be used to attenuate the sound signal. Thanks to this multi-window system, the initial dynamic range can be saved more accurately, while edge noise and soft signals can be separately reduced and amplified. Finally, if processing is performed in the logarithmic field, the calculated compression ratio can be linearized at 140 before the gain is applied to the input signal. Figure 3 shows a single band system 300' where one of the Dvc systems can apply the same gain to the parent left (L) and right (R) signals applied to the respective inputs 3〇4 and . In particular, as seen in Figure 3, the output of the DVC system 3G2 (provided by the logarithmic _ to 'linear signal converter 14 〇 17 201119422) can dynamically set the gain of each amplifier 308 and 31 各, respectively. Amplifying the respective left and right signals applied to the two inputs of system 300 to provide Lout and Rout signals at output 316 and 318 ° DVC system 302 can react to the entire frequency range of each L and R signal' or just for example Each of the selective band 'high pass filters 312 and 314 shown in FIG. 3 only passes the high frequency portion of the respective L and R signals to the DVC system 302, so that the latter is only available for each signal. High frequency content reacts. Alternatively, the multi-band system can be architected such that each of the selective frequency bands can be individually processed by its own DVC system to enable independent control of the L and R signals. As shown in FIG. 4, for example, dual channel system 40 uses two DVC systems 406 and 408, each for L and R signals' such that L and R signals applied to inputs 402 and 404 can enjoy independent gain. control. As shown, the L signal can be applied to the high flux filter 410 and the low flux filter 412 while the R signal can be applied to the high flux filter 414 and the low flux filter 416. In a dual channel system with a high and low frequency band diagram 4, 'by applying the output of each DVC system to the respective outputs of the high and low flux filters' DVC systems (406 and 408) can apply gain to high L and R signals in the band. In particular, the output of DVC system 406 can be applied to control the gain of each of amplifiers 418 and 420 that receive and amplify the high frequency output of high pass filters 41 and 412. Likewise, the output of DVC system 408 can be applied to control the gain of each of amplifiers 422 and 424 that receive and amplify the low frequency outputs of low pass filters 412 and 416. The outputs of amplifiers 418 and 420 can be added to signal adder 426 to produce an output 201119422 signal L〇Ut on output 428, while the outputs of amplifiers 422 and 424 can be added by δί1旒adder 430 to produce an output. Signal R0ut is on output 432. In another embodiment, a separate DVC system can be used for each of the L and R signals if the independent gain control of each of the L and R signals of the multi-band signal is desired. Furthermore, instead of a multi-band system, a high-throughput filter can be used to remove low frequencies from systems that do not respond to low frequencies, as shown in Figure 3. Regarding the use of filters in a multi-band DVC system, the crossover in frequency between each adjacent frequency band (a dual channel system for low and high throughput bands) can be adjusted. It is also possible to fix the cross on the frequency. One of the examples is based on the digital implementation of the resulting filter. The resulting filter is described in THAT Company Application Note 104 from THAT Corporation of Milford, MA and B〇hn, D. (Ed.), as in National Semiconductor Corporation, Santa Clara, CA 1976) § 5.2.4. In one example of the resulting damper implementation process, the crossover uses a second stage Butterworth LPF and the resulting HPF, which is summed to one, as shown in FIG. In another embodiment, the intersection is a conventional digital second stage having an HPF inverted Q = 〇.5 such that the frequency bands are summed up as a unit, as shown in FIG. In still another embodiment, the crossover is based on a Level 4 Linkwitz-Riley filter, which is summed up as a unit, as shown in FIG. In single-band volume control, the high-passput timer controls the input to the RMS detector. 201119422 Multi-Space Processing Protection (Mpp) SRS Stereo Ring Surround (Analog Surround) technology (for example, after the output of the road. This two-spot, etc.) is installed on the TV exterior wall of the two-channel TV sound. :3 can be carried out to the horn outside the TV or to enhance the presence of the "release; the surround technology will be manipulated by the good and the good can produce the sound section 22: usually 'this type of space to change the tension,: 匕Between the two. This is especially suitable for TV commercials that are enhanced to attract viewers' left endurance. When the sound and space are enhanced (for example, at the production point and at the sound of the TV, there will be obvious sound quality). Degraded. This heavy 'pre-processed sound tends to have significant L_R energy relative to L+R. The second, series phase of enhanced processing tends to increase the amount of L-R energy. Recent studies have shown that excess L-w _ is one of the important factors for the listener to fatigue. Similarly, there will be a significant volume increase. Thus, according to the present invention, an MPP system can be provided. In one embodiment, 'before the stereo enhancement technology of the television The MPP is a dual processing protection (DPP) system, which is a part of the television audio signal receiving and recording and then mMPP in the lower metro domain to surround the signal processor. The exemplary DPP system processes the sound signal, In order to minimize the difference (L - R) enhancement introduced at the production point (ie, the energy level of the minimum (L) signal relative to the combined (l + r) signal.) This will enhance the TV's 20 201119422 space. The technology uses the psychoacoustic field to listen to the sound of the listener's square wire. In the TV space, the Dpp system is connected in series before the sound processing, the basin system has been proved to be very effective in mitigating the secret effect of double space processing: in one implementation, DPP Μ Fully touched, money can be implemented in a soft assembler or digital hardware (HDL specification). It is understood that the DPP system can also be completely analogous, or analogous to the mixing of two components. The Dpp system reduces the L-R enhancement relative to the corresponding L+R level. This implementation (4) has less effect on the processing of multiple two-channel spatial effects. This - the embodiment of the system is sparsely shown in Figure 8. _. The left condition is applied to the input and purchase of the system 8 (8) by L and the right signal R. The L and R signals can be applied to the matrix represented by the two signal adders 8〇6 and 8〇8. The signal adder 806 and 808 constitute a combination (L+R) and difference l —R) The matrix of the signal. ^ , in the s (L+R) path, the signal is generally unaffected. The commencing number contains the sound content that is not localized. However, in an alternative embodiment The fresh contour is formed by the line to increase (4) as the 2 tone content of the dialogue is as shown, the signal will be multiplied by the center of the signal multiplier " ο before being supplied to the matrix of the signal adder 812-814 The constant constant. The central constant allows the center image (L+R) level to be adjusted. If you want to understand the dialogue, the left output signal L〇 on the input 816 is provided from the addition of the L-R signal. L + R minus L - R will provide a right turn signal R 在 at turn +818.

在第8圖所示之實施例中，大部分的處理會發生在DIF 201119422 路徑。可將L+R與L —R比較，以決定l —R訊號相對於 L+R的水平。在比較以前，這兩合與差訊號每一個均可通過各別的南通量滤器820與822 ’譬如在剩頻率回應不包括低頻率的環境中。L — RDIF訊號可進一步通過多頻帶等化器824 ’以強調耳朵的最敏感頻率，亦即中範圍頻率，以補償L — R訊號的感知噪音水平。等化器824使差聲道水平檢測得以依賴頻率。例如，當為了具有有限低音回應的便宜電視喇八而處理時，低頻率訊號可予以最小化。高頻率可予以最小化，以將回應限制在暫態聲音事件。基本上，耳朵最易感知到的中範圍頻率會予以相等化，以支配差水平檢測。一旦計算出差與合訊號水平的話，DIF/SUM比率則可予以決定。隨後，這些訊號的每一個會運行經過各別的訊號水平檢測器828與830。以上所陳列的檢測器可予以使用，譬如 RMS水平檢測器，雖然任何種類的水平檢測器（譬如以上所拖述者）可予以使用的話。同樣地，該處理均可全在對數範圍中進行，以藉由將它們通過對數領域處理方塊832 與834而來增加效率。方塊832肖834的輸出可予以施加到訊號加法器，盆中’可從所處理的差減減麵處理的合訊號。從在對數領域中的其他麵去-他號，其倾提供為處理合訊號之比率的訊號到在線性領域中差訊號者相同。—旦算出 L+R與L-R職水平雜’那麼l_r喊水平則可在水平檢測以前予簡等化，明加巾範，這兩訊號水 22 201119422 平則可藉由比較器838而與預設臨界840比較。這兩訊號之間的比率（（L —R) / (l+R))可藉由比較器838而與臨界比率比較，以便決定所推薦的L — R訊號增益調整。限制态階段842可使用來限制施加到L — R訊號之增益的數量與方向。所示的實施例限制增益在〇 dB，從而卻僅僅允許乙 —R訊號之減弱，雖然在一些應用中，放大L — R訊號是令人希望的。以相當長的時間常數，平均階段844會平均限制器階段842的輸出，以便避免DPP系統追縱短暫的暫存聲音事件。在藉由線性領域方塊846而轉換回線性領域以後，L — R訊號的水平可由訊號乘法器848相應地調整，以得到那目標比率。In the embodiment shown in Figure 8, most of the processing takes place on the DIF 201119422 path. L+R can be compared to L-R to determine the level of the l-R signal relative to L+R. Prior to comparison, each of the two and the difference signals can pass through the respective south flux filters 820 and 822', such as in the case where the residual frequency response does not include low frequency. The L-RDIF signal can be further passed through the multi-band equalizer 824' to emphasize the most sensitive frequency of the ear, i.e., the mid-range frequency, to compensate for the perceived noise level of the L-R signal. Equalizer 824 allows differential channel level detection to be frequency dependent. For example, low frequency signals can be minimized when processed for a cheap television with a limited bass response. High frequencies can be minimized to limit the response to transient sound events. Basically, the mid-range frequencies most easily perceived by the ear are equalized to dominate the differential level detection. Once the difference between the difference and the signal level is calculated, the DIF/SUM ratio can be determined. Each of these signals then runs through respective signal level detectors 828 and 830. The detectors shown above can be used, such as RMS level detectors, although any type of level detector (such as those described above) can be used. Likewise, the process can all be performed in the logarithmic range to increase efficiency by processing them through the logarithmic domain blocks 832 and 834. The output of block 832 Shaw 834 can be applied to the signal adder, which can subtract the summed signal processed from the processed difference. From the other side in the logarithm field, the signal is the same as the one that handles the ratio of the signal to the difference in the linear field. Once the L+R and LR levels are calculated, then the l_r shouting level can be simplified before the level detection, and the two signals can be used by the comparator 838 and preset. Critical 840 comparison. The ratio between these two signals ((L - R) / (l + R)) can be compared to the threshold ratio by comparator 838 to determine the recommended L - R signal gain adjustment. The limit state stage 842 can be used to limit the amount and direction of gain applied to the L-R signal. The illustrated embodiment limits the gain to 〇 dB, thereby allowing only the weakening of the B-R signal, although in some applications, amplifying the L-R signal is desirable. With a fairly long time constant, the average phase 844 averages the output of the limiter stage 842 to avoid the DPP system tracking short transient sound events. After being converted back to the linear field by linear field block 846, the level of the L-R signal can be adjusted accordingly by signal multiplier 848 to obtain the target ratio.

甚至在不存在複數階段的空間事先處理之下，目標（L R ) / ( L+R )比率可設定在低’以例如允許增加的節目對話理解性。用於雙重處理保護的另一方法與系統係為、、預言，，在乙 ~R訊號上所進行的事先處理並且補償源自該預言的事先處理。例如，假如SRS立體聲-環繞視為使用於l — r的話，那麼該訊號可相應地予以補償，以移除L_R增強。或者，訊號能量可予以超時監控，以演繹出在L_R訊號上所進行的事先處理。從此演繹，L — R訊號可予以補償，以移除任何此種L — R增強。事先處理可改變差（以及就此而言，合）聲道以及L一R/ L+R比率的頻率反應。事先處理器的反向濾器可予以施加到每一路徑，同時現存的L_R/ L+R比率調整仍可繼續使用。 23 201119422 再者，第8圖的DPP系統，其係顯示當做其中DIF訊號在可變增益控制放大器848以前予以測出，當做反饋系統’其中合與差訊號水平可在可變增益控制放大器亦有可能以後予以檢測。Even in the absence of spatial pre-processing of the complex phase, the target (L R ) / (L+R) ratio can be set low to, for example, allow for increased program dialogue comprehension. Another method and system for dual processing protection is, and is, pre-processed on the B-signal and compensates for prior processing from the prediction. For example, if SRS stereo-surround is considered to be used for l-r, then the signal can be compensated accordingly to remove the L_R enhancement. Alternatively, the signal energy can be monitored over time to perform prior processing on the L_R signal. From this interpretation, the L-R signal can be compensated to remove any such L-R enhancement. Pre-processing can change the frequency response of the difference (and in this case, the combined channel) and the L-R/L+R ratio. A pre-processor reverse filter can be applied to each path while the existing L_R/L+R ratio adjustments can continue to be used. 23 201119422 Furthermore, the DPP system of Figure 8 shows that the DIF signal is measured before the variable gain control amplifier 848, as the feedback system 'where the combined signal level is available in the variable gain control amplifier. It may be tested later.

結合DVC與DPP 因為DVC與MPP每一個均提供改善的聆聽經驗，所以這兩個會予以結合’以結合兩者的優點。有許多結合DVC 與DPP方塊的方式。一種有用的拓樸實例首先會放置DPP 方塊902，接著在串聯設計中的DVC方塊904，如第9圖所示。在此實施例中’ L與R訊號可予以施加到DPP方塊 902的輸入906與908。在輸出910與912上之DPP方塊 902的L’與R’輸出則予以施加到DVC方塊904的兩輸入 914與916〇DVC方塊的輸出918與920提供各別的輸出訊號Lo與Ro。串聯設計首先使DPP方塊得以移除差（L —R) 訊號增強，然後以DVC方塊來維持立體聲音節目的感知固定水平而無需周圍能罝的存在。拓樸的另一實例將DPP方塊1004放置在DVC方塊 1002的反饋路徑中，如第10圖所示。L與R輸入分別予以施加到輸入1006與1008。該兩訊號可予以施加到陣列（由訊號加法器1010與1012所代表），以便產生加（L+R)訊號與差（L — R)訊號。DVC方塊1002的輸出1〇14與1016 提供輸出訊號Lo與Ro。兩輸出1014與1016提供反饋路徑的兩反饋訊號。尤其是，Lo與Ro訊號可予以施加到關於訊號加法器1018與1020顯示的陣列，以致於Lo+R〇能 24 201119422 夠形成DPP方塊1004的一個輸入，且Lo — R〇能夠形成 DPP方塊1004的其他輸入。DPP方塊1004的輸出代表校正增益，其係隨後藉由訊號乘法器1022而施加到DIF訊號。後者呈可變增益控制放大器形式。應該理解的是，當結合DVC與DPP方塊的兩實施例顯示於第9與1〇圖時，其他的組合則是可能的。於是，本發明的實施例提供用於改善聲音訊號複製的性能，該訊號複製會減少在聲音節目編排中令人不希望的音量改變效果。已經讨論的元件、步驟、特徵、好處與優點僅用於說明。它們以及有關它們的討論並沒有意圖以任何方式來限制所保護的範圍。為數眾多的其他實施例亦可仔細考慮。此外，比起在此所清楚說明的，本發明的實施例具有較;、額外與/或*同的元件、步驟、特徵、好處與優點。這亦包括元件與/或步驟可不同排列與/或安排的實施例。除非以別的方式陳述，否則在本說明書所陳述，包括，以下申請專利範圍中的所有測量、數值、等級、位置、該、尺寸與其他規格，其係均為近似，非準確。它們意 =有合理的範圍，其係與它們所相_函數以及在該技藝中它們所習慣附屬的一致。 2發明中所引用的所有文章、申請專利範圍、申請 :糊應用與其他出版物，因此在此以引用的方式併入本文。假如並且當使用於申請專利範圍中_語、、意指，其 25 201119422 係意圖並且詮釋為包含已經說明的相應同物。同樣地’假如並且當使與其專這些用語物。缺乏限制t任1應結構、材料或行動，或它們的等= 全釋為用於任:3述=示=何f係予以打算或言全釋為專同公開，而無關是否陳述於申請專利範圍中/、者專受保護範圍僅僅受到以 ;=r釋為與使用接= 之申請專利範圍中之語言的-般= 致，並且包含所有結構性與功能性的等同物。— 圖式簡單說明】該等圖式揭露顯示性實施例。其他實施例則可另外匕們並沒有陳述所有實施必要的細節會予以省略，以地使用。顯而易見或不相反地，在不揭露所有細節^”更有效率地顯示。相同數字出現於不同圖式時，6丄可實施一些實施例。當步驟。匕意指相同或類似的元件或本揭露的態樣可在連同整地理解，在本質上，“Α圖來研讀時從以下說明更完式不-定按比例，其重點反(制性。該些圖該圖式中：疋放在本揭路的原理上。在 26 201119422 第1圖係為動態音量控制系統之一個實施例的簡化方塊圖；第2圖係為顯示一個節目改變檢測之操作之一個實施例的狀態圖；第3圖係為一個單一頻帶動態音量控制系統實施例的簡化方塊圖；第4圖係為一個多頻帶動態音量控制系統實施例的簡化方塊圖；第5-7圖圖式地顯示多頻帶動態音量控制系統的頻率回應；第8圖係為一個雙重處理保護系統實施例的簡化方塊圖；第9圖係為一個組合系統排列情形實施例的簡化方塊圖，包括動態音量控制系統與雙重處理保護系統兩者；以及第10圖係為組合系統排列情形之第二實施例的簡化方塊圖，包括動態音量控制系統與雙重處理保護系統兩者。【主要元件符號說明】 100 DVC系統 102輸入 104輸入 110 RMS水平檢測器 112對數變換方塊 114訊號平均AVG方塊 116目標訊號產生裝置 118上升臨界訊號裝置 120閘極臨界訊號裝置 122比率訊號裝置 27 201119422 124靜音鎖定裝置 316輸出 126訊號加法器 318輸出 128比較器 400頻帶系統 130比較器 402輸入 132訊號加法器 404輸入 134訊號加法器 406 DVC系統 136訊號乘法器 408 DVC系統 138上升/釋放選擇方塊 410高通量遽器 140對數-至-線性訊號轉 412低通量濾器換器 414高通量渡器 142訊號乘法器 416低通量濾器 144訊號乘法器 418放大器 200操作 420放大器 202靜音關閉狀態 422放大器 208靜音開啟狀態 424放大器 212靜音鎖定狀態 426訊號加法器 300單一頻帶系統 428輸出 302 DVC系統 430訊號加法器 304輸入 432輸出 306輸入 800系統 308放大器 802輸入 310放大器 804輸入 312高通量遽器 806訊號加法器 314高通量濾器 808訊號加法器 28 201119422 810訊號乘法器 906輸入 812訊號加法器 908輸入 814訊號加法器 910輸出 816輸出 912輸出 818輸出 914輸入 820高通量濾器 916輸入 822高通量濾器 918輸出 824多頻帶等化器 920輸出 828訊號水平檢測器 1002 DVC 方塊 830訊號水平檢測器 1004 DPP 方塊 832對數領域處理方塊 1006輸入 834對數領域處理方塊 1008輸入 838比較器 1010訊號加法器 840預設臨界 1012訊號加法器 842限制器階段 1014輸出 844平均階段 1016輸出 846線性領域方塊 1018訊號加法器 848訊號乘法器 1020訊號加法器 902 DPP方塊 1022訊號乘法器 904 DVC方塊 29Combining DVC with DPP Because both DVC and MPP provide improved listening experience, the two will combine to combine the advantages of both. There are many ways to combine DVC and DPP blocks. A useful topology example would first place DPP block 902, followed by DVC block 904 in a series design, as shown in Figure 9. The 'L and R signals can be applied to inputs 906 and 908 of DPP block 902 in this embodiment. The L' and R' outputs of DPP block 902 on outputs 910 and 912 are applied to the two inputs 914 and 916 of the DVC block 904. Outputs 918 and 920 of the DVC block provide respective output signals Lo and Ro. The tandem design first enables the DPP block to remove the difference (L - R) signal enhancement, and then maintains the perceived fixed level of the stereo sound program with the DVC block without the presence of ambient energy. Another example of topology places DPP block 1004 in the feedback path of DVC block 1002, as shown in FIG. The L and R inputs are applied to inputs 1006 and 1008, respectively. The two signals can be applied to the array (represented by signal adders 1010 and 1012) to produce an add (L+R) signal and a difference (L-R) signal. Outputs 1〇14 and 1016 of DVC block 1002 provide output signals Lo and Ro. The two outputs 1014 and 1016 provide two feedback signals for the feedback path. In particular, the Lo and Ro signals can be applied to the arrays displayed with respect to the signal adders 1018 and 1020 such that Lo+R can 24 201119422 form an input to the DPP block 1004 and Lo — R〇 can form the DPP block 1004. Other inputs. The output of DPP block 1004 represents the correction gain which is then applied to the DIF signal by signal multiplier 1022. The latter is in the form of a variable gain control amplifier. It should be understood that other combinations are possible when the two embodiments incorporating the DVC and DPP blocks are shown in Figures 9 and 1. Thus, embodiments of the present invention provide for improved performance of voice signal duplication that reduces undesirable volume change effects in sound programming. The elements, steps, features, advantages and advantages that have been discussed are for illustration only. They and their discussion of them are not intended to limit the scope of protection in any way. Numerous other embodiments can also be considered carefully. In addition, the embodiments of the present invention have more elements, steps, features, advantages and advantages than the ones described herein. This also includes embodiments in which the components and/or steps may be arranged and/or arranged differently. Unless otherwise stated, all measurements, values, grades, locations, dimensions, dimensions, and other specifications in the scope of the following claims are all approximations and non-accurate. They mean that there is a reasonable range, which is consistent with their _ function and the habits they are used to in the art. All articles, patent applications, applications, and applications of the present invention are hereby incorporated by reference. If and when used in the scope of the patent application, the language is intended to be interpreted as including the corresponding equivalents already stated. Similarly, if and when it is used specifically for these terms. Lack of restrictions on the structure, materials or actions of the 1st, or their etc. = full release for the purposes of: 3 statements = indication = what is the intention or full release of the disclosure, and whether or not it is stated in the patent application The scope of protection in the scope is only subject to the meaning of the language in the scope of the patent application, and includes all structural and functional equivalents. — BRIEF DESCRIPTION OF THE DRAWINGS These figures disclose illustrative embodiments. Other embodiments may not otherwise state that all implementations are necessary and will be omitted for use. Obviously or not, the details are shown more efficiently without revealing the details. When the same numbers appear in different figures, some embodiments may be implemented. When the steps are the same or similar elements or the disclosure The aspect can be understood together with the land preparation, in essence, "when the map is studied, the following descriptions are more complete and not proportional, and the focus is reversed (systematic. The figures are in the figure: 疋 in this In principle, on the basis of 26 201119422, FIG. 1 is a simplified block diagram of an embodiment of a dynamic volume control system; FIG. 2 is a state diagram showing an embodiment of an operation of program change detection; A simplified block diagram of an embodiment of a single band dynamic volume control system; Figure 4 is a simplified block diagram of an embodiment of a multi-band dynamic volume control system; and Figures 5-7 show a multi-band dynamic volume control system Frequency response; Figure 8 is a simplified block diagram of an embodiment of a dual processing protection system; Figure 9 is a simplified block diagram of an embodiment of a combined system arrangement, including Both the volume control system and the dual processing protection system; and Fig. 10 is a simplified block diagram of the second embodiment of the combined system arrangement, including both the dynamic volume control system and the dual processing protection system. 100 DVC system 102 input 104 input 110 RMS level detector 112 logarithmic transform block 114 signal average AVG block 116 target signal generating device 118 rising critical signal device 120 gate critical signal device 122 ratio signal device 27 201119422 124 mute locking device 316 output 126 Signal adder 318 output 128 comparator 400 band system 130 comparator 402 input 132 signal adder 404 input 134 signal adder 406 DVC system 136 signal multiplier 408 DVC system 138 rise / release selection block 410 high throughput buffer 140 logarithm -to-linear signal to 412 low flux filter converter 414 high flux 142 signal multiplier 416 low flux filter 144 signal multiplier 418 amplifier 200 operation 420 amplifier 202 mute off state 422 amplifier 208 mute on state 424 amplifier 212 mute lock status 426 No. adder 300 single band system 428 output 302 DVC system 430 signal adder 304 input 432 output 306 input 800 system 308 amplifier 802 input 310 amplifier 804 input 312 high throughput buffer 806 signal adder 314 high flux filter 808 signal addition 28 201119422 810 Signal Multiplier 906 Input 812 Signal Adder 908 Input 814 Signal Adder 910 Output 816 Output 912 Output 818 Output 914 Input 820 High Throughput Filter 916 Input 822 High Throughput Filter 918 Output 824 Multiband Equalizer 920 Output 828 Signal Level Detector 1002 DVC Block 830 Signal Level Detector 1004 DPP Block 832 Log Field Processing Block 1006 Input 834 Log Field Processing Block 1008 Input 838 Comparator 1010 Signal Adder 840 Preset Critical 1012 Signal Adder 842 Limiter Stage 1014 Output 844 Average Stage 1016 Output 846 Linear Field Block 1018 Signal Adder 848 Signal Multiplier 1020 Signal Adder 902 DPP Block 1022 Signal Multiplier 904 DVC Block 29

Claims

201119422 VII. Patent application scope: 1. A system for dynamically controlling the perceived volume of a stereo sound program including left and right channel signals, comprising: a dynamic volume controller, an architecture and arrangement to maintain the stereo sound program. Perceiving a fixed volume level; and a program change detector, structured and arranged to provide a program change signal indicating that the volume of the left and right channel signals has fallen below a critical level for at least a critical period in order to anticipate The possible change in the sound level of the left and right channel signals; wherein the dynamic volume controller is responsive to the program change signal. 2. The system of claim 1, wherein the dynamic volume controller comprises a compressor configured and arranged to have a rise time and a release time adjustment speed, wherein a program change signal can be If detected, then the rate of such rise and release times will increase. 3. The system of claim 2, wherein the dynamic volume controller is structured and arranged such that if a noise channel is changed to a silent channel, the increased release time allows for faster gain increase. To match a target sound output level, and if a silent channel changes to a noise channel, then the increased rise time will allow for a faster gain reduction to match the target sound output level. 4. The system of claim 3, wherein the dynamic volume control 201119422 is structured and arranged such that if the sound level rises above the critical level before the time deadline, then a program change cannot be detected. As in the system of claim 1, the critical period and the critical level are fixed. 6. For the system of claim 1, the critical period and the critical level may be adjusted. 7. The system of claim 1, wherein the critical period and the critical level are variable. . 8. The system of claim 1, wherein the program change detector is configurable and arranged to automatically react by detecting a silence condition during a channel change period. 9. The system of claim 1, wherein the program change detector is configurable and arranged to react when a user changes a channel by detecting a channel change from a user remote control. 10. The system of claim 1 of the 5th patent scope, wherein the program change detector can be structured and arranged to reflect a channel change condition communicated via the host processor. 31 201119422 11. A method for dynamically controlling a perceived volume of a stereoscopic sound program comprising one of left and right channel signals, comprising: dynamically controlling the volume of the left and right channel signals to change a signal in response to a program to maintain A perceptual fixed volume level of the stereoscopic sound program; and generating the program change signal to detect a program change signal indicating that the volume of the left and right channel signals has fallen below a critical level for at least a critical period for anticipation Possible changes in the sound level of the left and right channel signals. 12. A system for dynamically controlling the perceived volume of a stereoscopic sound program comprising one of a left and right channel signal, comprising: y a dynamic volume controller that is structured and arranged to maintain a perceptual fixed volume level of the stereoscopic sound program, The dynamic volume controller C includes at least one compressor that is critically responsive to a low rise and release ratio for deliberate silent, normal, and noisy perceived volume levels. 13. If you apply for a patent scope! A two-part system in which the compressor applies different compression ratios to the perceived volume levels of each of the silent, normal, and noisy. 14. As in the scope of claim 13 of the patent application, the compression ratio of the volume of the secret sound is better set to amplify the left and right signals of the material for the normal volume level (four) setting to save the 32 201119422 and other left and right signals, and the compression ratio for the noisy volume level is set to attenuate the left and right signals. 33