Analysis of problems in digital signal network cable transmission
Becausetherearemanyformatsofsignalsusedincurrentapplicationsystems,suchascommonVGA,Video,DVI,SDI,andHDMI,etc.,whenthesignalsareusedincombination,theformatunificationandconversionproblemswillinevitablyoccur,includingthefollowing:Severalconversionmethods. I.Conversionbetweenanaloganddigitalsignals TheanalogsignalscommonlyusedinthepastincludeVideo,VGA,Y/C,RGB,YUV,YPbPr,YCbCr,etc.Amongthem,VGAisusedintheITindustryandhasacorrespondingrelationshipwithDVI,whileothersarebasicallysignalsfromtheVideoindustryItcancorrespondtosignalssuchasSDI,HDSDIorHDMI.Wedonotconsiderthesituationofmanyprotocolsindigitalsignals,butonlyunderstanditfromtheconversionofthesignalitself.TheconversionfromanalogtodigitalisanA/Dprocess(AnylogToDigital).Itincludestwomainconcepts:1.Samplingrate2.Numberofquantizedbits.Thesamplingratereferstohowmanytimesthesignalissampledinonesecond.Forexample,ifthesampleistakenonceasecond,thesamplingrateis1HZ.Ifitistaken1000times,thesamplingrateis1KHZ.,96KHZ,192KHZ,etc.,correspondingtothesamplingrateofVideo,generallyYis13.5MHZ,U/Vis6.25MHZ,ifitisR,G,Bcolorgamut,all13.5MHZwillbeused,etc.Someintroductions.Theselectionofthesamplingrateshouldbebasedonthedesiredsignalbandwidth.AccordingtotheNyquistsamplinglaw,themaximumbandwidthoftheanalogsignalcanreach1/2ofthesamplingrate,whichcanbeunderstoodbycorrespondingtotheintroductioninthefirstpart.Theremustbeacorrespondingrelationshipwiththeline/field,andacorrespondingrelationshipwiththecolorloadwave.Therefore,valuessuchas13.5MHZ,74,25MHZ,etc.haveappeared.Thenumberofquantizedbitsreferstoanormalizedvalue,suchashowmanypartsarequantized.If1isdividedintotwo1/2s,thenumberofsampledbitsis1,whichis21=2.Itcandescribewhethertheamplitudeofasignalislargerorsmallerthan1/2,butiftheactualsignalislargerthan1/2andlessthan1,youcannotknowhowbigitis,onlythatitisbetween1and1/2.Thedescriptionisrough.Ifthenumberofsamplingbitsis2,ie22=4,thesignalisdividedinto4parts.Wecandescribethatasignalisbetween3/4and1/2,whichismoreaccurate.Ifthenumberofquantizationbitsis8,thenwhenwechooseasignal,itsaccuracycanreachanerrorof28=256,1/256.Ifitisa10-bitnumber,itserrorcanreach1/1024,andsoon.Atpresent,theaudiosamplequantizationcanreach12bit,16bit,andthevideosamplequantizationisgenerally8bit. D/A(DigitalToAnylog),thisisthereverseprocessofA/D. 2.LossinA/Dconversion IlearnedfromtheprocessofA/DandD/Athatdigitalsignalscanonlydescribeanalogsignalswithaslittlelossaspossible,butcannotbecompletelyrestored.Hereareafewquestions: 1.Lossofbandwidth.AccordingtotheNyquistsamplingrate,themaximumbandwidththatcanberetainedis1/2ofthesamplingrate.Thisdatahasalsobeenconfirmedinexperiments.TheattachedfigureisouropticalfibertransmissionequipmentfortestingVGAsignals.Measurementdata,theyellowlineisthespectrumcompositionoftheVGAsignal,andthebluelineisthespectrumobtainedthroughopticalfibertransmission(whichincludesasetofA/DandD/Aprocesses),becausedigitalsignalswillnotbetransmittedduringtransmission.Thereareotherlosses,sothelossescausedarebornebyA/D,D/A.Whenlookingatthe3dBpointoftheoriginalsignal,itisexactlyalossof6dB.AcompoundA/DD/Aprocessintheory,adescriptionof6dBofbandwidthloss. 2.Thelossofthesignal,becausethenumberofsamplingbitsislimited,thedescriptionofthesignalcannotbecomplete,andalotofinformationisalwayslost.Accordingtotheoreticalcalculations,thelosscausedby8-bitsamplingonthesignalcanreach52-54dB.Thelossof10dBcanreachabout60dB,(somesignallossisalsoakindofnoisefortheoriginalsignal),sonowthenumberofquantizationbitsisincreasing,from8bitto10bitto12bit,suchasDEEPCOLOURinHDMIandDPThisbringsanincreaseinbandwidth. III.Conversionbetweendigitalsignals Digitalsignalsneedtobeconvertedwhentheyaremultiplexedduetodifferentcodingforms.Inprinciple,becausetheyarealldigitalsignals,thereisnoA/DandD/Aprocess,andtherewillbenoseriousloss,buttheactualBecausevariousdigitalsignalsarebasedondifferentsignalsintheoriginalgeneration,suchasSDIandHDSDIarebasedonYUVcolorgamutspace,whileVGA,DVI,HDMI,DP,etc.areRGBcolorgamutspace.WeknowthatusingtheYUVcolorgamutcanuseUVbandwidthcompressiontoreducethebandwidth.Therefore,thesignalRGBcolorgamutbasedontheYUVcolorgamuthasasmallsignalbandwidth.TheYUVcolorgamutisconvertedtotheRGBcolorgamut.letter
MORE
Becausetherearemanyformatsofsignalsusedincurrentapplicationsystems,suchascommonVGA,Video,DVI,SDI,andHDMI,etc.,whenthesignalsareusedincombination,theformatunificationandconversionproblemswillinevitablyoccur,includingthefollowing:Severalconversionmethods. I.Conversionbetweenanaloganddigitalsignals TheanalogsignalscommonlyusedinthepastincludeVideo,VGA,Y/C,RGB,YUV,YPbPr,YCbCr,etc.Amongthem,VGAisusedintheITindustryandhasacorrespondingrelationshipwithDVI,whileothersarebasicallysignalsfromtheVideoindustryItcancorrespondtosignalssuchasSDI,HDSDIorHDMI.Wedonotconsiderthesituationofmanyprotocolsindigitalsignals,butonlyunderstanditfromtheconversionofthesignalitself.TheconversionfromanalogtodigitalisanA/Dprocess(AnylogToDigital).Itincludestwomainconcepts:1.Samplingrate2.Numberofquantizedbits.Thesamplingratereferstohowmanytimesthesignalissampledinonesecond.Forexample,ifthesampleistakenonceasecond,thesamplingrateis1HZ.Ifitistaken1000times,thesamplingrateis1KHZ.,96KHZ,192KHZ,etc.,correspondingtothesamplingrateofVideo,generallyYis13.5MHZ,U/Vis6.25MHZ,ifitisR,G,Bcolorgamut,all13.5MHZwillbeused,etc.Someintroductions.Theselectionofthesamplingrateshouldbebasedonthedesiredsignalbandwidth.AccordingtotheNyquistsamplinglaw,themaximumbandwidthoftheanalogsignalcanreach1/2ofthesamplingrate,whichcanbeunderstoodbycorrespondingtotheintroductioninthefirstpart.Theremustbeacorrespondingrelationshipwiththeline/field,andacorrespondingrelationshipwiththecolorloadwave.Therefore,valuessuchas13.5MHZ,74,25MHZ,etc.haveappeared.Thenumberofquantizedbitsreferstoanormalizedvalue,suchashowmanypartsarequantized.If1isdividedintotwo1/2s,thenumberofsampledbitsis1,whichis21=2.Itcandescribewhethertheamplitudeofasignalislargerorsmallerthan1/2,butiftheactualsignalislargerthan1/2andlessthan1,youcannotknowhowbigitis,onlythatitisbetween1and1/2.Thedescriptionisrough.Ifthenumberofsamplingbitsis2,ie22=4,thesignalisdividedinto4parts.Wecandescribethatasignalisbetween3/4and1/2,whichismoreaccurate.Ifthenumberofquantizationbitsis8,thenwhenwechooseasignal,itsaccuracycanreachanerrorof28=256,1/256.Ifitisa10-bitnumber,itserrorcanreach1/1024,andsoon.Atpresent,theaudiosamplequantizationcanreach12bit,16bit,andthevideosamplequantizationisgenerally8bit. D/A(DigitalToAnylog),thisisthereverseprocessofA/D. 2.LossinA/Dconversion IlearnedfromtheprocessofA/DandD/Athatdigitalsignalscanonlydescribeanalogsignalswithaslittlelossaspossible,butcannotbecompletelyrestored.Hereareafewquestions: 1.Lossofbandwidth.AccordingtotheNyquistsamplingrate,themaximumbandwidththatcanberetainedis1/2ofthesamplingrate.Thisdatahasalsobeenconfirmedinexperiments.TheattachedfigureisouropticalfibertransmissionequipmentfortestingVGAsignals.Measurementdata,theyellowlineisthespectrumcompositionoftheVGAsignal,andthebluelineisthespectrumobtainedthroughopticalfibertransmission(whichincludesasetofA/DandD/Aprocesses),becausedigitalsignalswillnotbetransmittedduringtransmission.Thereareotherlosses,sothelossescausedarebornebyA/D,D/A.Whenlookingatthe3dBpointoftheoriginalsignal,itisexactlyalossof6dB.AcompoundA/DD/Aprocessintheory,adescriptionof6dBofbandwidthloss. 2.Thelossofthesignal,becausethenumberofsamplingbitsislimited,thedescriptionofthesignalcannotbecomplete,andalotofinformationisalwayslost.Accordingtotheoreticalcalculations,thelosscausedby8-bitsamplingonthesignalcanreach52-54dB.Thelossof10dBcanreachabout60dB,(somesignallossisalsoakindofnoisefortheoriginalsignal),sonowthenumberofquantizationbitsisincreasing,from8bitto10bitto12bit,suchasDEEPCOLOURinHDMIandDPThisbringsanincreaseinbandwidth. III.Conversionbetweendigitalsignals Digitalsignalsneedtobeconvertedwhentheyaremultiplexedduetodifferentcodingforms.Inprinciple,becausetheyarealldigitalsignals,thereisnoA/DandD/Aprocess,andtherewillbenoseriousloss,buttheactualBecausevariousdigitalsignalsarebasedondifferentsignalsintheoriginalgeneration,suchasSDIandHDSDIarebasedonYUVcolorgamutspace,whileVGA,DVI,HDMI,DP,etc.areRGBcolorgamutspace.WeknowthatusingtheYUVcolorgamutcanuseUVbandwidthcompressiontoreducethebandwidth.Therefore,thesignalRGBcolorgamutbasedontheYUVcolorgamuthasasmallsignalbandwidth.TheYUVcolorgamutisconvertedtotheRGBcolorgamut.letter
