Hello , I try myself to realize MDMA-Based Dual-SHARC+ Parallel Pipeline Audio Talkthrough in SC573, here is my project. When I run the project , there is not output for 1962s
it printf in window:
Core1: Opening MDMA RAW streamCore1: Opening MDMA FILTERED streamCore1: Waiting for slave to install interrupt handlersCore2: MDMA RAW interrupt handler installedCore2: MDMA FILTERED interrupt handler installedCore2: SHARC link connection established
could you help me solve it ? it have spent some days.
thanks very much!4010.ADC DAC Audio Playback (SC573 SHARC).rar
Moving to CrossCore Embedded Studio and Add-ins
We understand that you have tried to port the application note project EE-383: MDMA-Based Dual-SHARC+ Parallel Pipeline Audio Talkthrough to ADSP-SC573.
In ADSP-SC589 EE-383 application, sequencing of events between the cores are controlled with help of using 32-bit variable stored in L2 memory.
That is #define sharc_flag_in_L2 (volatile unsigned int *)0x20087000. This is correct for ADSP-SC589.
But for ADSP-SC573,L2 memory space starts with 0x20000000 address. Please refer Figure 5. ADSP-SC57x/ADSP-2157x Memory Map in the following data sheet for ADSP-SC573.
For your application, variable should be stored in L2 memory( 0x20000000).
Hence the statement will be #define sharc_flag_in_L2 (volatile unsigned int *)0x20000000.
Kindly ensure that this change should be in SHARC_linkInterface.h files in both cores.
We have modified your code and tested it in ADSP-SC573 Ez-Kit. We can able to hear audio output at DAC side(1962).
Please try and let us know how you are getting on.
Best Regards,Santha kumari.K
thanks.I find the delay of MDMA-Based Dual-SHARC+ Parallel Pipeline Audio Talkthrough to ADSP-SC573 is longer than ADC_DAC_Playback_SC573_SHARC_Core. Is it possible to reduce the delay time of MDMA-Based Dual-SHARC+ Parallel Pipeline Audio Talkthrough to ADSP-SC573? Now I want the delay to be as small as possible of the Audio Talkthrough, What is the minimum delay？
Hello,MDMA-Based Dual-SHARC+ Parallel Pipeline Audio Talk-through project utilizes two SHARC cores and providing code flexibility that user can add their own filter algorithms. Once audio data is received in SHARC0,raw data is moved to SHARC1.In parallel,out of four channels two channels(0,2) are filtered in SHARC0. After getting raw data from SHARC0,now remaining channels(1,3) are filtered in SHARC1.Now filtered data(0,2) in SHARC0 is moved to SHARC1 via MDMA for merging with already filtered data(1,3) in SHARC1.Now entire data is moved to DAC.But playback project uses simple process buffer function to get audio data and copy it directly to DAC.More processes are not needed here.In case of more clarification needed,kindly find the EE-383 application note link belowwww.analog.com/.../EE383v01.pdf.Are you facing the delay while hearing audio at DAC side(lagging audio) or is it a kind of delay measurement/throughput in code?Kindly provide more details about the delay and how you are measuring it in the project to assist you better.Regards,Santha kumari.K