Copy Video Stream Several Times

[Altera_Forum]

Hello, 

I try to copy my incoming video stream several times. My first ideas were to split the signal before using the SOPC and use multiple clocked video inputs. This works fine with 2 inputs, not more. Then i tried to use the Color Plane Sequencer (CPS). CVI -> CPS -------> CPS -------> CPS -------> CPS...  |>Stream1 |>Stream2 |>Stream3 |>Stream4... This works, but if I do clipping, frame buffering and scaling with the streams, i only get a black screen. (The settings of clipper, frame buffer and scaler are correct, if i use only one stream with them, they are working). Does somebody has an idea, how i can split my stream?  Thank you!

[Altera_Forum]

Hello, 

 It seems like you should be using multiple CVIs connected to the same set of input wires. If you duplicate streams with the CPS you will not be able to process them completely independently because, with Avalon-ST, downstream issues on one path will overflow back to your input and the full system will collapse. If you use multiple CVIs, the datapaths will not interact with each other until you try to join them (assuming this is what you planned)    --- Quote Start ---  Hello, I try to copy my incoming video stream several times. My first ideas were to split the signal before using the SOPC and use multiple clocked video inputs. This works fine with 2 inputs, not more. Then i tried to use the Color Plane Sequencer (CPS). CVI -> CPS -------> CPS -------> CPS -------> CPS...  |>Stream1 |>Stream2 |>Stream3 |>Stream4... This works, but if I do clipping, frame buffering and scaling with the streams, i only get a black screen. (The settings of clipper, frame buffer and scaler are correct, if i use only one stream with them, they are working). Does somebody has an idea, how i can split my stream?  Thank you!  --- Quote End ---  

[Altera_Forum]

--- Quote Start ---  Hello,  It seems like you should be using multiple CVIs connected to the same set of input wires. If you duplicate streams with the CPS you will not be able to process them completely independently because, with Avalon-ST, downstream issues on one path will overflow back to your input and the full system will collapse. If you use multiple CVIs, the datapaths will not interact with each other until you try to join them (assuming this is what you planned)  --- Quote End ---    Hello, thats what I try to say with  "My first ideas were to split the signal before using the SOPC and use multiple clocked video inputs. This works fine with 2 inputs, not more." Is it possible that I use wrong FrameBuffer settings? They are: v Frame Dropping v Frame Repetition  Number of packets buffered per frame: 1 Maximum packet lenght: 10  A MM Width: 64 Write Only Master FIFO: 1024 Write Only Master Burst Target: 64 Read Only Master FIFO: 1024 Red Only Master Burst Target: 64  v Use sperate Clocks  Base Adresses of Frame Buffer are: First Buffer 0 Second Buffer 400000 Third Buffer 800000 Fourth Buffer C00000 Fifth Buffer 1000000...  My system is a Cyclone III Development Board with Bitec HSMC Digital Audio/ Video adapter.  Thank you!

[Altera_Forum]

Is there a better approach to cloning video output than using multiple CVIs at the video source? I have a system where a video input is processed (clip/scale, deinterlace, frame buffer) and then mixed with OSD and test pattern generator. I want to show the same mixed output video on two different interfaces VGA and SDI. I am thinking of these two approaches: 

 1. Use Color Plane Seq to split the AV-Stream and use frame buffer on each path; OR 2. Take the output from the VGA CVO block and feed it back into a CVI block to regenerate AV-Stream then send it to another CVO block connected to the SDI core. Not sure if I need a Frame Buffer though.  Any suggestions? Note, both VGA and SDI will have same resolution and frame rate to keep things simple.