STPC Consumer-II Video input processor choices

bpeterson · ‎2003-04-03

Posted on April 04, 2003 at 07:18

bpeterson · ‎2003-03-31

Posted on March 31, 2003 at 18:26

Can anybody tell me if there is a preference using the SAA7111A decoder over the SAA7113H decoder? Does one work better than the other with the STPC Consumer-II?

I understand the SAA7113H only supports CCIR-656 4:2:2 8-bit data stream and the SAA7111A supports both CCIR-601 and CCIR-656.

The price on both parts is very close and board space seems about the same given the difference in pin pitch (0.5mm vs 0.8mm).

Any ideas how these decoders stack up to the Micronas VPX322xE video input processors? I see this VIP has scaling capabilities, but would this be redundant with the capabilities of the STPC Consumer-II?

Thank-you!

thierry239955_st · ‎2003-04-02

Posted on April 03, 2003 at 06:28

In term of connection between the chip and the STPC, there is no problem uses either SAA7111, either SAA7113. CCIR-656 is easier to use and most of the design use SAA7111 in CCIR-656 (probably the reason CCIR-601 support have been suppress in SAA7113 version).

In term of features, the interest of SAA7113 compare to SAA7111 is the downscaling. STPC video overlay can do upscaling only. It is also more interesting to do downscaling at the level of the SAA to decrease the memory bandwidth (amount of data is decrease before to enter STPC). It depend if your application need downscaling or if upscaling is enough.

You should also maybe check for the perenity of the product. Will Philips continue to produce SAA7111 in the future and for how long?

I think SAA and Micronas have got almost the same level of features.

You should be carefull with one thing. Downscaling isn't done through the same way. Micronas put all the data at the beginning of lines and increase the size of the blank in consequence. It is directly compatible with STPC. SAA7113 add 0 values between pixels to suppress some of them. 0 value are invalid and the SAA7113 expect they will be suppress by the STPC but it is not the case. To be compatible with SAA7113 downscaling, you need to add a very small external logic to stop the pixel clock when an invalid value is generated (there is a signal which go out of the SAA7113 for that) so that the data isn't written into memory by the STPC. This logic is very simple but you need to think about this when you design your board.

Farfalla

bpeterson · ‎2003-04-03

Posted on April 03, 2003 at 14:37

Farfalla,

Thank-you very much for this insite into the video decoders. In our application we do need to down scale the live video and lay it over a 2D graphics screen. Kind of like a picture-in-picture mode.

I thought there were provisions to skip lines and pixels in the STPC to produce a down scale picture for overlay purposes. Off loading the scaling to the VIP does sound like it would speed up the process.

Sounds like the Micronas VPX322xE may be a better solution given it does not take as much glue logic to get the desired effects. If I understand your reply correctly, the Micronas still builds the same horzontal line buffer size, but the blank video data is at the end of the scan lines. To read out the smaller scaled picture, would you just use an offset in the frame buffer during the read cycle?

Thanks again, Pete

[ This message was edited by: Pete on 03-04-2003 18:11 ]

thierry239955_st · ‎2003-04-03

Posted on April 04, 2003 at 07:18

Hello Pete,

To be more precise, the increase of blank is done at the beginning of the line... as the Micronas can't process the data it didn't received yet!

It means a line like this:

BBBB*******************BBBB

will became

BBBBBBBBBB************BBBB

It is the case with the version of Micronas present on the STPC evaluation board. I don't know if it correspond to your reference but we can suppose all the Micronas use the same way.

Farfalla