UK Vintage Radio Repair and Restoration Discussion Forum - View Single Post

Karen O · 20th Jul 2011, 5:20 pm

Hi Kat,

It doesn't sound so bad but I've not done any Unix programming for a very long time. I like the idea of a standard program outputting a raw 377i file and an audio file, and then writing a small utility to handle the data repackaging and writing to the card. I think I could handle the latter.

My player format will use a sector to store one scan line. These will be in blocks of 188 or thereabouts, alternating odd and even fields. There's no need to store the blank lines as the PIC uses this time to issue the Read Sector(s) command to the card.

Audio is likely to be messy. My plan is that the last four bytes of each sector are reserved for audio, leaving 508 bytes for picture information, which is enough.

I plan to use an audio sample rate of 30.375kHz - three times line frequency. I'm a little concerned here - if the program responsible for decompression produces an audio file even slightly off this rate, there will be lip syncing problems after half an hour or so.

My pixel clock is 6.25MHz which, I think is very close to square pixels for 405 lines. Here's my maths I used to work this out:

Active lines = 377 = vertical pixels

At an aspect ration of 4:3, that means there are bout 503 horizontal pixels

These pixels are spread over an active line time of 80.3usec

So square pixel clock = 6.26MHz

Even if this is wrong, I'm still using a pixel clock of 6.25MHz!

20th Jul 2011, 5:20 pm	#73
Karen O Rest in Peace Join Date: Jul 2011 Location: Bridgnorth, Shropshire, UK. Posts: 787	Re: Simple memory card player idea Hi Kat, It doesn't sound so bad but I've not done any Unix programming for a very long time. I like the idea of a standard program outputting a raw 377i file and an audio file, and then writing a small utility to handle the data repackaging and writing to the card. I think I could handle the latter. My player format will use a sector to store one scan line. These will be in blocks of 188 or thereabouts, alternating odd and even fields. There's no need to store the blank lines as the PIC uses this time to issue the Read Sector(s) command to the card. Audio is likely to be messy. My plan is that the last four bytes of each sector are reserved for audio, leaving 508 bytes for picture information, which is enough. I plan to use an audio sample rate of 30.375kHz - three times line frequency. I'm a little concerned here - if the program responsible for decompression produces an audio file even slightly off this rate, there will be lip syncing problems after half an hour or so. My pixel clock is 6.25MHz which, I think is very close to square pixels for 405 lines. Here's my maths I used to work this out: Active lines = 377 = vertical pixels At an aspect ration of 4:3, that means there are bout 503 horizontal pixels These pixels are spread over an active line time of 80.3usec So square pixel clock = 6.26MHz Even if this is wrong, I'm still using a pixel clock of 6.25MHz!