Re: Apple II+ 80-col video conversion

[Michael J. Mahon <mjmahon@aol.com>]

Curtis Schroeder <publicpanther@gmail.com> wrote:
> On Sunday, March 24, 2013 1:21:51 AM UTC, Michael J. Mahon wrote:
>> Curtis Schroeder wrote:
>>> Again, what I am most interested in is recommendations for a PCIe 1x
>>> video capture solution or more detailed ][+ 80-col video information that
>>> might allow me to make an educated selection, i.e. do I need NTSC 4.43
>>> support, advanced de-interlacing, or both?  What feature(s) made the 15K RGB successful?
>> 
>> The fundamental problem for any video sampling solution (any non-analog 
>> solution) is the coherence of the video sampler with the Apple's video dot
>> clock--14.3MHz for most modes and 16MHz for IIgs SHR mode. 
>> 
>> If the converter dot sync range accommodates these frequencies, the output 
>> can be sharp and alias-free. 
>> 
>> For good 80-column text, 14.3MHz is the critical frequency. Sampling at any
>> integral multiple of 14.3MHz will result in crisp, non-aliased text. 
> 
> I found some more detailed Apple II video information (40-col, 80-col,
> HIRES, DHIRES, and IIgs) in this presentation:
> http://www.kansasfest.org/wp-content/uploads/2009-ferdinand-video.pdf
> 
> As I seek a better understanding of the ][+ 80-col video, I think there
> is more to it than just the 14.3MHz frequency.  The NTSC standard crystal
> frequency is 14.31818, which is a 4x multiple of the NTSC M color
> subcarrier, so the ][+ is on spec in this regard.  Of more importance may
> be a converter or capture card that can handle the Apple II's
> non-interlaced video and higher line bandwidth required by 80-col or
> DHIRES video (even though the bandwidth is technically still within NTSC
> 3.58 display spec)?
> 
> Is it possible the better converters treat the Apple II +/e/c video as PAL-M?
> https://en.wikipedia.org/wiki/NTSC#PAL-M_.28Brazil.29

All NTSC Apple II's have accurate color subcarrier frequency and a 4x dot
clock, since that is the design principle by which they create video
signals that are interpreted as NTSC color video. 

This fact does not in any way guarantee that a sampling converter will
*sample* at a phase-locked 14.3MHz. In fact, most converters seem to sample
the video at a lower frequency, even though they must also lock to the
3.58MHz color burst to recover the color information in the composite video
signal. 

Video sampling frequency is seldom described for converters, and is
probably determined by the chip they use and the amount of internal memory.


-michael - NadaNet 3.1 and AppleCrate II: http://home.comcast.net/~mjmahon