DCTV Question for Amiga 2000

[Sean Cunningham]

Someone should look inside and see if they can pull the s-video from the board before it is combined into composite.

I have one on the way and will do that when it arrives

S-Video isn't necessarily better than standard Composite video when connecting two devices.  It depends on which device has the better comb filter.  The signal is exactly the same before chroma and luma are split.

So, for instance, if you had a Pioneer Elite dvd player and fishbowl Magnavox television that just happened to have an S-Video input, you would connect S-Video, because for sure the high end Pioneer player will have better comb filters.  If you have a no-name $20 DVD player from China and a Sony XBR^2 TV, you would connect composite because the set will have better comb filters.  And if you have the Elite player and the XBR, it's likely a wash.

I remember taking the DCTV to a local TV station to lay off some work for them and, hooked up to the scopes, it was pretty clean.  At the very least it was as good as the Truevision VID I/O Box I'd previously been using for RGB->Composite.

[Sean Cunningham]

IMHO DCTV was a revolutionary device that could have done so much more and could have helped Commodore and Amiga greatly if further development had been done and,or more game programmers had written a version of their games in DCTV format.  Can you imagine all of our favorite Amiga games running at 24bit color depth and full speed on any Amiga model.

The DCTV was simply amazing.  It was like having a TARGA board, only you could do things with the DCTV that you couldn't do with the TARGA or any other truecolor boards for PCs or Macs at the time.  Or the VideoToaster.  

In 1990/1991 and for many years after truecolor framebuffers required single frame recording to very expensive video decks in order to achieve motion.  This was typically done with composite video to analog recording devices.  3/4" U-Matic was about the least expensive, industrial quality used at the time and it's a composite recording medium.  Betacam would be the next step up in quality at a major increase in price while still only offering 300 lines of resolution though it recorded a crude form of component video.  

It wouldn't be until a few years later with the introduction of the DPS-Personal Animation Recorder that you could get realtime playback of imagery with this sort of color and quality, though the PAR was only playback and didn't have any video paint support since it didn't allow that sort of access to its framebuffer.  DCTV Paint was head and shoulders better than the kludgy way Newtek implemented paint on the VideoToaster.

I had an Amiga3000 when DCTV came out and my buddy and I were running a little graphics and Amiga training outfit in Corpus Christi, TX and though I don't recall the job that paid for my DCTV it was likely the best $200 I ever spent on the Amiga (I could be wrong on price, I was still getting my old employee discount from when I worked at the local Amiga dealer in high school).  Combined with Pageflipper + Effects I could do 640x400 ANIM playback that didn't bog, recording to 8mm video at U-Matic quality and no need for single frame controllers.

Animations that I did in Hash's Animation:Apprentice and Imagine output full-color through the DCTV got me accepted into CalArts, which got me a job at an Oscar winning effects facility shortly thereafter and a pretty amazing career, and I credit a lot of that snowball to the DCTV (and the Amiga, of course).  I wasn't showing them HAM playback or single frames.  They were seeing images that looked as   good as what their MFA students were doing with Wavefront TAV and their TARGA recording setup.

[Sean Cunningham]

If I recall, there were similar rumors about Black Belt Systems that may or may not have been helped along by them.  I remember not being impressed with their box due to lower resolution.

[Sean Cunningham]

What for though?  For folks still using classic hardware?  Composite video has kinda had its day.  It's pretty horrible looking on flat panels.  Even VGA signals look bad on them now.

I wonder if we'll ever see some hipster start-up creating new CRTs.  I've still got my first 1080-HD CRT I ever bought, all 200lbs of it, and I'm not giving it up until it simply doesn't work anymore.

[Sean Cunningham]

Yes, it was uniquely Amiga.  Very clever.  I seem to remember the slow-scan digitizer functionality to work better than any other parallel port digitizer I'd tried.  Somehow the others all ended up with weird jaggies, even using a camera to digitize flat artwork there would often be errors with the others.

[Sean Cunningham]

Anyway...

[Sean Cunningham]

Yeah, the "Plus" sounds like an interpolation trick.  Definitely not something I'd have been interested in.  The basic 320x400 screen was good enough for most folks but it was too low rez for what I was working on.  Oversampling was either non-existent or really crudely implemented in all the rendering software back then.  Working in highres was a must and being able to playback animation in highres.

Composite is still the low end standard.  It ain't history yet..  ;-)

It just looks worse than it ever has, especially on flat panels, possibly due to A-to-D.  Composite video needs a CRT to look decent.  They keep it there because it basically costs nothing to add and there's likely a lot of kids out there with hand-me-down TVs that aren't digital.  Old people too.

[Sean Cunningham]

Welcome to video?

S-VHS, Laserdisc, etc. these S-Video capable formats didn't actually store to media the video as anything but composite.  Then either your player or your set comb filtered the signal and the quality of the filter at either end determined whether or not it was advisable to do it in the player or in the set.

It wasn't a function of whether the equipment could only output composite or S-Video (which is still composite video, since chroma is not separated into its component form) this is how the base signal being displayed is recorded and generated.  It's always composite and sometimes a mildly better composite.  

Whether there is a more opportune place to tap into the process when an RGB source is being converted to composite video would depend on the components used and whether or not existing RGB->Y/C in fact created discrete streams of Y and C or did a single stream conversion and then offered a second path through a comb filter.  In the case of DCTV this is a big "if" because its whole success is based on a hack of the composite signal affecting both streams.  Ultimately the tweak to composite only slightly increases luminance resolution since chroma stays abysmally low until you're working true component video in the style of BetaCAM and better.