Why did 405 look so good?

[peter scott]

With my current playing with generating 405 and 240 line content from a PC graphics card I was quite shocked when I came to view 625 on a Sony 9-90 using VHF.

With 405 my Sony doesn't manage better than 2.5 MHz.

   

but using the same source generating 625 lines  the Sony's 2.5 MHz bandwidth on VHF looks dire:

   

Peter

[ppppenguin]

A 2.5MHz limit on 625 lines will definitely look soft. It's under half the nominal system BW.

[Geordie McBoyne]

From the vintage-radio.net Forum: https://www.vintage-radio.net/forum/show...p?t=123614&highlight=Decca+DR123+dual+standard+TV&page=2

Your attention should be drawn to post #23. This was a topic about replacing the 405 line service with the 625 line standard using the existing bandwidth allocations for System A. The narrow bandwidth 625 line picture looked decidedly soft. Just like VHS.

Geodie McBoyne.

[ppppenguin]

David, that was a very instructive experiment.

Never mind VHS, what about colour? Standard CRTs had a coarse dot or stripe pitch which gave hardly anything above 3.5MHz. There was also a filter to get rid of chroma in the Y channel which disposed of everything above 4MHz. It was only later with comb filter decoders and finer pitch CRTs that colour sets approached the resolution of a decent monochrome set.

I think there was also another psycho-optical effect at work, rather like visible scan lines suggesting a sharp picture. The aliased fine detail caused by a shadow grille could make a picture seem sharper than it actually was.

[Hartley118]

David, that was a very instructive experiment........

I think there was also another psycho-optical effect at work, rather like visible scan lines suggesting a sharp picture. The aliased fine detail caused by a shadow grille could make a picture seem sharper than it actually was.

The impression of sharpness arising from aliasing could also account for a puzzling observation on PC screen resolution.

Once upon a time, many of us were limited to a 640x480 VGA screen resolution which, though limited, produced nice crisp graphics and text.  Today, though, with a typical PC monitor resolution of 1920x1080 (HDTV), if I should happen to display a VGA picture on it, it looks to my eyes decidedly soft and fuzzy, much worse than on a native VGA screen.  For text and graphics to look crisp on my HDTV monitor, it requires an  input of HDTV resolution.

So it does appear that the best subjectively sharp image is obtained if the native resolution of the display matches that of the video signal.  A higher resolution display merely softens the subjective effect, making it look unsharp.

Without wishing to drift too far OT, I don't think that there's any analogous effect with audio: widening the frequency response of the reproduction system doesn't cause a limited bandwidth audio signal to lose any subjective crispness.

Martin

[Doodlebug]

In my collection I have this CAT9 valve of the type used
in the last vision and sound stages of the TX.

[ppppenguin]

Lovely photo in post #56 but not sure about its relevance to this thread.

Martin's comment about displaying low res images on a high res screen is interesting. Obvioulsy a lot depends on the quality of the scaler in the display. Assuming the scaler is good then it's the inherent softness of a lower res image on a large screen. What looks good on an old 14" VGA display won't be so pretty on a 24" screen. Not all scalers are as good as they might be, espeically when de-interlacing is required.

If you've ever seen Fantasia 2000 at an Imax cinema you'll recognise the problem of old material on a bigger screen. All the new animation is wonderfully sharp on the Imax but the Sorceror's Apprentice sequence is soft. They did a lovely restoration job but what ain't there on the original just ain't there on the huge screen either.

[colly0410]

I'd forgotten all about the Kell factor, I suppose this helped the Belgian 819 line system look a bit better than theory says it should with a 5 MHz bandwidth. I've never seen a Belgian 819 line picture but I could imagine it looking a bit soft..

[ppppenguin]

Kell factor is a bit of a fudge. I don't think there's any way of properly caclulating it so it ends up as the ratio between nominal resolution and apparent resolution.

[Synchrodyne]

That Kell Factor “fudginess” comes out in the NTSC (1st) deliberations as recorded by Fink.  Initially the pertinent NTSC panel chose 441 lines, the 6 MHz channel width and 4 MHz vision bandwidth being givens.  But Fink himself took a second look and recommended 525 lines, which was accepted.  Whilst equal vertical and horizontal resolution was a nominal goal, it was known that there was a reasonable amount of latitude either way in which picture quality remained acceptable.  525 lines provided a flatter field than 441 lines, and was preferred on this basis even though it resulted in a horizontal definition that was not quite equal to the vertical definition, but nevertheless within the acceptable range.

Fink didn’t refer to the Kell factor as such, but noted that opinions varied as to ratio of vertical resolution to the number of active lines, over the range from 0.5 to 0.9.  Basis his own experience, he used 0.8, but acknowledged that values such as 0.64 and 0.7 has been recommended by other researchers.  The 0.8 factor pointed to a video bandwidth of 4.78 MHz for equal horizontal and vertical definitions in a 525/60 system.  So, avoidance of lininess took precedence over definition equality.  (NTSC video bandwidth started out at 4 MHz, but was increased to 4.2 MHz when colour arrived.)

Cheers,

Steve