[comp.sys.amiga] Interlace

bartonr@psu-cs.cs.pdx.edu (Robert Barton) (05/19/89)

In article <15500001@ux1.cso.uiuc.edu> phil@ux1.cso.uiuc.edu writes:
>Broadcast television does not suffer from flicker because the change from
>one line (scanned in one cycle) to the next (scanned in the other cycle)
>is very little.

  It just depends on what you're watching.  If you have cable, check out
CNN Headline News when they show one of their weather maps with the narrow
horizontal state lines and you can see flicker in the Midwest.

>High Definition Television, unfortunately, is continuing with the interlace
>system at about the same frame rate.

  The proposals I have seen for HDTV also call for 1050 scan lines instead
of the current 525 lines.

(* -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- *)


(* -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- *)

wn0e+@andrew.cmu.edu (William Nichols) (05/20/89)

In article <15500001@ux1.cso.uiuc.edu> phil@ux1.cso.uiuc.edu writes:
>Broadcast television does not suffer from flicker because the change from
>one line (scanned in one cycle) to the next (scanned in the other cycle)
>is very little.

Just look at special graphics on a tv screen, like those facsimilies
of credit cards on a commercial, or the team logos they show on
sporting event broadcasts.  Look at the graphics on The Computer Chronicals.
Anything with distinct horizontal lines of differing contrast.  They all 
flicker like crazy.  

Untill I got my Amiga, I never really understood why that flicker was
there, now I know.

On the other hand, many ham displays on the amiga look just like a TV screen,
i.e. do not seem to flicker at all.  

Bill N.

papa@pollux.usc.edu (Marco Papa) (05/21/89)

>In article <15500001@ux1.cso.uiuc.edu> phil@ux1.cso.uiuc.edu writes:
>>Broadcast television does not suffer from flicker because the change from
>>one line (scanned in one cycle) to the next (scanned in the other cycle)
>>is very little.

You must be joking! Get your facts, dude. Broadcast television suffers the
SAME flicker problems that the Amiga shows with images that have 1 scanline
tall horizontal lines. This is simply due to the NTSC standard. The flicker
on TV is very noticeable during the sports and weather reports.  Images
that have a lot of color variations don't flicker (on either a broadcast
TV or the Amiga) because the eye "averages" the colors.

-- Marco Papa 'Doc'
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
uucp:...!pollux!papa       BIX:papa       ARPAnet:pollux!papa@oberon.usc.edu
 "There's Alpha, Beta, Gamma and Diga!" -- Leo Schwab [quoting Rick Unland]
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

elg@killer.DALLAS.TX.US (Eric Green) (05/22/89)

in article <17324@usc.edu>, papa@pollux.usc.edu (Marco Papa) says:
>>In article <15500001@ux1.cso.uiuc.edu> phil@ux1.cso.uiuc.edu writes:
>>>Broadcast television does not suffer from flicker because the change from
>>>one line (scanned in one cycle) to the next (scanned in the other cycle)
>>>is very little.
> 
> You must be joking! Get your facts, dude. Broadcast television suffers the
> SAME flicker problems that the Amiga shows with images that have 1
> scanline tall horizontal lines. This is simply due to the NTSC
> standard. The flicker

You're both right. Marco, it's the low contrast of most broadcast
pictures that allow the eye to "average" the colors -- i.e., what the
guy above said. On the other hand -- I am sick and tired of seeing the
output of $10,000 studio character generators crawling across the
bottom of my screen. Anti-aliasing? They never heard the word! Nice
fonts? Uh uh, no way, Early Timex-Sinclair is the only font they do.

> that have a lot of color variations don't flicker (on either a broadcast
> TV or the Amiga) because the eye "averages" the colors.

The keyword is "Contrast", Marco. The number of colors doesn't enter
into it. 

--
|    // Eric Lee Green              P.O. Box 92191, Lafayette, LA 70509     |
|   //  ..!{ames,decwrl,mit-eddie,osu-cis}!killer!elg     (318)989-9849     |
|  //    Join the Church of HAL, and worship at the altar of all computers  |
|\X/   with three-letter names (e.g. IBM and DEC). White lab coats optional.|

tomb@hplsla.HP.COM (Tom Bruhns) (05/22/89)

> papa@pollux.usc.edu (Marco Papa) writes:
>>In article <15500001@ux1.cso.uiuc.edu> phil@ux1.cso.uiuc.edu writes:
>>>Broadcast television does not suffer from flicker because the change from
>>>one line (scanned in one cycle) to the next (scanned in the other cycle)
>>>is very little.
>
>You must be joking! Get your facts, dude. Broadcast television suffers the
>SAME flicker problems that the Amiga shows with images that have 1 scanline
>tall horizontal lines. This is simply due to the NTSC standard. The flicker
>on TV is very noticeable during the sports and weather reports.  Images
>that have a lot of color variations don't flicker (on either a broadcast
>TV or the Amiga) because the eye "averages" the colors.

Well, "dude", the last statement may indicate you need a bit of brushing
up on facts yourself.  If I build an image with "a lot of color variations"
by making the first field of a frame entirely different in intensity (as
well as hue and saturation) from the second field, I can about guarantee
you will be able to see flicker in it on most standard NTSC tv's.  The
least flicker is on images that are (1) very similar from one field to
the next (e.g., an image with no distinct lines but only gradual spatial
changes -- like most program images that are broadcast, except as noted
by previous posters.  And it probably helps that the images are usually
in motion.), and (2) dark.  Furthermore, because NTSC also interlaces the
color subcarrier with the horizontal (and therefore with the vertical)
rates, pictures with highly saturated colors, when displayed on a monitor
that doesn't try to play games with what it "thinks" should be luminance
and what should be chrominance, will show a lot of _color_ flicker.  This
is particularly noticable on sharp vertical edges between saturated colors.

There are perhaps some practical lessons in this for those who would like
to keep down flicker on interlaced (computer) screens:

 -- Since an RGB monitor doesn't use NTSC color encoding, you don't have
    to worry about the last problem.  Your eye may still perceive flicker,
    though, if one field has significantly more of one of your colors than
    the other(s).  So chose a set of colors for the screen that are not 
    too different in intensity.  There's an obvious tradeoff here with
    legibility of (text) information.
 -- Keep the intensity low.  This is easier to use if your room lighting
    is also fairly dim.  If you can't control the room lighting, you
    can achieve the same effect with dark glasses, leaving the screen
    itself fairly bright.
 -- Use text fonts that put the same number of "on" pixels on each field
    (that is, on alternate lines).  A test of this is to display a few
    lines of equal signs and/or minuses.  The minuses should be two lines
    high, and the equal signs should have two rows blank between the
    turned-on rows (or for larger fonts, something equivalent).
 -- If all else fails, de-interlace the screen (e.g., FlickerFixer).

Hope this is useful.
>
>-- Marco Papa 'Doc'

Tom Bruhns
tomb%hplsla@hplabs.hp.com

papa@pollux.usc.edu (Marco Papa) (05/22/89)

In article <8169@killer.DALLAS.TX.US> elg@killer.DALLAS.TX.US (Eric Green) writes:
>in article <17324@usc.edu>, papa@pollux.usc.edu (Marco Papa) says:
>> that have a lot of color variations don't flicker (on either a broadcast
>> TV or the Amiga) because the eye "averages" the colors.
>
>The keyword is "Contrast", Marco. The number of colors doesn't enter
>into it. 

Thanks for being more precise. I stand corrected.

-- Marco Papa 'Doc'

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
uucp:...!pollux!papa       BIX:papa       ARPAnet:pollux!papa@oberon.usc.edu
 "There's Alpha, Beta, Gamma and Diga!" -- Leo Schwab [quoting Rick Unland]
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

papa@pollux.usc.edu (Marco Papa) (05/24/89)

I just got a call from Rick Unland of CBM (most of Byte-by-Byte ray traced
images are his, though as usual never credited :-(, who told me that flicker
depends BOTH on contrast of luminance AND contrast of crominance (which
depends on color variations).  So I should'nt have been so quick to back off
from my statement.

-- Marco Papa 'Doc'
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
uucp:...!pollux!papa       BIX:papa       ARPAnet:pollux!papa@oberon.usc.edu
 "There's Alpha, Beta, Gamma and Diga!" -- Leo Schwab [quoting Rick Unland]
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=