hy all!


doubts,doubts....

first
if my dvd source is anamorphic 16:9, and if i use overscan(1,2 or 3) it give more compression and i can see more in the left/right of the image but...
(always have "but")...it encrease the size of the blackborders on the top/bottom, then....
what's the real gain if we loose more in the top/bottom then gain in right/left

second
encoding that source 16:9 without any resize but using only filters in the script, it will show like the original aspect radio dvd or "egg head" in tv

I "think" we can have in here a clear suggestion for diff. movie sources:

So IMHO the best compromise would be:

- In case of a 1.77:1/1.85:1 movie a "resized" overscan by 2(16px) for those movies with a width's down to 480 and a resized overscan by 1(8px) for those movies with a width of 352 .... no matter if PAL or NTSC.

- In case of a 2.35:1 (widescreen/Cinemascope) movie an "overlayed" overscan by 2(16px) also here for those movies with a width's down to 480 and also an "overlayed" overscan by 1(8px) for those movies with a width of 352 .... no matter if PAL or NTSC.

Because EVERY stream on Tv will be resized to 768x576 at PAL (625 Lines@50Hz) and to 640x480 at NTSC (525 Lines@~60Hz) .... thats called TV "Pixel"-AR.

So u can assume: the smaller a encoded stream size is (no matter if width or height) the more it will be sized up finally and by this also its applied overscan area IN THE SAME RELATION! So as a 352x480 stream will be resized on TV by 200% to each direction to become the needed 640x480 ... then also a 8px overscan Border will become its double size on your TV set afterwards! Youe see?

For assurance you can do a safety test by temporary! applying at the very end of your generated Avisynth script a "bicubicresize(640,480)" and preview it in Vdub, that will be the sate then how it will look on your NTSC TV set .... in case of PAL, well as expl. above do add temporary! "bicubicresize(768,576)".
If everything is ok ... delete that just added last Bicubicresize line


According to your second question ... I hope I did understand it well.
IF you encode an ANAMORPH movie still ANAMORPH then the setted flag "16:9" at encoding settings will add the needed information to be played back well on your SAP afterwards

great hints ink:
"applying at the very end of your generated Avisynth script a "bicubicresize(640,480)" and preview it in Vdub, that will be the sate then how it will look on your NTSC TV set .... "
i understodd your explanations and follw to see the preview in vdubmod before encode little samples to watch in tv!

then i encode a sample without overscan:
MPEG2Source("D:\DVD\toy story1.D2V")
LanczosResize(720, 366, 6, 0, 708, 480) # no overscan
AddBorders(0, 57, 0, 57) # no overscan
trim(10000,150000)

and another with overscan:
MPEG2Source("D:\DVD\toy story1.D2V")
LanczosResize(672, 334, 0, 1, 720, 47# overscan 3
AddBorders(24, 73, 24, 73)# overscan 3
trim(10000,150000)

results in all tvs: correct aspect radio but with overscan i get more details
in the right/left but got BIG blackborders on top/bottom.
i have gain in the right/left on the image but
loose the image size in top/bottom more than i gain in left/right cos the blackborders top/bottom are too big no matter if overscan is 1, 2 or 3!!!

in the preview in vdubmod show the same proportional differences in sizes with and without overscan but in tv is really big differences!
i can't see advantages in get 2 cms in right/left and loose ~10 cms in the image top/bottom
what is the real gain!?!?!?

the second answer i don't understood and i encode a sample without resize:
the result is anamorphic(egg head) in tv....please ink be patience and explain me again!
thanks !

First, we must be precise on the words.

There is only ONE definition for "anamorphic" : this is a 4:3 "egg-head" picture compressed vertically during the play (by the TV set or by the DVD standalone when it reads a DVD only ) to be 16:9.

Knowing taht, if you want to have a correct picture on your TV you must have :

1/ A KVCD in MPEG1 or MPEG2 *but* you MUST have a anamorphic compatible TV ! That means a TV that can do this vertical compression.

or

2/ A KDVD encoded in MPEG2 *but* you must ask to tmpgenc to set the "anamorphic" flag ON. This flag will tell to the TV that it has to compress the picture.
To activate the flag, you must chose "16:9 display" in the "destination aspect ratio" setting of tmpgenc.

oh my God,please,give me better english!

Phil,
i did this samples in CCE and not in tmpgenc
i know that you know but do me a favor:
load this 2 scripts in vdubmod and see the size of blackborders on top/bottom:

LanczosResize(720, 366, 6, 0, 708, 480) # no overscan
AddBorders(0, 57, 0, 57) # no overscan
and
LanczosResize(672, 334, 0, 1, 720, 47# overscan 3
AddBorders(24, 73, 24, 73)# overscan 3

see how the black borders encrease in size using overscan?
this is what i'm trying to show!
the gain on left/right is small comparing with the loose in top/bottom.

see that "fantastics" ilustrations:

with script 1 in vdub or tv:
---------------> blackborder top
---------------> blackborder top
no overscan


--------------> blackborder bottom
--------------> blackborder bottom



with script2 in vdub or tv
--------------> blackborder top
big blackborder
--------------> blackborder top
overscan

--------------> blackborder bottom
big blackborder
--------------> blackborder bottom

i'm still unclear?

Yes, but taht is normal

I hope you noticed that overscan 3 is TOO MUCH. You take thios only for your demonstration, do you ?

But you do not loose anything. The picture is smaller (black borders are bigger) but you still have all the pixels.

Qhat do you mean by "the loose" ?


A lot ! I didn't understand anythign to those lines !

suspeitei desde o principio...(i knew)

see that sizes inside this 2 "quotes":
yes Phil, this is only ilustrative but no matter if we choose 1, 2 or 3, the blackborders are always bigger than without overscan!

of course we loose,we still have all pixels but the picture is smaller and the blackborders encrease.
we see more of picture in left/right but the total (area)size(x,y) of the PICTURE is smaller with overscan!
we see less PICTURE inside the same size of the screen!

this is what i'm talking about, this means "loose" for me
what is the real gain if get centimeters on the right/left but loose much more in the top/bottom of the PICTURE?

i was a liitle clear now?

and more:
"egg-head" is where my brain try to live!

Jorel I really do not see your point !
You just discover what is the overscan or what

Overscan 1 = a border of 8 pixels IN ADDITION TO THE ONE THAT ALREADY EXIST.

overcan = 2 -> 16 pixels. overscan = 3 -> 24...

As the resolution of the TOTAL picture must be 704*480, the more you add pixels to the border, the less you have for the picture itself !

I really do not see what is new or complicate in this.

The gain is that this centimeters that you gain on the left/rigt are centimeters of PICTURE that you lost before.

This is YOU that always wanted to NOT lose ANY PIXELS of the picture ! I always told you than you should use OVERLAP that is A LOT BETTER than overscan when doing DVD -> DVD jobs.

Dodn't you remember this thread we already had ?


i was a liitle clear now?

and more:
"egg-head" is where my brain try to live!
[/quote]

changing the target:
yes Phil i remeber that thread but i only see more black border on right/left using overlap(2) like in this sample:

LanczosResize(720, 366, 6, 0, 708, 480)
AddBorders(0, 57, 0, 57)
LetterBox(0, 0, 16, 16) # overlap only put this line that give blackborders and take part of the picture then i loose 32 pixels on right/ left! why is better?

turning worse?

the actual frame size on tv is 768x576(Pal)/640x480(Ntsc), but you won't see all these 768/640 horizontal pixels because some of them fall into the overscan area of the tv.
So why should you keep these pixels, when you won't see them?
the real intention of overscan/overlap is to cut the area of the movie which you won't see on tv and add black borders instead. That saves bitrate for the visible movie area.

just downsizing the movie won't give you any advantages, but having a smaller picture and bigger black borders top and bottom...
so you have to crop more of the left/right in your script to get the wanted effect!

i'm not an expert, so plz don't blame me when anything is wrong, just correct me :P

crackhead wrote:
"just downsizing the movie won't give you any advantages, but having a smaller picture and bigger black borders top and bottom... "
exact what i wrote

we got wrong informations about aspect radios,ntsc and pal,borders,etc.
if you all want a good reading,please follow that links (too big):
http://www.uwasa.fi/~f76998/video/conversion/

see the last picture in the bottom of that page....is exact what i try to show...to see -270 pixels you need a big blackborder to got a perfect AR:
http://www.hometheaterhifi.com/volum...vember-99.html

and more:
ntsc and pal differences are always posted wrong!
B,G/PAL - 625 - 50 (Europe)
M/NTSC - 525 - 59,94 (and have differences in Japan)
M/PAL - 525 - 59,94 (Brasil)
N/PAL - 625 - 50 (Argentina.Uruguai,Paraguai)
see that PAL is the color system and not the STANDARD and in STANDARD M we call as 60 but the truth is 59,94..then our reference is wrong.
the differences between PAL and NTSC are the STANDARDS and the COLORS SYSTEMS!

take a look here:
4.5 It still does not make any sense. For starters, all the 525/59.94 equipment I have only works in 720×480, not in 720×486 (and definitely not in 711×486)! How do you explain that?
http://www.uwasa.fi/~f76998/video/conversion/#4.7

then we always got wrong informations but what i really want is in the next post!!!

@ Phil,ink, Kwag,Boulder, V(gold star), ....all,

help me please....tv was my work and i understand all like in the links that i posted....i can read(i knew it all a long time) but i can't write, understand me?


what i really want is a script that show the exact blackborders in tv
exact like i see the dvd source in tvs, no more or less blackborders
...nothing more!

of course...one script for 4:3 and another for 16:9 to do my KDVDS

can you ..please?

So do NOT use ANY resize, crop or add border line of any kind !

Remove all lines concerning that points and you will have EXACTLY the same picture that with your DVD.

Thats exactly what I told you since WEEKS in telling you to just use overlapping ! The overlap is just to put black borders on the part that AREN'T DISPLAYED by the TV. Thats DOES NOT MAKE YOU LOSE ANYTHING as this part is never saw on your DVD but you will gain in compression.

Phil wrote:
" The overlap is just to put black borders on the part that AREN'T DISPLAYED by the TV."
(¡Es exactamente lo que iba yo a decir!)
and
"Thats DOES NOT MAKE YOU LOSE ANYTHING as this part is never saw on your DVD ..."
this i knew . (¡Lo sospeche desde un principio!)

"...but you will gain in compression."
this i don't knew !!!

Phil wrote:
"So do NOT use ANY resize, crop or add border line of any kind !
Remove all lines concerning that points and you will have EXACTLY the same picture that with your DVD. "
ok..i encode samples loading only source to watch in tv.
( ¡No contaban con mi astucia! )

the toy story2 (4:3) show perfect AR without resize in tv

the toy story1 (16:9) show WRONG AR without resize in tv
the original dvd HAVE blackborders and in the encodes without resize is "egg head" without blackborders.
(¡Todos mis movimientos están fríamente calculados!)

this is the point Phil, what i do in this case

Jorel, jorel ,jorel

We DID NOT GET wrong informations when doing our resizing and as you see No matter if you call it NTSC or PAl ... what counts is the ACTIVE PIXEL AREA of the movie! And that active pixel area of the movie is defined ALSO by mainly the frequency the TV set builds sthe lines and in your case it is approx. 60hz = 60 fields = approx. 30fps which is correct said 30 fps at 525 lines @ 60hz by 1000/1001 = 29.97 fps!

You cant say we did wrong as Moviestacker AND fitCD took THAT site above as reference!

SO what mainly counts is the PixelAR and NOT the Movie AR, cause the resizing (if the Movie AR is keeped) will base totally on these Pix AR's

So as said a 768x576 or 640x480 got a Pix AspectRatio of 1:1 means it will be shown as it is as the pixels at 1:1 state are in TV square pixels state, that means you can just resize that 768x576 to 702x576 which does make the image QUADRA Pixel based and that at a Pix AR of NOT 1:1 but therefore 128/117 (in case of PAL i.E.) ... check that in your quoted link and you'll understand.
So example:

768x576 = 1:1

702x576 = 128/117 (=1,09401)
so the TV does the following ... 702 x (128/117) = 768!

Nope I didnt mistyped the 702! as that would be most correctly but as mpeg bases non MOD sizes it was MODrounded up to 704

Now lets do another calculation:

480x576 gots a Pix AR or 128/78 !

So IF you resize from 768 width to SVCD size, the ACTIVE pixel area wont be 480x576 !!!!! Nope ..:
768 / (128/78 ) = 468 !!!

So the EFFECTIVE PIXEL area of a PAL SVCD Stream IS 468x576 ! (proofed in the link you posted above) Means you got 6+6 Pix in the Width which wont be seen on TV no matter if overscan setted or not (even if a PAL tv wouldnt have an overscan area) cause these will go out of the 768 width, means they will out off TV size offset!

If you would take a stream of 480x576 (whole width gots movie info)
Then the movie info of that stream would be resized on tv like this:

(480 x (128/78 )) x 576 = 787 x576 ... you see 6px on each width border/edge encoded movie information for NOTHING!


So you would ask ... "Why can FIT CD or Moviestacker do output in case of SVCD Target 480x576!!!"

Very easy ... the stream resizing calculation is like this:

First resize (internal, not inavisynth as we talk about claculation in here ) to Correct Pix AR, means to 468x576 .... so the user WANTS the 480 width, ... means calculating the image PROPORTINAL to 480x592 and cropping then the top and bottom so it gets 480x576 again

Means in case of a captured PAL 704x576 input:

DO use 468x576 incl. 6px+6px black borders if no overscan is choosen cause THIS AVOIDS vertical scale interpolation! And even with all this you still would see only the 468x576 ... so its NOT worth!

Now you say whats about 720x576 (PAL DVD Sources) ...
ok, lets calculate

720x576 gots a Pix AR of also 128/117

So lets get that baby to final 1:1 Square Tv pixels:

720 x (128/117) = 787,69 means 788 x 576
And as our PAL Tv only can display 768x576 .... you can assume why it DOESNT make sense to re-encode at 720x576 for TV purposes!

And thats also the reason WHY a direct simple rescaling to 480x576 out od the 720x576 does NOT output a wrong AR. But still these 6+6 px in your SVCD Stream wont be seen defenitely even if your TV wouldnt have an overscan area!

So a total correct "effective" resizing from 720x576 to 480x576 would be:

taking the 720 and crop on each side 8px as you saw above that these px will be OUT OF TV area. So that would end up in 704x576 by keeping the correct 128x117 Pixel AR. Then we do simple resize these 704x576 to 468x576 and add 6+6 px black borders at the side, voilá.

You can compare all that using the link above you posted AND for shure if doing these resizing in FitCD. But should set (only for testing) "No Cropping" and modify the resized height that much till the "error:XXX" will be as small as possible ... then you really see how it resizes internally AND WHAT will be cropped in effective


Now my fingers do hurt!!!!

Jorel, read again my post from 4:11 am. I explained you all there.

You missed that point :
I explain again :

- you have a TV that is 4:3 and you are playing a DVD that is anamorphic . Then the DVD PLAYER compresses vertically the picture, adds borders , and send to the TV a picture with a correct A/R.

- when you want to play it ON A KVCD the player DOES NOT ADDS THE BORDER because these feature works ONLY when you are playing a DVD, not a VCD.

If you do not have a TV that can do the job that is not done by the DVD player then YOU CAN'T have a KVCD that respect exacly what you see on your original DVD.

See the post of Inc to understand the reason of this impossibility.

in parts cos your post is too big ink:
you wrote:
"We DID NOT GET wrong informations..."
i was not talking about "ACTIVE PIXEL AREA" but was talking about AR!
from the link that i posted:
1.1 The dirty little secret revealed

Tutorials and manuals usually tend to keep very quiet and secretive about the finer technical details of digital video, particularly when it comes to the topic of (pixel) aspect ratios and image geometry.

Even if converting (resampling) video clips to other resolutions is discussed, the accompanying explanation is usually troublingly simplistic and vague – often inaccurate and misleading – and sometimes the suggested methods are just plain wrong. It is not uncommon that the examples only deal with arbitrarily chosen ("x pixels by y pixels") frame dimensions and use ideal frame aspect ratios such as 16:9 or 4:3 as the basis for calculations – not the actual pixel aspect ratios – which is usually a good indicator that the writer may not actually take the real image geometry into account at all.

It is almost as if the whole aspect ratio issue was considered some sort of dirty little secret of the video industry; black magic you could not even begin to explain to mere mortals in reasonable terms. This is a shame. In this case, there is really more to it than meets the eye. Confusing people with incomplete and watered-down explanations does not do any good to the industry.

Now that you have read this far, it is time to reward your effort with The Third Big Revelation about aspect ratios and frame sizes - the one that is usually left unsaid:
Not a single one of the commonly used digital video resolutions exactly represents the actual 4:3 or 16:9 image frame.
Shocking, isn't it? 768×576, 720×576, 704×576, 720×480, 704×480, 640×480...
none of them is exactly 4:3 or 16:9; not even the ones you may conventionally think as "square-pixel" resolutions.

So there. Now you finally know the truth. Let's find out what it actually means.

end of the "copy"

this is one of the "things" ink ! more will come ,let me read all in your post

ink, seems OT but take a look:
using the same source (toy story 1 - 16:9) and the scripts from:

MovieStacker
mpeg2source("D:\ts1\1.D2V")
BicubicResize(720, 366, 0, 0.6, 6, 0, 708, 480)#from moviestacker
AddBorders(0, 57, 0, 57)#from moviestacker

FitCD
mpeg2source("D:\ts1\1.D2V")
BicubicResize(720,368,0,0.6,8,0,704,480)# from fitcd
AddBorders(0,56,0,56)# from fitcd

what script is wrong,what is right (if any) and why

kiddin:
from my signature...
you have more than problems,
you have jorel questions

The SQUARE Pixel TV state resolutions got EXATCLTY 4:3! So no shocking at all

768/576 = 1.333333....
640/480 = 1.333333...
------------------------
4/3 = 1.333333...


16/9 = 1.7777777...

So 576x1.777777 = 1024! at 1:1 Square Pixel Aspect Ratio!... Means an Anamorph 720 DVD Source on a 16:9 Tv will have 1024x576 pixels ... thats the resolution freaky Divx backup makers do use in case of archiving max divx Movieresolutions at 1:1 PixelAspectratio.

576/1.3333 = 432 = the new height of an anamorph to non anamorph resized stream where 72+72 px borders will be added (letterboxing).

Jorel, to me it seems you still misunderstand the difference between Movie AR and Pixel AR

Movie AR is how the Movie will have its proportion afterwards on Tv, so just Fullscreen 4:3 (1.33:1 filled out), or in a VIEWABLE 16:9 (1.78:1 / 1.85:1) proportion (black bars top/bottom) ... or even Cinemascope (2.35:1 .. within a 4:3 Tv Set means even bigger Borders).

ALL other Resolutions you showed above do have a different PIXEL AspectRatio compared to 1:1 square Pixels, but as Phil explained, the SAP will "stretch" them to correct 1:1 .... and by this you will get that squeezed KsVCD shown correctly on your TV.



According to your scripts... did you performed in DVD2AVI a cropping of the orig black borders??? As the resized Movie height where borders will be added to me on the first (fast) view seems to be a bit confusing.

ink wrote:
"Jorel, to me it seems you still misunderstand the difference between Movie AR and Pixel AR "
no,i don't ...maybe my english show it but i know the differences!
and your explanations are cool!


ink wrote more...and this is important:
"did you performed in DVD2AVI a cropping of the orig black borders...
...on the first (fast) view seems to be a bit confusing. "

for me is more than confuse ink......
the vobs in dvd2avi DON'T have black borders( is egg head)
but the original dvd in tv HAVE big black borders(no egg head).
script without resizes don't have blackborders(are egg head) like the vobs in dvd2avi
scripts from fitcd/moviestacker(little differences between progs) show the blackborders like(seamless) the original dvd(no egg head).
than my big doubt is if i use the script With resizes to show the image in tv like the original dvd OR without resizes like the preview in dvd2avi that don't have blackborders!?!?!?
it's not a question of my "taste" and this is driving me crazy!!!
i'm lost and don't know where i am, don't know where to go and don't know what i choose.....resize or not resize?
i'm in the middle of "don't know what to do".....not kiddin!
what i'm missing ink thanks in advance .

Because the movie in the theater was originally in 1.85. Even if it should have a little border (16 pixel) but perhaps they cut it to do the DVD in 1.77 (true 16:9 A/R).
Because the DVD player adds it
Did you read my posts or only Inc's ones ?

grrrrrrrrrrrrrr..............

Do you have a TV that can add the borders itself YES OR NOT ?
In other words, does your TV handle anamorphic pictures or not ?
In other simpler words : can you force your TV in 16:9 mode or not ?

If not, for sure you MUST add these borders during encoding because they won't appears by god wish.

Because you don't read all the answers you received !

My two previous posts