Encoding: Pal dvd to ntsc dvd - how will it look?
 

I have ripped and converted a pal dvd to ntsc dvd and i'm wondering how it looks on a real ntsc tv. The converted ntsc dvd looks ok on my pal multisystem tv. If somebody is interested in helping me with this problem can download a short test file here: http://212.214.94.37/pmk/

vcd4ever.

Looks ok on WinDVD, and viewed with BitRate viewer, the frame rate and everything looks correct. But your bit rate is very low for a 720x480 MPEG-2 encode, and there are visible artifacts on the clip :!:
Action scenes will look very blocky with the settings you used.

-kwag

Kwag,

I noticed some artifacts on the original dvd :( when ripping and encoding the testfile, so i decided to use a lowbitrate for reducing the filesize. The quality was not important for me because i was mostly wondering if i had used the correct aspectratio when converting the pal dvd to ntsc dvd. The artifacts i noticed when playing the original dvd on my standalone player was some sort of crawling pixels around objects and also some artifacts that looked like the famous "staircase" and "combo artifacts". I don't know if it's possible to remove this artifacts and because the quality was not important for the test i uploaded the testfile. The quality of the testfile when it comes to the artifacts should have been much better if i had encoded it as a standard vcd 8O because the resizing from full d1 to sif is very good at reducing artifacts. The only format i'm using is standard vcd :oops: so i don't know why i uploaded this crappy testfile :evil: . I have therefore ripped and encoded the same clip to a ntscfilm vcd. I hope it's possible for you to download the new testfile and that the quality is better, if the quality or anything else isn't good, please let me know.

Many thanks to you Kwag for helping me with this problem.

vcd4ever.

I see you're not using any AviSynth filters :?:
As far as the clip, it does play correctly.

-kwag

The long answer why i'm not using any avisynth filters.
 

Kwag,

The first time i tested using avisynth filters was in 1998 and the filters avaible then was not imo that useful. I therefore decided in 1998 to do all my vcd encoding tests without using avisynth. I began my test encodings for acheving the best possible quality when capturing and encoding to vcd in 1998 and 3.5 years later i had achived the vcd quality that was acceptable for me. The only thing i'm not so happy about is the fact that it did take me 3 years to come to the conclusion that the only way to achive acceptable quality for me on vcd's was probably to use filters. In the remaining 6 months, 2 months was for finding out if it was possible to remove or reducing artifacts by using filters, the answer was yes so the remaining 4 months was for finding out what filters to use. I have now used my secret filterchain since 2001 when it was finnished and the filterchain is still used when capturing and encoding to vcd. I have lately been thinking about trying the latest version of avisynth because i haven't tried any new avisynth version since 1998, i was also thinking about trying the motion adaptive filter that i have read about here on kvcd but i decided that it's more of a challenge for me trying to improve my filterchain without using avisynth or the motion adaptive filter. I have downloaded all the samplefiles that has been encoded with the use of the motion adaptive filter and played them on my standalone dvdplayer and the quality is amazing.

Thanks Kwag for the help, you have been so helpfull that i decided to post my complete quality research story here on kvcd instead of dvdrhelp.

vcd4ever.

Hey vcd4ever..

hehe.. I remember you, ..your 40mb test clip that is.

* (AVI) - (120702) - (by vcd4ever) - (test clip of) - darkAngel.avi


http://www.digitalfaq.com/archives/error.gif

-- -- -- -- -- -- -- -- -- DARK ANGEL -- -- -- -- -- -- -- -- -- --


..the fun of that filter challenge.

I still have that clip :) mind you. During my on-going seeking of the
"ultimate" quality.. I never just throw out a clip. They're sort of
a timecapsole of my video progress, and your file was just a reflection
(reminder) of that. Also, I just re-checked it out. And, I happy to
say, it's still good quality :) Did you remember what setting you
used on that test clip, coming from PAL land ??

The "holy-grail" is close, but not close enough hehe :lol:

Be good.
-vhelp

vhelp,

I'm suprised that you still have the samplefile and that the quality is still good. My first impression after reading your post was, if the quality is still good my filterchain couldn't have been that bad :D . This also makes me very happy because i'm still using the same filterchain. I haven't yet managed to improve the vcd quality by enhancing my old filterchain :( but i believe that the "holy grail" is somewhere in the neighborhood. I have also been thinking that a possible explanation why i haven't managed to improve my filterchain and vcd quality is perhaps that i have reached the quality level for the vcd format when it comes to encoding captured materials to vcd :roll: .

Now for the answer to your question:

The darkangel sample was captured from tv at 336x288 with mjpeg and then frameserved from virtualdub to tmpgenc. The settings used in tmpgenc for pal was the default pal template and for ntscfilm the default ntscfilm template. The only modifications done on both default templates was this, motion search: high quality(slow), image positioning: center, gop structure: I=1 P=4 B=2, sequence header interval=1 and no detect scene changes. The only difference between the pal and ntsc sample was the resizing and framerate conversion needed for encoding the pal capture to ntscfilm.

vcd4ever.

Hi vcd4ever..

Actually, what I ment was the the overall quality after the filter clean up.
That pic (above) is straight from the original .AVI file you shared. If you
magnify it, you'll experience "daja-voo" (or see the line nose) This has ben
plauging me since I upgraded to an ESC K7S5A mobo, and I'm now comming
to the belief that the whole cause of noisy captures is due to the mobo and
chipet/cpu, among other things. See below link for an exhaustive research
that has not ben resolved yet, since Aug 17, 2002 over at vcdhelp:

* Line Noise - mostly from VCR sources...not quite solved yet.

.
.
and your capture sample .AVI file showed the exact same noise as in mine.
That's my ATI-TV Wonder pci, in my ESC K7S5A mobo setup. So, i know
how you feel, and I've ben their w/ lots of filter-chains.. I've got an army
full of them scattered across my harddrive. And, in order to keep track of
which is which and which does what etc, I've had to resort to naming them
w/ a brief desc. of filter(s) used etc. That's for vdub's filters, but for those
w/ AVIsynth, they are a little eaiser to catalog etc. But, at the moment, I'm
a bit messy w/ my files, and I coudn't find some as quickly as I'd like.

But, right now, I mostly use my ADVC-100 for just about everything. And,
it has ZERO noise. When I'm bored or need a break from things, I give
this "noise" thing some attention.

Well, it's a good thing we love what we do.

Sorry if I went too far off topic here :roll:

Have a great weekend all.
-vhelp

vhelp,

Sorry for not understanding your question, but i assumed that you where in NTSC land and wanted to know what settings i have used when encoding the avifile to ntsc mpeg because you had posted a picture from the avifile and asking the question if i remembered what settings i have used on that clip coming from PAL land. I know believe thanks to your reply that the question was what settings i have used for cleaning up the avifile before encoding to mpeg. I didn't know that i also was having this "line noise" in my captured avifiles because i borrowed 4 different tv-cards some years ago for testing and the quality from these 4 tv-cards was exactly the same and therefore i came to the conclusion that this quality is what tv-cards can produce. The tv-card i'm using now had exactly the same quality as the 4 tested tv-cards and because of this i considered this tv-card a good tv-card to buy. If the filters i was using before encoding the avifile to mpeg was reducing or removing the "line noise" i must do some encoding tests because the only noise reduction used is tmpgenc's noise reduction with the default settings. Is it possible for you to capture and uploading a short avi testfile that is not having this "line noise" for me to download and use for testing purposes. I hope this post is more useful than the previous one.

vcd4ever.

hi vcd4ever..

Too bad I can't do a capture w/ that same "dark angle" episode, because
that would really help you to understand, from YOUR point-of-view in PAL
land. Anyways.. Then, I would capture w/ my ATI-TV Wonder (ATW) card
and ADVC-100 and post here, the results for ya to experiment on.

But, as far as Line Noise in your captures, trust me.. that's what you
have in that clip from 12.07.02 for sure. I had (still have) the same noise
when I use my ATW on my main pc (esc k7s5a) though sometimes, I seem
to get a little better capture quality (less noise) if I fiddle w/ my BIOS and
lower the clock speed to 100/100 vs. my 133/133. It seems to help, but
I'm sure 100% sure it is, cause I don't have any "gauging" tool to use in
this. Anyways..

If I can do you a test clip or two I will. But, I don't want to confuse this
topics title :wink: Anyways..

I'll have a lookseez at what I can come up w/ later on maybe. I'm just soo
busy w/ LOTS of fund stuff here, and I was planning on shutting down my
pc (causng me some issues w/ weirdness)

See you later perhaps.
-vhelp