digitalFAQ.com Forums [Archives]

digitalFAQ.com Forums [Archives] (http://www.digitalfaq.com/archives/)
-   Video Encoding and Conversion (http://www.digitalfaq.com/archives/encode/)
-   -   Kvcd: change to 352 instead of 528 for horizontal resolution? (http://www.digitalfaq.com/archives/encode/5651-kvcd-change-352-a.html)

ong001 09-18-2003 12:27 PM

Kvcd: change to 352 instead of 528 for horizontal resolution?
 
I hope Kwag won't kick me out for using this subject title.

But i hope this post will lead kvcd to another new height.

Let me outline the 2 steps process:

Step 1.
start as usual for kvcdx3 mpg1, then change to 352 instead of 528 for horizontal resolution both in the script and in TmpgEnc. Set the CQ to 100.
Encode as usual.

Step2.
Load the output (either m1v or mpg) as input to TmpgEnc.
Load the standard Video CD template.
Put in the audio source
Use Highest quality to encode.

Finally burn the mpg file as standard compliant vcd and view on TV.

Please Post your comments on what you see on your TV

hi jorel, you are on line. Are you going to be the first to reach 3,000 posts?


abc

jorel 09-18-2003 12:40 PM

hy abc!

me?...first to reach 3000?
you forgot Kwag,he have 3 times more than me!
:lol:

abc,...i don't understand your target?!?!?
encoding 2 times?!?
what's the reason?
:?

ong001 09-18-2003 12:54 PM

Hi jorel,

Someone told you the first to reach 3,000 posts to buy something right?

Most people (ALL?) thought that encoding mpg videos again lead to lost in quality.

But k2vcd would give better quality! :roll:

The advantage is that the resulting vcd can be played on all vcd players.

For clean good source, the result is : see for yourself :!:


Ong

kwag 09-18-2003 01:00 PM

Sorry :!:
You're double encoding.
Compare your first encode to the second encode.
You'll clearly see the first encode far better that the second.
Just blow up the images in vdub, and you'll see what I mean ;)

-kwag

ong001 09-18-2003 01:13 PM

Hi kwag,

i think now a lot of people in some forums still don't believe KVCD gives quality output.

They keep argue and argue but willl not try out kvcd.

Yes, it is double encoding. :lol:

The first encode is better than the second encode.
But the first encode is using CQ100 :!:
and no one is using that for kvcd.

But k2vcd will give 80 minutes of good and even excellent quality
video that will play on ALL vcd players. (ALL DVD players also?)

abc

BobNET 09-18-2003 01:29 PM

Quote:

Originally Posted by abc
But k2vcd will give 80 minutes of good and even excellant quality
video that will play on ALL vcd players. (ALL DVD players also?)

Why not use one of the KVCD scripts, but when encoding select the standard VCD template with the Notch matrix? That will produce something that'll play on all VCD players, and should look better than encoding directly to VCD. And I can almost guarantee that it'll look better than encoding to MPEG twice...

This is what I use to make VCDs out of TV shows; I know I'll want to play them at friends' places, but I seem to be the only one with a DVD player that'll play KVCD... :?

kwag 09-18-2003 02:48 PM

Quote:

Originally Posted by abc

But k2vcd will give 80 minutes of good and even excellant quality
video that will play on ALL vcd players. (ALL DVD players also?)

But that's exactly the same thing, if you set MIN=1,150 MAX=1,150 and CQ=100.
There's no difference, other than a slight VBR instead of CBR.
The quality will be identical if you encode as above, or if you encode 1,150Kbps CBR.

-kwag

ong001 09-18-2003 08:22 PM

Hello BobNET,

I think i will try your method to see the result.

kwag, i'll try to set MIN=1,150 MAX=1,150 and CQ=100. and see the results also.


But i have been doing double encoding for quite a long time and
the results are good. Actually the first pass can use some other
template than kvcdx3 but with higher bitrate.

For a good source, on my 25" TV, i can have near dvd (or dvd?) qualities.
I think one pass encoding to standard vcd can never do that!

No matter how good is the source, direct encoding to standard mpg1
would produce mosquito noise. TmpgEnc would add blockiness to static scenes also.

k2vcd is practically free of that. It also produces smoother playback


kwag, i can send you one of my samples on vcd.
Seeing is believing.
That's why i said view the results on your TV.


abc

kwag 09-18-2003 09:13 PM

Quote:

Originally Posted by abc

No matter how good is the source, direct encoding to standard mpg1
would produce mosquito noise. TmpgEnc would add blockiness to static scenes also.

Of course, because you're smoothing out the picture, and removing the high frequency details.
Sorry, but the results of ANY double encode are not worth watching, if you see the results on large TVs or on a HDTV.
There's too much detail lost :!:
Quote:


kwag, i can send you one of my samples on vcd.
Seeing is believing.
That's why i said view the results on your TV.


abc
No need. Already know the results ;)
Remember the basic rule: ANY third generation encode, will always be worse than the original.
There's just no way around it :!:

-kwag

ong001 09-20-2003 12:43 PM

Hello kwag,

Thanks for your patience to read my posts.

When i said "i hope this post will lead kvcd to another new height. "
I was thinking this way:
if k2vcd can produce standard vcd with much better video quality
then maybe someone can come out with better ideas and push the
quality of kvcd to even higher standards.

k2vcd is for standard vcd, so it can't be as good as kvcd because kvcd
can use higher resolutions. But if more people look at it and help,
maybe it can be closer to kvcd.

At "http://www.kvcd.net/dvd-models" you say kvcdx3
would give 75 minutes "Full Screen" on a 80 minute CD-R.
As k2vcd will give 80 minutes on a 80 minute CD-R, with this constraint,
would k2vcd be able to get closer?

I think some time ago there was a long discussion on
"double encoding" and want to look at it, so i do a search
and found that MrTibs is doing something similar with good results.

By the way if you don't mind, where is that long discussion on "double encoding"?

(my typing is very slow, so i'll explain more in my next post)

abc

kwag 09-20-2003 02:49 PM

Quote:

Originally Posted by abc
At "http://www.kvcd.net/dvd-models" you say kvcdx3
would give 75 minutes "Full Screen" on a 80 minute CD-R.
As k2vcd will give 80 minutes on a 80 minute CD-R, with this constraint,
would k2vcd be able to get closer?

That was a long time ago, before using motion adaptive scripts. We now can get ~2 hours at 528x480 on a single CD-R, with quality far above any VCD :)
Quote:


I think some time ago there was a long discussion on
"double encoding" and want to look at it, so i do a search
and found that MrTibs is doing something similar with good results.

By the way if you don't mind, where is that long discussion on "double encoding"?

It's on the AviSynth forum. I saw your post there, so I guess you already found it ;)

-kwag

MrTibs 10-29-2003 09:32 PM

Well, not to go against Kwag here but I think the double encode also has one other issue that may be improving the compression: motion estimation. The theory is the motion estimation is acting as a motion adaptive temporal smoother. Although it is true that overall a third encode will be worse, I think that on very noisy sources, there may be an advantage to the double encode.

I won't even suggest that I can improve on Kwag's DVD based encodes but I'm still trying to improve my noisy cable captures. I have some testing to do still, so we'll see but I'm still at it. (I'm still thinking of a motion estimation temporal filter.....)


All times are GMT -5. The time now is 05:28 AM  —  vBulletin © Jelsoft Enterprises Ltd

Site design, images and content © 2002-2024 The Digital FAQ, www.digitalFAQ.com
Forum Software by vBulletin · Copyright © 2024 Jelsoft Enterprises Ltd.