digitalFAQ.com Forums [Archives]

digitalFAQ.com Forums [Archives] (http://www.digitalfaq.com/archives/)
-   Video Encoding and Conversion (http://www.digitalfaq.com/archives/encode/)
-   -   DVD2ONE vs KDVD? (http://www.digitalfaq.com/archives/encode/3270-dvd2one-vs-kdvd.html)

kwag 04-25-2003 04:30 PM

Quote:

Originally Posted by nicksteel
Thanks! That fixed my screen size problem.

Bye the way, what value do I use for DVD+ 4.7GB to establish sample size? For example, I use 800 - MP2 size for SVCD, what do I use for DVD, 4700 - MP2 size or 4380 - MP2 size?

Use (4.3GB - MP2 size) as a safety factor ;)

-kwag

nicksteel 04-27-2003 12:06 PM

Now trying an anime, having prediction problems.
 
I'm using KDVD DVD full to create an R+ compliant DVD. I am using the following script and setting GOP in TEMPGEnc to 15.

16:9, DVD2AVI ForceFilm

LegalClip()
GripCrop(720, 480, dest_anamorphic=true)
GripSize(resizer="BicubicResize")
STMedianFilter(10, 50, 0, 0, 10, 50)
unfilter(50,50)
temporalsmoother(1,2)
mergechroma(blur(1.58 ))
mergeluma(blur(0.3))
GripBorders()
LetterBox(16, 16, 16, 16)
LegalClip()

My problem is predicting final size of the M2V file. The d2v file in FitCD has 176K frames, runs 122 minutes and Headac3he generates a MP2 file of 200MB.

When I use the prediction formula:

176,000 / 122 / 15 = 96.17
(4,300-200) / 96.17 = 42.633 sample size

Encode sample @ 89 CQ = 42,551

Full encode @ 89 CQ = 2,530,000,000, not approx 4,100,000,000.

Obviously I'm not predicting properly.

Also, is there any quality gain in using Highest Quality (very slow) in TMPGEnc?

kwag 04-27-2003 08:54 PM

Re: Now trying an anime, having prediction problems.
 
Quote:

Originally Posted by nicksteel

Encode sample @ 89 CQ = 42,551

Full encode @ 89 CQ = 2,530,000,000, not approx 4,100,000,000.

Obviously I'm not predicting properly.

The problem is that for that particular movie, you can even raise the CQ to 100, and you will still be under prediction :D. That's KVCD's matrix working in all it's glory :!:
You just won't get any better quality, unless you raise the MAX bit rate, and I doubt you'll see a difference.
Quote:


Also, is there any quality gain in using Highest Quality (very slow) in TMPGEnc?
No. Just leave it on "High Quality".

-kwag

vhelp 04-27-2003 10:06 PM

@Kwag..

>> That's KVCD's matrix working in all it's glory

I thought it was refered to as: Q.Matrix or -Notch :?:

Yes, Kwag.. I agree.. it's a great template !! Bar None !!

The Matrix is Bar None!!

-vhelp

kwag 04-27-2003 10:43 PM

Quote:

Originally Posted by vhelp

The Matrix is Bar None!!

-vhelp

Thanks vhelp :D
It took a lot of days and nights to get it to what it is today 8)

-kwag

vhelp 04-27-2003 11:04 PM

@Kwag..

well, from one person who's ben there and done it, ..those late nights
and fun trial-n-error encodes.. I know where you're comming from he he..

I'm working w/ it in my VHS projects and I do love how they come out. I
only wish I didn't suffer from the Noise when using an Analog cap card. I
REALLY like the new one I have, and passed all the tests I threw at it, xcept
one. Anyways..

I just got the WS VHS version of Contact (I have the FS VHS, and the
DVD) but this is a hobby of mine (and yours) and as such, having a bunch
of (owned and purchased) copies and doing what we want w/ it is our pleasure,
and challenge.
Anyways.. also got The 6th Day that I wanted to see. I'll try and do
a quick capture and encode and author to CDR this week, using the my
favorite, KDVD.352x480.D1 template of yours !!

Thanks again for your hard work. It's MUCH appreciated.. oh, Bar none!!
hehe..

Later pal.
-vhelp

kwag 04-27-2003 11:11 PM

Hey vhelp,

You might want to play with the STDMedianFilter parameters, specially for your VHS captures :idea: If you peek the spatial values (not the temporal ones), you should get a very good noise/spike reduction from your (VHS) material :!:

-kwag

vhelp 04-27-2003 11:41 PM

@Kwag..

lucky for you (ME) their was an older version as well. I'm currently trying
it out now, but I didn't notice anything. I'm using the search here, to see
what values were used. I do read these forums here, and remembered
seeing many of your testings where you include your suggestions. So, I'll
give it a go now. Thanks.

Well, I did an author to CDRW disk, even though you can see the noise, its
only sligte, on my 13" TV, but I must say, that coupled w/ your KDVD
template, it's great looking none-the-less. So far, I love the quality that the
test CDRW is giving me. 8)

Now, let me see what I can do to help the noise out just a little.
-vhelp

vhelp 04-27-2003 11:49 PM

pfew!!

I really have to WUSIWYG picSwitch (for .AVI captues) cause it's getting
harder for me to notice anything w/ two pics in vdub, and I'm growing
tired of copying to and saving each pic to scan in picSwitch. But, heck, at
least it's better than juding while using vdub's time-line window !!

I found some settings to try out.. saw Jorel's post here:
http://www.kvcd.net/forum/viewtopic....9d69094a86725f
but I still didn't notice anything. I'm loading up picSwitch now..

-vhelp

nicksteel 04-28-2003 08:06 AM

Re: Now trying an anime, having prediction problems.
 
Quote:

Originally Posted by kwag
Quote:

Originally Posted by nicksteel

Encode sample @ 89 CQ = 42,551

Full encode @ 89 CQ = 2,530,000,000, not approx 4,100,000,000.

Obviously I'm not predicting properly.

The problem is that for that particular movie, you can even raise the CQ to 100, and you will still be under prediction :D. That's KVCD's matrix working in all it's glory :!:
You just won't get any better quality, unless you raise the MAX bit rate, and I doubt you'll see a difference.
Quote:


Also, is there any quality gain in using Highest Quality (very slow) in TMPGEnc?
No. Just leave it on "High Quality".

-kwag

:?: I assume my prediction formula (using 15 rather than 24) is correct for anime and that the GOP change for KDVD full to 15 is correct for anime?

Also,

:?: When I use IFOEdit to generate VOB's, I am using Authoring from the toolbar. Do I need to do anything else prior to creating image file? I seem to remember seeing a post somewhere that says you have to use IFOUpdate to correct a 3:2 pulldown error in IFOEdit created VOB files.

:?: IFOEdit allows you to save a chapters text file. Is there any way to directly use this file with the 23fps created by DVD2AVI from the original 29fps DVD? I assume the location pointers are different due to the change in fps.

:D I really appreciate all the advice.


All times are GMT -5. The time now is 04:50 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.