Im encoding American History X right ( ToK (MPEG-1 704x576 25fps CQ 90)) now, on 2 CDs and Its gonna take my at least 10 hours!!1


What do u mean with "clocking it" ???

Plautzer

Ask Jorel about the clocking deal he will explain it better :P

i bolive 704x576 pal would make a differencs in time , movie just
finish now I heard the ding by Tmpeg... hehe i was encoding at 528x480
that would be much faster than 704x576. :wink:

i am having the same problem.. with tok it took 3 tries to get my cq but it took 12 minutes what the heck is that. what do i need to do to clock it?? ( and what exactly is clocking it) thanks ??

They are talking about "Overclocking". Not clocking. It's when you increase the CPU clock on your machine over it's designed limit. That's why it's called OverClocking.

-kwag

He's talking about overclocking his processor to get more speed. That's a whole nother can of worms. www.anandtech.com

Aight ill gonna ask jorel,
but I have another problem, which i never had when I was using the old avisynth!!
During the encoding, when I go off the internet.. my computer reboots!!
That piss me off, cuz soemtime its right before the ende!!

Did anybody have a problem like that??

Thx guys

Plautzer

Right Kwag, i dont have the board nor the memory speed to over joy :hihi: this thing

Based on my results so far with today's script I feel like I just got a CPU upgrade for free :) :)

Based on your results so far, I feel like I need to pay for a CPU upgrade :!: :mrgreen:

-kwag

Sound like a setting in windows that when the computer is about to crash it will reboot instead. :idea:

I apologize for this "divide by zero" error, I should have check this.

But this point me to an other problem that the solution found by Kwag introduces :

round(YDifferenceToNext()+1) always gives values >= 1

So round(1/nf) and round(3/nf) will produce only TWO values : 0 or 1.

I'm sorry, I didn't saw that before :BangHead:

I think that the better solution (for speed and accuracy) is :

(until Kwag give us its advice on temporalcleaner ;-))

Latest changes = less compression?
 

I just made a test using the latest changes to the optimal script. I have a 21 second clip for testing. (It is the opening "couch" sequence from an episode of The Simpsons).

The end result is going to 352x240 but using the KVCD3 template. I only change the resolution.

The latest script as of Sunday produced a file size of 2,206KB. I ran the exact same project but with the changed script as of today and the file size was 2,384KB.

I did not notice an perceptable quality difference in VirtualDub or by playing them on my TV. (Using Nero and burning as VCD).

I understand the script is really optimized for DVD movies not animation. I just though the file size change was interesting.


Grantman

Re: Latest changes = less compression?
 

Yes, that a logical result of the problem I mentioned just above : with my old suggestion, temporalcleaner works with only two values : 0 and 1. In other words : does'nt have a great impact.

Try with the ultimate modifications I did in my post of 12:03 am, and tell us.

Script corrected. I guess I wasn't that tired after all earlier this morning, but I wasn't too awake :)
First I saw the error, then I thought It was me. As a rule, I shouldn't do math early in the morning :lol:

Thanks Phil ;)
-kwag

@Dialhot,
I can't get it to work. Could you please look at my script and tell me what's wrong?

When I load this script in ToK and flip to the video tab it says:
And I have no clue to what that means. 8O

The function fmin() is expecting only integer values. You are sending floating point values in the temporalcleaner function :!:
You need to "round" the values in the function call. Just like in the "unfilter" line.

-kwag

8O Say what? Could you please show how the line is supposed to look? As I said earlier: I'm no programmer... :wink:

It should look like this:

-kwag

Thanks a lot kwag! I really wanted to try the MAS with TemporalCleaner but I couldn't figure out what was wrong. You're my hero! The Man, The Myth, The Legend - KWAG! :D

editing:

Phil post this lines and works fine,seems with more clean colors:

ScriptClip("nf = YDifferenceToNext()"+chr(13)+ "nf >= 2 ? \
unfilter( -(fmin(round(nf)*2, 100)), -(fmin(round(nf)*2, 100)) ) : \
TemporalSoften(4, round(1/nf) , round(3/nf) ,0, 2) ")


ps:
i was posted only part of this lines,now are complete!
:)