New CQmatic user
 

Ive been making vcds for a while. done tons of reading. Now Ive gotten into the kvcd thing which rocks by the way. I noticed sometimes my movies need to be bigger so I use the whole cd. Im using this cq matic and think Im doing it right. Read a guide. Anyway how long is this prediction cycle supposed to take. Its now been 3.5 hours. Says its on the 3rd cycle. The CQ started at 60 the 2nd cycle is 31 now its says its using 16.5. Is this wuts supposed to happen. Is it supposed to take this long? Any suggestions?

Welcome pySHtr,

No way 8O
Unless you have a slow machine, Pentiun III @500Mhz or so :!:
That's just too long. What processor do you have :?:
I have a P4 @1.6Ghz, and each prediction cycle using the Motion Adaptive script, takes between 4 to 6 minutes, depending on the length of the movie.

-kwag

Gots me a athlonxp 1800. It usually only takes a couple hours or so to encode a full movie. Now this cqmatic has been on all nite and on cycle 5. It says "using cq of 5.63".It started at 60. Iknow its not working for me. But I dont have a clue why. Its a 90 minute movie. Im using a 528x 480 mpeg template from kvcd. I ripped the audio first with virtual dub. Set up evrything in tmpge. saved it as text Used a calculator for average bit rate. Pluged it all in cqmatic and wanted to try again. Ive tried it a few times in the past few days with same prob. I must be missing something.

Thats the guide i went off of. But skipped the DVD2AVI cause my movies are already divx/xvid. And if the audio is not ac3 i didnt use headac3he. Am i supposed to follow this guide completely? Or does this stuff not work with divx. Now I have to learn about scripts too? Im lost

Hi pySHtr,

There's you problem already.. your source type!! ..divX

Its ben ruled #1 pain in the aX. It's very slow, due to the KEY and scan
offset (seek times) and all.

That's one.

The other, could be other issues within your script or *.TPR file.

* If you're using a script, to rule out the slowness, eliminate the .AVS
...script and use directly ie, *.d2v source into TMPG.

* If not using .AVS script, have a better look inside your *.TPR file, by
...examining your TMPG settings.. ie, framerate, Interlace, IVTC, 3:2 pulldown
...etc etc.

I had an issue very similar to yours, and took the whole night to do, but
my source was not the "X" source, rather it was a DVD rip (via VFAPI)
and it still took forever. But it was just that one movie, and I had done
something or overlooked something that caused it. Unfortunately, I just
can't seem to remember what it was :( but, I think you might be suffering
it also.. but your source is "X". Also, it's not recommended to use "X"
sources, nor is it chearfully accepted because of the great pain that it is
for seeking :wink:

I'm too lazy to search right now, cause I've got crap-full of fun projects
and things going on and I'm in the middle of a search and destroy mission
for a project I'm working on in Delphi. So, if you want, try a search on
my past issues :)

Have a great day all,
-vhelp

K vhelp. I will look. So if divx/xvid is the problem, is there a work around for that or m I hooped. Ive been having this same prob over and over. Like with another movie I wanted on two cds. It to was divx. So I gave up on the cqmatic and just encoded it 3 times till I got the file size I wanted. I guess I got alot more reding to do now. :evil:

Hi pySHtr,

The best solution, forget about DivX :D

-kwag

Now ive just started to read about using tok for divx coversions. Is that better ya think. Its hard to forget about divx :(

Hi pySHtr,

no, there's not much you can do, if your source type is "X", but to add, it
is not chearfully accepted if its a source you D/L'ed off the net. So, please
do bare that in mind when you have issues and things, and need help..,
responses to D/L'ed "X" sources will most likely go unanswered here :)

Good luck in your endeavors,
-vhelp

Re: New CQmatic user
 

When you are making your project in TMPGEnc, make sure you select video only, and delete the audiosource field, if there is anything in your Audiosource CQMatic will behave as you describe...

I seem to be having the same problems trying to encode an AVI file as pySHtr.

Instead of 'sampling' the movie during the prediction stage, CQMatic decides to encode the whole movie, then adjust's the CQ up or down accordingly and then encodes the whole movie again, and so on until the correct CQ is found.
Where am I going wrong? :banghead:

I am using CQMatic version 1.2 and the optimal MA script, and TMPGEnc Plus 2.53

For DivX sources, I think it is better to use manual prediction, since CQMatic needs a lot of seeking and seeking in MPEG4 content takes a lot of time... Manual prediction will also be slower with divx sources comparing to DVD sources, but not so slowwwww as you described.

[]'s
Vmesquita

I find its more than common to take upto and over 1 to 1.5 hours for my predictions regardless :( its kewl but takes so long, im running a 1.8 gig athlon 1 gig of ram, with over 200 gigs hard drive space at disposal. I dont think my pc is at fault.

or are these times normal? or an i (as usual) doing something wrong ?

edit:all sources are dvd

Hi TLTw,

How long does the full encode take :?:
Because prediction can take anywhere from 5 minutes to ~20% of a full encode :!:
Depends on what the final "range" CQ falls in.

-kwag

The work im doing at the moment takes around 5 hours for 1 hour of video, im using the faerydust filter though, however in the past i have had similar lengths of time for prediction without this filter or similar time eating filters, on full length movies.

I have to admit i havent kept or remember exact times ill do a test encode and get the figures back to you. :oops:

90% of my encodes are anime, could that be part of the reason? :roll:
what do you mean exactly by CQ 'Range' ?

The linear curve of TMPEG. For example, if your final CQ will be below ~50 or above ~80, it will take way longer to predict, than if the CQ will fall between those numbers.

-kwag

Ok, ill get off my soap box now :oops:

I just did a prediction With faerydust in my script, and it only took 4 tries@ 43 mins total. CQ was 89 so i guess its just the luck of the draw when doing these things. Total Encode was 5 hours, so i guess ill just hope to get CQ above 80 from now on :lol:

cheers for the help :drink: