QuEnc 0.72 [Archive] - Page 3

Rockas

12th August 2006, 12:51

hmm the page has had over 150 visitors since this new release, not bad since its not announced on the main doom9 page or other places(videohelp etc). surprised i havent got any more comments about if its ok etc yet though. guess it must be working ok since nobody is jumping up and down complaining.

Sorry mate!
I've been extremely busy on the last weeks (months?)... I really didn't had the chance to test it yet... but I will... as soon as my vacations start :)
But... knowing you... I'm pretty sure that it can't be worst than the last release.

keep it UP

A_T

12th August 2006, 21:44

I always use QuEnc to re-encode my dvds. The results are always of terrific quality but the final output size can vary greatly. No too much of a problem when it's undersized but a pain in the backside if it's even a lttle too large.

But thanks anyway. :D

dragongodz

13th August 2006, 01:46

I'm pretty sure that it can't be worst than the last release.
umm thanks, i think .... that was a compliment wasnt it ? ;)

Yeah, when something is broken, everybody jumps at your throat, but not much feedback if everything just works!
haha ye the wonderful world of free software. :)

Rockas

13th August 2006, 11:14

umm thanks, i think .... that was a compliment wasnt it ?
Of course!
You know that I'm a big fan of your devotion to the "cause"! :)

haha ye the wonderful world of free software.
See what I mean?

People that develop freeware applications only have two ways!

1st - They're completly out of their minds

2nd - They're great people that can't support the idea of not sharing knowledge and that everything worths money...

On this one... I'll let you choose LOL LOL LOL

ljpp

14th August 2006, 20:32

Okay, QuEnc is about to get a new home - right next to HC Encoder here at http://quenc.bitburners.com/ . The page(s) are still very rough around the edges, as they were constructed in extreme hurry.

However, the binaries are there waiting for downloaders.

dragongodz

15th August 2006, 06:06

On this one... I'll let you choose LOL LOL LOL
i am sure i would get votes for both. ;)

updated the first post with bitburners page link. thanks to ljpp for providing hosting.

easyfab

15th August 2006, 20:38

Thanks for this tool.

For info is it multithreads optimized? or possible ?
Because it's based on ffmpeg and i see in ffmpeg a "-threads 2" option.

I tried with setmtmode(2) in avs script and it give me only ~2fps faster 24fps->26fps.

Boulder

15th August 2006, 23:16

easyfab

16th August 2006, 10:59

Thanks :)

dragongodz

17th August 2006, 08:50

For info is it multithreads optimized?
currently QuEnc is not multi-threaded.

note i have now uploaded the ffmpeg source used by QuEnc on the bitburners page. i have used a different archiver which compresses even better than 7zip. if you have any problems then try the version on my googlepages but also please report here aswell.
note the one on bitburners is meant to be only for windows 2000/xp.

Nic

17th August 2006, 10:30

It's a sad state of affairs when I'm far more interested in that KGB Archiver than I am the new version of QuEnc :(
Thanks for keeping QuEnc alive Dgz :)

-Nic

EDIT: Ahh, KGB is based on PAQ6. Nothing new & interesting there then :(

dragongodz

18th August 2006, 07:06

Thanks for keeping QuEnc alive Dgz :)
not a prob and nice to hear from you Nic.

Ahh, KGB is based on PAQ6. Nothing new & interesting there then :(
well its open source and compressed the ffmpeg source files down more than 7zip(about 600k smaller using medium setting). so less for me to upload on a slow upload speed account.
of course it is as slow as all hell archiving and even decompressing which is why i only used it on the ffmpeg source.

kyuuketsuki

18th August 2006, 15:34

dragongodz,

I was looking for your email but can't find it anywhere. Anyway, just wanted to let you know that I belong to a community that has been using QuEnc for a long time. I personally think it is the best free encoder out there. I would not be exaggerating to say that thousands of us are using it. Since we are also using dvd2svcd and QuEnc became incompatible with 0.70, we've been anxiously awaiting the next version.

Thanks for the fixes and upgrades.

Boulder

18th August 2006, 15:40

Doesn't the latest version work? At least manolito reported in the DVD2SVCD section here that it does work (he had created a batch file that was a workaround for the problem).

kyuuketsuki

18th August 2006, 20:58

I haven't actually tried it yet since I am still at work, but I will try tonight. I assume that it does work though since it was reported by manolito.

Sorry for any confusion. When I said "next" version, I meant next after 0.70. I should have said "this version".

MrC

29th August 2006, 16:40

It seems that 0.71 version brings back some file undersizing problems (solved in previous versions) when 2pass encoding.

Feeded with the following AVS script:

Video = DirectShowSource("clip.avi")
Audio = DirectShowSource("clip.avi")
AudioDub(Video, Audio)
ConvertToYV12
LanczosResize(720,576,0,0.5)

(clip.avi is a 640x480 1114kbps 25fps DivX5 + 128kbps CBR 2ch MP3)

and with following settings:

VBR on, HQ on, 2pass on, Extreme off, GOP off, Matrix off, 4700 avg bitrate, 6700 max bitrate, DC 10, 4:3, Audio AC3 192kbps, DVD mux profile

Theroretical -> 1,580,204,032 byte
QuEnc 0.70 -> 1,489,737,728 byte
QuEnc 0.71 -> 940,064,768 byte

:confused:

Bye
________
head shop (http://headshop.net)

GeorgeT

2nd September 2006, 21:57

Dragongodz
QuEnc0.71commandline refuses my mpeg1 option. Would you please recover it? Thanks.

Mr_Odwin

2nd September 2006, 22:32

Dragongodz
QuEnc0.71commandline refuses my mpeg1 option. Would you please recover it? Thanks.

The command line arguments have changed. You need to use -mpeg1mux instead. Run quenc /? to see all the commands.

GeorgeT

3rd September 2006, 07:20

The command line arguments have changed. You need to use -mpeg1mux instead. Run quenc /? to see all the commands.:) :)
Mr_Odwin, Thanks for your reply.
Because QuEnc0.61 acccept mpeg1 but can't accept mpeg1mux, so I never try mpeg1mux on QuEnc0.71.
Thank yoy very much. Mr_Odwin, you did a great help!:D :D

tamahome

3rd September 2006, 14:44

which is the command line to use a matrix custom?
-cmatrix and?
an example of command please :)

Mug Funky

3rd September 2006, 16:00

-cmatrix "drive:\path\matrix.xcm"

or -nocmatrix (for MPEG-standard)
or -qlb

dragongodz

4th September 2006, 01:21

QuEnc0.71commandline refuses my mpeg1 option
which is the command line to use a matrix custom?
good thing i didnt waste my time writing text files such as QuEnc_changelog.txt and QuEnc_settings.txt as i have a feeling nobody would read them if i did. :devil:

GeorgeT

5th September 2006, 14:32

Hi! Dragongodz, Ask for your help! Please!
I got twinkle-like images; when I used QuEnc(any version) convert the enclosed sample DVD clip to VCD. May I ask how to setting or any methods to solve this problem. (In mpeg-1, cbr, 1150kbps, and maxbitrate 1150)
If somebody can help me to handle this, I am appreciating for your help too.
Sample DVD clip. Name: DVD.7z Size: 9MB (Choose one of two)
http://www.sendspace.com/file/g7q2ng
http://files.filefront.com/DVD7z/;5466148;;/fileinfo.html
#-------------------------------------------------------------------------
LoadPlugin("C:\Program Files\AvPack\AviSynth 2.5\plugins/DGDecode.dll")
MPEG2Source("DVD.d2v")
Sharpen(0.2)
LanczosResize(352,240)
ConvertToYV12()
#-------------------------------------------------------------------------

DDogg

6th September 2006, 01:30

dragongodz, as you might remember, back in the dark-ages, I had a habit of deriving the bitrate from a 1% CCE OPV encoded at an empirical Q (say 28) using (((Sample_Size_In_Bytes*(100/Sample_Size_Percentage))*8)/1000)/(Total_Frames/Frame_Rate)

This DABR gave me some fingerprint data on the complexity of the source and allowed some general projections of the bitrate requirements needed to contain the source complexity.

I would like to do something similar with QuEnc0.71 for experimentation purposes, but I am having a little trouble with the docs. Could you recommend your command line syntax that would be appropriate?

I do realize CCE's opv mode is unique to their proprietary product, but generally I am looking for a constant quant type of thing that will allow the source complexity to control the bitrate within some low/high legal limits.

I only need it to create the 1% sample, not the full encode. I'll use 2 pass for that.

I've been away from this stuff a long time, so I hope you can bear with my post being less than precisely worded. I also understand my theories linked in my sig, if you could call them that, were not universally accepted. However, they worked for me and I would like to see if they could also do so with this rather extraordinary thing called Quenc.

Rhetorically speaking, who would have ever thought a free mpeg encoder would rival multi-thousand dollar software? Real nice job you guys have done.

Mug Funky

6th September 2006, 04:00

hmm. that would work, but:

- i'm not sure if quenc supports floating point constant quant, so you'd have to round upward to avoid oversizing (FYI, enter any bitrate lower than 31 and it'll go into constant quant mode).

- max bitrate isn't enforced in constant quant mode, meaning you're likely to get incredibly high peak bitrates unless your CQ value is really high.

manolito

6th September 2006, 14:01

Hi DDogg,

long time no see...
max bitrate isn't enforced in constant quant mode, meaning you're likely to get incredibly high peak bitrates unless your CQ value is really high.
I believe that this is the main obstacle. HC does have a CQ mode which does enforce the max bitrate setting, but QuEnc does not. You're asking for major trouble in the muxing / authoring stage.

I suggest to check out QuEnc's one pass VBR mode. I am not sure how this mode is implemented, Nic was not very clear about it. Probably start with a fixed quantizer and at certain time intervals analyze file size and adjust the quantizer accordingly? All I know that this mode always reaches the target file size, and the tweaks DGZ made in the current version make for a surprisingly good rate control. So far version 0.71 never gave me an encode that Muxman would refuse.

Cheers
manolito

DDogg

6th September 2006, 14:12

(FYI, enter any bitrate lower than 31 and it'll go into constant quant mode). That's what I needed, thanks very much.- max bitrate isn't enforced in constant quant mode, meaning you're likely to get incredibly high peak bitrates unless your CQ value is really high. Shouldn't be a problem with what I am doing as it is used only as a general indicator. Say you have 4 encodes to put on one dvd - The simple way is to calc a group bitrate for all the sources to hit the size target. But, for conversation purposes, say 2 of the 4 sources are Matrix and SPR. SPR requires 3 times the BR of Matrix. That is going to be a problem. If we can do the sample and calc the DABR of the samples we then have a number to indicate how to better allocate bitrate to SPR and steal some from Matrix. We can still use a bitrate based final encoding method to keep the size under control, but we also have a way to generally control the viewing quality of the disparate sources.

/add: clarification - each individual sources might use a different bitrate, but the average of those individual bitrates used will still be the same as the average that was calculated originally to fit all 4 sources on the DVD. The total bitrate available is simply better distributed to where it is needed.

Hopefully this will work with QUenc as it does with CCE. I'll play around with it and see.

Mug Funky, thanks again for your reply. I've learned a lot from your posts in the past.

DDogg

6th September 2006, 14:30

Hi DDogg,

long time no see...

... HC does have a CQ mode which does enforce the max bitrate setting, but QuEnc does not. You're asking for major trouble in the muxing / authoring stage.
manolito ...Hey buddy! Yep, life has a way of getting in front of my hobbies :)

Actually I want the QUenc encoding method that most triggers a higher bitrate when it encounters a higher source complexity, even if the rates are crazy. This gets me a sample DABR that tends to mimic the source complexity. I was actually worried the CQ mode might have been throttled which would mess me up. Any suggestions? These non bitrate based methods don't seem to be documented too heavily. I suppose it is a form of self protection for the authors :)

manolito

6th September 2006, 16:05

@DDogg
Hey, I should have reread the link in your signature before answering...

CQ mode in QuEnc really is hardly documented at all. Probably because you cannot really use it for making compliant DVDs. But for your method it seems to be ideal.

I normally use a method very similar to yours when I am not sure if I will be able to get a decent quality for a very long encode. The actual conversion is done with QuEnc one pass VBR, but before this I use D2SRoBa with CCE 2.70.02.10 trial to find out which Q value I would get. This CCE version has no logo embedded in the output, so the result should be comparable. If my Q is not satisfactory, then I can do something about it, and if it is, I go ahead encoding with QuEnc. Certainly not scientifically correct (because I assume that CCE and QuEnc have similar characteristics), but so far I always had success with this method.

Cheers
manolito

MrC

11th September 2006, 12:37

It seems that 0.71 version brings back some file undersizing problems (solved in previous versions) when 2pass encoding.

Feeded with the following AVS script:

Video = DirectShowSource("clip.avi")
Audio = DirectShowSource("clip.avi")
AudioDub(Video, Audio)
ConvertToYV12
LanczosResize(720,576,0,0.5)

(clip.avi is a 640x480 1114kbps 25fps DivX5 + 128kbps CBR 2ch MP3)

and with following settings:

VBR on, HQ on, 2pass on, Extreme off, GOP off, Matrix off, 4700 avg bitrate, 6700 max bitrate, DC 10, 4:3, Audio AC3 192kbps, DVD mux profile

Theroretical -> 1,580,204,032 byte
QuEnc 0.70 -> 1,489,737,728 byte
QuEnc 0.71 -> 940,064,768 byte

Am I really the one having this kind of problem? Or it is well known and will be fixed with next release?

:confused: :thanks:

Bye
________
cheap volcano vaporizer (http://vaporizers.net/volcano-vaporizer)

manolito

11th September 2006, 12:50

Am I really the one having this kind of problem? Or it is well known and will be fixed with next release?

:confused: :thanks:

Bye
No, you are not the only one. Take a look here:
http://forum.doom9.org/showthread.php?p=864296#post864296
And in your case the reason for undersizing could be the same as in the other thread. Your source was already heavily compressed, so QuEnc really had no use for the much higher bitrate you specified. Suggestion: look at your quantizer values. If the average quant is 1 (or close to 1) then you already have the maximum quality QuEnc can deliver for this source with this matrix. The only way to avoid undersizing would be padding, but this is not desirable.

Cheers
manolito

MrC

11th September 2006, 14:59

Thanks Manolito for your clear answer. :)

One more thing I cannot figure out: why QuEnc 0.70 was closer to target in filesizing than QuEnc 0.71? There is really a big difference! 0.70 did padding?

:confused:

Bye
________
hemp marijuana (http://marijuanahemp.com)

manolito

11th September 2006, 16:07

changes from 0.70 to 0.71
-------------------------
.small 2 pass ratecontrol tweaks
This is something you will have to ask DGZ about. I mostly use QuEnc in 1-pass VBR mode (quality is amazing, on my standard CRT TV I could never see a difference between 1-pass and 2-pass), so I'm not affected by these ratecontrol tweaks.

Cheers
manolito

Mr_Odwin

13th September 2006, 14:24

I'm not sure if this is a bug - to me it is.

This:

http://img241.imageshack.us/img241/930/nopullqt8.jpg

and this:

http://img145.imageshack.us/img145/5958/pullqe1.jpg

result in different files. (All options are the same except that pulldown is greyed-out selected in the lower image.)

As a consequence of the above, I think either pulldown should be turned off when interlaced encoding isn't selected, or provide a command line option for it, -nopulldown perhaps?
I noticed this because Muxman was saying that the files I was inputting weren't dvd compliant. (GOP too large, 45 instead of 36.)

dragongodz

19th September 2006, 06:42

Am I really the one having this kind of problem?
it only happens that bad some times. you should notice though that some bad frames/low quality issues were fixed even at the lower average rate. yes i will look at this when i get time to readjust the bitrate distribution.

I'm not sure if this is a bug
yes a very small one. it was already reported so i do plan on fixing it.

P.S.
just so people know, i have been trying to update LAVC + LAME + FAAC. LAVC and LAME both compiling fine and LAME being found etc. however FAAC doesnt want to compile properly even after i updated all of cygwin (158mb download). comes close but still no go. so i may give updating a miss for now because it is driving me batty.

Mntneer

5th October 2006, 17:43

Is there a reason why QuEnc would stop encoding after the output file reaches 1GB? I'm trying to go from HDTV to DVD, following the instructions on the "HDTV to DVD" guide, http://eswrite.50megs.com/hdtv2dvd.html, and have tried running the encode twice, but both times the encoding stops part way through.

Could it be a problem with the source MPG?

Any suggestions on going from .TS to DVD compliant .MPG?

Mug Funky

6th October 2006, 08:13

there's no size-limit reason, so it might just be a crash in your file.

try open the part-finished file in virtualdub or similar, seek to the last frame, and then open up the script and seek to just before the same spot. advance a few frames and see if the program crashes... (do not hit play - virtualdub might disappear and leave an overlay window stuck in the screen with no way to close it... happens to me a fair bit and i have to reboot to fix).

Affar

6th October 2006, 08:25

Hi

I have little problem with this command line:

quenc.exe -i "C:\avs001.avs" -o "C:\sample.mpg" -b 1600 -maxbitrate 2500 -dc 10
-priority 4 -1 -mpeg2mux -nohq -novbr -noscene -notrell -nocgop -nointerlaced -noextreme
-gopsize 15 -maxbframes 2 -nocmatrix -aspectratio 4:3 -mpeg2mux separatepases -audiocodec mp2 -audiobitrate 160
-auto -close

If i use -close setting, quenc only compress video.

Thanks :)

Mr_Odwin

6th October 2006, 09:29

Looks like you have separatepases instead of separatepasses.

Mntneer

6th October 2006, 14:14

GeorgeT

8th October 2006, 20:55

maxfu

26th October 2006, 11:16

Any batch processing GUI tool for QuEnc.71?

Affar

3rd November 2006, 03:28

The -mpeg2mux appearance twice.

Yes, but if i try this line, i have same problem

quenc.exe -i "E:\avs.avs" -o "E:\1.mpg" -b 1600 -maxbitrate 2500 -dc 10 -priority 4 -1 -hq -vbr -scene -notrell
-nocgop -nointerlaced -noextreme -gopsize 15 -maxbframes 2 -nocmatrix -aspectratio 4:3 -mpeg2mux separate
-audiocodec mp2 -audiobitrate 160 -auto -close

Only compress video. But if i exclude -close option, audio and video are compressed ok.

Thanks

Mr_Odwin

3rd November 2006, 11:50

You have separate instead of separatepasses.

Affar

4th November 2006, 03:01

You have separate instead of separatepasses.

Command Line options in Quenc 0.71

-mpeg2mux *** = MPEG-2 Mux Profile( dvd, svcd, novideo, noaudio ,separate)

Really problem is in -close. without this option, the compression of audio and video is correct.

Thanks

dragongodz

4th November 2006, 04:26

problem is in -close. without this option, the compression of audio and video is correct.
confirmed. will look at it when i get time. had absolutly no time recently though so can not say when.

thanks for finding it though. :)

Affar

4th November 2006, 19:52

confirmed. will look at it when i get time. had absolutly no time recently though so can not say when.

thanks for finding it though. :)

Thanks for your work with Quenc :)

maxfu

5th November 2006, 08:57

Any batch processing GUI tool for QuEnc.71?

Finally got one. Named KuuEss.

Heini011

11th November 2006, 00:02

Hi,

i have a undersize problem with QuEnc 0.71 too.

my .avs script: 720x576p, 25 fps for pal dvd. 328586 frames total

the following QuEnc job should deliver 7,20 GB mpeg2 video file, but the result was only 3,94 GB and dosn't look good.

hc encoder with the same bitrate and matrix hit the target size and finished with an average quant of 3,1.

D:\DVD-Tools\codecs\QuEnc\QuEnc.exe -i v.avs -o v.m2v -b 4708 -maxbitrate 9500 -dc 9 -priority 4 -2 -hq -vbr -scene -trell -nocgop -nointerlaced -extreme -gopsize 15 -maxbframes 2 -cmatrix D:\DVD-Tools\video_enc\matrix\heini-dvd.xcm -aspectratio 16:9 -mpeg2mux dvd

PS: chapter support (keyframe times via text-file import) would be great too...