Experiences with HC Encoder

[cmsoliveira]

No problem, shall create a new command which will disable the end code output, will PM you about it.

That's good news :)

Even so, i did run a few tests just to make sure that it wasn't RB fault.

This 2 tests were made using the rebuild disc:

First I demux the problematic movie and remux with muxman (with all audio and subtitles) -> stutter remained
Then I re-author the movie with dvd-shrink (movie only) -> when the movie changed chapter, the image freeze seconds later but the sound continues.

Using an original dvd:
Dvd-Rb with no-compression - movie played perfect
Dvd-rb with cce - movie played perfect
Dvd-rb with HC - movie stutter (very little) between chapters

The stutter was only detected on my sony, in another dvd player everything went fine.

So the problem could very well be the end sequence (i hope it is).

[rendez2k]

Glad to know this will be sorted for us Sony owners! Just out of interest, how many passes approx. in CCE are related to the Normal, Good and Best profiles in HC?

[onesoul]

Originally posted by rendez2k
Glad to know this will be sorted for us Sony owners! Just out of interest, how many passes approx. in CCE are related to the Normal, Good and Best profiles in HC? That's hard to say because the encoders apparently work in different ways. Someone said in another post that CCE delivers better quality at 1 pass vbr, well with some tests I have made I tend to agree on that or at least they are as good.
Providing 1 pass vbr analysis is used in order to find the right quantizer which then would deliver the most constant quality output (same Q throughout the movie). Ddogg defended OPV (1 pass vbr), and I was stubborn at some point to accept it but I believe it is the way to go with CCE.

But I still prefer HC :D

[edeus]

im tempted to test CCE versus HC again in the following (if someone doesnt say it has been done)

Normal/Grainy/Action/Cartoon/Anime/Animation video:
CCE 2, 3, 4 and 5 pass
vs
HC in normal, good and best modes

Why do scene rippers use CCE @ 5p if supposedly doesnt get better after 4p?

Basically I just do 3p or 4p and can do a DVD/4h which if it can equal HC's good or near its best mode - I can see it being a good tradeoff for quality vs time.

[jdobbs]

Try this test:

1. Buy CCE and encode a movie.
2. Download HC, encode a movie, and go out for a nice dinner.

They both cost about the same. :D


(Don't get me wrong. I LOVE CCE, and it is well worth the $58, but let's make it apples-to-apples). To be honest with you -- QuEnc is really close to these two encoders as well, and it is almost as fast as CCE.

[onesoul]

I didn't take in account other possibly factors when posting above. I believe that 1 pass vbr encode is sufficient when having enough bitrate. But if you have low bitrate (lets say around 3000kbps or lower) then you can take advantage of multipass and it's bias setting which should get higher the less bitrate you have available (bias 25 it's good compomise). (Not to mention the obvious size encoding precision but that wasn't the point anyway.)

And for multipass I wouldn't go above 2-3 passes (including vaf file creation), don't see any improvements with right settings chosen.

[Fishman0919]

I'm in the middle of doing a quality test with different encoders and I was trying to narrow down what ver of CCE to use (CCE SP 2.50, 2.66, 2.67, 2.70...). Well I started with CCE SP 2.70, I did a quick test of a few short clips at diff. bitrates (2000-2750)with OPV, 2 passes, 10 passes and (LOL) 100 passes to see if there was any Quality diff. between them (100 passes w/ CCE SP 2.70.02.00 never again, biggest waste of time). Well, with 3 of the same Mits 65" HD TV's sitting next to each other (I work at an Audio/Video retail store, makes it a little easier to do that) and with 3 of the same Denon DVD players playing the same clips at the same time from OVP, 2 passes and (hell why not) 100 passes, not one person who I asked could see any difference... well, with me knowing what was what...I couldn't tell. They all looked the same.

Hey, I did have a point to this at one time but for me 2 passes is great... hit my target size all the time and to me...it looks the same as 100 passes.

Edit: Experiences with HC Encoder....sorry, a little OT

[johnhamler1]

i have increased my tagetsize value a bit higher...a lot higher, and now I am at 250 mb under , when before it was 450 mb undersized.
so it is better now with HC.

with QUenc, i see artifacts, there are some parralel artifacts on everything straight, like the edges of houses.this is on my beamer, so u have to zoom to see .

Seem I have better result with HC, except it is still undersized.will do more projections in the futur.
HC is the best free encoder so far.I dont have the one at 58$, so I can not the difference, should I buy it???

[rendez2k]

Any news on the updated HC encoder for the Sony problem?! I'm egger to re-test! :D

[hank315]

Just implemented a command to skip the end code and fixed some other bugs.
Needs some testing... so please be patient :)

BTW, if you still have the m2v files, just use a hex editor and edit the m2v files, delete the last 4 bytes (00 00 01 B7), rebuild it again, that's all.

[rendez2k]

Arh, i don't have the m2v files, so I'll be patient and wait for the new version :D

[cmsoliveira]

BTW, if you still have the m2v files, just use a hex editor and edit the m2v files, delete the last 4 bytes (00 00 01 B7), rebuild it again, that's all.

That's what i did.

I took the rebuild dvd and run it thru rebuilder in "no compression" (3 click mode), i.e., prepare and encode.

Next download a little tool called ReStream and load all the segments that are a part of the movie (or all the segments that are giving you trouble - don't know if the last segment need to have the end sequence code, so i let it stay) and process them (don't forget to check "Remove Seq. End Codes")

After that you will have all the m2v files without the end sequence code.

Load again rebuilder, press rebuild and it´s all done.

I tried that and in the parts were the jitter was, is gone :)

I still need to see the whole movie but i'm sure that it's perfect.

I only this this because of the encoding time of HC is too much (i have a AMD Tbird 1.3 MHZ), but i continue to say,

Great encoder, hank315 :)

[lamster]

@jdobbs - selecting Help / Display Versions doesn't display the version of HC. (HCbatch_012a.exe contains an internal version of 1.0.)

[jdobbs]

Unfortunately all versions of HC show as being the same... (v1.0.0.0) -- the only way I could check would be with a filesize, and I wouldn't want to do that until after it stabilizes.

[johnhamler1]

watched king arthur last night on my beamer DLP.
what I can say is, in action scenes, it is less blurry than dvd schrink, and the movie is watchable in kung fu sequences.

HC is a great tool. the guy selling the 58$(CEE(?))should now review his price.HC will be probably better than CEE. if it is not already!

[rendez2k]

Are there future plans to sell or take donation for HC?

[lamster]

Originally posted by jdobbs
the only way I could check would be with a filesize
Actually, I think it would be very useful if you would include the version number, file's timestamp, and filesize for each of the files that you're checking. (I'd also like to see an auto-version option, which would automatically include the version info in the log file at the start of the Prepare phase. This could log only the executables that are being used for the current rebuild - including that of Rebuilder itself.)

[rendez2k]

I assuming the new 0.82 corrects the 'sony' issue with HC?
- Added code to removed sequence end codes that are added as a default on some encoders. The end code at the end of a segment can cause problems on some players.

[jdobbs]

That's the intent... someone will have to confirm it though.

[hank315]

Unfortunately all versions of HC show as being the same... (v1.0.0.0)Will update it in the next version.

[quantum]

Originally posted by rendez2k
I assuming the new 0.82 corrects the 'sony' issue with HC?
- Added code to removed sequence end codes that are added as a default on some encoders. The end code at the end of a segment can cause problems on some players.Has anyone confirmed this is fixed with 0.82?

.

[Fishman0919]

Just re-did Elektra w/ DVD-RB .82 pro, first time was with DVD-RB .80 Freeware had the skip/shudder at chapter points... went in with a Hex Editer and removed end code of the .m2v files and ran rebuild again, it worked fine. 2nd time with DVD-RB .82 pro worked fine also.


Edit: Sorry, 2nd time encoding the movie... first time doing the movie w/ DVD-RB .82 pro worked fine.

[jdobbs]

:cool: That's good to know!

[JohnG]

Originally posted by jdobbs
:cool: That's good to know!
I second that! After harshly judging Sony in an earlier post, I feel a bit foolish now(although I still won't buy Sony). I just played a bunch of backups in my primary player (I usually test them on my secondary player which is more discriminating) and discovered that they ALL skip at chapter points. It is a Samsung which usually plays anything I've ever thrown in it. My test one is a $40 Magnavox set up for the kiddies (which happens to play the HC encodes flawlessly :rolleyes: ) Oh well, I guess I can live with them...I'm not about to Restream them and burn 15 new DVD+R's. I have a Toshiba DVD player on the way...hopefully it can handle these better than the Samsung :)

[Vanderlow]

So with the new 80 free version and HC encoder 12a I see a max bitrate of 8552 (upper right corner of of HC) even though I have an ini line of max bitrate of 8000.

*maxbitrate 8000
*matrix qlb
*profile best
*logfile C:\Program Files\DVD-RB\HC\test.log

[cmsoliveira]

So with the new 80 free version and HC encoder 12a I see a max bitrate of 8552 (upper right corner of of HC) even though I have an ini line of max bitrate of 8000.

I believe that bug was fixed un v0.80, where all the encodes were done with a maxbitrate of 9000 kbps.

Now,rebuilder sends an argument with the right maxbitrate, and i think that arguments have priority over the ini file.

If i'm wrong,please correct me.

[manolito]

Originally posted by cmsoliveira
Now,rebuilder sends an argument with the right maxbitrate, and i think that arguments have priority over the ini file.

If i'm wrong,please correct me.
No, you are absolutely right.

Cheers
manolito

[jdobbs]

Originally posted by JohnG
I second that! After harshly judging Sony in an earlier post, I feel a bit foolish now(although I still won't buy Sony). I just played a bunch of backups in my primary player (I usually test them on my secondary player which is more discriminating) and discovered that they ALL skip at chapter points. It is a Samsung which usually plays anything I've ever thrown in it. My test one is a $40 Magnavox set up for the kiddies (which happens to play the HC encodes flawlessly :rolleyes: ) Oh well, I guess I can live with them...I'm not about to Restream them and burn 15 new DVD+R's. I have a Toshiba DVD player on the way...hopefully it can handle these better than the Samsung :) The were all done with HC?

[JohnG]

Originally posted by jdobbs
The were all done with HC?
Yes. With rbv080Pro and 081. I always mark the process on each finished DVD. The ones where CCE2.5 and OPV were used, there are no problems.

[jdobbs]

@JohnG

You can rebuild any discs that may have the glitch without having to reencode. Just run the backup through DVD-RB again using "No Compression" as the mode and DVD-RB will correct it upon REBUILD. It doesn't take too long... You can even do it straight from the DVD (because the backup isn't encrypted) without ripping it to the hard drive -- although that might slow it down.

[JohnG]

Originally posted by jdobbs
@JohnG

You can rebuild any discs that may have the glitch without having to reencode.

I am just stingy...it's not the process that concerned me, just the cost of the new discs. ;)

[jdobbs]

Ok, cool... but maybe someone else in a similar circumstance will read this post and say "Hey! That just saved me a whole lot of time."

[Vanderlow]

I thought the max bitrate for DVD was suppose to be 8000. I thought that was the standard?

[JohnG]

9800 (video & audio combined) I think? :confused:

[Sir Didymus]

Hi guys!

Here is what the standars say (source Dvd Demystified):

http://img107.exs.cx/img107/1411/image15ep.th.gif (http://img107.exs.cx/my.php?loc=img107&image=image15ep.gif)

Hope it is clear...
Anyway to summarise:
- Video should stay below 9.8 Mbps
- The sum of all of the audio tracks should stay below 6.144 Mbps
- The sum of all of the subpicture tracks should stay below 3.360 Mbps
- The sum of the Video, all of the audio, all of the subs, and the PCI (real time program control info) should be less than 10.08 Mbps

Cheers,
SD

[Vanderlow]

OH ok, so all is well I quess. Thanks.

[jdobbs]

Originally posted by Vanderlow
I thought the max bitrate for DVD was suppose to be 8000. I thought that was the standard? That is incorrect. The DVD standard is up to 9800 for video... BUT.. the total for all audio, video, and subpictures combined can't be more than 10.08Mbs -- so DVD starts at 9000 (to be safe) and subtracts the audio streams to get the maximum. Frankly the maximum almost never gets hit anyway, so it isn't much of an issue.


[ADDED] I guess I should have read further before blabbering on. Sir Didymus already answered this one :D

[Sir Didymus]

Originally posted by jdobbs
...Sir Didymus already answered this one :D

Well, the truth is that life, for a poor bug-hunter like the writer, is very hard in the last weeks, since RB is becoming almost perfect... :angry:

So I have nothing else to do, for spending my time, than providing some links to some nice people looking for that info...

:p

[johnhamler1]

max bitrate set at 8000 give me 8150 when encoding.
except this, rb0.8 with HC works perfectly...

just let you know.did 4-5 tests last week.set my targetsize, and it gives me a iso file at 4.06G(max is 4.35Gig ,correct???). I still have a marge of 280 mb.

doesnt it sound correct?

[jdobbs]

You're not setting the HC "One Pass CQ VBR (with analysis)" option are you? That will severely limit the accuracy of the output.

[johnhamler1]

no no, I have 2 pass, best quality, and the matrix.
run a test last night, HULK PAL, this time it overshoot the size of 250 meg.

with king arthur (pal) I was around 300mb below the 4.7 gig, and with the same settings with HULK I am around 250 mb above 4.7.

there are around 500mb difference in the final size between the 2 movies.why????
what I did for HULK (oversized), I used DVD shrink(after HC encoding) to resize the extra.

[Vanderlow]

My last encode was 4.27GB in size with .80 free.

[johnhamler1]

the size doesnt matter, it is the difference between sizes that bother me.

[jdobbs]

But the fact that they are different movies is exactly the reason they differ. That's why the default TargetSectors value in DVD Rebuilder is set to what it is.

You're not using OPV by any chance are you?

[SAPSTAR]

Originally posted by johnhamler1
the size doesnt matter, it is the difference between sizes that bother me.

That's a known pb with HCBatch, hank is working on improving the bitrate control. It`s not related at all with DVD-RB

[Vanderlow]

Originally posted by johnhamler1
size doesnt matter,

you really believe what your girlfriend been telling you?

[onesoul]

Originally posted by Vanderlow
you really believe what your girlfriend been telling you? lol, good joke there :)

[johnhamler1]

is there a fix at this time???

what is OPV???

[jdobbs]

OPV is One Pass VBR encoding.

[dragongodz]

johnhamler1 - you should try the new HC 0.13. its both faster and has better sizing. so restore the TargetSectors back to default and see how you go.