Problem with Troy and my newly donated 1.07

[kRuZaDeR]

I just installed my 1.07 and checked to make sure my setup/configuration
was correct. I just left all the other extra features on 'Default' and will
be tweaking that later. I proceeded to BATCH Rebuild two DVD's in my collection,
"KING ARTHUR" and "TROY" and as usual I do my rebuilding
overnight since I do a 5-pass, which takes about 5+ hours in my P4-2.8gh.

The next morning when I checked on it at 5 AM, It halted on the "TROY"
rebuild and the rebuild goes as far as, I believe, reencoding of the menus
which is a new thing with 1.07 and not done by the 'Free' versions. The
software interface is still up but the CCE is not encoding and stops at VOB_0
something "Creating M2V for VTS_01 segment 0". I checked the files
and it stopped at 2 AM.

I had to go to work so I closed everything and will work on it this evening
when I get home. The "KING ARHTUR" rebuild seems to have been completed,
but at '346' minutes seems to be longer than most of my rebuilds averaging at
305 minutes (using NAN's DGDECODE.DLL version 1.1.0).

Will post more this evening. Maybe any of you guys encountered this
problem and has a solution. Want to get a head start as soon as I get
home. Thanks.

[Fairhope]

jdobbs has/will be releasing an updated version which might help you, especially if Troy has ILVU.

[jdobbs]

You may also want to consider trying DGDECODE v1.4.5 if you run into a "hang" problem -- the main reason I implemented it was because of older bugs in v1.1.0 in which it could hang (although it seems to be very rare):

- Added support for DGDECODE v1.4.5. This update
includes many enhancements/bug fixes and makes
DVD-RB's support consistent with Neuron2's most
current release. Support for MPEG2DEC3DG and
DGDECODE v1.1.0 remains. This update includes
a fix which corrects a rare problem in which an
encoder may "freeze" while trying to encode a
segment. The DVD-RB Installer now includes and
installs DGDECODE v1.4.5 by default.

As mentioned, in addition there was a bug introduced in v1.07 related to freezing on ILVU that has been corrected in v1.07.1

v1.07.1 is being released today and all VIPs should get it within the next few hours.

[kRuZaDeR]

Thanks for the quick response!

I did notice that there was interleaving and angles during
time I was inputting the source. That could be it.

As far as the DGDECODE issue. I did not change the
Dgdecode.dll ver. 1.4.5 that came with the software.

All I was mentioning was the comparison of length of time
it took to rebuild with this new DVD-RB 1.07 w/ Dgdecode 1.45
versus DVD-RB v0.96 + NAN's Dgdecode.dll ver. 1.1.0, which I
have used the last few months to rebuild most of my collection.

It seemed to me to be longer (346 minutes) but then again this is
just one DVD. I was averaging around 275-300 minutes per DVD on
the Free version plus NAN's Dgdecode.

Will check my e-mail this evening. Thanks.

[jdobbs]

You may want to also experiment a little with the IDCT settings (under the Options tab). Different processors perform better with different settings.

If you want, you can also still use the NaN decoder... just point to it under the SETUP dialog. Remember, though, that the D2V format is different -- so you'll have to redo any PREPARE you may have done already.

I just feel bad that I didn't find that bug in testing on v1.07... I did several ILVU discs.

[m1ckran]

I just feel bad that I didn't find that bug in testing on v1.07... I did several ILVU discs.

Don't feel bad. Microsoft have made an art of producing buggy software. You've got a long way to go to compete with their record.

Errors are bound to creep in occasionally, considering your prolific output. It's the quality of the corrective action that distinguishes the wheat from the chaff. Besides, it's true that only people who do nothing never make mistakes!

[jdobbs]

Ain't that the truth... hmmm... maybe that's my goal -- to do nothing and make no mistakes!

[m1ckran]

Too late!
I'm already there.
I can even do nothing and still make mistakes, so stick with what you do best.

Think I'll take my pills...

[kRuZaDeR]

Just reinstalled 1.0.7.1 and rebuilding "TROY" in progress. No problem so
far! Still left all the other extra features on the 'Default' setting with only
one slight change. I used 32bitSE2/MMX on my IDCT setting. I believe
this is the setting used for Pentium 4's ??? I started at 7:56 and now it
is 8:57 and it is at 20% done. So with 20% per hour, estimate is 5 hours.
Just about right.

[NaN]

@ kRuZaDeR: If you use my DGDecode-build, I would recommend idct=6 for your P4. It gives another 5% encoding speedup (you can do this with DGDecode 1.4.5 too, however in my build the IEEE1180 compliant version gives more accurate results).

Cheers, NaN

@ jdobbs: unfortunately I don't have the time in the moment, however I will look into your issue, would be nice if you could test the fix then!

[jdobbs]

The only disc I've ever personally seen it happen on is "Reefer Madness, The Musical", NTSC, R1.

[kRuZaDeR]

I did an experiment just to find out the difference in speed
between the DGDecode version that came with DVDRB 1071
and NAN's DGDecode ver 1.1.1 I used in DVDRB 096.
I used my King Arthur DVD (CCE, 5-pass) and the results are:

1. DGDecode (version 1.4.5 I believe) that came w/ DVDRB1071.

a) with IDCT on "Default" (don't really what CPU it indicates)

Total time is 347 minutes

b) with IDCT set to 32bitSSE2MMX

Total time is 339 minutes! (improvement over default)

2. NAN's DGDecode ver 1.1.1 and DVDRBv096 Free.

Total time is 327 minutes!!! (WOW)

Also there is a problem with the Menus when encoded
with the donated DVDRB1071. The highlighted cursor
to pick the various options in the Menu is MISSING???

The Menu is normal with the FREE DVDRBv096!!??

Just want to let you know.

[jdobbs]

Something is very definitely flawed in your test. Almost all the time during encoding is spent in the encoding phase... and the encoder can't change speeds between versions. There was more to encode with the Pro version -- it was reencoding the menus. If you turned menu encoding off it would probably be exactly the same time. In addition -- you could just have easily used NaN's decoder with the Pro version...

Also -- the bug with menu encoding is known and a notice was posted here on DOOM9. It has been fixed and the new version is being sent out today. Of course the menu worked on the freeware version -- it is an exact copy of the original.

[kRuZaDeR]

That makes sense (Menu encoding takes up time). I apologize,
I did not mean to sound reproaching or overcritical. I am still
trying to learn much about the new features of this wonderful
program. I appreciate you pointing out that we can disable
menu encoding. I'll set some time today to look over the new
features and play around with the new settings. I know that
there is an FAQ or HOWTO out there the explains the various
technicalities of different options. I'm really EXCITED!
Thank you very much. And you are really cranking out those
fixes! You ROCK!

[jdobbs]

I'd like it a lot better if there weren't any fixes to mail out... but -- I guess that's just life. :p

[Rippraff]

@kRuZaDeR

And when you post such kind of comparison which is welcomed please give the essential information of your CPU as this has a big influence (specially NaN's build). ;)

Cu Rippraff

[kRuZaDeR]

Abit IC7-G2
Intel 2.8gh processor
Corsair 2x1GB Dual Channel PC3200 DDR400 2GB memory
ATI Radeon 9800 256mb
Maxtor SATA 120GB
Seagate SATA 120GB

[kRuZaDeR]

I manage to test KING ARTHUR this weekend with NAN's DGDecode 1.1.1
set to IDCT = 3 (32bitSSE2MMX), 5-pass CCE, DVD-RB 1.0.7.2 w/ Menu
encoding, P4 system that I have and the results was surprising:

364 minutes vs the DgDecode 1.4.5 set to IDCT = 3 comes up with 339 minutes!

Hmmmm....

Previous Post:

I did an experiment just to find out the difference in speed
between the DGDecode version that came with DVDRB 1071
and NAN's DGDecode ver 1.1.1 I used in DVDRB 096.
I used my King Arthur DVD (CCE, 5-pass) and the results are:

1. DGDecode (version 1.4.5 I believe) that came w/ DVDRB1071.

a) with IDCT on "Default" (don't really what CPU it indicates)

Total time is 347 minutes

b) with IDCT set to 32bitSSE2MMX

Total time is 339 minutes! (improvement over default)

2. NAN's DGDecode ver 1.1.1 and DVDRBv096 Free.

Total time is 327 minutes!!! (WOW)

Also there is a problem with the Menus when encoded
with the donated DVDRB1071. The highlighted cursor
to pick the various options in the Menu is MISSING???

The Menu is normal with the FREE DVDRBv096!!??

Just want to let you know.

[jdobbs]

The difference in time between the new time and the old one is simply the menu portion being encoded.

On my Athlon XP 3200+ system I get almost the exact same speeds from NaN's version and v1.4.5... I don't know why it acts differently with different processors -- I assume its the SSE2 code.