DGMPGDec 1.4.4 Final! - Page 3

neuron2 · 22nd September 2005, 04:30

I have it fixed and could give you a new version, but I have this issue:

http://forum.doom9.org/showthread.ph...739#post714739

Do you want to wait for that or should I put out an intermediate version?

EDIT: I have that issue fixed too, so I'll make a release tomorrow.

neuron2 · 22nd September 2005, 14:13

Here's the fixed version:

http://neuron2.net/dgmpgdec/dgmpgdec144rc2.zip

* A deficiency in the implementation of pulldown in DGDecode was repaired.
The result is that decoding is now resilient to changes of field order. This
makes Fix D2V superfluous, so it has been removed.

* Colorimetry was reported incorrectly for MPEG2 in some cases. This has been fixed.

mg262 · 22nd September 2005, 18:51

Quote:

* ... This makes Fix D2V superfluous, so it has been removed.

Brilliant! That eliminates some amount of manual labour. Thank you very much.

Fotis_Greece · 22nd September 2005, 19:48

Hi from Greece, It's a great prog which I use a lot. So far it happened to use it with movies with ac3 audio. I just used it with dts and there seems to be a problem there. The dts bitrate is full 1536 and when demuxed I can't play it with no software. Not either powerdvd plays it and reports it as mpeg 2.0 256 audio and can't hear anything apart from hissing.
Needless to say that I have many dts sounds files in my pc which all play fine.
I trancoded the demuxed dts to 6 mono wav files and reencoded back to dts with bitrate 768. The file plays fine .

I authored the dvd using both dts soundtracks (the demuxed and the reencoed by me) and both audio tracks play fine.
So what is wrong there?

neuron2 · 22nd September 2005, 19:56

I don't have any DTS audio VOBs. Can you supply a fragment of one that you say fails?

Fotis_Greece · 22nd September 2005, 20:50

"There is a house in New Orleans" Animals sing.
Just a small part of it in dvd, authored by me with the most expensive encoders (dts 1536) and authoring software
test it. It plays fine, but when you demux it you can't hear the dts file

http://www.interchefs.com/club/testdvd.rar

512 kb

If you need longer vobs just tell me

guada2 · 22nd September 2005, 22:59

Very good optimization.
Thank you very much Neuron2.

hank315 · 23rd September 2005, 00:41

Just did some tests, colorimetry reported OK now, thanks.

neuron2 · 23rd September 2005, 01:37

Quote:

Originally Posted by Fotis_Greece

If you need longer vobs just tell me

What I need is for you to tell me what I need to play DTS files. I tried AC3Filter. Nothing. I tried VLC Player. Nothing. I just get no audio. So, how do I play a DTS file? I have downloaded good DTS files from the internet. I can't make them play.

EDIT: Never mind. Media Player Classic plays them.

@Hank

Thank you.

neuron2 · 23rd September 2005, 01:50

@Fotis_Greece

Everything works fine for me. I demultiplexed the DTS elementary stream from your VOB using DGIndex 1.4.4 RC2 and the resulting DTS file plays just fine with Media Player Classic 6.4.8.4. This process also works for all the other DTS files that I downloaded.

Esc · 23rd September 2005, 04:25

neuron2,
how hard would it be to mark the selection start with Home key and the selection end with End key (like in VDub)?

neuron2 · 23rd September 2005, 05:09

Quote:

Originally Posted by Esc

how hard would it be to mark the selection start with Home key and the selection end with End key (like in VDub)?

Let's just say that it is so easy that I've already coded, tested, and documented it. It will be in the next release. Thank you for your suggestion.

neuron2 · 23rd September 2005, 05:27

Quote:

Originally Posted by mg262

That eliminates some amount of manual labour. Thank you very much.

It's always interesting to hear the user perspective. Here I am sitting here thinking "Oh my Darwin, I've been wrong all along about these transitions. Pulldown has been broken since the first release of mpeg2dec3dg (later mutated to DGDecode)!" But you guys see it as "Ho hum, things are a bit easier now." Believe me, I lose sleep over this stuff. Lemme tell you, I'm just glad I discovered all this before someone told me.

I'm a bit surprised (but relieved) that no one has reacted to the paradigm shift that has occurred.

TripleA · 23rd September 2005, 11:03

Quote:

Originally Posted by neuron2

I'm a bit surprised (but relieved) that no one has reacted to the paradigm shift that has occurred.

That's probably because the changes are invisible to the user. Except as a small reduction in work-load, in some cases.

Marius-the-Mad · 23rd September 2005, 11:34

Quote:

Originally Posted by neuron2

* Colorimetry was reported incorrectly for MPEG2 in some cases. This has been fixed.

I just want to say "Thank you!" to everyone involved.

It was a really weird case.

So: thank you all!

mg262 · 23rd September 2005, 11:54

Mmmm. This is my broad perspective: (for me) there isn't time to understand everything related to video work; so I deliberately choose to focus on understanding things that can improve video quality (AVISynth filters and also capturing for analogue material) and not those that can't (MPEG decoding*, IFOs, CCE, XVid, H.264, Matroska, etc.). For the latter I read the manuals, FAQs, guides, and some posts and then stay close to default settings; when problems occur I deal with them. I am happy (and grateful) that other people write free software that lets me work this way; and as I chose not to put in the time to understand the technology properly, I'm hardly in a position to criticise anything, let alone the suboptimal two-stage (Create+Fix) implementation of a procedure where one stage sufficed. I am also happy that you (via software) choose between x.d2v or x.d2v.bad as opposed to my deciding myself -- partly because it's less work but also because the decision is much more likely to be correct.

Not having 52 extra x.d2v.bad files created is also a bonus

.

*Nearly all my sources are PAL, so no telecine to worry about.

Inventive Software · 23rd September 2005, 12:22

Great! With the amount of new releases there are I'm updating your fine product almost every week!

Esc · 23rd September 2005, 15:34

Quote:

Originally Posted by neuron2

Let's just say that it is so easy that I've already coded, tested, and documented it. It will be in the next release. Thank you for your suggestion.

Happy, happy, happy! Thanks a lot. Now it does everything I ever wanted.

midnightsun · 24th September 2005, 05:16

Hello neuron2!

I think I've spotted a minor mistake in the documentation:

I was reading the dgdecode user manual and noticed that in the mpeg2source() filter syntax section, it reads:

Code:

i420: true/false (default: false)

Output I420 Colorspace.
Possibly required by some legacy applications.
Ignored if the input is not YV12 (4:2:0), or if upConv=true.
- true: Output I420
- false: Output YV12

but in fact upConv no longer takes true/false as an argument.

Is that correct or am I missing something there?

neuron2 · 24th September 2005, 13:38

Yes, that's a mistake. Thank you. I'll fix it for release 1.4.4.

22nd September 2005, 04:30	#41 \| Link
neuron2 _ Join Date: Jan 2002 Posts: 16,634	I have it fixed and could give you a new version, but I have this issue: http://forum.doom9.org/showthread.ph...739#post714739 Do you want to wait for that or should I put out an intermediate version? EDIT: I have that issue fixed too, so I'll make a release tomorrow. Last edited by neuron2; 22nd September 2005 at 06:51.

22nd September 2005, 14:13	#42 \| Link
neuron2 _ Join Date: Jan 2002 Posts: 16,634	Version 1.4.4 RC 2 Here's the fixed version: http://neuron2.net/dgmpgdec/dgmpgdec144rc2.zip * A deficiency in the implementation of pulldown in DGDecode was repaired. The result is that decoding is now resilient to changes of field order. This makes Fix D2V superfluous, so it has been removed. * Colorimetry was reported incorrectly for MPEG2 in some cases. This has been fixed.

22nd September 2005, 19:48	#44 \| Link
Fotis_Greece Registered User Join Date: Dec 2002 Location: Athens, Greece Posts: 124	Hi from Greece, It's a great prog which I use a lot. So far it happened to use it with movies with ac3 audio. I just used it with dts and there seems to be a problem there. The dts bitrate is full 1536 and when demuxed I can't play it with no software. Not either powerdvd plays it and reports it as mpeg 2.0 256 audio and can't hear anything apart from hissing. Needless to say that I have many dts sounds files in my pc which all play fine. I trancoded the demuxed dts to 6 mono wav files and reencoded back to dts with bitrate 768. The file plays fine . I authored the dvd using both dts soundtracks (the demuxed and the reencoed by me) and both audio tracks play fine. So what is wrong there? __________________ Xaser III 1000, Gigabyte 8KNXP Rev2,Herolchi PowerOP 525 Watt, P4 2.8 800 @3.4, 2GB DDR400 HyperX, 1TB 8 HDs, Pioneer 106 DVD-RW, Nec 3520,Plextor Premium, Radeon 9800 Pro, Audigy2 ZS Plat. Pro, Cinergy 600TV, Megaworks 5.1 550

22nd September 2005, 20:50	#46 \| Link
Fotis_Greece Registered User Join Date: Dec 2002 Location: Athens, Greece Posts: 124	"There is a house in New Orleans" Animals sing. Just a small part of it in dvd, authored by me with the most expensive encoders (dts 1536) and authoring software test it. It plays fine, but when you demux it you can't hear the dts file http://www.interchefs.com/club/testdvd.rar 512 kb If you need longer vobs just tell me __________________ Xaser III 1000, Gigabyte 8KNXP Rev2,Herolchi PowerOP 525 Watt, P4 2.8 800 @3.4, 2GB DDR400 HyperX, 1TB 8 HDs, Pioneer 106 DVD-RW, Nec 3520,Plextor Premium, Radeon 9800 Pro, Audigy2 ZS Plat. Pro, Cinergy 600TV, Megaworks 5.1 550 Last edited by Fotis_Greece; 22nd September 2005 at 20:52.

23rd September 2005, 12:22	#57 \| Link
Inventive Software Turkey Machine Join Date: Jan 2005 Location: Nottingham, UK Posts: 1,952	Great! With the amount of new releases there are I'm updating your fine product almost every week! __________________ On Discworld it is clearly recognized that million-to-one chances happen 9 times out of 10. If the hero did not overcome huge odds, what would be the point? Terry Pratchett - The Science Of Discworld

22nd September 2005, 19:56	#45 \| Link
neuron2 _ Join Date: Jan 2002 Posts: 16,634	I don't have any DTS audio VOBs. Can you supply a fragment of one that you say fails?

22nd September 2005, 22:59	#47 \| Link
guada2 Registered User Join Date: Jun 2005 Location: Lyon Posts: 187	Very good optimization. Thank you very much Neuron2.

23rd September 2005, 00:41	#48 \| Link
hank315 Registered User Join Date: Nov 2003 Location: Netherlands Posts: 445	Just did some tests, colorimetry reported OK now, thanks.

23rd September 2005, 01:50	#50 \| Link
neuron2 _ Join Date: Jan 2002 Posts: 16,634	@Fotis_Greece Everything works fine for me. I demultiplexed the DTS elementary stream from your VOB using DGIndex 1.4.4 RC2 and the resulting DTS file plays just fine with Media Player Classic 6.4.8.4. This process also works for all the other DTS files that I downloaded.

23rd September 2005, 04:25	#51 \| Link
Esc Registered User Join Date: Jan 2005 Posts: 69	neuron2, how hard would it be to mark the selection start with Home key and the selection end with End key (like in VDub)?

23rd September 2005, 11:54	#56 \| Link
mg262 Clouded Join Date: Jul 2003 Location: Cambridge, UK Posts: 1,148	Mmmm. This is my broad perspective: (for me) there isn't time to understand everything related to video work; so I deliberately choose to focus on understanding things that can improve video quality (AVISynth filters and also capturing for analogue material) and not those that can't (MPEG decoding, IFOs, CCE, XVid, H.264, Matroska, etc.). For the latter I read the manuals, FAQs, guides, and some posts and then stay close to default settings; when problems occur I deal with them. I am happy (and grateful) that other people write free software that lets me work this way; and as I chose not to put in the time to understand the technology properly, I'm hardly in a position to criticise anything, let alone the suboptimal two-stage (Create+Fix) implementation of a procedure where one stage sufficed. I am also happy that you (via software) choose between x.d2v* or x.d2v.bad as opposed to my deciding myself -- partly because it's less work but also because the decision is much more likely to be correct. Not having 52 extra x.d2v.bad files created is also a bonus . *Nearly all my sources are PAL, so no telecine to worry about.

24th September 2005, 05:16	#59 \| Link
midnightsun Registered User Join Date: May 2004 Posts: 59	Hello neuron2! I think I've spotted a minor mistake in the documentation: I was reading the dgdecode user manual and noticed that in the mpeg2source() filter syntax section, it reads: Code: i420: true/false (default: false) Output I420 Colorspace. Possibly required by some legacy applications. Ignored if the input is not YV12 (4:2:0), or if upConv=true. - true: Output I420 - false: Output YV12 but in fact upConv no longer takes true/false as an argument. Is that correct or am I missing something there? __________________ Sorry I had to go see about a girl

24th September 2005, 13:38	#60 \| Link
neuron2 _ Join Date: Jan 2002 Posts: 16,634	Yes, that's a mistake. Thank you. I'll fix it for release 1.4.4.

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode