DVDx 2.6 is out!

[Jim@labDV.com]

We're proud to announce that we succeeded in migrating the whole DVDx project to the latest Windows development platform: Microsoft Visual C++ 2500 Professional with Intel C++ 9.1 Compiler.

This has immediate benefit for the end-user by fixing bugs introduced by previous compiler, such as when DVDx silently crashes when exiting Input Settings with 44.1KHz or 48.0KHz (no conversion) option.

This will have great benefit for developers, so end-users, by providing new productivity features such as navigation across C++ objects declarations and definitions with one right-click. Intel 9.1 C++ compiler will provide latest greatest Hyper-Threading and Multi-Core optimization. And x64 cross-compilation will be facilitated.

Debug and DebugP4 versions are back, so we'll be able to debug DVDx mush more quickly.

Download:


all versions and binaries (including Debug, DebugP4 and NoOptimization) from labDV.com as a VIP user
standard (ripper free) version and source code from labDV as a non registered user
standard (ripper free) version and source code from Source Forge (will be up-to-date in a few days)


What's new?

Packaging:


Ultra Edition now contains a Tools menu with all DVDx binaries (Debug, DebugP4, NoOptimization and RipperFree), so any user will able to run a debug version to tackle bugs.
Ripper Free package has no change.
"No upgrade" package is deprecated.


Bug fix:


DVDx silently crashes when exiting Input Settings with 44.1KHz or 48.0KHz (no conversion) option. So now you can try these modes to workaround audio/video synchronization issues.


Debug builds:


thanks to the new compilation platform, both Debug and DebugP4 configurations can be built (and are included in Ultra Edition installer).


Compilation:


update compilation platform to Visual Studio 2005 Professional (VS8) and Intel C++ Compiler 9.1
new reference compilation platform is Visual C++ 2005 Pro with Microsoft Platform SDK for Windows Server 2003 R2, DirectX SDK October 2006, Intel C++ 9.1.28
Visual C++ Projects have been converted to Intel C++ Projects
all configurations compile successfuly (including Debug and DebugP4)
added a new open source project: BuildNumber to manage build auto-increment
removed VC6 build auto-increment macro


DVDx offical website at labDV.com (http://www.labdv.com/dvdx)

[setarip_old]

Hi!

1) For those of us who are not familiar with your product, perhaps you can briefly describe here what "DVDx" does?

2) Is it freeware?

[Jim@labDV.com]

Sorry, DVDx is very popular so I didn't know I've to remember what it does.

DVDx allows you to convert DVD to VCD2.0 or SCVD1.0 or DivX or WMV in one step (including multiplexing, splitting). It produces very good quality movies in MPEG1/2 format and you don't need to have 5GB or more free on your hard disk. DVDx has been designed especialy for unexperimented users, it is a simple and intuitive program. You only need Nero to burn your VCD/SVCD. DVDx has evolved in a very powerful program, the fastest DVD backup solution ever and is now adopted by a lot of people including advanced users.

DVDx is an encoder which converts any MPEG/DVD movie in a smaller file such as DivX, AVI, WMV, VCD, SVCD or WMV.

DVDx is an open-source software, hosted as a Source Forge project, official website is hosted at labDV.com.

DVDx installers are all freeware, available for numerous places. Due to legal constraints, the Ultra Edition (which includes the DVD ripper) is available from a fewer location.

The best place to download DVDx is the official website (http://www.labdv.com/dvdx/download.php) or Source Forge (http://prdownloads.sourceforge.net/dvdx) and of course Doom9 which deserve to be updated to 2.6 release (http://www.doom9.org/)

[setarip_old]

Thanks for the information ;>}

[Jim@labDV.com]

U're welcome :)

[745aardvark]

The DOOM9 site is without a doubt an excellent site. It
is very unfortunate that DOOM9 is bound by others, and I
mean DVDx, as to when a pgm is available.

The problem is that LabDV announces a version is available
when it is NOT. This refers to DVDx 2.6 which has to be
purchased

[745aardvark]

The DOOM9 site is without a doubt an excellent site. It
is very unfortunate that DOOM9 is bound by others, and I
mean DVDx, as to when a pgm is available.

The problem is that LabDV announces a version is available
when it is NOT. This refers to DVDx 2.6 which has to be
purchased.

DVDx is either freeware, shareware, or NOT either. It is tricky and
unethical to announce something is available when it is not.

DVDx should either be commercial, freeware, or shareware and
not something in between.

[setarip_old]

@Jim@labDV.com

As pointed out by "745aardvark", it seems that there's some confusion here. Please be good enough to answer a couple of questions directly:

1) As I asked before, is DVDx, in its entirety and all different "flavors" (including "Ultra Edition"), freeware? (As your previous post appears to indicate)

2) Why is it that SourceForge indicates v.2.5 to be the most current available for download?

3) Why is it that your links (both homepage and download) have resulted in multiple timeouts?

4) Can you provide a DIRECT link to DVDx on the Doom9 site?

[Chetwood]

And most important: how does it compare to the completely free AutoGK?

[dragongodz]

This refers to DVDx 2.6 which has to be
purchased
this is incorrect. to download the ultra version(which includes decss so you can rip from dvds) from LabDV you must be a member. membership to the whole LabDV website and access to the files server is what you are paying for. that is perfectly allowed under the GPL. it has been discussed in the past.
i know Jim said
standard (ripper free) version and source code from labDV as a non registered user
but that obviously hasnt been done yet.

setarip_old - that should answer your first question aswell.

DVDx is either freeware, shareware, or NOT either. It is tricky and unethical to announce something is available when it is not.
it IS freeware. just because you have to be a paying member to get it from 1 place doesnt change that. once its available at other sites, which it always is, its up to them if they wish to charge you access or for free. many are free.

2) Why is it that SourceForge indicates v.2.5 to be the most current available for download?
sourceforge usually takes a little while to be updated to the latest version. also the decss version binary will not be available from there ofcourse.

And most important: how does it compare to the completely free AutoGK?
and you have been here long enough to know you should decide that for yourself. :)

[setarip_old]

it IS freeware. just because you have to be a paying member to get it from 1 place doesnt change that.So, it's "free", except you have to pay to get it from the one place that presently makes it available;>}that is perfectly allowed under the GPLWhether it is allowed under GPL is not the question. The posts to this thread by "Jim@labDVD.com" (presumably the author) appear to be intentionally vague regarding any cost involved to obtain any version of this software.

I would hope that "Jim@labDVD.com" wil be good enough to directly respond to my previously asked four questions, as well as those of the other posters to this thread, so there can be no misunderstanding or ill will...

[dragongodz]

So, it's "free", except you have to pay to get it from the one place that presently makes it available;>}
basically yes. you have to either be a current member or pay to become a member of LabDV. it still doesnt not make the actual program freeware. also as stated the ripper free(meaning decss) version will be available from sourceforge soon, that just usually takes a little while later.
should i mention other software such as FairUse which you can download the light version for free(restricted to making no larger than 700mb avis) but have to pay for the full version unless you get it elswhere ?

Whether it is allowed under GPL is not the question. The posts to this thread by "Jim@labDVD.com" (presumably the author) appear to be intentionally vague regarding any cost involved to obtain any version of this software.
not at all. i have quite clearly spelled it out to you. as for author, look at the names on the main screen sometime and you will see me there. i was one of the authors for a very long time, contributing code to Jean Luc Pons(the original author) and it was me that released DVDx 2.3 alone. however with my other interests(rejig and QuEnc) DVDx took a backseat. so Jim and others have taken over updating it from 2.4 on.

so there can be no misunderstanding or ill will...
any misunderstanding is coming from you guys. DVDx has been around for many years and its distribution is not breaking the GPL so really you have nothing to complain about. ok it would be nice if the version sourceforge was done at the same time but surely waiting just for a little while is not going to kill anyone.

[setarip_old]

@dragongodz

As I said previously, I would hope that "Jim@labDVD.com" (the threadstarter) will be good enough to directly respond to my previously asked four questions, as well as those of the other posters to this thread, so there can be no misunderstanding or ill will...

[LoRd_MuldeR]

At least version 2.5.1 can be found here:
http://prdownloads.sourceforge.net/dvdx/?sort_by=date&sort=desc

DVDx 2.6 is not there (yet?)

But who needs DVDx anyway when MEncoder can do the job very well ???

[dragongodz]

As I said previously, I would hope that "Jim@labDVD.com" (the threadstarter) will be good enough to directly respond to my previously asked four questions
the fact is i have already answered several of those questions. it should not take Jim to answer them for you to believe or understand them.
i will however spell them out so its crystal clear.

1) As I asked before, is DVDx, in its entirety and all different "flavors" (including "Ultra Edition"), freeware? (As your previous post appears to indicate)
YES. all versions are freeware. the fact you have to be a paying member of a particular website to initially get the ultra version does not change that at all.

2) Why is it that SourceForge indicates v.2.5 to be the most current available for download?
sourceforge usually gets updated with the latest ripper free version a little while after. this is even mentioned on the LabDV DVDx download page(not main DVDx page). the links on the download page are working though.

3) Why is it that your links (both homepage and download) have resulted in multiple timeouts?
works perfectly fine here. to download the standard(ripper free) version go to the download page
http://www.labdv.com/dvdx/download.php
go down to the section headed "Standard Version and Source Code". from there you can download the standard version binary and the source code without being a member.

But who needs DVDx anyway when MEncoder can do the job very well ???
hmm and who needs thread hijackers and derailers ? if you are interested in promoting another program then open a new thread for it. otherwise you are just being offensive to the people that have put in their free time and effort to provide another alternative free program.

there is a thread about free pie, seems some people should read it.
http://forum.doom9.org/showthread.php?s=&threadid=7770

[blutach]

Thanks for the work and updates guys. Bout time someone actually said that. :) KUTGW

Regards

[setarip_old]

@dragongodz

Even if all "Jim@labDVD.com" says in direct response to my questions is, "Everything 'dragongodz' said is correct", that would be fine - but please don't suggest that I must accept your explanations for questions I've posed specifically to the threadstarter.

Just as you are free to offer your explanations, I'm free to ask for a response from the threadstarter, who likewise is free to respond or not...

[DDogg]

Straight download here (http://www.labdv.com/content/free-hosting/dvdx/files/bin/DVDx_2_6_setup.zip)

[dragongodz]

but please don't suggest that I must accept your explanations for questions
what i said was
it should not take Jim to answer them for you to believe or understand them.
so please do not say i said you MUST anything.

Just as you are free to offer your explanations, I'm free to ask for a response from the threadstarter
of course. you are perfectly free to not believe someone(me) who actually knows something about the subject just because they didnt start the topic. :)

the link DDog gives by the way is the one i mentioned is on the download page(i gave the link to that). if that fails for you aswell then there is a problem with your connection to LabDV and not LabDV itself. if so say and i will do a quick mirror for that if you want.

[OCedHrt]

Too bad code generated by Intel's compiler runs slower than code generated by either GCC or Microsoft's compiler with basic O2 or O3 optimization on AMD Athlons.

[_xxl]

Did you patch the Intel Compiler?
http://www.swallowtail.org/naughty-intel.html

[buzzqw]

on sourceforge is still missing...

BHH

[Jim@labDV.com]

Hi all,

sorry that I haven't replied before but I didn't receive the reply notifiction mails!

Yesterday I've release new 2.7 release at labDV where 2.6 and 2.7 standard version ("ripper free") are available for all for free at http://www.labdv.com/dvdx/download-arch.php

A new DVDx website page because I've also revamped this website, for the benefit of users and developers.

For some reason my friend Starbuck has to redo his Source Forge access setup, but I guess 2.6 and 2.7 will be uploaded there in a couple of weeks.

BTW, source code is also available from the DVDx web page I mention above.

And the previous remark about Intel 9.1 compiler bersus GCC is pure assumption but one who didn't test... MS and Intel compilers are very good.

Cheers

[OCedHrt]

Thanks for the update.

I dunno specifically about ICL 9.1, but ICL FFDShow builds I have tried recently have always been slower than their VC7 counterparts on my system.

[Jim@labDV.com]

Thanks for the update.

I dunno specifically about ICL 9.1, but ICL FFDShow builds I have tried recently have always been slower than their VC7 counterparts on my system.

DVDx speed is reinforced by lot of assembly for MPEG loops (iDCT etc) and by multi-threading.

New compiler won't change assembly and will do better optimization, the builds are done with "global optimization" setting so is long but result has a lot of nice loop optimization.

And is the better way for x64 porting (DVDx experimental x64 has a lot of new x64 assembly and is 25% fatser than x86, however, it's incomplete because still a lot of assembly code to migrate).

Hope we'll got good feedback for DVDx 2.7, nd it runs good on my PCs.

Jim

[OCedHrt]

That's great. I shuold give 2.7 a try when I have time.

But does ICL allow the use of all instructions in AMD64? I would think that it would only use stuff in Intel's x64 instruction set.

[Jim@labDV.com]

To be honest I don't know.

LJames, aka DVDx-DEV, has done the x64 migration, which was assembly code changes.

Anyway, the only differnece I've seen between x86and x64 code is that registers are larger so the CPU can move larger blocks of data from memory to registers and in between registers. Then CPU operations (which always occur inside registers) process more data in a single instruction, so in same clock timing.

You can see that by doing a diff in between x86 and x64 code.

And I guess these "mov" instructions we're talking are the same in Intel and AMD processors.

And these changes are efficient (so far measured +25%) because MPEG code has a lot of loops to move, process and store data (DCT, iDCT, resize, etc).

About DVDx roadmap, we're thinking to balance benefit of assembly code we've to maintain and compiler optimization we need to follow on Dual Core, Quad Core, 64 bits evolutions. I guess we're doing good now even if we're not doing perfect - so room for improvement for new developers to join us.

[Jim@labDV.com]

Too bad code generated by Intel's compiler runs slower than code generated by either GCC or Microsoft's compiler with basic O2 or O3 optimization on AMD Athlons.


Sure, if you run x86 code on a x64 machine, the thunks (WOW stuff) will slow down the program. I don't have a x64 machine here to test and optimize, even to successfuly compile DVDx x64. If AMD would donate a machine I'd be pleased to work on it.

[Jim@labDV.com]

Sorry that I've missed this thread, I'm deep in DVDx code for a while and not regularly at Doom9 forums.

First the good news: as asked in other Doom9 forums, I'm working on a better support of Xvid and it's very encouraging, Xvid source code (from xvid.org) is very nice (the best free code I've seen for years in video processing), third-party developers are more welcome by Xvid team compared to DivX so I guess I will emphasise Xvid support in the future, changing DVDx roadmap. It would be rather perfect if an Xvid API documentation, even draft, was available.

Second: yes, I admit we've been bad at SourceForge uploads for a while, it's due to the SF SVN migration which we missed and CVS is a real pain in the ass. So as of now I guess we'll update SF file releases on time (and forget about CVS - which as no added value to DVDx project).

Third: one shouldn't be mistaken but the English word "free". As defined by software developer community, "free" stands for "freedom" not for "gratis". So freeware for free software means you are free to use, to get source code and modify it, assuming you respect license - GPL means you must publish your modification.

Therefore DVDx project and labDV team follow on carefully Open Source and Free Software policies, this is a clear strategy.

I would like to thanks all people who contribute(d) in a way or another to DVDx project, contributors as well as users who pay(id) little fees to labDV. DVDx project isn't a profitable business at all, but I don't care because I'm retired - I'm not so old btw - and I really enjoy it. I still need a lot VIPO contributions because I'm not rich enough to pay all the hardware, software and ISP fees.

Because Jean-Luc, DVDx creator, has no longer time to participate to the project, I've taken over coding and after 2-year learning (and with a university background in digital image processing) I'm not so bad now as improving DVDx code and features. And I'm thinking to DVDx 3 - but it will take more than a year to get a public beta - while I'm still improving DVDx 2.

To people like 'setarip_old' , you would be welcome to contribute in any way (coding, testing, writing docu) to any open source, free software project...

Please check this about free software: http://www.fsf.org/licensing/essays/free-sw.html

Jim

[setarip_old]

one shouldn't be mistaken but the English word "free". As defined by software developer community, "free" stands for "freedom" not for "gratis".
If you read my posts in this thread CAREFULLY, I asked if the program, in all of its iterations is FREEWARE, not simply "free" - and "dragongodz", cliaming to speak for you, stated that it is "FREEWARE".

Read the Wikipedia definition of "FREEWARE" at the following link:

http://en.wikipedia.org/wiki/Freeware

[glb]

If you read my posts in this thread CAREFULLY, I asked if the program, in all of its iterations is FREEWARE, not simply "free" - and "dragongodz", cliaming to speak for you, stated that it is "FREEWARE".

Read the Wikipedia definition of "FREEWARE" at the following link:

http://en.wikipedia.org/wiki/Freeware


Can someone else play this game?
Read the Merriam-Webster definition of "FREEWARE" at the following link:

http://www.m-w.com/cgi-bin/dictionary?book=Dictionary&va=freeware

[Jim@labDV.com]

I may answer just once to this because it may not be very valuable.

I disagree with Wikipedia definition, you can go back there and read:

This article or section does not adequately cite its references or sources.
Please help improve this article by adding citations to reliable sources. (help, get involved!)
This article has been tagged since October 2006.

...which send you there: http://en.wikipedia.org/wiki/Wikipedia:Attribution

It's not because it's in Wikipedia of other web site that's it's the truth and I guess it's a good example for all.

My point is that "free-software" definition came from the free (as freedom) and open source community efforts which are now consolidated and promoted by FSF, the Richard Stallman group. Nevertheless, free-software and open-source software disagree, read "why free software is better than open source": http://www.fsf.org/licensing/essays/free-software-for-freedom.html

For being an old man who's lived the whole www creation, from my prospective, "freeware" came from websites which wanted to distribute software for free, so "shareware" term was created as well as "freeware", but the later as a contraction of "free software" understood as "free of charge software" by the public.

And I've to recognize that when I use freeware I don't encourage the comprehension of "free to use software" so I contribute to this burden.

Finally, my point is to argue against all the people who use only "free of charge" and "cracked" software, who never contribute in anyway to www community nor developpers and who complaint when they're asked a small fee for a software which make them saving a lot of money.

These consider that developper work ins'nt worth it, and they contribute in that people, usually students, who started nice software initiative then gave up after a few months... I guess it would be better to encourage them. Which is done by the majority, thanks to them, but a few people I'm arguing here are enough to spread negative opinions and this won't encourage the "free" community work, it will do the good job for Bill G and if we don't argue and battle, we'd have only $$ Billouware available in a very short term.

At the moment, I'm thinking at the best deal we can do at labDV to contribute to WWW/Dev community with free-of-charge & free-to-use & open-source software but keeping some technical advantage to avoid commercial software stoling us.

So, and after reading Richard statment "Why you shouldn't use the Library GPL for your next library" ( http://www.fsf.org/licensing/licenses/why-not-lgpl.html ), I guess I'll come to the conclusion that I'll publish some code as L-GPL to benefit to the "others" assuming the "others" will contribute to keep a leading edge code. While I'll keep some propriatary code to avoid what I see for years now: the DVDx rip-off sold by spam, providing disappointed users and which piss me off!

This will be difficult because some DVDx libraries are GPL so they can't be changed to L-GPL.

Sorry if this become too technical and annoying for standard users but it's a very important question, so important that FSF is working on a GPL-3 to avoid the new attacks from U.S. majors to patent every piece of code they create and want to kill the whole "free" communiy initiatives.

For example the more successful free software is Linux and would be bad to kill it.

And it was reported that Billou has used some Unix/Linux code in NT (easy for a coder to copy-paste the C code of TCP/IP stack) - sure they changed it when they were discovered. But imagine that after learning TCP/IP form open-source, they rewrite part of it and then patent it, then anyone wanting to code a TCP/IP stack would have to ensure not infringing patent or may being sued!!!... Sure nobody will then write public open code...

At my little level, I'm just trying to make DVDx survive the author give up, taking over the coding and protecting the work against commercial ripp-off and well as trying to get revenues to make DVDx coding a balanced accountancy.

I've already succeeded because I got a lot of pleasure coding DVDx and it's becoming my full-time occupation. And I don't need money for myself from DVDx.

And I really enjoy this work, I'm ordering smart (and expensive) technical books from Amazon to make the best digital video code. I'm like a student preparing a PhD, reading and learning a lot. More than ever in my whole career. I love this!

I can afford these books, but at the moment I can't afford a x64 Vista machine with a HD-DVD and Blu-ray drives which I need, to keep DVDx on the edge.

I hope new supporters to become labDV members providing more VIP revenues, and I'd be pleased to see a company donate us such a PC for our work. Intel and AMD did this in the past (to demonstrate their CPU speed at encoding MPEG) but now they aren't interested ny longer, I guess they prefer the Billou company.

Jim

[OCedHrt]

Sure, if you run x86 code on a x64 machine, the thunks (WOW stuff) will slow down the program. I don't have a x64 machine here to test and optimize, even to successfuly compile DVDx x64. If AMD would donate a machine I'd be pleased to work on it.

That's probably true, but I was referring to that fact that Intel's compiler disables key optmizations on non Intel-Genuine processors such as AMD's.

[Jim@labDV.com]

Are you suggesting I provide a x86 binary compiled by the MS compiler to run it in AMD's machines?

[OCedHrt]

As mentioned by drevil_xxl at the top of this page, the ICL libraries can be patched.

It seems like the FFDShow folks are patching their ICL libraries now.

[anatase]

Jim@labdv.com - can you provide a quick scenario in what technologies are involved from at least vob to wmv?

I'm simply concerned about speed and amount of muxing (or whatever the term is).

Basically I just dont want to hear that the product converts vob to dvix or some avi then to wmv.

[Jim@labDV.com]

DVDx uses DVD libs from DVD2AVI projects to extract frames from VOBs to uncompressed RGB or UYV frames (with nice MPEG-2 decoder).

For MPEG, these frames are then encoded to MPEG-1/VCD or MPEG-2/SVCD with the internal MPEG1/2 encoder (or external frame-serving from Flask/VirtualDub/Premiere).

For AVI these uncompressed frames are encoded by the AVI codec.

For WMV, these frames are encoded by native WM SDK code (DirectShow COM interfaces).

Therefore there's no intermediate picture quality loss.

Jim