|
Looking For Some Volunteers
|
|
Staff Member
2 product reviews
|
10. May 2004 @ 11:58 |
Link to this message
|
|
At the request of a friend, I've been doing some quality testing of CCE Basic (using DVD Rebuilder) against DVD Shrink. While I was working on this I decided that I'd like to do a more comprehensive comparison that includes other compression programs. The problem is that I've never seen results from one of these programs that impressed me enough to buy it, so unless I want to spend the money for them now I have to find other people to help me out. When I'm done I'm planning to write an article of some kind on the results that will hopefully be a good resource for someone who's trying to decide what method to use. Yeah, I know, dual layer burners are right around the corner, but a lot of people won't be upgrading for a while, so I still think this can be useful information.
My methods so far are this. I start by re-authoring the Main Movie and compressing with DVD Shrink. Then I repeat the process, except that I leave it uncompressed, and then back up the reauthored disc with DVD Rebuilder. Once both backups are finished, I use AviSynth and a plugin called SSIM which compares each frame of the final backup with the uncompressed re-author with the same frame from the compressed version and gives a value between 0 and 1, based on the pixel by pixel differences and creates a text file that can be read by Excel. After I've done this for each of the compressed backups, I load the text files into an Excel spreadsheet and find the frames in each that are furthest from the original and make a list of them. Finally, using the lists I made, I compare the worst frames from each backup to the corresponding frame in the other (uncompressed and alternate compression method) backups, visually inspecting them for differences.
So far, with just DVD Shrink and CCE to compare, I've already made some interesting observations. The most interesting thing (to me at least) is that apparently Shrink only compresses 2/3 of the frames. Every 3rd frame in the Shrink backup I did is identical to the original. I've also noted that in the Shrink version, most frames are actually closer to the original than the CCE version, but the worst frames (in either backup) invariably looked better when re-encoded with CCE.
If you'd like to help and you have a copy of one of the commercial DVD-9 to DVD-5 compression programs and can follow simple instructions, PM me so we can discuss it further. You don't need to know anything about AviSynth or spreadsheets to help.
If anyone has any comments, suggestions, or questions, either post here or PM me.
Edit: Since I've heard from a couple of people, I realized that I forgot a couple of crucial points here. First, if you contact me about helping out make sure to mention what program(s) you have to test with and whether you're in PAL-land or NTSC country. Brian100 has generously volunteered to help me with CCE (DVD Rebuilder) testing for PAL sources so I can make sure to get a comprehensive test.
I'm also going to need some suggestions for movies in each of 3 categories. I plan to run 1 test with a high action / high bitrate source, another test with a long, but moderate bitrate title, and a third with animation (which I expect to be the hardest to decide on a movie). If you have suggestions for any of the 3 tests please include that information as well.
This message has been edited since posting. Last time this message was edited on 10. May 2004 @ 17:43
|
|
Advertisement
|
|
|
|
AfterDawn Addict
|
10. May 2004 @ 21:40 |
Link to this message
|
Sounds like an interesting test. I hope that running everthing through Shrink first doesn't goof things up. DVD Rebuilder's author was a bit concerned about this method as he felt Shrink might be throwing something else into the mix, obviously even uncompressed, Shrink is changing the original structure somewhat.
I would suggest that the tests be done with very long movies as many have said that transcoders can actually do a better job than CCE on shorter movies. Lord of the Rings: The Twin Towers would probably be a good test choice. Also using DVD Rebuilder as the only test bed for CCE may not be a complete test.
I've also played around with DVD2DVD-R 1.81 which is a Movie Only program that uses CCE, but the way it goes about things is different, Rebuilder authors differently and encodes with CCE chapter by chapter. There is also DVDREasy which is Movie Only using CCE, but I've never gotten that one to work (hangs at the CCE phase). The nice thing about both these programs (and InstantCopy 8) is that they don't have to be stripped or run through Shrink first.
I'll be PM'ing you for some details. I'd like to test out DVD2DVD-R 1.81, Intervideo 1.2 and InstantCopy 8. I'd also like to see if the results from Recode 2 are the same as Shrink.
I'm NTSC. As a side note although I've gotten very nice results with DVD2DVD-R that play nice in my JVC I have chapter stop problems in my Sampo.
|
|
grogey
Suspended due to non-functional email address
|
10. May 2004 @ 23:43 |
Link to this message
|
|
I've got a couple of suggestions for your animated choices vurbal.
These two are of course computer animations:
1)Final Fantasy - The Spirits Within and
2)Shrek.
Both of these movies are one's that both children and adults could watch a few times over without getting bored to tears.
For the ultimate Disney experience my pick would be Fantasia2000.
Sound Mix: DTS / Dolby Digital / SDDS
Runtime: 75 min
Also Known As:
Fantasia '99 (1997) (USA) (working title)
Fantasia 2000 (1999) (USA) (alternative spelling)
Fantasia Continued (1996) (USA) (working title)
My other suggestion would be
Grogeybear
My best advice for NEWBIES is: "use the search function.
THE BEST REVENGE IS TO LIVE A GOOD LIFE
Aim for Charm not Perfection
|
AfterDawn Addict
|
11. May 2004 @ 02:00 |
Link to this message
|
|
I hate volunteering lol, but it sounds like a good idea...
|
Staff Member
2 product reviews
|
11. May 2004 @ 05:36 |
Link to this message
|
MysticE:I'm also a little concerned about whether Shrink does anything else to the movie when you reauthor. It occurs to me that while I'm getting things organized I could compare Shrink results against the same movie run through IFOEdit. If the movie is actually changed at all with Shrink, IFOEdit is the obvious solution to get around it since it's also free, and it's not difficult to use with decent instructions. Now I just have to dust the cobwebs out of my brain and remember exactly how to do it with IFOEdit. It's been a long time since I have.
As far as movie length, I agree that the transcoders should theoretically be better for those, which is why I'm planning to run 1 test with a moderate bitrate, but long, movie. That was the type of test I started out with, and after seeing how Shrink does things (leaving 1/3 of the frames untouched) I'm beginning to question that theory. Hopefully Shrink will prove to be the exception. On the other hand I'm counting on a lot of surprises (maybe even from CCE), so it should be interesting either way.
Oh, and I'm glad you have all of those programs. I was hoping for results from all of them. I'm particularly interested in Recode after seeing how Shrink works, but I've seen a lot of good things about IC8 and Intervideo so I'm pretty excited to have them represented. I hadn't considered comparing the other CCE methods, but it seems like an excellent idea. After all I'm looking for the best results from the 1-clicks so I should be looking for the same from CCE. I mostly picked RB because you can use CCE Basic with it, and I'm mostly cut off from the copy of SP I used to use.
|
Staff Member
2 product reviews
|
11. May 2004 @ 05:41 |
Link to this message
|
|
grogey: I hadn't thought of Shrek or Final Fantasy, but they might be the best bet for animated movies that people without kids might have.
Oriphus: I was actually going to see if you wanted to contribute some webspace for the results at the end, so you can volunteer without all the headaches of volunteering. 8D
|
Staff Member
2 product reviews
|
11. May 2004 @ 05:57 |
Link to this message
|
|
For anyone else who might like to help out, I currently have the following programs covered:
PAL:
Pinnacle Instant Copy 7
DVD2One
NTSC:
DVD Rebuilder
Pinnacle Instant Copy 8
DVD2DVD-R
Intervideo DVDCopy
DVD X Copy (I need to find out which version still)
If your software of choice isn't included here, feel free to help out. I'm hoping for the most comprehensive test possible, and as far as I'm concerned, every piece of software is the same until we prove otherwise.
To everyone who's already volunteered help or suggestions I appreciate the help. I'm betting that there will be some results that surprise everyone, myself included.
This message has been edited since posting. Last time this message was edited on 11. May 2004 @ 05:58
|
AfterDawn Addict
|
11. May 2004 @ 06:36 |
Link to this message
|
|
Wel,, you know all of your efforts are very welcome on my webspace Vurbal, so of course that can be arranged ;-)
I prob will help out with the tests as well though....;-)
|
Staff Member
2 product reviews
|
11. May 2004 @ 09:26 |
Link to this message
|
Oriphus: I didn't figure you'd have a problem hosting it. I just thought I'd be polite and ask. I can be polite when I want something. Did I actually type that? Oops ;)
Let me know what you'd like to do and I'll add you to my growing list. You're in the UK aren't you? I'm trying to keep the PAL and NTSC testers straight in my head.
MysticE: I'm in the middle of running the test I mentioned using IFOEdit and DVD Shrink, and so far it seems that they produce identical output. All the files from both are identical size (down to the byte) and I'm processing them both, using the same quality test I plan to use for the overall testing. There doesn't seem to be any difference between them, so I'd say that unless there's other problems with Shrink's authoring (which IFOEdit would probably have to be duplicating), it looks like Shrink works fine, at least for movie only. If this doesn't end up being true for 100% of frames I'll probably go with IFOEdit instead. I was surprised to find that I didn't need a refresher to remember how I used to reauthor with it. I guess that's what endless repetition and another year plus of experience does for you.
I don't have any problem with using other programs to do the reauthoring just as long as we can still correlate frame numbers to visually compare them. I'm thinking I'll add ReJig to the test and run it myself because it may have some bad code from the older versions of DVD2AVI/MPEG2DEC3 that sometimes drops frames. If I remember correctly, Nic (ReJig's author) was also responsible for some of the development of those tools and I don't know if he uses that code or the new code, or Donald Graft's fixes, or something else entirely.
|
AfterDawn Addict
|
11. May 2004 @ 09:45 |
Link to this message
|
|
Yeah, im in the UK, but im kinda busy this week with stuff. Got a couple of females i need to find more out about and try and get in to them for one ;-) lol. But seriously, i've crappy coursework for University im working on (Bloody game in Shockwave) but im happy to participate.
|
|
grogey
Suspended due to non-functional email address
|
11. May 2004 @ 10:08 |
Link to this message
|
vurbal Are you intending on testing Nero Vision Express and Nero Recode - I'm not sure if I have the expertise but I'm certainly willing to give it a go. I also have Sonic Record Now and if I dig deep in my software tray I should be able to find Ulead dvd workshop (I think that's the title) if that is a help.
Just let me know.
As for the movies themselves - when I was thinking about what we've got here so I could make some suggestions I thought the same Quote:
they might be the best bet for animated movies that people without kids might have.
I also considered what would be bearable (no pun intended lol) for adults to play around with - if anyone dares suggest to use The Brave Little Toaster I'm off to hide in the cave...........although I'm not sure it is on DVD but my eldest son played this particular video sooooo many times that I just couldn't take it anymore and hid it in a top cupboard in the laundry when he was 6. Last weekend he was getting down his sleeping bag - and you guessed it - he found it and to my horror played it. If I wasn't sentimental and wanted to keep it for his children one day - I'd have to do some serious damage to the tape!!
Bugger I forgot to mention that my son will be 12 next month.
Grogeybear
My best advice for NEWBIES is: "use the search function.
THE BEST REVENGE IS TO LIVE A GOOD LIFE
Aim for Charm not Perfection
This message has been edited since posting. Last time this message was edited on 11. May 2004 @ 10:11
|
Staff Member
2 product reviews
|
11. May 2004 @ 10:09 |
Link to this message
|
|
No problems on the schedule Chris. I don't expect to finish getting testers rounded up until later this week, and then we still have to pick out movies to use and I need to make sure I get good instructions to everybody. Fortunately, I seem to be assembling a group of tester's who should make it pretty easy.
|
|
grogey
Suspended due to non-functional email address
|
11. May 2004 @ 10:16 |
Link to this message
|
I'm back again vurbal...........can you explain something for me please? How do you work out whether something is high or moderate bitrate? Right out of my area but now you've raised the subject I wanna know more.
Grogeybear
My best advice for NEWBIES is: "use the search function.
THE BEST REVENGE IS TO LIVE A GOOD LIFE
Aim for Charm not Perfection
|
Staff Member
2 product reviews
|
11. May 2004 @ 10:23 |
Link to this message
|
grogey: I'd love to have tests for any tools out there. Nero, in particular, is a important for me since that seems to be the burning program of choice for the majority of informed people. I don't have the latest version personally, but that has to do with issues Nero 5.5.x (sometimes) had when I started burning DVDs. I started using RecordNow Max 4.5 and never looked back or even updated it.
Are you in Australia or am I thinking of someone else?
Edit: Good question about the bitrates (see that's why I ask for input). I have 2 programs I use to evaluate bitrates. The first is Bitrate viewer. There's a free version of this, but I don't completely trust it because it's not supposed to be as accurate as the pay version. My personal choice is a program made by Elecard called MPEG Stream Eye, which I think gives you more accurate results. In a pinch, you could also do a rough estimate by opening a movie in DVD Shrink, removing the audio tracks, making sure it's set to no compression, and dividing the resulting size by the number of seconds. That's not completely accurate since it doesn't take muxing overhead into account, but it would still work for our purposes. Of course to use that method you'd need to start with a movie you know has to have a fairly high bitrate, like the last 2 Star Wars movies or the Matrix movies, so you could tell what's high and what's not. As a last resort, you can generally tell from looking at the movie's content from an encoder's standpoint. Things that use a lot of bitrate are small details (especially when they move), edges, high contrast, and probably some other things that don't immediately come to mind.
I can't find where I downloaded MPEG Stream Eye from, but I can send the evaluation version to you if you'd like (PM me). It should be on my wife's old PC (that I just took off the network to set up the new one I gave her for Mother's Day) You can download the free version of Bitrate Viewer from this page:
http://www.tecoltd.com/bitratev.htm
This message has been edited since posting. Last time this message was edited on 11. May 2004 @ 10:59
|
Staff Member
2 product reviews
|
11. May 2004 @ 11:23 |
Link to this message
|
In order to elicit more comments on the whole process, and let everyone who's already helping get an idea of what they're in for, this is a rough breakdown of what I'm going to need testers to do for each movie.
1 Make a movie only backup with no compression. I'd prefer that you use either DVD Shrink or IFOEdit to do this with since I'll be using one of them and they give identical results. BTW MysticE, I finished the test and DVD Shrink should be fine.
2 Using whatever program you're testing, make the highest quality backup you can from the uncompressed movie only backup.
3 Use DVD2AVI to make a project file so AviSynth can read the VOBs.
4 Load an AviSynth script provided by me (you'll just have to edit a text file) into VirtualDubMod and play it. Because of the plugin the script will use it will take longer than playing the movie normally. On my PC (P4 2.4GHz, 768MB RAM) it takes about 1 1/2 times the play time of the movie. When it's done, you'll need to close the video file before exiting VDubMod, and it will write a CSV (Comma Separated Values) text file for you.
5 Send me the CSV file so I can analyze it.
6 After analyzing all the files I'll send you another AviSynth Script (or maybe several depending on the results) that you can open with VDubMod. The script(s) will trim out everything except for the frames that different programs had trouble with from the movie.
7 This is the part I haven't completely worked out yet. I'm going to need the frames sent to me, either as uncompressed AVI from VDubMod or as individual screenshots, so I can put them together in some organized fashion. I may have to temporarily put up an FTP server for that or I may come up with another solution. We'll just have to see how much there actually is.
8 I'll take all the material and try to write a comprehensive summary of the strengths and weaknesses of each program, as well as contrasting how each one did on frames that gave others problems.
I hope this makes the whole project a little clearer.
|
AfterDawn Addict
|
11. May 2004 @ 16:21 |
Link to this message
|
|
It ounds like a good plan Vurbal...
Recommended Movies:
Final Fantasy - Spirit Within - definitely
The Matrix Revolutions - very digitised at times
Star Wars in Digital Re-mastered versions
|
|
grogey
Suspended due to non-functional email address
|
11. May 2004 @ 21:45 |
Link to this message
|
Sounds like a plan indeed!
vurbal I have the latest version of Nero - I always update or upgrade as new versions become available so I can help out there.
make note to self to actually install Nero & Nero vision express updates that I downloaded last week.
Thanks for bitrate info - I'll have a look at that later on after I get some housework done!! lol
I'll pm you in a day or so if you like to find out any more relevent information.
Grogeybear
My best advice for NEWBIES is: "use the search function.
THE BEST REVENGE IS TO LIVE A GOOD LIFE
Aim for Charm not Perfection
|
Staff Member
2 product reviews
|
12. May 2004 @ 06:44 |
Link to this message
|
|
For anyone who might try to contact me today, I'm not ignoring you. After having problems last night with my ISP's name servers, my PC decided that it would rather not be on a network, and after several hours of troubleshooting I've had to re-install, so I'll be busy with that for much of the day.
|
Staff Member
2 product reviews
|
13. May 2004 @ 06:45 |
Link to this message
|
My computer and I are back. I think we need to figure out the first movie to test. I've gotten a few suggestions, so I'm just going to make a list here of the suggestions (plus any of my own) so we can figure out if we all have any of them.
High Bitrate:
Star Wars Episode 2
Aliens
The Two Towers
The Matrix (any of them)
Lower bitrate / Long Movie
The Green Mile
Animated
Shrek
Final Fantasy
Fantasia 2000
Make sure that if you're suggesting a movie it's a good transfer. I personally buy a lot of movies from the bargain bin, and they aren't always the highest quality video. I think that if we're trying to get the highest quality video you should assume a high quality source.
|
|
grogey
Suspended due to non-functional email address
|
13. May 2004 @ 07:04 |
Link to this message
|
No good chick flicks in there vurbal lol
I can get my hands on any of the LOR or Matrix ones in section 1.
Think you should choose a good chick flick instead for section 2.
I have Shrek & Final Fantasy for section 3.
Did I just add to to the confusion? Now for the medium bitrate we don't want anything with too much action huh? Here are a few that I have that might fit the bill: Robin Williams Live On Broadway, The Last of His Tribe, Sweet Home Alabama, Shakespeare In Love, Intolerable Cruelty, Fairy Tale - A True Story. They are just a few from my rack that I haven't already backed up yet so I'd get a side benefit out of the testing.
Grogeybear
My best advice for NEWBIES is: "use the search function.
THE BEST REVENGE IS TO LIVE A GOOD LIFE
Aim for Charm not Perfection
|
Staff Member
2 product reviews
|
13. May 2004 @ 08:06 |
Link to this message
|
|
I guess I should have added that the I have the Matrix movies, the first LOTR (I should buy the 2nd one anyway), Aliens, and Shrek.
[g]grogey: Have you checked to see how big movie only backups would be on those movies? I have a friend with the Robin Williams disc, and I think he did have to compress it to make a backup, but I don't personally know about any of the others. I'm looking for something with a significant amount of compression required for a movie only backup. I'd prefer at least 20%, which I guess would be 80% if you're looking at in Shrink.
|
|
grogey
Suspended due to non-functional email address
|
13. May 2004 @ 09:31 |
Link to this message
|
Hey vurbal you lookin' for big compression rate huh? Well I think this should suffice:
Screenshot of Robin Williams in DVD Shrink. I was very tempted to hit the back-up button but refrained!! lol
Grogeybear
My best advice for NEWBIES is: "use the search function.
THE BEST REVENGE IS TO LIVE A GOOD LIFE
Aim for Charm not Perfection
This message has been edited since posting. Last time this message was edited on 15. May 2004 @ 08:07
|
|
brian100
Suspended due to non-functional email address
|
13. May 2004 @ 10:26 |
Link to this message
|
|
Talking about movie selections
Ive mentioned to Vurbal that the problem we have is finding movies from different REGIONS that have EXACTLY the same content.It would not be too difficult with the "Movie only" option but the "Full Disk Backup" will be meaningless unless we can track down IDENTICAL titles. I honestly would not know where to start on this task. If anyone has any input or suggestions it would be greatly helpful.
Would it be a possibility to "circulate" the 3 movies between interested parties?.This would eliminate any possibility of region "variation" and allow us the chance to really source 3 excellent examples. I know this would invlove a lot of trust between all testing parties, but I am sure it would be the best option to all concerned. I am not sure if any copyright infringement would be breached. Whadya all think? This option would also allow vurbal more time to analyse the results rather than being bombarded with files left right & centre.
_
Looking for my old AD This message has been edited since posting. Last time this message was edited on 13. May 2004 @ 13:44
|
|
grogey
Suspended due to non-functional email address
|
13. May 2004 @ 13:38 |
Link to this message
|
|
Definite food for thought Brian100 - I hadn't thought of the differences in the actual content between regions.
Grogeybear
My best advice for NEWBIES is: "use the search function.
THE BEST REVENGE IS TO LIVE A GOOD LIFE
Aim for Charm not Perfection
|
|
Advertisement
|
|
|
Staff Member
2 product reviews
|
14. May 2004 @ 04:29 |
Link to this message
|
|
We should probably take this conversation to email so we can avoid cluttering the boards just talking among ourselves. If you haven't already sent me your email address, PM me with it and I'll get everybody a message today so we can hopefully get this jump started again.
|
