Is there a 'best' DVD backup program?

What’s the best program to backup my DVD’s?

Sounds like a simple enough question, lets see if we can provide some answers. Believe me, I’m no expert; I’m in an intermediate range when it comes to DVD transcoding. But let’s face it there seems to be unnecessary division among the ranks as to good…..better…..best, which makes for tough, biased choices if you’re a newbie. We all have, and are entitled to our own opinion, but can we safely shoot our mouth off about our ‘treasured applications’, without shooting ourselves in the foot? Is their really a best program, and method? Are they all the same, or at least so similar it doesn’t matter? Is it maybe just subconsciously we prefer a certain interface, status bar and catchy icon on the desktop?

To begin I downloaded some of the most popular available trials- DVD2one, Clone and Any DVD, Intervideo’s DVD Copy 3, Nero Recode and lets not forget those freeware beauties, DVD Decrypter and DVD Shrink. Most of these programs require Any DVD or a similar Decryption driver to work correctly, or conversely you can simply rip with DVD Decrypter then encode. (Believe me all you DVD RB Fans I did want to include this app, but it involved a little too much, considering that it all comes down to the encoding engine used, and being that there are 4 different encoders to choose from, 2 freeware and 2 shareware. So I gave it a miss, but am currently studying up on it.)

What The Figures Below Mean

To understand what the figures below mean we first need to understand the terminology. Bitrate: defines how much physical space one second of audio or video takes in bits. (Not to be confused with bytes*) Therefore the Peak Bitrate is mostly inconsequential to the quality of picture. It’s the Average Bitrate were looking for to define the overall quality. So the higher amount of useable data there is to read per second will equate to a better quality picture. So what we want is a higher Average Bitrate.

The two types of bitrate are: CBR or Constant Bitrate, and VBR or Variable Bitrate. CBR: Means that the bitrate doesn't vary during the video or audio at all, but is the same throughout the clip. This means however that sometimes bitrate is wasted, as it will unnecessarily make parts of a clip a higher bitrate than they need to be (eg credits or black parts of a film) and others lower than what they should ideally be. (eg bright outdoor scenes and digital special effects) It does however make it easy for an appliction to accurately predict how much space the total clip will take up and will generally encode more quickly. VBR: Means the bitrate fluctuates to whatever bitrate the application calculates is required at that particular moment, therefore being as close to the original as possible, through use of a “Deep Analysis” function.

“The best way to understand why this is used is to think of a movie -- when there are shots that are totally, absolutely black, like scene changes, normal 1-pass CBR encoding assigns the exact same amount of data to that part as it would for complex action scene. By using VBR and multi-pass, the encoder "knows" that certain piece’s of the clip are Ok with a lower bitrate and so that extra bitrate can then be used for more complex scenes, thus creating better quality for those scenes that require more bitrate.” (taken from Afterdawn.com’s Glossary)

Quantization is somewhat complicated, and does have an impact on overall picture quality. “Quantization techniques generally compress by compressing a range of values to a single quantum value. By reducing the number of discrete symbols in a given stream, the stream becomes more compressible.” (Taken from datacompression.info) This number can describe the quality level. From my research I’ve found that the higher number declares higher compression - worse quality, the lower number means less compression - better quality. Some encoders use the standard linear Quantization tables, and others use their own custom, non-linear tables. But nevertheless while different encoders use different algorithms and tables to compress, the main compression factor depends on the Quantization table stored in the originals stream. So with the Quantization figures on the table below, we’re looking for a lower value.

Time Taken

To some this can be an important factor, as no-one wants to wait for extended periods of time waiting to transcode one backup after another. But are they really that slow? Or are us Human Beings just getting more impatient? Shown below are the times taken by each application on my system**. As we have found though DVD Shrink is quick for some and slow for others so please take all these times with a grain of salt, and treat them all as estimates. One thing I did find is DVD Shrink can be significantly slower when reading directly from disc. (Even with Any DVD) By ripping with DVD Decrypter and reading from the Hard Drive you can reduce time taken, although the time taken to rip almost negates the saved time itself (LOL). Left out are burn times, as we all have varying speed burners, with times that are pretty much set in stone.

Nuts and Bolts

What did Bitrate viewer reflect of samples provided by each product in question? The samples, were the entire movie uncropped. No Menus or special features were included. Cropping a movie is the extreme enthusiast’s way to surrender more physical space for a higher Bitrate. By removing start up promotional footage, intro’s and end credits, more space can be utilized for picture quality. This plus removing of subtitles and unnecessary audio streams can free up a few hundred Mb’s as well.

(note: all apps were current versions)

Using the movie: “Titanic” (title only, high compression, roughly 50%, with no removed audio and no removed subtitles) Total size= 7250Mb

BP-Bitrate Peak
BA-Bitrate Average
QP-Quantization Peak
QA-Quantization Average

DVD2ONE (in conjunction with Any DVD)
VBR BP-5599 BA-2076 QP-8.69 QA-3.97 Time-17:48

DVD2ONE (in conjunction with Any DVD)
CBR BP-3652 BA-2011 QP-8.91QA-3.97 Time-19:27

Clone DVD (in conjunction with Any DVD)
BP-6297 BA-2024 QP-9.33 QA-4.33 Time-18:09

DVD Copy 3 (in conjunction with Any DVD)
BP-4410 BA-1847 QP-8.33 QA-3.24 Time-24:12

DVD Shrink (in conjunction with Any DVD, using Deep Analysis)
VBR BP-4558 BA-2001 QP-8.80 QA-3.65 Time-49:27

DVD Shrink (in conjunction with Any DVD, without Deep Analysis)
CBR BP-4167 BA-1994 QP-8.95 QA-3.64 Time-25:08

Nero Recode (in conjunction with Any DVD, using Deep Analysis)
VBR BP-4613 BA-2006 QP-9.78 QA-4.72 Time-42:18

Nero Recode (in conjunction with Any DVD, without Deep Analysis)
CBR BP-4167 BA-2023 QP-8.95 QA-3.63 Time-16:48

Uncompressed Movie- BP-7899 BA-4644 QP-6.88 QA-3.23

1 Click DVD (in conjunction with Any DVD)
BP-5458BA-2133QP-8.56QA-3.93Time-25mins

Clone DVD/DVD X Studios
BP-4899BA-1887QP-8.05QA-3.22Time-30mins

DVD Fab Express
BP-6038BA-1852QP-8.86QA-4.06Time-21:01

DVD Decrypter Entire Disc Rip Time-14:37


DVD Rebuilder
DVD Rebuilder is a very different program indeed. I’ve read from various posts It makes use of an ‘encoding’ engine as opposed to a ‘transcoding’ engine.* It can use single or multipass VBR, (dependant on user settings) first analyzing the DVD then stripping it down and encoding each individual cell # according to what the analysis has determined. It then rebuilds the cells and structure of the DVD.
Setting up DVD RB isn’t as hard as it used to be. You should go here first and download Avisynth. You’ll need to install it before DVD RB which you can get from Here This current DVD RB install pack pretty much sets itself up. You only need to specify which encoder you wish to use. This installer comes with QuEnc and Rejig included and ready to go.

Nuts and Bolts using DVD Rebuilder

DVD Rebuilder (in conjunction with Cinema Craft Encoder Basic, 2 pass)
BP-5738BA-2018QP-7.49QA-3.09Time-369 Min

DVD Rebuilder (in conjunction with Cinema Craft Encoder SP, 4 pass)
BP-3546BA-1859QP-9.10QA-5.89Time-780 Min

DVD Rebuilder (in conjunction with QuEnc, 2 pass)
BP-5875BA-1944QP-7.76QA-3.19Time-484 Min


Overview

These results above may make you gasp at first. (as I did) I then looked at comparisons from a transcoder to an encoder in a different light. I looked at the graphed results, as well as the numbers. Look below at the Bitrate to Quantizer scale using CCE Basic compared with Clone DVD (Elby) and DVD2One as examples.

Green- Quantization Scale
Yellow- Bitrate Scale







DVD Rebuilder using CCE Basic

Clone DVD

DVD 2 One(using VBR)


So what we can gather from this is DVD Rebuilder via an ‘encoding’ engine is a much more precise tool than any of the one click ‘transcoding’ solutions. The conformity of the graph from DVD RB via CCE Basic shows how accurately it compresses the data dependant on the bitrate it selects at a certain point. These two lines follow each other almost perfectly from start to finish. Looking at Clone DVD and DVD2One the graph is very messy, the two lines cross each other at many points and have enormous gaps between at some points. And not to forget higher peaks make for higher averages, and the peak is generally meaningless. What we could assume from this is: even though DVD Rebuilders average Bitrate is lower, and its average Quantization level is higher, the quality of picture should be better. This is probably only discernable with a High Definition TV. It would however seem from the numbers above that comparing ‘Transcoders’ to ‘Encoders’ is most likely pointless.

*Definition of Encoding according to Videohelp.com- Encoding is the process of changing data from one form into another according to a set of rules specified by a codec. The data is usually a file containing audio, video or still image. Often the encoding is done to make a file compatible with specific hardware (such as a DVD Player) or to compress or reduce the space the data occupies.

Definition of Encoding according to Afterdawn.com- Encoding is opposite of decoding. Encoding means that a file, whether it is an audio, video or picture file, is compressed to another format that normally takes up less physical drive space than the previous format.

Definition of Transcoding according to Videohelp.com- On this site generally another name for encoding. A more technical term would be "The reformatting of content, without changing the source, to another type of content - most often of a different format than the original (but does not have to be)"

Definition of Transcoding according to Afterdawn.com- Transcoding or more specifically Compressed-domain Transcoding means normally a re-encoding process that changes the video or audio features, such as resolution or bitrate, by changing parts of the a/v content, but not by reconstructing the content again (which is the case in encoding process). Compressed-domain transcoding also maintains the format of the file same as in the original file.

#Cell- A 'Cell' is a small segment of a chapter (or part). It is the smallest resolution at which DVD navigation commands can act (e.g. 'Jump to Cell 3 of Part 4 of Title 2'). Typically one chapter contains one Cell but on complex DVDs it may be useful to have multiple Cells per chapter.



* Bit = a zero or one, that makes up data and commands in the Computer world. Short for ‘Binary Digit’.
Byte = a group of 8 bits, a byte is the common measurement of space and data in the
Computer world.

** My system: XP2600 @ 2249Mhz, MSI K7N2 @ 346FSB, 1Gg DDR RAM, IDE ATA133 HDD, NVIDIA FX5600 375/625 128Mb. Reader Sony DDU 1612.

Quality Verdict

I’ll now leave that up to you. This test was not done with the aim of selling you a product or telling you what to buy. It was done as an unbiased study to present some facts that may help you understand all the parameters that come into play, therefore allowing you to make up your own mind.

Time Verdict

Again the verdict is up to the reader. Times can vary quite a bit from system to system, so realistically the times I’ve posted here may be far from what you might get.

Price

Lets not forget the all important dollar, or whatever your currency might be.

DVD Decrypter- Free!

DVD2One- $51.00 USD

Clone DVD- (combined with Any DVD) $59.00 USD (Separate: Clone and Any DVD $39.00 USD each)

DVD Copy 3- $99.95 USD

DVD Shrink- Free!

Nero Recode- $50.00 USD

1 click DVD- $99.95 USD

Clone DVD from DVD X Studios- $99.95 USD

DVD Fab Express- $39.95 USD

DVD Rebuilder- Free!

QuEnc Encoder- Free!

Cinema Craft Encoder Basic- $58 USD

Cinema Craft Encoder SP- $1950 USD


In conclusion, what can I say? I did this simply as an exercise for myself, but felt maybe I should share it with a forum that has shared so much with me. Remember I’m no expert, if I’ve missed something please point it out, if you disagree don’t hold back. I researched as well as I could, all readings were taken twice and if necessary an average was made, to be sure there were no major anomalies. Standard DYS3 firmware was used.

I hope I’ve been of some help.

Pollushon.

(To view the results as a table in word.doc format click the link)

---

Last edited by Dee-ehn; 25-03-2005 at 00:26.

DVD rebuilder with CCE has come along pretty far lately. Unfortunately, you still need to do all of your preprocessing before putting it into DVD-RB, so it can't be considered a complete backup solution (yet). But, I have to say, the quality is simply amazing. A drawback is the sheer amount of time it takes to encode. Two pass mode takes about 2.5 hours on an AMD64 3000+. Again though, other programs are needed to remove titles and such.

Pollushon > Nice. Only thing i would have maybe said along with price...you might have added which have good or bad or no support. Thanks for the info.

good work pollushon.Should give newbie's and freaks alike something to think and read about.

Nice job! From your tests, it's difficult to declare a clear winner. It will be harder to distinguish with real time playback. Personally, I would re-use Titanic instead of Scream 3 for the 2nd round. This test should reflect what an experienced user can do to tweak each application to yield the BEST possible picture.

-Remove ALL video titles except for the main video title.
-Remove ALL audio titles except for English Dolby AC 3/6 or 5.1.
-Remove ALL subtitles.
-Remove the lead-in at the beginning of the movie (Fox, Universal, Sony logos, etc...)
-Remove the end credits at the end of the movie.

The goal is to free up as much space as possible for the main video title. This should dramatically reduce the compression level of the main movie. Note that that the audio is not compressed. Shrink and Recode should come out on top because these applications can edit the start and end of the main movie title. DVD Copy 3 should rank at the bottom because this application does not remove the unselected audio stream from the final compilation.

Nice Thread pollushon...Great Read!!!

Have to agree nice work pollushon this type of thread is far better than the usual "my program is better than your program" at least here you have supplied info and facts...

Good job. IMO I'll save the $50-$100 and keep using what has worked for me.
Dvdshrink and Dvddecrypter

Great post and very beneficial to a newbie just waiting for my dvd-rw to get shipped. I had my doubts about the divide between using the freeware and software you paye for. I will feel alot better now using shrink/decrypter now, thank you pollushon

*third time i wrote this...that 'forum jump' address box being highlighted as default is real annoying as it jumps off the page when you scroll down (habit of scrolling)

Excellent job. I'm very interested to see your results with rebuilder CCE. I'm getting a 50" HD tv this week and curious to see how much better the encoders are than the transcoders on the big screen. Please don't be so humble, would love to hear your opinion on video quality.

Yes by all means it was a most excellent report.

However with quality DVD blanks now at $0.30 each, and artifacts from any compression visible on large HDTV screens (for those with fussy eyes), I simply use two DVDs where necessary. Burns are fast, complete, and getting up partway through the movie to change disks gives me an excuse to grab another beer.

@pollushon

Good idea for a sticky, but a couple of comments:

If you are going to examine DVD Rebuilder, you should include Hank's new HC encoder along with CCE and Quenc. From informal first looks, it is looking awfully impressive going up against CCE.

Probably a more visually complex (fast moving action) movie should be used instead of or in addition to Titanic and Scream 3 since since transcoders seem to have more trouble with that type of material.

And last but not least, perhaps you should bold face or otherwise draw more attention to the quality conclusion area. Plenty of noobs will just see numbers and draw unwarranted conclusions about their relationship to overall quality.

@ furballi

I disagree about ripping everything. Plenty of people rip minimally or not at all- either because it's easier or they want the 'full experience'.

And more importantly, these transcoders should be pushed to the limit. Now that there are more and more two-disc releases with all extras relegated to a separate disc, not to mention superbit releases where the studio itself tries to maximize bitrates for hardcore videophiles, it is easily forseeable that even after ripping out everything you listed, you might still need compression down to 60% or 50% of the original size. It would be a rude awakening for someone who picked their transcoder based on 90% or 80% tests.

Interesting work with the quality stats with different programs. Although I would say that what people regard as "best" is too general to be definitive. Reminds me of a random quote from my mIRC quit messages...

"All generalizations are dangerous, even this one." Quoted in Evan Esar (ed.), Treasury of Humorous Quotations - Alexandre Dumas 1803-1870

i agree that purposely trying to minimize compression isn't going to prove much at all in terms of seeing which transcoder produces the best video quality. i'm sure that all of the transcoders will produce excellent results when compression is minimal. only higher levels of compression will reveal which transcoders are most capable of reproducing the original image.

as for a more complex movie, i think that Saving Private Ryan would probably be the gold standard when talking about movies that are hard to compress.

Well, realistically I'm looking more for the encoder method vs transcoder method than putting encoders head to head, maybe later but QuEnc and CCE Basic/SP will more than suffice, plus they are widely used.(although only the rich can use CCE SP)

Yea, I agree, I'm going to get rid of Scream 3, the main reason for that choice was average compression, and the fact its a dark movie. I'm thinking Lord of the rings would have been a good choice. Titanics a good piece because it has it all: high compression, vivid bright scenes, fast moving light and dark scenes, and I've noticed that water and reflections can really put a transcoder to the test.

If thats the case it means they havnt read the whole piece, I hate to quote myself:
If newbs cant read that and then apply that to the figures above.....well they really should give transcoding a miss And that is the exact point of this exercise.
An excellent suggestion, similar compression to titanic with plenty of action.
Thanks a lot for all your comments. Stay Tuned

One problem is that the intentional grainy look that Spielberg chose gives transcoders and encoders fits, but is not present in very main mainstream movies. Might be a good test of abilities, but the ability to extrapolate and generalize from the results would be in doubt.

The point of your exercise has been explained already in your post but you've produced a test that focussed on information about bit-rate versus time taken, and cost. That will not measure how well individual transcoding engines handle different video content. Also the very nature of an individual's requirements means that you need much more information in order to produce a truely useful comparison.
Once you have a table with feature-sets on these different programs; discussions about ease for beginners versus features useful for intermediate to advanced users; discussions on how well the transcoding engine coped with different types of scenes along with frame samples and measured output over a broad range of movies, only then people will be able to draw better conclusions about "best". I'm certainly not trying to disparage your work but point out how these tests can be expanded to be of reference in a more comprehensive work.
Having said that, of course I commend you on your tests and hard work as there are few other comparisons such as this (probably due to level of commitment and time constraints) Trouble is, once you start something like this, it's going to be annoyingly difficult to stop!
One last thing: these programs change with each new version, so it's immensely difficult to keep up-to-date with all the different solutions.

Don't get me wrong mate, you've done a sterling job and if this is just the start I'm really looking forward to more pollushon reviews!

Take care

TZ

Kind of, I just thought I'd add time and cost in, to me though time is never a factor, cost can be to a degree.
What your talking about here are things that can be deep and advanced.(not for total newbs) I get the feeling maybe some advanced experienced users missed the point I was trying to get across with this exercise. Or its probably your thirst for advanced facts and figures that leaves you feeling a bit dissapointed. More movies, reviews of interfaces, ease of use, cropping, customer support, bugs etc.... would have definately made it a full review, but the absolute and only point at this time that I was trying to make and a point that will always be the case is as Furballi put it:We have had wars, sometimes flame wars on the "best" transcoding software, when in short there is none as far as what they do. Higher Bitrate equals Higher Compression and vice versa. I want newbs to look at these figures and draw the conclusion: 'its what I feel most comfortable using'. I think maybe 'best' was the wrong word to describe the exercise, because as your stating best would entail all possible factors, plus the addition of final results across a broad spectrum of films. But yes I do believe everything your saying is completely correct, and yes my time constraints are really starting to limit me. I start a new job on the 29th and am studying an MCSE 2003, but given time I think I will expand explore and test, because the review your describing is the review I would really love to do. But for now its plain and simple for newbs, to hopefully draw the conclusion: They all do the same thing, as far as quality there is no best, I'll use the program that appeals to me.

*wipes brow* Phew! Thought I was going to get nailed to the ceiling by the proverbials for my post. I wish you the best of luck in this project and will offer as much support as I can. Speaking of which, ever thought of delegating some of the work to others?

LOL, who am I to nail proverbials for honest, constuctive feedback. I think what I'm seeing is a growing number of bright minds coming into this area of the forum.(DVD Backup) As far as delegation...I'm just a member not a mod or admin, so anyone (hint Furballi) feel free to PM me any facts or figures (unbiased of course) you think should/could be added to the first post up top. I have an insider who can insert it. Dont worry you'll get your name up in lights

it's exactly for this reason that it'd be a "true" test of a transcoder's/encoder's abilities. it's basically a worst case scenario that would make artifacting much more pronounced.

My gold standard test for transcoding software is the movie Gladiator. Very long, lots of action, and beats transcoders to a pulp. Saving Private Ryan is certainly a good one, too. Any of the Lord of the Rings movies would also be good choices. Very good info though! Nice job!!

Bearing in mind that I have only the foggiest grasp of how graininess makes life hard on transcoders in the mathematical sense, I always considered Saving Private Ryan more of an outlier test- akin to testing airbags by colliding a vehicle with an M1 tank, or testing hiking boots on Mars (worst case scenarios as you put it). If it aces the test, then yes, it will handle damn near everything you throw at it. But if not, then you really can't draw conclusions about it's performance against lesser obstacles since the test was so outside the realm of normal occurence. My understanding (and correct me if I am totally off-base) is that different problems (excessive grain, fast motion, low starting bit-rate) are different in character regarding how they challenge the transcoder/encoder; and that since excessive graininess is so rare in movies nowadays (excepting the unique cases when it is intentional), how a transcoder/encoder deals with that particular obstacle is somewhat less relevant than if there were inherent correlation between a transcoder's abilities to handle those disparate problems.

$.02 deposited.

The more grain in video, the harder it is to encode because grain = noise. Noise is random in nature and random data is difficult to compress. Random data has few repeating sequences that compression techniques rely on. The most extreme type of random video data has to be video noise; like when a TV without auto-non-signal detection has the source feed cable pulled out. It's a real pain to encode. Scenes that feature, for example, a scene transition with video noise require such high bitrates that an mpeg2 encoder like CCE will often balk unless sufficient bitrate is allowed for in the encode.

the point of using Saving Private Ryan wouldn't be to actually try to successfully reproduce the quality of the original, but to simply see which transcoder/encoder comes the closest (although i suspect the overall results would be pretty bad). it's for the very reason that it's an extreme case that it will correctly gauge how disparate the transcoders/encoders are in terms of quality. the "bad" programs will produce totally horrible results, while the "better" programs will produce more watchable results.

as for why grain is harder to compress, TigerZai basically explained why, and in fact, ANY compression problem is pretty much related to grain/noise. the actual type of scene being compressed has no direct relation to how hard it is to compress; it's just that fast motion happens to be grainy because of motion blur, and therefore it's harder to compress. testing how well a transcoder/encoder handles intentionally grainy scenes is no different from seeing how it handles fast motion scenes.