We Need Better Objective Opinions

It is really hard to compare the quality of a particular method by converting an entire video then viewing. Generally you’re going to think it looks very good whether you use CCE, DVD2one or InstantCopy. Even if you try all three methods and put them in your player one at a time you will probably not detect much difference. All three produce reasonable quality though not identical and also not identical with the original. A key point is there could be a fairly large difference in quality that is not discerned by the typical process of playing a video in your standalone.

Except in cases where there are obvious distortions (e.g. pixilation), the observer will not be able to define quality issues and will therefore not notice differences in quality in the time it takes to remove a DVD and reinsert one made by a different process. It is much the same situation when selecting audio equipment. You can’t really compare the quality of, say, audio speakers from memory of the sound between speakers when played at different times. Basically, you need to compare the performance in real time or as close to real time as possible.

For audio equipment the general solution is to A/B two speakers. This means the listener can rapidly switch between two audio components to determine which he prefers. If you simply listen all the way through and then again with the other equipment, the job of determining a preference is much harder and less accurate.

Many on this board are saying that InstantCopy must be better than DVD2one because it takes longer to encode and longer must be better. That is a poor argument which has little if any merit. By similar logic a bigger computer box must be better than a smaller computer box. Well I can tell you the size of many components continues to shrink at the same time that performance increases (Moore’s Law is a statement of a similar observation). This is not to say that I believe DVD2one is equal, better or worse than InstantCopy. I’m not sure yet.

To accurately compare results from CCE, DVD2one and InstantCopy (perhaps also the original) we really need all three versions playing simultaneously (i.e. 3 monitors). Bring in a random sample of objective viewers and ask them to rate each video versus the others that are playing simultaneously. Essentially, this is what you do when you buy a new television ... you compare the picture quality between sets in real time.

[Tilandra]

Your method is still flawed. What if one of the TV sets has their sharpness or color settings way out of whack? And "quality" viewed with the naked eye is still subjective... two people can still have very different opinions about what is a quality picture.

Me, I put my original in my Xbox, my copy in my DVD player, start both at the same time and flip between the two with my stereo receiver. Pretty simple and I'm using the same TV so I know it's not the set if I see any differences. The only problem with that is using two different players, but hell, I'm not *that* picky. Plus I can switch the copies and compare again just to be sure, if I want.

Why is everyone so emotionally invested in proving their favorite software is "the best"? Use what works for you and let others do the same. The only people who should care are the ones who want us to buy the software, and they should be listening to complaints and improving their products instead of trying to argue us into believing they're the best option. And if they're not providing demos so people can compare before purchase, shame on 'em.

[furball6969]

Another way would be to run a short, difficult-to-encode clip through each of the encoders. Using all of the same settings you should be able to take a snapshot at exactly the same frame then you can compare the results using a pic viewer like ACDSee.

@Tilandra
Your method may be very good. Why not post some results especially if you can arrange for someone to make judgements that does not know where the source came from.

[Tilandra]

While I'd love to under normal circumstances, I can probably only provide one person's opinion in a blind study, and that's not a good statistical sample. (My husband, and he's not very concerned with picture/audio quality. I can barely get him to sit down and watch a movie! Heh heh)

Anybody have two exact-same model players set up on an A/V receiver with a big group of videophile friends?

Here is a frame by frame comparison I found of DVD2one, InstantCopy and the original

The original DVD was 6,709,553,152 bytes and has 3 titles. InstantCopy was run with default settings, it produced a result that was 4,131,458 kb in size. Dvd2One manually on each piece so that the end result would be 4,543,627,264 bytes. The results were captured using DVD2AVI. The frames captured are not exactly the same but very close.

Read the thread where I found this:
http://forum.doom9.org/showthread.php?s=&threadid=46502

Or go straight to the comparison pictures:
http://members.cox.net/starfleet93/compare.htm

What do you think? Too close to call?

I am also curious. There was a lot of discussion on the Shimmering or sharp/blurry flickering that people observed with DVD2one. For many, this was the single biggest issue with DVD2one. Has anyone reported whether InstantCopy does the same thing?

[Marco33]

Another way would be to run a short, difficult-to-encode clip through each of the encoders. Using all of the same settings you should be able to take a snapshot at exactly the same frame then you can compare the results using a pic viewer like ACDSee.

This sounds like the best way.

I think a good movies to test is one that tests the limits of the disk space.
Like "Casino" 3 hours long. Not heavily graphic but long. I did it through DVD2DVDR with CCE and BOY... it had a rough time. It looked only marginaly better than VHS quality. But, 2 hour movies are cream cheese for CCE.

[Doggiedaddy]

Tilandra,
Nice to see on the dvdrhelp.com site, your input was always very interesting and helpful in the DAZZLE forum.

Dd
Strength and Honor
Good Luck
dvd9to5 me

http://www.dvd9to5.com

[Tilandra]

Golly DD, how'd you recognize me? :P

Good to see familiar "faces" in a new forum.

this comparison of DVD2one, InstantCopy and the original is impressive. Its in German but the charts and video are there for comparison. The bitrate chart for The Matrix shows that InstantCopy does a fairly good job of following the variable speed (higher motion) that was in the original. Compare the bitrate chart for King Kong and you see that DVD2one has almost completely lost the variability of bitrate while InstantCopy is looking quite good. Looking at the still pictures all l can say is that I prefer the original in each case.

Perhaps other opinions?

http://www.videoxone.net/mb1/instantcopy7/instant5.htm

This German review/comparison of DVD2One and InstantCopy shows there is no comparison based on bitrate use. IC is following the motion and DVD2one is not ... its flattening the whole bitrate curve and not allocating based on motion and complexity. Take a look how DVD2one is allocating bitrate for King Kong. It is bouncing up and down frame by frame and this is, I suspect, the root cause for the many previously reported "shimmering" effects seen with DVD2one films.

[Will Hay]

By similar logic a bigger computer box must be better than a smaller computer box.

With all due respect, that is just a terrible, terrible example.
Only an idiot would assume such, the comparisons for video quality is a concept far deeper than pure ignorance.
What you and so many others so easily fail to mention is one critical synopsis:
'personal preference'.
Many of us don't have 62" plasma-thingamybob screens, we don't wish to sit in front of our monitor screens with illegal copies of CCE and worry about the colour of subtitles.
We don't pretend to be video editors, some of us have an audience who prefer to chew dvd's rather than watch the extra's so a four second 10% pixelation in three scenes over three hours is not the end of the world.
I quite agree you in principle regarding you objective request, there are far too many posts containing the phrase 'this is crap, u r stupid to use dvd2one/instantcopy/whatever" etc. but these people are morons.
You can't educate a block of wood.
Every time I seem to post nowadays my signature seems to have more and more relevance
Will

[jopereira]

With all due respect, that is just a terrible, terrible example.
Only an idiot would assume such, the comparisons for video quality is a concept far deeper than pure ignorance.

I will not take a part of the discussion about A is better than B, but...

"Terrible example"?!?!? I would say "real example" !!!

Is that sufficient to choose A over B? That's a personal question everyone have to put to themselfs...

Most of the movies have 95-99% that don't care much about the encoding program but the >5% left make a REAL difference between progs...

I have seen movies where the original is already "crap" and the worst is that the dual layer disk isn't full...
"Emperor Grove ..", don't know the real english name, from Dream Works is a good example...

@silky31
I don't take offense to your comments but I hope you will recognize that I am only trying to be objective not take sides. In the end it is your decision and your decision alone what method to follow. I for one am trying to understand the differences between the methods.

Maintaining the variable nature of the bit rate should improve quality. What this means is that the encoder, transcoder, etc. is using more bits when there is a lot of motion or high complexity to a sequence of frames. When hardly anything is moving fewer bits should be utilized. This is what you see when you use bitrateviewer on an orignal film. So far, what I have seen DVD2one squashes this down so there is little if any variable bitrate left. Even worse it is jumping up and down frame by frame. This means that some frames have a lot of bits and the next has few. It is not proof, but I will hypothesize that this is what people have been describing as a flicker with DVD2one (also sometimes called "shimmering" or "blurry/sharp/blurry ..."). That is not to say that DVD2one can't get rid of this issue, they probably can. It is also not to say that InstantCopy is my preferred tool. I'm not taking sides, I being scientific and putting my emotions aside.

[Will Hay]

@silky31
I don't take offense to your comments but I hope you will recognize that I am only trying to be objective not take sides. In the end it is your decision and your decision alone what method to follow. I for one am trying to understand the differences between the methods.

Having re-read my posts I hope you didn't think that "only an idiot would assume such...(with regards your example)" was aimed at you personally.
I was trying to demonstrate that to determine a better quality picture you would have to take time to investigate. Anyone assuming a big computer box is better than a smaller one is basing their opinion on ignorance (perhaps) and niavety.
None of my other points where aimed at your initial post in the slightest, I simply find that all posts regarding the differences are not considering this simple premise of 'what's no good for you maybe fine for me' and thank you for post (I think it was something I'd tried to hint at in other posts but failed to articulate as well as you have).
Will Hay

In another forum int 21h
http://forum.doom9.org/showthread.php?s=&threadid=46526
gave the following speculation

About differences between InstantCopy and DVD2One*:

I believe InstantCopy is decoding frames all the way down to the macroblock level, and then re-encoding frames (MV and all), and then splicing in the new frames (by complex references and such in the original stream). I believe it uses some sort of critera to know when to splice in new frames.. possibly motion.. possibly original frame size, I haven't investigated enough yet to know. But this would definitely explain the differences in bitrate distribution between InstantCopy and DVD2One.

What leads me to believe this? Well consider for a moment the speeds we are all getting with the program... I'm getting about realtime on a P4 1.5ghz, which is about the same speed I get with CCE.

I did some research on VOB Computer Systems. This is the company Pinnacle acquired and gained right for software now sold as Pinnacle InstantCopy.

http://www.softlandmark.com/InstantCDDVD.htm

Encoding

VOB offers the following MPEG Audio/Video encoding DLLs for licensing. These encoders are also used in our InstantVideo application.

MPEG1 encoding (Video and audio) - compatible with VideoCD
MPEG still picture encoding - compatible with high-res VideoCD
MPEG2 encoding (Video and audio) - compatible with SuperVideoCD
DVD encoding (Video and LPCM audio) - compatible with DVD Video specification
This encoder is implemented as a DLL with a proprietory interface. The input format is DIB (any kind of video input can be used and converted into DIB) and the output is a standard compliant VideoCD stream.

The DLL works by call back functions and can therefore easily in integrated into other applications. The multiplexing of the stream is done on-the-fly.

Notice their proprietary technology for MPEG 1 and 2 encoding as well as their "on-the-fly" multiplexing of the stream.

Sounds to me that int 21h may be on to something.

int 21h goes on

What makes DVD2One so fast? Two things... assembly and Compressed Domain encoding. DVD2One doesn't take the video all the way down to the macroblock level, it doesn't decode MotionVectors (the most time consuming part of Mpeg-2 transcoding), instead it reuses them, and just increases the compression of the frames (It does this by some criteria that is unknown to me at this time, but it appears that it does it more often than InstantCopy, possibly by some type of math it performs to figure out how much to reduce).

so far it looks like the two companies are taking a fundamentally different approach to reducing the video size.

Found this nice comparison of DVD2one / InstantCopy on the doom9 board

http://forum.doom9.org/showthread.php?s=&threadid=46630

very close on stills but this doesn't analyze the "shimmering" issue.

also take a look at this thread

http://forum.doom9.org/showthread.php?s=&threadid=46502

another good quality study for instantcopy
http://forum.doom9.org/showthread.php?s=&threadid=46652

recommendation is to say at 70% of original size as a minimum, otherwise there is a fairly noticeable degradation in quality. The link lets you make your own judgement. Good stuff

[wulf109]

To be valid an A/B comparison should be done with one monitor with the DVD players hooked up thru a video/audio switcher. But you also need to take into account that different DVD players will effect the playback comparison. I see shimmering and macro blocking when I play with one DVD player but my other two players show no sign of these artifacts when I play the same disk.

[letgoofmyears]

I have three Apex 1100s, the 1 meg version, but I don't have them connected to the same receiver. I'll do this sometime soon, and also get a group of people "who couldn't care less" as test guineas. I'm off for the next three days, so I'll do this sometime before then and let you know.

[Nelson37]

Tests and comparisons should also be done without knowledge of which method produced which clip so that any prejudices for or against particular programs are eliminated.

It is difficult or impossible to evaluate a series of moving pictures based on a few milliseconds of still pictures.

I personally do not care in the slightest for technical explanations which purport to "prove" which method is, or should be, best. While these can be illuminating and informative, and even useful, they "prove" nothing. The best method is the one which produces video that I find more pleasing, technical "proofs" be damned.

Is it possible to "prove" that blueberries taste better than strawberries? The only one who would attempt to do so would be someone who has blueberries to sell.

[mattcham]

There was a lot of discussion on the Shimmering or sharp/blurry flickering that people observed with DVD2one. For many, this was the single biggest issue with DVD2one. Has anyone reported whether InstantCopy does the same thing?

From my own comparisons of DVD2ONE and InstantCopy7,
the shimmering blocks on InstantCopy7 are worse.
They are pretty bad in DVD2ONE, but they are worse
with InstantCopy.