'Disney's Hunchback of Notre Dame' R1 vs. R2
Bjoern Roy , April 7th, 2002

First thing i have to say: I still love this movie. Didn't loose any of its charm. The songs are just beautiful and the story is pure magic.

I think most people will be happy with the PQ on this disc. Yet, i have a few complains.

Non-videophiles should probably stop reading here.

The main problem, which induces every other, is grain. There is simply too much for an animation type movie. Its not intentional here, like in AI or Pi, its simply due to the age or quality of the print used.

Put into perspective:

Pocahontas (1995): a grain desaster, the transfer is terrible
Hunchback (1996): less grain than Poca, more than Hercules, though
Hercules (1997): not perfect, but quite ok actually
Mulan (1998): about the same as Hercules
Tarzan (1999): non-issue
New Groove (2000): non-issue

So the grain seems to get less, the newer the release. Thats to be expected, but neither of these are 'old' movies. I mean 1996?

Although i prefer the clean look of Tarzan and New Groove, I found the grain in Mulan and Hercules acceptable. The amount of grain in Hercules is borderline.

The grain is not so bothersome in itself. Its enough to lead to bitrate starvation, though. Meaning: grain is random noise, it sucks up bitrate without encoding any actual information. A 7.0 Mb/s transfer from a grainy source is going to have more compression artefacts than a 4.0 Mb/s transfer from a clean one. See Fast and the Furious, low bitrate, no grain -> no compression artefact.

I didn't measure the bitrate but its very obvious that its simply not nearly appropriate. Apart from a few scenes, there is only a bit of vertical edge enhancement in the transfer and almost no horizontal, great. Yet, the edges are far from being clean. They are soft and often undefined. The bitrate starvation leads to lots of compression artefacts in many scenes.

Many edges (and traditional animation has lots and lots of edges) have so much mosquito noise that it looks like EE ripples to the untrained eye, yet they are not in this case.

Many of the solid colored areas (and traditional animation has lots and lots of solid colored areas), like faces in semi-shadow show rather severe blocking artefacts in form of posterization and banding.

From 1.4 times screen width away, i found this very difficult to watch at times. So i went into the second row of seats (~2 times width), and many of the issues got much more acceptable. At 2.5 times width this might actually look like a good transfer, so i think most casual viewers will find this transfer just fine.

Every other aspect of this transfer is almost perfect. I found the colors to be dead on, the shadow delineation is great, so is the contrast.

Overall, i slightly prefer both Mulan and Hercules to this transfer, although they are non-anamorphic. As i said, Hunchback could be a tad more detailed, Mulan and Hercules are not far behind in actual detail. Yet they have a fair bit less grain and AND probably more bitrate to fight that grain, since they have less data to compress (non-anamorphic), and less extras on the disc.

Neither of these transfers are even close to Tarzan and New Groove.

So my rating would actually be in the order of their age. Again, i find this rather puzzling, because i don't consider 5-6 years to be a long time for a film print. Maybe there where considerable improvements over the last years in the animation to film techniques that lead to less grain even on the negatives. Thats my best guess anyway.

What makes me angry is the fact that the video bitrate is so obviously insufficient, yet they include a DD5.1 track, a DTS5.1 track, 2 DD2.0 tracks (if i remember correctly) and a commentary track. And then they even cram all the extras in.

It would have been better to have a DD5.1 track and the commentary only, with problematic material like this. And the extras on a second disc, even if its not enough to actually fill it. Thats what i like about some of the Vista Series titles. They use a second disc for extras, even though they aren't that plentyful. Bitrate considerations on the first disc being a primary concern.

Even better, they should have used a better film print, if available anywhere at any cost, or even the negative.

Anyway, apart from the compression issues and with a proper seating distance, or less revealing system, this will satisfy most.


R1 vs. R2 screenshots comparison

I was curious to see whether the R2 (german) DVD release exhibits the same problems with grain as the R1.

Here we go.

The R1 shots are always on the left or at the top. The R2 shots on the right or at the bottom.


Scene 1:

As you can see, there is some hefty edge enhancement in the opening logo in the R1 version (left), while the R2 is a lot smoother (yet not less detailed!). The EE is not nearly as bad in the movie itself. In comparison to the grain and compression artefacts, the EE is almost a non-issue. You can also already see here, how the grain leads to some hefty compression artefacts, especially mosquito noise around edges. The R2 only has a bit of posterization in the blue background, but looks fine otherwise.


Scene 2:

Same thing here. There is a LOT of grain in the R1 (top). The bitrate is not enough to handle it, so the grain doesn't look very natural. If you look close enough, you can see that the grain is exaggerated to lots of quantization noise. The R2 looks a lot smoother in comparison.


Scene 3:

The red arrow points to the 'mosquito noise' around the hand. Look at the silhouettes of both characters and note that there is more noise around them as there is in the plain sky, for example. Whenever there is more noise around edges or fine detail, its a compression artefact that is called 'mosquito noise', which is caused by insufficient bitrate to handle the detail in the picture. You can also see the mosquito noise around the 2 guys in the R2 version, but at an much lower overall level.

The yellow arrow points to a row of vertical dots. See how they are much finer resolved on the R2 version (bottom), thus indicating that the R2 does indeed have more vertical detail than the R1. So the lesser amount of grain on the R2 transfer is not simply achieved through heavy low-pass filtering of fine detail. The R2 is more detailed AND has less grain. The result is a more film-like image.


Scene 4:

Same thing here. There is a huge amount of mosquito noise around every edge that is in the picture. The bitrate simply isn't nearly enough to fight the grain.


Scene 5:

Again, same thing here. But this scene is a good example that the bitrate on the R2 is not really sufficient as well. While the R1 has the typical noisy look, the R2 uses such a low bitrate in this scene that there is severe posterization in the green shirt. There are a few scenes on the R2 version, where this posterization in darker areas is quite apparent, more than on the R1. But its not nearly as much of a problem as the graininess in EVERY scene on the R1.

Here, I increased the color saturation of the shot, to better demonstrate what the posterization looks like. The inverse 'F' on the R2 (bottom) is due to color quantization, which leads to a 'posterization' effect. Note that there are also some blocking artefacts in the R1, but at a finer level, due to the noise.


Scene 6:

As i wrote in the review above, a large part of the bitrate on the R1 is simply needed to encode the grain (which is random noise) and not actual picture content. This means that there isn't enough bitrate headroom when there is a large amount of motion in the scene. This bitrate starvation leads to horrendous compression artefacts on some scenes. The scene where this was most apparent is when people start to throw fruits and other stuff at Quasi. Every compression artefact know to men can be seen in some of the shots. Look at the example above. YIKES!



Conclusion and final thoughts

If the lesser amount of grain in the R2 version is due to heavy use of digital noise reduction (DNR), which usually is just as bothersome EE is, then its actually a wise decision in this case. The bitrate on the R1 simply can't handle the grain, period. I didn't notice too many DNR artefacts on the R2 apart from the posterization in some scenes, which could also simply be due to the low overall bitrate. The only way they could have handled the grain better on the R1 would have been to get rid of the extra soundtracks/commentaries and all of the other extras on the first disc and to put them on a second disc.

Or they should have not transfered the movie from a film print at all and use a direct digital transfer, since most if not all animation pictures are said to be archieved in the digital domain in Disney's vaults anyway. Maybe doing that is a little more 'uncommon' or even more expensive to do, but the result would have been a much better transfer of one of my favorite animation movies of all time. Too bad.

So, will i keep the R1 or the R2 DVD? I will keep the R1. The 4% PAL speedup of the R2 version is killing me. These are some of the most beautiful musical acts ever produced. I take the somewhat grainy picture on the R1 DVD, which is much less of an issue in my second seating row (~2 times screen width), over the ruined pace and tone of the R2 DVD. If this had been a normal animation movie without musical performances, i might have chosen the R2, which i do very seldomly.


Best regards
Bjoern Roy



Back to Review Main Page

WEBCounter by GOWEB WEBCounter by GOWEB