A Few Words About A few words about... The Wizard of Oz (1 Viewer)

[dpippel]

The screencaps from the new version definitely look warmer *and* softer to me than the screencaps from the current release.

 

[Mark B]

This is the first time I've seen Garland's skin tone look right. She's always been too pink in the past.

 

[Robert Harris]

If I might suggest a slight format change. Images of the same size would be more easily compared for detail and registration. And this makes little sense anyway on a web site. To properly view registration anomoies, one really must be viewing in 2 or 4k Cineon files, each of which could be over 100mb. What we're seeing here are highly downrezzed images, and detail has already been lost.

The black and white sequence is rather irrelevant to the discussion. It is what it is.

I'll also make another point here which will undoubtedly set professional calibrators to light their torches, but the overall color really isn't that important. The point to consider is scene to scene correction, which in this DVD is superb.

Not only does the eye adjust to a point or yellow or red, but unless a setup is professionally calibrated on a continuous basis, it will tend to drift.

What this means is that in one's home viewing environment, the image is apt not to match what one is seeing on a computer screen.

The bottom line here is whether the image is pleasing in an overall sense, outside of a highly calibrated professional environment. I believe the anwer is "yes."

RAH

 

[Brandon Conway]

Which is sepia.

Just playin' with you, Mr. Harris.

I see a very familiar discussion along the lines of Ben-Hur coming later in this thread....

 

[DaViD Boulet]

I think what RAH means is that the BW segements are not comprised of 3 discrete merged/aligned elements and so there's no registration issue and therefore no gains to be made using the "ultraresolution" process which only has meaning when applied to 3-strip technicolor film.

Agreed we need to see the screen caps the same size (why are they different anyway...all (NTSC) DVD is 720 x 480..shouldn't they already be the same???) and often I find that screen-captures can be misleading...it's seeing those images *in motion* on a wide-angle high-resolution display that really tells you what they look like. I'll gladly watch the 4K digital master if anyone wants to send me the file and a 4K projector to review it on...

 

[Wes]

Just another closer look and comment/question, why do you suppose the Witches eyes are bloodshot in the new pics and white in the old?

Wes

 

[Robert Harris]

Because the image has a couple of points of yellow, and a bit of additional chroma. Standard brew for witch eyes.

Try holding a soft yellow filter over the smaller image. Color correction...

which is probably a misnomer here as finding what is "correct" is not easy. Pleasing is more likely.

As a hint to finding a higher resolution image, look at the highlights in eyes.

They've done a beautiful job.

RAH

 

[ScottR]

RAH,
I know we were talking about Academy Ratio cropping in another thread. I noticed a little here, as well. Would different theatres have different maskings or cropping, making this a non-issue?

 

[Mark B]

I wish they'd just windowbox academy ratio dvds so that nothing is cropped....

 

[Robert Harris]

There is no such thing as a standard theatre cropping in the real world. There are SMPTE standards, which can be viewed by returning to one of my earliest Bits discussions.

Beyond that fact, there is ALWAYS cropping. Three-strip Technicolor productions were imbibed from their matrices in a standard flat 1.37 format, with slight difference from camera aperture to camera aperture.

Printed over that image was a hard matte in black and white to cover production artifacts which were not meant to be seen, as well as the splice lines.

Beyond that, and unless a theatre had an absolutely perfect throw, one had optical problems of one sort or another.

Imagine a theatre with a twenty degree downward angle from the booth, and a 100 foot throw.

In order to project a rectangular image, an aperture plate would have to be cut in the shape of a trapazoid to alleviate the problem of the angular throw.

Even without windowboxing, the standard DVD image, no matter how cropped (within reason) is far better than what one would have seen in a move palace.

To experiment with this, simple cut a rectangular matte, tape it over the end of a flashlight, and angle your image.

RAH

 

[rich_d]

Robert,

If overall color isn't that important, they why bother obtain paint samples for Madeleine's green Jaguar in Vertigo?

All,

Certainly color continuity is important but I see no advantage for not doing the best job one can on getting the color right.

Viewers may adapt to a film but they still notice when colors they expect to see are not what they expected. Green velvet should look like green velvet, etc.

Agreed that the images comparisons should be the same size image as others have noted.

Herb,

Were the images captured off the same computer and with the same software?

 

[ScottR]

RAH, thanks for that explanation.....I can't wait to get this!

 

[Wes]

I know, I know my computer monitor is not calibrated but the more I look at these pics the more I dislike the new transfer.
Look at a few more areas, The Lions hair is a nice rich brown color in the old shots and has a golden highlights in the new. His face is almost yellow in some areas where the old is alot more natural (as a man painted to look like a lion can be I guess).

Back to the house with the ruby slippers, The old has a definite tint of Blue were as the new has a slight feel of yellow and green. I am not sure what it should be like but I would say somewhere in between the two. The old looks like there was a slight blue filter on the lens and the new they lowered the blue level which in return brings out the green, yellows and reds.

Again just my opinion which really does not matter!

Wes

 

[Roger Rollins]

I have been lucky enough to see dye-transfer 35mm prints of OZ several times, most notably one made in 1955.

The color hues in the new DVD transfer more accurately reflect what was in the dye-transfer prints than the previous (very nice) DVD transfer. (As I mentioned in a posting several days ago, I was fortunate enough to see the new DVD at a friend's home).

Indeed, there was no "blue tint" in the dye-transfer prints, but definitely a greater sense of yellow and green.
This was especially true in scenes such as Munchkinland and
the arrival at the Emerald City.

The frame grabs posted here by Herb reflect more detail, sharpness, and clarity than that of the previous disc, although the differences aren't that shockingly different because the old disc was very nice, indeed.

As a kid, I bought the first video release on Betamax, then upgraded to the 1983 laserdisc, both of which came from muddy print. Then came the 1989 release, which taken from a true dye-transfer release print, finally gave us a better rendering of Oz in terms of color. However, release prints, especially dye-transfer prints, are not a good source for telecine. The resulting image was lovely, but fuzzy (although far superior to previous video versions).

The 1993 "Ultimate Oz" LD reflected the use of a new Interpositive made from the original YCMs. A significant improvement. That transfer was used for MGM's 1997 vanilla DVD.

I asssume that same element was used for the far-improved transfer found on the 1999 WB DVD. It still didn't have the real feel of true dye-transfer hues, but was a very fine job.

It is my own personal opinion that this new transfer, is the most faithful representation of the original creators' intent, and is something to celebrate.

I trust others will feel the same joy I did when they get to see what the wizards of Warner have done this time.

Cheers to all

 

[Mark Lucas]

The new transfer seems to have a lot of yellow compared to the previous version, which might've had too much blue. The shot with Dorothy and the munchkins behind her is confusing though. The previous edition seems to have more pleasing colors.

I don't know what RAH is talking about either. The overall color isn't important?

 

[Herb Kane]

Rich, I don't know what happened there with respect to the sizes of the images. I used the same PC, done at the same time - absolutely identical for each version...



Anyway... I'd be the first to agree that one shouldn't base an absolute opinion - one way or the other on a screen cap - especially those relatively compressed such as these. However, the obvious leaps out in this case; improved texture and better colors. These improvements were quite obvious projected on a large screen.

The HTF Review is up and posted here.

 

[DaViD Boulet]

thanks Roger Rollins for your excellent information.

Herb...off to read your review!

 

[Robert Harris]

I'm not saying that color isn't important.

Color is important, and should be captured in the negative as accurately as possible, hence the search for automotive paint chips, costumes, etc.

What I AM saying is that once that color is encoded to the negative, it is very difficult to get two prints made on two consecutive days to match each other. Each will be slightly different from the other, but both will get the point of the color across and will pleasing to the eye.

The original prints of Oz that I've seen, have also not ever been cool. They always tended toward warm tones.

RAH

 

[Tim Glover]

Thanks for the great points RAH. I was beginning to prefer the older disc. However, if the ones who know the accuracy of the film say the new version is more accurate then I'm good with that.

 

[Mark Lucas]

I just want to know why her dress looks so dull compared to the previous edition. I'm starting to think that neither edition is really correct.

Edit: The comparisons on DVDBeaver.com are better than what was posted in this thread. Her dress seems more accurate and the colors are even more saturated than before. It just goes to show that you need accurate full sized screenshots to show the difference.