Best 3D Blu-ray Deals

Best Blu-ray Movie Deals, See All the Deals »
Top deals | New deals  
 All countries United States United Kingdom Canada Germany France Spain Italy Australia Japan
Harry Potter Hogwarts Collection (Blu-ray)
$97.23
 
War for the Planet of the Apes 3D (Blu-ray)
$9.99
8 hrs ago
Alita: Battle Angel 4K + 3D (Blu-ray)
$29.96
 
Aquaman 3D (Blu-ray)
$32.21
 
The Great Gatsby 3D (Blu-ray)
$11.34
14 hrs ago
Pokémon: Detective Pikachu 3D (Blu-ray)
$29.99
 
Walking with Dinosaurs: The Movie 3D (Blu-ray)
$9.99
8 hrs ago
Dawn of the Planet of the Apes 3D (Blu-ray)
$9.99
 
Men in Black 3 3D (Blu-ray)
$8.92
 
Predator 3D (Blu-ray)
$9.99
22 hrs ago
Blade Runner 2049 3D (Blu-ray)
$19.99
 
Justice League 3D (Blu-ray)
$19.99
 
What's your next favorite movie?
Join our movie community to find out


Image from: Life of Pi (2012)

Go Back   Blu-ray Forum > 3D > 3D News and General Discussion


Reply
 
Thread Tools Display Modes
Old 06-27-2016, 02:01 AM   #1
bavanut bavanut is offline
Senior Member
 
bavanut's Avatar
 
Jun 2011
Burbank, California
1
Default Grain in Over-and-Under 3-D

I wish to present a logical argument that I believe proves a fact that runs contrary to the received wisdom on the subject of over-and-under 3-D. It is my position that an over-and-under 3-D film need only look grainier than common 2-D film formats if it is projected onto a wider screen at greater magnification; projected onto a screen of common width but variable height, over-and-under 3-D will not necessarily look any grainier than common 2-D formats.

The size of film grain onscreen is directly proportional to the power of magnification. That is to say, film grain becomes larger (and more apparent, and possibly more objectionable) with greater frame magnification. Grain will be more obvious with a frame magnification of 300X than it would be at 150X, for instance.

Assuming a constant screen height, a two-perf frame is being magnified more than a comparable three-perf frame, which in turn is being magnified more than a comparable four-perf frame. On a screen three meters high, a two-perf frame is being magnified roughly 300X (to a width of about 7.5 meters), a three-perf frame roughly 250X (to a width of about 5.5 meters), and a four-perf frame roughly 185X (to a width of about four meters).

But assuming a constant screen width, the power of magnification is constant. On a screen 5.5 meters wide, screen magnification for all three formats (two-perf, three-perf and four-perf) is 250X. Assuming similar quality negative stocks and similar quality release prints, one would expect the same apparent size of grain for two-perf, three-perf and four-perf motion pictures of the same vintage.

Therefore it follows that over-and-under 3-D will look grainier than common 2-D 35 millimeter film formats only when it is magnified to fill a wider screen at a constant screen height.

I have anecdotal reason to believe that 3-D films were not usually shown on the very widest of wide screens. Several theater managers and projection engineers I spoke to over the years told me that their strategy during the early 1980s was to open their side-movable screen masking out to 1.85 (not 2.35), then project 3-D films within that screen area-- sometimes with improvised bottom masking, sometimes without. To a man, they seemed to regard it as folly to project over-and-under 3-D onto a full-sized CinemaScope-format screen, partly owing to diminished brightness, and partly owing to objectionable grain and image softness. It is not unreasonable to conclude that numerous other cinema managers and projectionists took the same view as those persons I had occasion to speak to.

In any event, when an over-and-under 3-D title is presented on a 16x9 television, it is being held to a constant screen width in relation to any 1.85:1 2-D title also presented in 16:9. Therefore it follows that, given similar quality original elements of a similar vintage, a well-transfered 2.4:1 over-and-under 3-D title ought not have larger (or worse, or more objectionable) visible grain on Blu-Ray than a well-transferred 1.85:1 title. Assuming the same type and speed of film stock in the original elements, any discrepancies in image sharpness, contrast, or resolution are surely owing to the quality of the lenses employed.

Thoughts?
  Reply With Quote
Thanks given by:
Paul H (06-27-2016)
Old 06-27-2016, 02:22 PM   #2
3Dfan 3Dfan is offline
Active Member
 
Oct 2011
Default

On my passive 3dtv, the channels that used the over/under format inclusing ESPN 3D looked horrible. While on the same tv, things that used the side by side format looked great.

So I really don't get where this idea that over/under looks better on a passive tv than side by side comes from. It doesn't. At least not in my experience. It's the opposite. The side by side looks MUCH MUCH MUCH better on a passive 3Dtv than "top/bottom" "over/under" format does.
  Reply With Quote
Thanks given by:
bavanut (06-27-2016)
Old 06-27-2016, 03:03 PM   #3
bavanut bavanut is offline
Senior Member
 
bavanut's Avatar
 
Jun 2011
Burbank, California
1
Default

I did not properly anticipate discussion of over-and-under video. I really appreciate your observations, 3D Fan, but I should clarify for others that my remarks above apply to 35mm over-and-under 3-D, like Space-Vision, Stereovision, Optimax III, Marks 3-Depix, and ArriVision.
  Reply With Quote
Old 06-27-2016, 08:28 PM   #4
Interdimensional Interdimensional is offline
Special Member
 
Interdimensional's Avatar
 
Nov 2014
Default

Quote:
Originally Posted by 3Dfan View Post
On my passive 3dtv, the channels that used the over/under format inclusing ESPN 3D looked horrible. While on the same tv, things that used the side by side format looked great.

So I really don't get where this idea that over/under looks better on a passive tv than side by side comes from. It doesn't. At least not in my experience. It's the opposite. The side by side looks MUCH MUCH MUCH better on a passive 3Dtv than "top/bottom" "over/under" format does.
Interesting observation, having never heard that said or expressed that idea, or made any direct comparisons of the two approaches.

However, on a passive 1080p 3dtv, one would expect over/under to give better results, at least theoretically. When you watch 3-D on a passive set, the two views are represented on alternating horizontal lines. As such, you're watching two images that are each 1920 x 540 merged into one 1920 x 1080 image.

If those images were initially presented side by side, each would be 960 x 1080 to begin with. To then merge them into an image that can be viewed in 3-D, you'd be further reducing the resolution. The final image would be 960 x 540. On a hd passive set you're already halving the vertical resolution, and by starting with a side by side image, you'd be halving the horizontal resolution as well.

If you begin with over/under images, the resolution will begin and end at a 1920x 540 image per eye, which is the maximum the screen is capable of displaying 3-D at.

Of course this is irrelevant to active 3dtvs and 4k 3dtvs.
  Reply With Quote
Thanks given by:
bavanut (06-27-2016), Paul H (06-27-2016)
Old 06-27-2016, 08:39 PM   #5
bavanut bavanut is offline
Senior Member
 
bavanut's Avatar
 
Jun 2011
Burbank, California
1
Default

I can at least add one little nugget of information to the discussion:

A few weeks ago, I made two experimental over-and-under video files for my LG passive television. These video files were burned into discs that would read on my Blu-Ray player. Speaking very subjectively, I thought that the results were a bit better than comparable side-by-side video files I tested.

I wonder if the poor appearance of ESPN and other over-and-under format channels is owing to peculiarities of cable or satellite transmission. But all this is only a guess.
  Reply With Quote
Thanks given by:
Interdimensional (06-27-2016), Paul H (06-27-2016)
Old 06-27-2016, 09:25 PM   #6
Interdimensional Interdimensional is offline
Special Member
 
Interdimensional's Avatar
 
Nov 2014
Default

Quote:
Originally Posted by bavanut View Post
I wish to present a logical argument that I believe proves a fact that runs contrary to the received wisdom on the subject of over-and-under 3-D. It is my position that an over-and-under 3-D film need only look grainier than common 2-D film formats if it is projected onto a wider screen at greater magnification; projected onto a screen of common width but variable height, over-and-under 3-D will not necessarily look any grainier than common 2-D formats.

The size of film grain onscreen is directly proportional to the power of magnification. That is to say, film grain becomes larger (and more apparent, and possibly more objectionable) with greater frame magnification. Grain will be more obvious with a frame magnification of 300X than it would be at 150X, for instance.

Assuming a constant screen height, a two-perf frame is being magnified more than a comparable three-perf frame, which in turn is being magnified more than a comparable four-perf frame. On a screen three meters high, a two-perf frame is being magnified roughly 300X (to a width of about 7.5 meters), a three-perf frame roughly 250X (to a width of about 5.5 meters), and a four-perf frame roughly 185X (to a width of about four meters).

But assuming a constant screen width, the power of magnification is constant. On a screen 5.5 meters wide, screen magnification for all three formats (two-perf, three-perf and four-perf) is 250X. Assuming similar quality negative stocks and similar quality release prints, one would expect the same apparent size of grain for two-perf, three-perf and four-perf motion pictures of the same vintage.

Therefore it follows that over-and-under 3-D will look grainier than common 2-D 35 millimeter film formats only when it is magnified to fill a wider screen at a constant screen height.

I have anecdotal reason to believe that 3-D films were not usually shown on the very widest of wide screens. Several theater managers and projection engineers I spoke to over the years told me that their strategy during the early 1980s was to open their side-movable screen masking out to 1.85 (not 2.35), then project 3-D films within that screen area-- sometimes with improvised bottom masking, sometimes without. To a man, they seemed to regard it as folly to project over-and-under 3-D onto a full-sized CinemaScope-format screen, partly owing to diminished brightness, and partly owing to objectionable grain and image softness. It is not unreasonable to conclude that numerous other cinema managers and projectionists took the same view as those persons I had occasion to speak to.

In any event, when an over-and-under 3-D title is presented on a 16x9 television, it is being held to a constant screen width in relation to any 1.85:1 2-D title also presented in 16:9. Therefore it follows that, given similar quality original elements of a similar vintage, a well-transfered 2.4:1 over-and-under 3-D title ought not have larger (or worse, or more objectionable) visible grain on Blu-Ray than a well-transferred 1.85:1 title. Assuming the same type and speed of film stock in the original elements, any discrepancies in image sharpness, contrast, or resolution are surely owing to the quality of the lenses employed.

Thoughts?
Anything I've said on the matter wasn't received wisdom so much as what I would infer from the design of the system as I understand it. Surely it stands to reason that you reduce the potential image quality when reduce the negative size by half?

I hope I haven't spoken beyond my knowledge on the subject.

In ideal conditions, I'm sure competent people could work with the limitations of this system to produce impressive results, but they are always in short supply, and conditions are often far from ideal.

From what you shared here, it seems clear people would've been aware of the limitations of the system and had various workarounds to address these issues
  Reply With Quote
Thanks given by:
bavanut (06-27-2016)
Old 06-27-2016, 09:35 PM   #7
Interdimensional Interdimensional is offline
Special Member
 
Interdimensional's Avatar
 
Nov 2014
Default

Quote:
Originally Posted by bavanut View Post
I can at least add one little nugget of information to the discussion:

A few weeks ago, I made two experimental over-and-under video files for my LG passive television. These video files were burned into discs that would read on my Blu-Ray player. Speaking very subjectively, I thought that the results were a bit better than comparable side-by-side video files I tested.

I wonder if the poor appearance of ESPN and other over-and-under format channels is owing to peculiarities of cable or satellite transmission. But all this is only a guess.
Yeah, I'll admit I don't know all the variables that may have come into play in the ESPN example. But for creating video files to play back on passive hd 3-D tvs, such as my own, I would expect over/under to produce better results.

One thing that occurred to me is that even though there's currently no 3D spec for UHD Blu-ray, and even if one is never announced, it'd still be possible to present 3D material on a UHD disc in over/under format and benefit from the increased colour gamut, doubled horizontal resolution and high framerate capability. Unlikely to happen perhaps.
  Reply With Quote
Thanks given by:
bavanut (06-27-2016)
Old 06-27-2016, 10:03 PM   #8
bavanut bavanut is offline
Senior Member
 
bavanut's Avatar
 
Jun 2011
Burbank, California
1
Default

Quote:
Originally Posted by Interdimensional View Post
Surely it stands to reason that you reduce the potential image quality when reduce the negative size by half?
I think you have pinpointed the spot where we 3-D buffs (myself included) have always made an incorrect inference. A two-perf frame meant to be projected through spherical optics does not invite direct comparison with a four-perf frame meant to be projected through anamorphic optics. The more apt comparison is with a 1.85:1 image projected through spherical optics.

So many persons who were there have told me they never, ever projected two-perf frames onto a full-sized CinemaScope screen. They projected instead onto a 1.85:1 screen, with reduced screen height to match the 2.4:1 (-ish) aspect ratio. So the apparent grain would really be comparable to the 1.85:1 image.

To cite a hypothetical example that I think clarifies my position, if you were to project a 1.85:1 film through a special aperture plate that cropped it down to 2.4:1, but made no further enlargement of the picture-- that is to say, the image onscreen stayed the exact same width, but now had a reduced height-- the overall appearance of grain would remain the same.

In other words, given the way I surmise those films were often projected, the grain level in a "stacked frame" 3-D film would be similar to the grain levels in an ordinary 1.85:1 film. And likewise, assuming a quality scan of a decent negative or other suitable film elements, the grain in a "stacked frame" 3-D film on Blu-Ray ought to compare favorably to the grain in a 1.85:1 film of the same vintage photographed on similar stock.

In a theatrical setting, the grain would only become a huge issue if one attempted to project that two-perf frame onto a really, really huge wide screen. This may have been attempted-- many times, for all I know-- but it was really inadvisable given reduced light levels and concomitant image softness. And, as I say, most persons I have spoken to personally who projected over-and-under 3-D in theatrical venues in the 1980s were adamant that they always projected the image onto the 1.85 screen, with reduced image height and (sometimes but not always) suitable improvised masking along the bottom.
  Reply With Quote
Thanks given by:
Interdimensional (06-29-2016)
Old 06-28-2016, 02:09 AM   #9
Zivouhr Zivouhr is offline
Blu-ray Duke
 
Zivouhr's Avatar
 
Dec 2011
USA
3
125
Default

Quote:
Originally Posted by 3Dfan View Post
On my passive 3dtv, the channels that used the over/under format inclusing ESPN 3D looked horrible. While on the same tv, things that used the side by side format looked great.

So I really don't get where this idea that over/under looks better on a passive tv than side by side comes from. It doesn't. At least not in my experience. It's the opposite. The side by side looks MUCH MUCH MUCH better on a passive 3Dtv than "top/bottom" "over/under" format does.
I prefer side by side 3D to top/down 3D also.

Probably because I rarely see anything on blu ray that is formatted as top/down 3D. Plus most modern professional blu ray 3D discs will use MVC format, already blended and ready for the glasses without having to set the TV to side by side.
For youtube 3D though, there are some more options.
  Reply With Quote
Old 06-29-2016, 02:54 AM   #10
Interdimensional Interdimensional is offline
Special Member
 
Interdimensional's Avatar
 
Nov 2014
Default

Quote:
Originally Posted by bavanut View Post
I think you have pinpointed the spot where we 3-D buffs (myself included) have always made an incorrect inference. A two-perf frame meant to be projected through spherical optics does not invite direct comparison with a four-perf frame meant to be projected through anamorphic optics. The more apt comparison is with a 1.85:1 image projected through spherical optics.

So many persons who were there have told me they never, ever projected two-perf frames onto a full-sized CinemaScope screen. They projected instead onto a 1.85:1 screen, with reduced screen height to match the 2.4:1 (-ish) aspect ratio. So the apparent grain would really be comparable to the 1.85:1 image.

To cite a hypothetical example that I think clarifies my position, if you were to project a 1.85:1 film through a special aperture plate that cropped it down to 2.4:1, but made no further enlargement of the picture-- that is to say, the image onscreen stayed the exact same width, but now had a reduced height-- the overall appearance of grain would remain the same.

In other words, given the way I surmise those films were often projected, the grain level in a "stacked frame" 3-D film would be similar to the grain levels in an ordinary 1.85:1 film. And likewise, assuming a quality scan of a decent negative or other suitable film elements, the grain in a "stacked frame" 3-D film on Blu-Ray ought to compare favorably to the grain in a 1.85:1 film of the same vintage photographed on similar stock.

In a theatrical setting, the grain would only become a huge issue if one attempted to project that two-perf frame onto a really, really huge wide screen. This may have been attempted-- many times, for all I know-- but it was really inadvisable given reduced light levels and concomitant image softness. And, as I say, most persons I have spoken to personally who projected over-and-under 3-D in theatrical venues in the 1980s were adamant that they always projected the image onto the 1.85 screen, with reduced image height and (sometimes but not always) suitable improvised masking along the bottom.
I see your point. Although in the example of Jaws, which is where this came up for discussion, the previous films in the series would've been shot anamorphically on 4-perf 35mm film.

No one should expect the quality of over/under to hold up to that.

Even comparing to a 1.85:1 non-anamorphic image, the size of the grain relative to the size of the picture will make that image look better.

Instead we should expect similar image quality to films created using the techniscope process. While they do have a bit of a reputation for being grainy, some digital transfers have been a revelation.
  Reply With Quote
Thanks given by:
bavanut (06-29-2016)
Old 06-29-2016, 12:55 PM   #11
cinerama cinerama is offline
Special Member
 
cinerama's Avatar
 
Dec 2011
Default

The Bubble is another over-under 3D title that does not look as bad as Jaws 3-D.
  Reply With Quote
Thanks given by:
bavanut (06-30-2016)
Old 06-29-2016, 01:06 PM   #12
Zivouhr Zivouhr is offline
Blu-ray Duke
 
Zivouhr's Avatar
 
Dec 2011
USA
3
125
Default

As far as film grain, The Bubble looks nice in 3D, I agree. Some natural film grain is a good thing IMO. For the conversion of Jurassic Park 3D, they tried to smear out the grain of the faces. It got rid of some of the details, though not as heavily as Predator 3D, as with Carl Weathers close up face for example, which looks smoothed out compared to the original film's graininess from what I remember.
  Reply With Quote
Old 06-30-2016, 04:03 AM   #13
JoeMarus JoeMarus is offline
Member
 
Oct 2015
7
468
261
Default

I am no film expert, but I believe that there are several different grades of film stock. The cheaper film stock I would imagine could have a natively larger grain structure than more expensive film stock. Also film stock has probably improved in quality over time. So it could be case of just the type of film stock used?

It could also be a question of what film elements were actually available. The highest quality source would be the camera negative, but what if this doesn't exist anymore? They might have had to settle for using an actual distribution print shown in theaters. I've heard those prints vary in quality because of the optical printing process and the care (or lack of care) put into making the print, so they could possibly have worse grain than the negative.
  Reply With Quote
Reply
Go Back   Blu-ray Forum > 3D > 3D News and General Discussion


Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT. The time now is 03:56 PM.