A Few Words About A few words about...™ Elizabeth -- in HD (1 Viewer)

[Robert Harris]

I'm afraid I have to get back to my earlier concept -- that some viewers are looking way too closely at discs.

So the point from my perspective is not that someone is seeing problems, but how many people also seeing the same problems.

When I screened Elizabeth, I noted nothing that made me take note. I'll have to take another look at the suggested areas and see if I can find it, but the more major area of the discussion really is "who" is seeing the problems.

I'm not saying that they might not be there, but at a certain point, they don't matter.

RAH

 

[Douglas Monce]

Are we sure that this edge enhancement comes from this encode of the movie? Do we know for sure that it doesn't come from an original digital intermediary? I don't know if Elizabeth used one or not but its possible. The Director may have asked for some shots to be sharpened in post production.

Do we know for sure that the shots in question aren't in actually a digital effects shot and artifacts are creeping in from that? Are we sure that this isn't in fact flaws in the optics of the camera lenses or some kind of filtration used on the lenses?

I haven't seen this disc yet but the point is there might be many sources that could cause something that looks like edge enhancement.

Mr. Harris' point is well taken. If you look closely enough at any visual product your going to find flaws, even at the cinema.

Doug

 

[Stephen_J_H]

3.5 m screen from 4.5 m? I would disagree, since the ideal viewing distance would be 2 to 2 1/2 times screen width, meaning you should be @ least 7 m from your screen. At any distance closer than this, artifacts (digital or otherwise) are bound to crop up.

 

[Ken Horowitz]

50 years ago, my Dad was working at Kodak, in the Color Technology group. When he and Mom would go to the movies, Dad would often point out to Mom the "cyan shadows" or "magenta highlights" that he found so annoying -- there was no way he could not see them. (They've been happily married now for over 50 years, so Dad obviously learned pretty quickly to stop telling Mom about this stuff.)

Decades ago, someone pointed out cue-marks to me, and ever since I've been unable to ignore them. Until that point, I'd never noticed even the brilliantly-colored Technicolor cue-marks.

I think it's important here to keep in perspective everyone's backgrounds and areas of expertise. And to consider same for the general public as well -- what constitute reasonable expectations in order for this business (which it is) to survive?

 

[Michel_Hafner]

Ideal meaning what? Ideal for me is to see detail up to 1080p resolution. How close you have to go for that depends on your vision. Ideal to me also means to have a good immersive experience and cinema feeling. From experience I know for a fact that from my viewing distance there is no EE visible to me when there is no EE on the disc. And when there is I see it. That is how it should be. From my viewing distance I do not see aliasing on well antialiased sources nor compression artifacts on well compressed discs. Again as it should be. To not see the EE on Elizabeth I would have to sit at least 3 times further away. Neither possible in my room nor attractive at all.
See http://www.myhometheater.homestead.c...alculator.html
Recommended for my situation is 17.7 feet. I use 15 feet, only little less.

 

[Michel_Hafner]

Elizabeth is a pre DI film. No digital sharpening on film elements. So when I see EE it's either only the disc or my chain adds it. My chain does not add it (verifiable with clean sources). So it's on the disc. If what I see is not EE but an optical effect/artifact from original photography I would be mistaken objecting to it. But what I see looks like EE and not lens artifacts.

 

[Michel_Hafner]

If we did an experiment and showed the same HD disc (Elizabeth or whatever) to
- me
- Mr. Harris
- a top compressionist
- a top director of photography
- a top director
- a sfx expert
in the same room on the same screen from the same chair and then asked each of them what he liked or disliked, what technical errors he saw or what was wrong or right about this disc, would we all say the same? Of course not. We all are shaped by past experiences and personal likes and dislikes plus individual specific knowledge we need for our specific jobs. The compressionist would see compression issues the others would likely not see or find marginal at best when pointed out (assuming this is a well compressed disc). The DOP would probably object a bit to some color details or highlight or shadow details he would have liked to see a bit 'better' or differently, and this out of focus shot he could not reshoot. The director might like what he sees but still be annoyed that he had to cut this scene or that take xy was having continuity problems and he could not reshoot. The sfx expert would maybe complain about lack of grain to hide some CGI seams or bad CGI lighting not fitting in with the live footage. Or some blue screen spill in the lead actress' hair. What Mr. Harris would say he might answer best himself. I would complain when I see anything that looks not as filmlike as I want it to look if I'm convinced it could be done better (e.g. avoidable digital artifacts). And we all would be right, and no one would have the complete picture (pun intended).

 

[Vern Dias]

Sorry but I have to disagree here. These guidelines are OK for upscaled SD DVD, but are woefully out of date for HD. I sit at 1.1 times the screen width of my 14' wide 2.76:1 screen and actually would be comfortable sitting even closer. At that point, the image about fills my horizontal field of vision.

This has always been my preferred seating distance for 35mm film, and HD with a good 1080 projector fully preserves that experience.

If you are sitting 2.5x the screen width from the screen, you might as well save money and buy a 720 projector, because you are not taking full advantage of all the the resolution a 1080 source can provide.

Just my $.02, YMMV, etc etc

Ted

 

[Douglas Monce]

2 to 2 1/2 times screen width is actually a fairly well accepted practice in the cinema as well. That is in fact where the mixing console sits in every mixing theater I've been in.

Doug

 

[Stephen_J_H]

I respectfully beg to differ. As Mr. Monce has pointed out, 2-2.5x is a widely accepted standard not for SD DVD, but for theatrical presentation. I also find that sitting any closer would induce motion sickness on certain films for me.

 

[Michel_Hafner]

Mixing is for sound, not critical image evaluation.
At 2 times (wide)screen width you need better than 20/20 vision to resolve 1080p detail. Not ideal for critical image evaluation for non eagle eyes.
http://www.myhometheater.homestead.c...alculator.html
16:9 screen -> maximum viewing distance for 1080p 1.8 screen widths

 

[Douglas Monce]

The mixing console is set at the optimal viewing distance from the screen, which by no coincidence is also the optimal listening point in the theater.

Sitting too close to the screen tends to revel things such as the holes in the screen that allow the the sound to pass through. With HD you start to see the individual scan lines.

Doug

 

[Michel_Hafner]

Again, optimal for what? A distance where you no longer resolve 1080p detail is NOT optimal for critical image evaluation. Holes in the screen are bad anyway for image evaluation as they dither the picture and create potential aliasing issues and can mask image artifacts as well. My screen has no holes, by the way, but lets the sound through nonetheless. A high quality screen has no holes that interfere with image content. And yes, cinema screens have too large holes for optimal results due to economic reasons. Film sourced HD has no scan line issues unless you watch on (interlaced) CRT. Digital HD is based on a pixel grid. You can have aliasing issues. Seeing these when present is again proper for critical image evaluation or you can no longer tell properly antialiased and aliased content apart.

 

[Douglas Monce]

HD still has scan lines. Thats why its called 1080, its the number of scan lines.

Doug

 

[Michel_Hafner]

The term is misleading for 1080p material from film sources. Of course every pixel grid has lines and columns. Stressing lines makes sense if lines are indeed different from columns as in interlaced systems where lines are treated differently from columns and therefore the line structure (as opposed to the column structure) of images is visible regularly in the form of line twitter, deinterlacing artiafcts and the like. On film sourced 1080p material lines and columns are equally affected if pixel structure shows up as aliasing. It can happen in every direction. If it's properly antialiased it's not an issue. Such material is not pin sharp, though, as that requires more than 1080p resolution.
Anyway, I don't see 'scanlines' on HD from where I sit if the material is properly antialiased film sourced 1080p. And I would see the EE in Elizabeth even sitting twice as far away as I do now.

 

[Douglas Monce]

I'm not talking about scan lines from the software, or interlacing, I'm talking about the actual lines in the display device. You sit close enough and you'll see them. In the days of CRTs they called them the individual phosphor. They are smaller now on HD but they are still there and if you sit close enough you'll see them.

Doug

 

[Michel_Hafner]

That's again the pixel structure of any digital device. Only with CRT you have 'scanlines' since there is an electron beam scanning a phosphor surface. So it's not a source but a display thing. All digital projectors have imaging chips/panels that are simply pixel grids. How well you see the pixels (lines and columns equally) depends on the technology and image content (degree of aliasing in the picture). D_ILA/SXRD/LCOS have high fill factor and the pixel grid in itself is invisible from 'normal' viewing distances. LCD and DLP have less fill factor and the grid structure is more pronounced.

 

[PaulDA]

Well, I'm with the OP on this one. I just had a chance to watch this (a favourite of mine and one I've used as a teaching tool more than once) on HD DVD and it looks spectacular. Is it perfect? Nothing is perfect. But I have no complaints with this disc at all. I guess some people's eyesight is too good (I have 20/17 in one eye and 20/15 in the other, but it's not enough for me to see anything wrong, of note, with this presentation). This will definitely get a lot of repeat viewing.

 

[ManW_TheUncool]

Paul,

Since you use a 720p FP, IIRC, maybe the downconversion has obscured the issues being pointed out by some. For instance, that seems likely to be the case for at least one HTF reviewer judging from the EE/DNR thread elsewhere.

Certainly, downconversion (or rather downscaling) of images can hide/reduce issues like noise, which is pretty well known in the world of digital still photography. In fact, some consumer level digicams employ just that trick to help hide high ISO noise. And Fuji even employs a variation on that trick to reduce high ISO noise in their semi-pro/pro level digital SLRs.

_Man_

 

[Sam Posten]

If one of you wants to replace my 720P projector with a nice 1080p one I'll not say no =p