What is the purpose of anamorphic 1.66:1? (1 Viewer)

[Robert Dunnill]

Not many, but it certainly matters for some, e.g. Warner's Horror of Dracula.

Indeed! On Tombs of the Blind Dead, the subtitles were cut off by the RP-91's one-size-fits-all scaling.

 

[Ken_McAlinden]

What I meant was that most “flat” widescreen films should be mastered at 1.78:1 (with 16x9 enhancement).

Granted, some movies were framed with 1.66:1 in mind. However, my contention is that cropping most of these movies slightly (to 1.78:1) in the telecine phase would not be a disaster (See Dr No, From Russia With Love and Goldfinger as examples).

I actually would apply almost the same reasoning to come to the conclusion that since the difference is slight, why bother to crop anything? The black bars will be small and unobtrusive.

Regards,

 

[Tom_Bechet]

Can anyone answer this question:
Do we actually gain anything from 1.33 anamorphic enhanced DVDs??
Yes they exist basically what they do is use a 4.3 original ration (mainly from documantaries) then add black bars on the sides so that the whole picture fills a 1.77:1 ratio.
But I wonder will that give better quality on a widescreen telly?
I imagine on a 4:3 telly and a DVD player that shows letterboxed image it could lead to problems.

 

[Damin J Toell]

Do we actually gain anything from 1.33 anamorphic enhanced DVDs??

No. You lose resolution. A normal 1.33:1 transfer will use every pixel in DVD's 480x720 resolution. An anamorphic 1.33:1 transfer will only have a resolution of 480x540. Picture quality is lowered for everyone, no matter what kind of TV you have.

DJ

 

[Bill Burns]

Deane wrote:

There's one advantage. They get one additional sale. Me.

Two.

Tom -- the key is to remember, as Dave was good enough to explain at length on another thread (it was the first detailed examination of the issue I'd found), that both 4x3 and 16x9 transfers use the exact same number of pixels for DVD -- the pixels are shaped differently, but they number precisely the same. Thus, if you shrink a 4x3 film to fit inside of the 16x9 window, you lose pixels.

The numbers work out somewhere around 1.5:1, to my understanding. Anything narrower than 1.5:1 finds its highest resolution in a 4x3 DVD window, while anything wider than 1.5:1 (possibly including that ratio, I'm not certain) gains resolution by being placed within a 16x9 DVD window.

Those with widescreen televisions also enjoy an image that more or less fills their set without the need for zooming when a 1.66:1 transfer is presented in a 16x9 window; if the set has no appreciable overscan, small black bars will be visible to the sides. If a 1.66:1 transfer is made in a 4x3 window, the bars to the sides are much thicker, as you're only watching that 4x3 window with a 1.66:1 transfer inside of it, and thus using less of the television's native resolution (1.33:1 inside of 1.78:1, rather than 1.66:1 inside of 1.78:1). There are also, now, black bars to the top and bottom! You may be able to zoom into that image to make it better fill the screen, but the loss in resolution becomes all the more evident. On a set such as mine (a 4x3 CRT with a 16x9 mode, compressing scan lines into a 16x9 frame), line structure is much more evident at 4x3 than 16x9, and so the need to sit further from the set is another factor. I rarely buy letterboxed 4x3 as a result -- the loss in resolution, coupled with the greater scan line visibility, usually makes for a decidedly video-like presentation, ruining the illusion of film (one of the only exceptions I've found is Warner's Giant, whose "hard cut" sections -- don't ask -- look absolutely fantastic, no equivocation necessary ... though, of course, the film would look better still if 16x9 encoded, nevertheless). This is not so for quality native 4x3 presentation (films shot for and presented in 4x3), because the native resolution is the same as it is for the 16x9 window when filled -- scan lines remain more visible, but the resolution helps alleviate that factor, and presentations remain very filmlike when transferred properly.

It's important to note that HiDef downconversions add a lot to the appearance of film on a video format such as DVD -- if working from a HiDef master, a DVD is going to look better than if working from a low def video master, and so there may be cases out there where a HiDef downconversion, at 4x3 letterboxed, looks better than a standard def 16x9 formatted version of the same film (perhaps in an alternate territory). I don't know of any examples off-hand, but someone might run into this. In such a case, the HiDef remaster has improved quality quite a bit, but that HiDef downconversion would be much improved itself, if 1.66:1 or wider (I don't know of any 1.5-1.65 films off-hand), were it transferred in 16x9.*

As to the question of 1.66:1 and 1.78:1, there is an important difference on films composed with care for one or the other. Another example that leaps to mind is Branagh's Henry V, overmatted on DVD around 1.78:1; opening credits are partially cut off (vertically) as a result. Any transfer that takes slices from the side of a hard-matted master to reduce 1.85:1 to 1.78:1, on another tack, would of course lose potentially valuable bits of material. Opening a matte to reduce a ratio, if done to only a minor degree (1.85:1 to 1.78:1) isn't significant, as greater variances are found in theatres. But whenever a matte is increased, or whenever visible film image is reduced to alter a ratio, the potential to lose details intended by the filmmaker, or to change the overall visceral impact of the frame from that intended by the filmmakers, is something that should be carefully considered. This latter concern is the primary reason fully open matte presentations must be considered inferior to properly matted widescreen presentations, as well, when the filmmakers composed their images for that wider frame.

It's a big field. But if studios understand that a dedication to meeting the highest quality standards does, in fact, increase sales, the likelihood that any given film will find itself released to such standards is greater. Filmmaker insistence is also an important factor, I'm sure -- those who are informed about such things and oversee the home video presentations of their films can, of course, directly ensure that the format is used to the best of its potential.

* (nudge nudge) WB, M-G-M (nudge nudge)

 

[Seth--L]

I actually like the look of 1.66:1 at lot. For a documentary I'm making I matted it to this because I did not want the box look of TV, but 1.85:1 was too dramatic.

 

[Ike]

I've only seen one 16x9 1.66:1 film so far (That Obscure Object of Desire), and I don't know if it was because of overscan, but it had no bars on the image of anykind. I paid attention to the framing for a bit, but found it not to be cramped in any way, so I didn't really worry about it.

 

[DaViD Boulet]

Ike,

do you have any WS Disney animated titles? Lots of those are 1.66:1 16x9.

In any case, your experience is like most folks...they don't see any pillarboxing bars due to overscan.

Warner seems to be pushing for 4x3 1.66:1 as a test of will. My guess is it's a "personal" thing with someone in charge there and we won't see 1.66:1 DVDs properly 16x9 encoded until there's a change of attitude/or this individual(s) moves on...

 

[Ike]

Ike,

do you have any WS Disney animated titles? Lots of those are 1.66:1 16x9.

No. I've got The Discreet Charm of the Bourgeisie, which is supposed to be 1.66:1, but I think it was actually released 1.78:1.

Okay, checking DVD Profiler, I have My Metier, Cries and Whispers, Gertrud, In the Mood for Love, Last Tango in Paris, and Peeping Tom that are all listed as anamorphic 1.66:1. I'll check one of these out to see if it was overscan or not. (It's worth noting that all but one of those is a Criterion title.)

 

[Jon Robertson]

I think Peeping Tom and In The Mood for Love are cropped (My Metier isn't anamorphic, and is framed rather bizarrely with a thick bottom bar and a barely existant top one).

I have to say that minor alterations to aspect ratios don't really bother me. Unless the framing looks grossly inaccurate (which it very rarely does cropping from 1.66:1 to 1.78:1), I'm too caught up in everything else the movie has to offer to notice.

 

[DaViD Boulet]

Agreed,

Though it's hard for me to get "caught up" in Babette's Feast when I have to choose between cutting off actors faces or cutting off the subtitles...

dave

 

[Craig Robertson]

I'd love to have an anamorphic 'Nightmare Before Christmas' and gladly take black sidebars!!!

you and me both!

 

[Mattias_ka]

Granted, some movies were framed with 1.66:1 in mind. However, my contention is that cropping most of these movies slightly (to 1.78:1) in the telecine phase would not be a disaster (See Dr No, From Russia With Love and Goldfinger as examples).

Well, are you a OAR guy or a joe sixpack delux that want the movie to fit your tv-set?

 

[David Von Pein]

Not many, but it certainly matters for some, e.g. Warner's Horror of Dracula.

I concur here (some severely cropped heads).
Plus -- The "Horror Of Dracula" DVD is matted even further to 1.85:1, not just 1.78:1, making it even "tighter".

I wonder WHY any studio would jump TWO ratios up (wider)--from an OAR of 1.66 all the way to 1.85, skipping the more-pleasing 1.78 AR. ??? (Esp. considering the fact that 1.78:1 is the perfect, matching ratio for a 16x9 monitor.)

Anybody understand why Warner did this for "Horror"? Seems unwarranted.

And -- Placing the usual "Matted to preserve the ratio of its original theatrical exhibition" blurb on the packaging appears to be disingenuous in this case, as well. Correct? (Or did some theaters actually show "HOD" in the cramped confines of 1.85:1?)

 

[David Von Pein]

Thus, if you shrink a 4x3 film to fit inside of the 16x9 window, you lose pixels.

But .... If it's an "Anamorphic 1.33" transfer (which I've really never heard of before, because I always thought the term "Anamorphic" was inherent to "Widescreen" only), wouldn't this mean that the "Pillarboxed Bars" created for such a DVD transfer would not be created by utilizing picture image scan lines? Thus, how can a 1.33 Anamorphic image produce LESS resolution than a standard 1.33:1 transfer?

What am I not understanding here? I always assumed "Anamorphic" meant more resolution / better PQ. ???

 

[Damin J Toell]

What I am not understanding here? I always assumed "Anamorphic" meant more resolution / better PQ. ???

A bad assumption, but you're certainly not alone in making it.

 

[David Von Pein]

An "anamorphic" 1.33:1 transfer will only have a resolution of 480x540, as the rest of the pixels will be used to display black bars on the sides of the screen. With aspect ratios narrower than around 1.5:1, you no longer gain resolution by making the transfer anamorphic.

Which is exactly the opposite of what "Widescreen Anamorphic" accomplishes, right?

Meaning -- that with WS Anamorphic, NO scan lines are taken away from the image to create the black bars (the bars are simply "empty space", correct?).

But with 1.33 Anamorphic (rarely used, granted) we apparently have the opposite scenario occurring, right? With actual scan lines REMOVED from the Anamorphic picture in order to create the "Pillarbox" bars. Which, it would seem, totally negates the reason for producing said transfer in Anamorphic in the first place. Why not just keep it non-Anamorphic Full-Frame?

Have I got this accurately assessed?

If so...my natural question would be...Why can't the 1.33 "Side Bars" be created just like a WS Anamorphic transfer does (just "empty" space, using no scan lines, keeping ALL the resolution/scan lines within the actual video image)?

Or --- Does the answer have something to do with the fact, with 1.33 "Pillars", we're dealing with "Vertical" Bars, rather than the "normal" WS horizontal black bars?

 

[cafink]

As Damin said, the resolution of a DVD is 720×480 pixels. These pixels can be arranged into a 4:3 rectangle (so that each pixel will be a little taller than it is wide), or a 16:9 rectangle (each will be a little wider than tall), but the number of pixels never changes, only their shape.

If the material on the DVD happens to be in either the 4:3 or the 16:9 aspect ratio, then it can be stored utilizing every single pixel available. Otherwise, there has to be some empty space — the "black bars" — stored around the edges (either top & bottom or left & right).

When a 16:9 encoded DVD is played back on a regular 4:3 television, the "black bars" are put there by the DVD player, and are not actually encoded on the disc.

When you watch a 4:3 DVD on your 16:9 television, a similar phenomenon occurs, except that the "pillar bars" are put there by your television instead of the DVD player. Again, a 4:3 transfer, encoded without "anamorphic enhancment" already utilizes the full resolution of the DVD format. There's no resolution to be gained by employing any kind of aspect-ratio trickery.

It's certainly possible to imagine a 16:9 video format in which 4:3 material could be encoded as you describe, but DVD just doesn't work that way.

 

[Robert Dunnill]

If so...my natural question would be...Why can't the 1.33 "Side Bars" be created just like a WS Anamorphic transfer does (just "empty" space, using no scan lines, keeping ALL the resolution/scan lines within the actual video image)?

Because (AFAIK) as per the DVD specification, our favorite format is 4:3-oriented. The raw DVD content is a color samples map, not an image; it is intended to be displayed using a pixel aspect ratio of 72/79, which when applied against the 1.5:1 aspect ratio of the color samples map, produces an image of roughly 4:3 dimensions. When such an image is sent to a 16x9 display operating in native mode, it is "stretched" sideways to fit the screen (which is why in some cases the 16x9-enhancement is referred to as "anamorphic" enhancment).

The flexibility you describe wasn't designed into the standard; it could be implemented in a de facto manner through the use of scalers or scaling players with special ratios, but the contents of such discs would display in a distorted fashion on standard players.

I hope that made sense.

RD

 

[DaViD Boulet]

If so...my natural question would be...Why can't the 1.33 "Side Bars" be created just like a WS Anamorphic transfer does (just "empty" space, using no scan lines, keeping ALL the resolution/scan lines within the actual video image)?

Because that *is* a 4x3 encoded image.

My DVD player has the ability to take that 4x3 image (that's using all 720 x 480 pixels for the 4x3 picture) and slap electronic black-bars on the left/right sides, put it in a 16x9 frame, and send it to my 16x9 TV. This can be acomplished without downconversion because unlike the vertical axis for TVs which is fixed (my TV has 480 veritcal scan-lines), the horizontal axis can pack as many pixels that the CRT guns of my tube can show...so the original horizontal 720 pixels are still there in the 4x3 box inside the new 16x9 frame. This is a neat trick that my DVD player (panny rp91) can perform and is important to me because my TV locks into "16x9" mode when it senses a progressive-scan input.

The word "anamorphic" misleads folks.

In reality there IS NO ANAMORPHIC DVD...DVD can simple record images in two basic shapes...4x3 or 16x9. Each shape uses 720 by 480 pixels, and if the image you want to scan into that shape doesn't exactly fit...then the extra area is filled with black.