AVS Forum banner
1 - 19 of 19 Posts

·
Registered
Joined
·
5,355 Posts
Discussion Starter · #1 ·
I'm wondering if this misnomer adds to the confusion many people seem to have with this simple concept.


If I have a 4:3 screen (and PJ), then my set up is 100% constant width. Every aspect ratio is the EXACT same width. This is true on a 16:9 screen if I never watch 4:3 or 1.66:1 content. But even then, with a 16:9 screen, your set up is constant width for all ARs greater than 16:9 and not-constant width for ARs less than that.


OK. The vast majority of constant height set ups are no where near that. There is much more variability of height than there is width in a CW set up. 16:9 content and 1.85 content are different heights on most "constant height" set ups. 2.35:1 and 2.76:1 content are different heights on a "constant height" set up. 2.2:1 content will either lose some content at the top and bottom, or be yet another height!


What most "constant" height set ups are are really "dual width" setups. You are basically fixed width at either 16:9 or 2.37:1.


The only people approaching constant height would be those that leave the lens in place all the time and scale to not use all the pixels (and do things like slight vertical and horizontal stretching of 1.85:1 so that it is actually wider than 1.78--but this type of set up is not possible with a movable lens). But, even they don't have constant height with 2.76:1 content.


So, why do we use the misnomer "constant height" when to this day, I've never actually seen a true constant height system in a home? "Dual Width" seems to be a much more accurate description of what most people have.
 

·
Registered
Joined
·
28,939 Posts

Quote:
Originally Posted by Person99 /forum/post/14174862


I'm wondering if this misnomer adds to the confusion many people seem to have with this simple concept.

You're way overthinking this.
 

·
Registered
Joined
·
3,069 Posts
I don't know but to me it seems perfectly obvious. If I watch 16:9 on my 2.35 screen it is 51" high by 91 wide, If I watch 2.35 it is 51" high and 120" wide 4:3 on the rare ocassions that I watch it are also 51" high.


Now true I use my projector to vertically stretch the 2.35 picture and my lens to horizontally stretch it to fit the screen but it is that height non the less. any descrepancy is corrected by zooming in or out a tad. So the name makes perfect sense to me.
 

·
Registered
Joined
·
24,952 Posts

Quote:
Originally Posted by McCall /forum/post/14175101


I don't know but to me it seems perfectly obvious. If I watch 16:9 on my 2.35 screen it is 51" high by 91 wide, If I watch 2.35 it is 51" high and 120" wide 4:3 on the rare ocassions that I watch it are also 51" high.


Now true I use my projector to vertically stretch the 2.35 picture and my lens to horizontally stretch it to fit the screen but it is that height non the less. any descrepancy is corrected by zooming in or out a tad. So the name makes perfect sense to me.

This is the explanation that I would give. It is no more complex than that.


Of course there are exceptions like Ben Hur etc but these are relatively few as are 2.55:1 films.
 

·
Registered
Joined
·
390 Posts

Quote:
Originally Posted by Person99 /forum/post/14174862


I'm wondering if this misnomer adds to the confusion many people seem to have with this simple concept.


If I have a 4:3 screen (and PJ), then my set up is 100% constant width. Every aspect ratio is the EXACT same width. This is true on a 16:9 screen if I never watch 4:3 or 1.66:1 content. But even then, with a 16:9 screen, your set up is constant width for all ARs greater than 16:9 and not-constant width for ARs less than that.


OK. The vast majority of constant height set ups are no where near that. There is much more variability of height than there is width in a CW set up. 16:9 content and 1.85 content are different heights on most "constant height" set ups. 2.35:1 and 2.76:1 content are different heights on a "constant height" set up. 2.2:1 content will either lose some content at the top and bottom, or be yet another height!


What most "constant" height set ups are are really "dual width" setups. You are basically fixed width at either 16:9 or 2.37:1.


The only people approaching constant height would be those that leave the lens in place all the time and scale to not use all the pixels (and do things like slight vertical and horizontal stretching of 1.85:1 so that it is actually wider than 1.78--but this type of set up is not possible with a movable lens). But, even they don't have constant height with 2.76:1 content.


So, why do we use the misnomer "constant height" when to this day, I've never actually seen a true constant height system in a home? "Dual Width" seems to be a much more accurate description of what most people have.

Uh, it's called constant height because the height is constant. In other words, the height of the image never changes, just the width. For example, I have a 2:35 screen that is 35.75" x 84" when playing 2:35 material and 35.75" x 64" when playing 16:9 material.


Is this really that complicated?
 

·
Registered
Joined
·
114 Posts
i thought Person99 had an interesting point and technically he is right

the usual CIH setup has only constant height with 3 ARs 1,33:1 | 1,78:1 and 2,35:1, every other AR has a different height and there are some almost equally common 1,85:1 and 2,40:1

surely the 1st three are the most common, but constant height is technically wrong
 

·
Registered
Joined
·
5,588 Posts

Quote:
Originally Posted by Theron2 /forum/post/14237049


i thought Person99 had an interesting point and technically he is right

the usual CIH setup has only constant height with 3 ARs 1,33:1 | 1,78:1 and 2,35:1, every other AR has a different height and there are some almost equally common 1,85:1 and 2,40:1

surely the 1st three are the most common, but constant height is technically wrong

No, it's technically correct if it is completely implemented.
If you do a complete implementation using either an HTPC or a scaler and a fixed anamorphic lens, then ALL AR's can be projected at the same height.


If the installer / owner chooses to incorrectly / incompletely implement CIH in their HT for whatever reason, then that is their choice and does not invalidate the term.


For example, in my HT, any AR including, but not limited to 1.20:1, 1.33:1, 1.66:1, 1.75:1, 1.77:1, 1.85:1, 2.00:1, 2.20:1, 2.35:1, 2.40:1, 2.55:1, and 2.66:1 uses exactly the same image height.


You can't get much more CIH than that.



Vern
 

·
Registered
Joined
·
1,210 Posts
With the prism setup of a Prismasonic you can actually vary the width of the picture to any size you want (well within reason). If you make the image the same height for all ratios and you know what the correct width should be you can adjust the prisms to match the proper width. This should maintain the correct aspect ratio for all films. I do this for my oddball films like Woodstock (2.20:1 on the DVD) etc..


Laters,

Jeff
 

·
Registered
Joined
·
5,588 Posts

Quote:
Originally Posted by eq_shadimar /forum/post/14237259


With the prism setup of a Prismasonic you can actually vary the width of the picture to any size you want (well within reason). If you make the image the same height for all ratios and you know what the correct width should be you can adjust the prisms to match the proper width. This should maintain the correct aspect ratio for all films. I do this for my oddball films like Woodstock (2.20:1 on the DVD) etc..


Laters,

Jeff

No this won't work without a scaler or HTPC. You can't simply change one dimension of an image without messing up the proportions of the image and turning circles into eggs....



Vern
 

·
Registered
Joined
·
1,210 Posts
Humm well I thought I was doing it without a scaler or HTPC..maybe not? I use the Vertical stretch option on my projector (ok I will give you that this is a scaler of sorts) to nuke the black bars and project the image full height without any black bars (52" in my case). I dunno how the projector figures this out but it seems to work without clipping off any image information on all the oddball films I have tried it with. Then I just figure out what the width should be using simple math and I make the image that width using the knobs on the Prismasonic lens. All of the circles and squares appear to be correct when I do this. I dunno.


Laters,

Jeff
 

·
Registered
Joined
·
114 Posts
@ Vern

what AR is your screen?


i suppose not wider than 2.40:1?

if you really watch all ARs with a constant image height, then you would cut off the sides on movies wider than 2.40:1, i admit there are only a handful


But i wouldn´t call a moveable lens "not completely implemented" it´s just a slightly different philosophy against a fixed setup.

A fixed setup introduces the lens artefacts and scaling degredation even in ARs that could be displayed with CIH without scaling/lens (every AR including and under 1.85:1 assuming with a movable lens you zoom a bit for 1.85:1)

Further you use less panel pixels and therefore have lower light output for every AR including and under 1.85:1, which is exactly the opposite most ppl with a lens setup want

A movable setup allows you to watch the most common ARs (2.35:1 and everything under and including 1.85:1) in the best possible quality with image degredation through scaling and lens only when it´s really necessary
 

·
Registered
Joined
·
5,588 Posts
Theron2:


My screen is 60" x 160". I use a 1.5x expansion Isco Cinema DLP anamorphic lens and a Qualia 004 with the medium throw prime lens.


As far as your arguments for a movable lens, they are incorrect on all three points:


A movable lens corrupts video level calibration, requiring two different calibration settings to maintain a correct black level, one for the lens in the light path and one for it not in the light path.


With a fixed anamorphic, the light level on the screen stays exactly the same for all AR's because the reduction in illuminated pixels exactly matches the reduction in illuminated screen area for each AR.


A high quality anamorphic does not cause image degradation visible to the human eye or visible artifacts other than a slight pincushion that remains consistent for all AR's. This enables a consistent and accurate pincushion correction with screen curvature. This is not possible with a movable lens setup. Scope will have pincushion, flat will not. Short of using two screens, you will have to deal with both cases.


Jeff:

Effectively, you are scaling in both dimensions, horizontally with the lens, and vertically with the projector. Of course, it's a little awkward to do it this way when compared to selecting any one of a number of AR's on the fly by pressing a button on a remote or a hot key combination on an HTPC.... A good example: supplemental materials are often in a different AR than the feature.


Vern
 

·
Registered
Joined
·
7,335 Posts

Quote:
Originally Posted by Vern Dias /forum/post/14240902


Theron2:


My screen is 60" x 160". I use a 1.5x expansion Isco Cinema DLP anamorphic lens and a Qualia 004 with the medium throw prime lens.


As far as your arguments for a movable lens, they are incorrect on all three points:


A movable lens corrupts video level calibration, requiring two different calibration settings, one for the lens in the light path and one for it not in the light path.

I would be surprised to find that a lens in path has a visibly noticable color shift. An instrument color shift? Quite possble.

Quote:
With a fixed anamorphic, the light level on the screen stays exactly the same for all AR's because the reduction in illuminated pixels exactly matches the reduction in illuminated screen area for each AR.

OTOH, some of us like to get every pixel of resolution we can get. So taking the lens out of the light path is a worthwhile endeavor. You're going to lose information when you do the squeeze for any aspect ratio besides 2.35:1.


You pick your poison, I'll pick mine, thank you very much




Quote:
A good quality anamorphic does not cause visible artifacts other than a slight pincushion that remains consistent for all AR's. This enables a consistent and accurate correction with screen curvature. This is not possible with a movable lens setup. Scope will have pincushion, flat will not. Short of using two screens, you will have to deal with either case.

The curvature of the screen is such that the pin cushioning is not noticable on 1.85:1, 1.78:1 or 1.33:1 material.


Quote:
Effectively, you are scaling in both dimensions, horizontally with the lens, and vertically with the projector. Of course, it's a little awkward to do it this way when compared to pressing a button on a remote or a hot key combination on an HTPC....


Vern

Perhaps it's just semantics, but I don't see the anamorphic lens stretch as scaling, but rather as optical correction to a distorted image. There's no massaging of the pixels which is what I consider scaling to be.


Like I said, probably semantics.
 

·
Registered
Joined
·
5,588 Posts
John, color shift is not the issue. The addidtional light loss caused by the anamorphic is the issue.


As far as increased resolution without the lens, this is a common misconception among those who advocate movable anamorphic lens approach.


In my experience, it's not an issue:


A. With a 1080 projector

B. If you are sitting at an optimum seating distance where you can just no longer see the pixels with the anamorphic lens in place.


In this case, viewable resolution is determined by the pixels on the screen and the visual acuity of the human eye. If you can't distinguish the individual pixels with lens in place, then removing the lens simply makes the pixels smaller and will have no effect on the perceived resolution.


The effective resolution is directly controlled by visual acuity and you are already sitting at the point where any improvements in on screen resolution due to reduced pixel size are lost due to the resolving capabilities of the human eye. Of course, if you have a 720 projector, or you sit close enough to make out the individual pixels with the lens in place, the above logic doesn't apply.


Vern
 

·
Registered
Joined
·
7,335 Posts

Quote:
Originally Posted by Vern Dias /forum/post/14241190


John, color shift is not the issue. The addidtional light loss caused by the anamorphic is the issue.

Okay, but you're losing light and resolution at 1.85/1.78 and 1.33.


You have

You have 1440x1080 for 1.78:1 material instead of 1920x1080.

You have3 1080x1080 for 1.33:1 material instead of 1440x1080.


So you're willing to sacrifice 1/3 of the resolution available to keep your system pseudo "constant brightness". I'm not.

Quote:
As far as increased resolution without the lens, this is a common misconception among those who advocate movable anamorphic lens approach.

See above, it's quite simple mathematically.

Quote:
In my experience, it's not an issue:


A. With a 1080 projector

B. If you are sitting at an optimum seating distance where you can just no longer see the pixels with the anamorphic lens in place.


In this case, viewable resolution is determined by the pixels on the screen and the visual acuity of the human eye. If you can't distinguish the individual pixels with lens in place, then removing the lens simply makes the pixels smaller and will have no effect on the perceived resolution.


The effective resolution is directly controlled by visual acuity and you are already sitting at the point where any improvements in on screen resolution due to reduced pixel size are lost due to the resolving capabilities of the human eye. Of course, if you have a 720 projector, or you sit close enough to make out the individual pixels with the lens in place, the above logic doesn't apply.


Vern

I sit at 1.1x for 2.35:1 content and at 1.5x (give or take) for 1.85:1 content. I see the loss in resolution.


There is one factor you aren't considering, which is individual visual acuity. It doesn't stop at 20/20. Personally, I have correction which puts me at 20/18 and there are others out there with 20/15 vision. 20/20 isn't perfect, it's normal.


Put up a resolution pattern with and without the anamorphic lens and tell me they're identical on screen. You can't because you've lost resolution in the scaling process. No, you don't watch test patterns but they can tell you what to look for in regular viewing.


It's a matter of choice in this not right or wrong, and each is an equally valid choice.
 

·
Registered
Joined
·
114 Posts

Quote:
Originally Posted by Vern Dias /forum/post/14240902


A movable lens corrupts video level calibration, requiring two different calibration settings to maintain a correct black level, one for the lens in the light path and one for it not in the light path.

alright, but what´s the problem to calibrate the PJ 1 times more, many PJs have 2 custom settings in addition to the presets, so that´s not really a good argument from my POV, you have to push a button to switch the scaling, i push a button to change the custom settings

Quote:
Originally Posted by Vern Dias /forum/post/14240902


With a fixed anamorphic, the light level on the screen stays exactly the same for all AR's because the reduction in illuminated pixels exactly matches the reduction in illuminated screen area for each AR.

the most common ARs have the same brightness with a movable setup, with the mentioned 2 calibration settings as well, because the reduction of screen and panel area usage is the same in relation to a fixed setup, only ARs between 1.85:1 and 2.35 and above 2.35 would differ in brightness, but over 2.35:1 you shouldn´t have the same brightness aswell since the screen area is getting bigger but the panel pixels are already all in use @ 2.35:1, if you spread this brightness on a wider area to avoid cutting off the sides, then your "above 2.35:1" pic should be dimmer aswell or am i missing something?


Quote:
Originally Posted by Vern Dias /forum/post/14240902


A high quality anamorphic does not cause image degradation visible to the human eye or visible artifacts other than a slight pincushion that remains consistent for all AR's.

but scaling does have a negative effect on the pic and my philosophy is always if you don´t HAVE to temper with the original picture quality, then don´t

btw not everybody is able to afford a high end lens and for them i believe the movable setup would result in an better pic for the rest of the ARs that don´t need scaling per se, i have a Isco 3 though but still i prefer the movable setup, each his own i guess

Quote:
Originally Posted by Vern Dias /forum/post/14240902


This enables a consistent and accurate pincushion correction with screen curvature. This is not possible with a movable lens setup. Scope will have pincushion, flat will not. Short of using two screens, you will have to deal with both cases.

i give you that, constant pincushion correction with the lens/curved screen combo is a plus, but still as i see it on the expence of overall pic quality for the "no scaling needed ARs"
 

·
Registered
Joined
·
5,588 Posts

Quote:
You have 1440x1080 for 1.78:1 material instead of 1920x1080.

You have 1080x1080 for 1.33:1 material instead of 1440x1080.

Yep, and you have a corresponding reduction in screen size as well, 60" x 106" for 1.78:1 and 60" x 80" for 1.33, so the perceived resolution remains constant.
Quote:
I sit at 1.1x for 2.35:1 content and at 1.5x (give or take) for 1.85:1 content. I see the loss in resolution.

I sit at 1.2 for 2.40:1 and I don;t see any loss of resolution because the horizontal viewing angle is being correspondingly reduced.
Quote:
Put up a resolution pattern with and without the anamorphic lens and tell me they're identical on screen. You can't because you've lost resolution in the scaling process. No, you don't watch test patterns but they can tell you what to look for in regular viewing.

Yes, you can see the (slight) difference on test patterns, but, show me any movie that contains alternating black and white vertical lines one pixel wide.
Quote:
the most common ARs have the same brightness with a movable setup, with the mentioned 2 calibration settings as well, because the reduction of screen and panel area usage is the same in relation to a fixed setup,

No, they don't. Take a 5' high screen for example and a movable anamorphic lens: In the 1.77:1 case without the lens, all the pixels are illuminating a 60" x 106" screen area (6360 sq "), while in the 2.40:1 case with the lens all the pixels are illuminating a 60" x 144" screen area (8640 sq"). Add the slight light loss through the lens, and the issue is even more severe. As a result the 2.40:1 image is always going to be dimmer than the 1.78:1 image. While a calibration can compensate for the black level change, it can't do anything about the decrease in white level.
Quote:
only ARs between 1.85:1 and 2.35 and above 2.35 would differ in brightness, but over 2.35:1 you shouldn´t have the same brightness aswell since the screen area is getting bigger but the panel pixels are already all in use @ 2.35:1, if you spread this brightness on a wider area to avoid cutting off the sides, then your "above 2.35:1" pic should be dimmer aswell or am i missing something?

In my case, using a 1.5x anamorphic lens. All 1920 panel pixels are used to fill the 2.66:1 screen, not a 2.40:1 screen.


Other thoughts:


Granted, people have differing visual acuity, but for someone who has better than 20/20 vision, they will have to increase their viewing distance to avoid seeing the pixels that compose the image.


Also, this setup is clearly not for everone. We all make choices for out HT setups and live with the compromises that result.


Having spent close to 50 years of my life as a professional projectionist, I want maximum flexibility with no perceivable loss in image quality. Noting looks as tacky to me as watching someone change AR's with a slide and watching as the image slowly slides off the screen and on to the walls.


I have been doing CIH in my HT for over 20 years now, and I have played with many combinations of projectors and lenses and I would never go back to a movable anamorphic setup. Just my $.02.


It's been a good lively discussion, but I think I'm going to go watch "Mutiny On The Bounty" in it's full CIH 2.65:1 Ultra Panavision glory (with no black bars top and bottom).


Vern
 

·
Registered
Joined
·
114 Posts

Quote:
Originally Posted by Vern Dias /forum/post/14247162


In my case, using a 1.5x anamorphic lens. All 1920 panel pixels are used to fill the 2.66:1 screen, not a 2.40:1 screen.

oops yeah, completely forgot, damn i´m not getting younger


Quote:
Originally Posted by Vern Dias /forum/post/14247162


while in the 2.40:1 case with the lens all the pixels are illuminating a 60" x 144" screen area (8640 sq"). Add the slight light loss through the lens, and the issue is even more severe. As a result the 2.40:1 image is always going to be dimmer than the 1.78:1 image. While a calibration can compensate for the black level change, it can't do anything about the decrease in white level.

true, my mistake, the 2.35:1 image is dimmer, while 1.78:1 is as bright as can be, but from a different POV a movable setup allows you to increase the brightness for 1.78:1 (sort of positiv thinking
) and i would see that increase as a plus, instead of the change in brightness as a negative

on the other hand with a fixed setup every AR has the same brightness, but on the brightness level of your biggest AR and could have a brighter pic with some ARs that wouldn´t need scaling/lens

Quote:
Originally Posted by Vern Dias /forum/post/14247162


I have been doing CIH in my HT for over 20 years now, and I have played with many combinations of projectors and lenses and I would never go back to a movable anamorphic setup. Just my $.02.

don´t get me wrong, i don´t want to nag and i totally respect your decision for your setup, i sometimes just need to "say things out loud" to better understand and/or be corrected to learn new stuff



Quote:
Originally Posted by Vern Dias /forum/post/14247162


I'm going to go watch "Mutiny On The Bounty" in it's full CIH 2.65:1 Ultra Panavision glory (with no black bars top and bottom).

that sounds pretty cool, do you care to post a pic when you have the time? If you don´t want it public you can send me a PM, would be interesting to see, i guess i won´t often get the chance for something like that
 
1 - 19 of 19 Posts
Top