I would like to clarify something,
I would not like the readers to interpret by the way the comment is made that the zoom method is for videophiles and the CinemaScope with anamorphic lens is not, which is not true.
In fact, I could present it as that the zoom method is actually not for videophiles regarding quality when is not making use of the full resolution and light of the projector chip, as the CinemaScope system does "when using top quality scaler/lens/projector", but that is not the purpose of my article, and I intentionally stayed away from comparisons.
I mentioned that there is a place for both methods and the decision is in the hands of consumers, and the choice is mainly driven by money and the quality of the CinemaScope system, but also based on operational complexity of manual adjustments, and the projector's functionality, done in a repetitive manner.
There are people that will definitely not spend what they spent on their 1080p projector on adding a CinemaScope system, and prefer to go for the less costly zoom method, temporarily or permanently.
There are also people that cannot implement the zoom method practically, even though they might want to save the money of a CinemaScope system, because their ceiling-mounted projectors are accessible only by bringing the ladder from the garage and adjusting the non-motorized zoom and focus manually every time they switch the aspect ratio of the movie they watch, a task that many times needs two people to be done correctly. Not to mention doing this in front of the HT guests.
Again, this article is to cover 2.35:1 CinemaScope using the anamorphic lens implementation that is starting to be used widely in the projector industry, and I intentionally stayed away from analyzing and comparing the pros and cons of each method precisely to avoid confusing consumers within a series of articles that is complex enough already.
Best Regards,
Rodolfo La Maestra
CinemaScope™ HDHT - Part 3 - Screens and Aspect Ratios
-
aaronstout
- Member
- Posts: 40
- Joined: Thu Oct 21, 2004 6:46 pm
Well, right after I posted my reply, it dawned on me that 2.40 material is still only going to be encoded on the HD disc as 1920 horizontal pixels regardless and the vertical resolution likewise will be less than 1080, so scaling would have to be done in both directions anyway.
Although I think it might be a "cleaner" way to do it, you would still have scaling issues to deal with. Oh well, for the foreseeable future I will just watch the wider aspect ratio films with a smaller vertical height.
I understand the zoom method as well, but both seem somewhat awkward to implement. I suppose if everything is automated in the Cinemascope arrangement, it would greatly reduce the "hassle" of going between formats. Unfortunately this drives up the cost of such solutions quite a bit.
I do think it's terrific that the cost of all these components is going down however, so hopefully soon a fully automated "package" with projector, scaling processor, lens and masking screen will become more affordable for mainstream consumers...
AaronS
Although I think it might be a "cleaner" way to do it, you would still have scaling issues to deal with. Oh well, for the foreseeable future I will just watch the wider aspect ratio films with a smaller vertical height.
I understand the zoom method as well, but both seem somewhat awkward to implement. I suppose if everything is automated in the Cinemascope arrangement, it would greatly reduce the "hassle" of going between formats. Unfortunately this drives up the cost of such solutions quite a bit.
I do think it's terrific that the cost of all these components is going down however, so hopefully soon a fully automated "package" with projector, scaling processor, lens and masking screen will become more affordable for mainstream consumers...
AaronS
-
Rodolfo
- Author
- Posts: 755
- Joined: Wed Sep 01, 2004 8:46 pm
- Location: Lansdowne VA
aaron,
2.40:1 will have 1920 horizomtally, as yoiu said, but there is nothing to scale in that direction, the full chip is used on that axis.
Only the vertical would be scaled (as 2.35:1), plus the optical stretch by the anamorphic lens to make the image wider, but that is not part of the scaling function.
The 2.40 would still show with two very small bars (like the 1.85 would on 16:9), but a minor overscan adjustment could take care of that. I personally leave it as is because the 2.35:1 image is so poweful that two small bars do not affect me.
As you said prices are coming down for high quality products.
Best Regards,
Rodolfo La Maestra
2.40:1 will have 1920 horizomtally, as yoiu said, but there is nothing to scale in that direction, the full chip is used on that axis.
Only the vertical would be scaled (as 2.35:1), plus the optical stretch by the anamorphic lens to make the image wider, but that is not part of the scaling function.
The 2.40 would still show with two very small bars (like the 1.85 would on 16:9), but a minor overscan adjustment could take care of that. I personally leave it as is because the 2.35:1 image is so poweful that two small bars do not affect me.
As you said prices are coming down for high quality products.
Best Regards,
Rodolfo La Maestra
-
aaronstout
- Member
- Posts: 40
- Joined: Thu Oct 21, 2004 6:46 pm
Rodolfo,
I guess I wasn't being clear when I talked about horizontal scaling. I meant in the hypothetical case of a native imager that had 1080 X 2592 or similar "native" pixels, the HD content from an HD DVD or Blu-ray source would still have to be electronically scaled horizontally (to eliminate the anamorphic lens) as well as scaled vertically (since the image on disc would not have a full 1080 pixels of vertical resolution) to fill the entire imaging chip.
I realize also that some of the imager pixels would be "wasted" on other image ratios, but nothing is really "lost" in that scenario. We always have the highest resolution the image is capable of displaying with minimal distortion of the image. Of course it might be that optically stretching the image would cause far fewer artifacts than electronically doing so. Unfortunately I don't have the exposure to the good "toys" to know personally, so I'll trust your reviews and experience in this regard. There is no substitute for firsthand experience!
That is what I was referring to. I do understand the principles of CIH, but sometimes I think people mis-state things when they say that you lose "vertical resolution" watching the movie with the black bars at the top and bottom. The resolution is not in the image to begin with. You do lose image size and brightness, since much of the imager is not reflecting or passing light (depending on technology). I think it is largely the size difference that is most dramatic, but again with a larger image, the lost brightness is only compounded, so both are important factors.
In any case, I hope you don't get the wrong idea about my posts, I was just brain storming a bit and thought it might make for a good exchange. Presonally at some point in time, I would venture to guess we might see larger pixel count imagers in consumer front projectors. I think I have even read of a Cinemascope width ratio rear projection set for consumer use. I don't remember who was making the prototype, but I'm thinking it was probably JVC. I don't know how they were achieving that though. It may not have been with a larger pixel count imager.
Regards,
AaronS
I guess I wasn't being clear when I talked about horizontal scaling. I meant in the hypothetical case of a native imager that had 1080 X 2592 or similar "native" pixels, the HD content from an HD DVD or Blu-ray source would still have to be electronically scaled horizontally (to eliminate the anamorphic lens) as well as scaled vertically (since the image on disc would not have a full 1080 pixels of vertical resolution) to fill the entire imaging chip.
I realize also that some of the imager pixels would be "wasted" on other image ratios, but nothing is really "lost" in that scenario. We always have the highest resolution the image is capable of displaying with minimal distortion of the image. Of course it might be that optically stretching the image would cause far fewer artifacts than electronically doing so. Unfortunately I don't have the exposure to the good "toys" to know personally, so I'll trust your reviews and experience in this regard. There is no substitute for firsthand experience!
That is what I was referring to. I do understand the principles of CIH, but sometimes I think people mis-state things when they say that you lose "vertical resolution" watching the movie with the black bars at the top and bottom. The resolution is not in the image to begin with. You do lose image size and brightness, since much of the imager is not reflecting or passing light (depending on technology). I think it is largely the size difference that is most dramatic, but again with a larger image, the lost brightness is only compounded, so both are important factors.
In any case, I hope you don't get the wrong idea about my posts, I was just brain storming a bit and thought it might make for a good exchange. Presonally at some point in time, I would venture to guess we might see larger pixel count imagers in consumer front projectors. I think I have even read of a Cinemascope width ratio rear projection set for consumer use. I don't remember who was making the prototype, but I'm thinking it was probably JVC. I don't know how they were achieving that though. It may not have been with a larger pixel count imager.
Regards,
AaronS
-
Rodolfo
- Author
- Posts: 755
- Joined: Wed Sep 01, 2004 8:46 pm
- Location: Lansdowne VA
Aaron,
I agree with the statements regarding original resolution that was never on the image, but I believe you misread my statements regarding which resolution we loose.
We loose the "projector's chip" resolution capability when we put that projector to display 30% of black bars. One clever way the CinemaScope system found to gain back those lines of resolution of the chip (and with them the brightness we agree we need, anticipating we will make the image larger at the end) is by having "a quality scaler" mapping 1080 from the 800+ lines of the image (the part that is not black bars).
Do that scaling job cheaply and one can destroy the image.
We never would have 2592 (horizontal pixels) of resolved image because the telecine transfer is done for a target of 1920 of consumer level media, so that is another problem I see when people start thinking how good would be if manufacturers of consumer products would do 2.35:1 chips with a larger width and more horizontal resolution while maintaining the vertical 1080 (because at times one would need to display a 16x9 fully resolved image). To use the full potential of that imaginary dream chip, the content has to be also with that resolution.
Would that be the Super Blue-laser wide format? Who knows, but as you see we need all the pieces in the image chain to make it work clean, the content, the disc format, the transfer technique, and the chip, AND MAINLY USEFUL FOR 2.35:1 content, systems, people, market, manufacturers; quite an uphill.
This brainstorming is good because it allow us to explore the constraints of having a true 2.35:1 system from the content to your screen; and we could appreciate the benefits a CinemaScope system when playing with all those constraints and still offer a theatrical image.
Regarding "the toys" you referred to, it depends how often you view 2.35:1 movies those toys might become a "must have and cannot live without them".
You have to see the faces of people when I click one button and the system opens the image to a wall to wall 2.35:1 shape while they still sit at the same chair, it fills all the peripheral vision and produces the theatrical impact. Add to it 16 speakers with clean and powerful amplification, and a shaking room, and they never want to leave.
Put a cheap scaler, a cut-corner anamorphic lens, a low light projector, and the image might be big but pitiful, and it might have been better a smaller image with black bars.
So as with any component claiming to reach the level of theater for large screen purposes, each piece is not a toy, it has to be a quality piece, and that usually does not come cheap.
However, the good news is that while this type of theatrical environment used to cost on the several tens of thousands, you have now a way to get some decent 1080p quality of the video pieces under $20K, that is exactly the purpose of this project and the articles.
One good part of this kind of system is that one could experiment with components and eventually upgrade them to better quality all and each individually, a projector, a scaler, the lens, a screen; a situation not possible with large RPTVs for example, where everything is in the same box.
Best Regards,
Rodolfo La Maestra
I agree with the statements regarding original resolution that was never on the image, but I believe you misread my statements regarding which resolution we loose.
We loose the "projector's chip" resolution capability when we put that projector to display 30% of black bars. One clever way the CinemaScope system found to gain back those lines of resolution of the chip (and with them the brightness we agree we need, anticipating we will make the image larger at the end) is by having "a quality scaler" mapping 1080 from the 800+ lines of the image (the part that is not black bars).
Do that scaling job cheaply and one can destroy the image.
We never would have 2592 (horizontal pixels) of resolved image because the telecine transfer is done for a target of 1920 of consumer level media, so that is another problem I see when people start thinking how good would be if manufacturers of consumer products would do 2.35:1 chips with a larger width and more horizontal resolution while maintaining the vertical 1080 (because at times one would need to display a 16x9 fully resolved image). To use the full potential of that imaginary dream chip, the content has to be also with that resolution.
Would that be the Super Blue-laser wide format? Who knows, but as you see we need all the pieces in the image chain to make it work clean, the content, the disc format, the transfer technique, and the chip, AND MAINLY USEFUL FOR 2.35:1 content, systems, people, market, manufacturers; quite an uphill.
This brainstorming is good because it allow us to explore the constraints of having a true 2.35:1 system from the content to your screen; and we could appreciate the benefits a CinemaScope system when playing with all those constraints and still offer a theatrical image.
Regarding "the toys" you referred to, it depends how often you view 2.35:1 movies those toys might become a "must have and cannot live without them".
You have to see the faces of people when I click one button and the system opens the image to a wall to wall 2.35:1 shape while they still sit at the same chair, it fills all the peripheral vision and produces the theatrical impact. Add to it 16 speakers with clean and powerful amplification, and a shaking room, and they never want to leave.
Put a cheap scaler, a cut-corner anamorphic lens, a low light projector, and the image might be big but pitiful, and it might have been better a smaller image with black bars.
So as with any component claiming to reach the level of theater for large screen purposes, each piece is not a toy, it has to be a quality piece, and that usually does not come cheap.
However, the good news is that while this type of theatrical environment used to cost on the several tens of thousands, you have now a way to get some decent 1080p quality of the video pieces under $20K, that is exactly the purpose of this project and the articles.
One good part of this kind of system is that one could experiment with components and eventually upgrade them to better quality all and each individually, a projector, a scaler, the lens, a screen; a situation not possible with large RPTVs for example, where everything is in the same box.
Best Regards,
Rodolfo La Maestra
-
rpeckrpeck
- Member
- Posts: 11
- Joined: Thu Aug 04, 2005 6:14 am
direct-scan laser will make this moot before most folks care
If I had bags of money I'd definitely go for a sled-based anamorphic 1080p fp system today, but Joe six-pack doesn't have any clue.
IMO, before most folks care about constant-height or anamorphic 2:35:1 we will have direct-scan laser projectors (projectors which generate the
image by directly scanning one or more lasers onto the screen, rather than using a microdisplay device with fixed resolution).
Such a projector will have no need to scale a 2:35:1 image vertically to 16:9 and then to shrink the image back to 2:35:1 using an anamorphic lens. . .
And whlie I'm throwing around predictions, I expect that such projectors will pretty quickly be used with motion-adaptive scalers to show scaled 2k or 4k images from 1080p sources.
IMO, before most folks care about constant-height or anamorphic 2:35:1 we will have direct-scan laser projectors (projectors which generate the
image by directly scanning one or more lasers onto the screen, rather than using a microdisplay device with fixed resolution).
Such a projector will have no need to scale a 2:35:1 image vertically to 16:9 and then to shrink the image back to 2:35:1 using an anamorphic lens. . .
And whlie I'm throwing around predictions, I expect that such projectors will pretty quickly be used with motion-adaptive scalers to show scaled 2k or 4k images from 1080p sources.
-
raka
- New Member
- Posts: 1
- Joined: Mon Oct 18, 2004 6:32 pm
another option?
a 2.35 chip would obviously the best - but due to manufacturing expense & limited market, probably not any time soon.
however, my projector ALREADY has a decent scaler AND a lens - the 2 required ingredients, so it should be a lot simpler (compared to custom chip) to do the following:
1- ADD an option to the projector's built-in scaler to fill the chip with a (distorted) 2.35 image using all pixels
2- ADD a built-in lens element inside the projector that can slide in & out of the light path to stretch the 2.35 image. since this element will work with only 1 dedicated specific lens, it can be quite small & compact - maybe featuring a simple lever sticking out of the projector that can be manually used. (or make it motorized & auto-sensing for more $$$)
however, my projector ALREADY has a decent scaler AND a lens - the 2 required ingredients, so it should be a lot simpler (compared to custom chip) to do the following:
1- ADD an option to the projector's built-in scaler to fill the chip with a (distorted) 2.35 image using all pixels
2- ADD a built-in lens element inside the projector that can slide in & out of the light path to stretch the 2.35 image. since this element will work with only 1 dedicated specific lens, it can be quite small & compact - maybe featuring a simple lever sticking out of the projector that can be manually used. (or make it motorized & auto-sensing for more $$$)