[Thunder]

Hey Guys, I have a Lumagen HDP Pro. I have a in a CIH set up and am looking into getting a HD player. My projector does not do vertical stretching so at minimum, the VP will be retained to do this function. Question is though, are there any advantages or disadvantages of sending a 1080P signal to the VP to just have it stretch it? Let's assume for the purpose of this question that the deinterlacing and upscaling (in the case of DVD) of the source is as good if not better than the Lumagen.


Other question is, will the Lumagen have to completely re-process the signal in order to stretch it or is it a simple step that will not negate quality deinterlacing for the source?


Hope you can help demystify this for me. Thanks.

 

[usualsuspects]

The Lumagens will not accept greater than 1080p24/i60 on the inputs, so none of the current HD-DVD players (that I know of) will be able to send anything other than 1080i60 to a Lumagen. That is fine, the Lumagen will de-interlace 1080i60 back to 1080p24 and scale as necessary. This has worked great for me for quite some time. Three issues here: scaling, de-interlacing, and image enhancement. The question is - what device does each thing the best? Scaling: toss up between Toshiba and Lumagen - your call. De-interlacing - unknown - good de-interlacing is known to happen on the Lumagens, unknown how well the Toshibas do it. Image enhancement - reports are that the Reon on the XA2 is outstanding, but does it work at 1080i60? Conflicting reports here, but with a promised 1080p24 output soon on the XA2. Ah, the fun of bleeding edge equipment

I like it! Roll the dice, I did! XA2 on the way!

 

[Barry928]

Assuming you can get a BD/HDVD player to output 1080p I would not use any processing or optical stretching at all. The only advantage is 20% more light output by filling the 16:9 chip. Since neither disc format is offering an anamorphic option there is no resolution gain and running the native 1920 x 810 image through a processor and then an optic conversion will most likely degrade the image. I am re-thinking my 2:35 plans to go instead with a 16:9 screen with vertical masking because of the new discs.

 

[usualsuspects]

Ben there, done that. Try it and see. Theory and practice are two different things.

 

[D_B_0673]



Quote:
Originally Posted by Barry928 /forum/post/0


Assuming you can get a BD/HDVD player to output 1080p I would not use any processing or optical stretching at all. The only advantage is 20% more light output by filling the 16:9 chip. Since neither disc format is offering an anamorphic option there is no resolution gain and running the native 1920 x 810 image through a processor and then an optic conversion will most likely degrade the image. I am re-thinking my 2:35 plans to go instead with a 16:9 screen with vertical masking because of the new discs.

Why is there no resolution gain? If you stretch to fill the 16x9 screen and then use a lens to recompress it into the 2.35 size you are using all the vertical pixels (1080) I am only speaking about a constant width setup


Am I missing something?

 

[Barry928]

If you take 810 pixels and stretch them taller to map across 1080 pixels then optically squish them back down to the original shape no resolution is gained.

 

[Thunder]



Quote:
Originally Posted by usualsuspects /forum/post/0


The Lumagens will not accept greater than 1080p24/i60 on the inputs, so none of the current HD-DVD players (that I know of) will be able to send anything other than 1080i60 to a Lumagen. That is fine, the Lumagen will de-interlace 1080i60 back to 1080p24 and scale as necessary. This has worked great for me for quite some time. Three issues here: scaling, de-interlacing, and image enhancement. The question is - what device does each thing the best? Scaling: toss up between Toshiba and Lumagen - your call. De-interlacing - unknown - good de-interlacing is known to happen on the Lumagens, unknown how well the Toshibas do it. Image enhancement - reports are that the Reon on the XA2 is outstanding, but does it work at 1080i60? Conflicting reports here, but with a promised 1080p24 output soon on the XA2. Ah, the fun of bleeding edge equipment

I like it! Roll the dice, I did! XA2 on the way!

Usualsuspects, thanks for the very informative post. To clarify a few things:

* Does a VP deinterlace first, then scale, then vertically stretch? Or is it in a different order?

* When you say 1080p24, is that the same as 1080 24sf?


To confirm:

*When the VP receives the 1080p24 signal, all it would do is the required stretch and not re-process it in any way, retaining the image enhancements made through the source. Correct?

*When VP receives an interlaced signal, it will need to deiterlace and when it does, any image enhancement from the source will be lost (this assumes that the source can even provide image enhancements through an interlaced signal - as you point out, this is uncertain).

 

[Thunder]



Quote:
Originally Posted by Barry928 /forum/post/0


If you take 810 pixels and stretch them taller to map across 1080 pixels then optically squish them back down to the original shape no resolution is gained.

But you're missing the advantage of CIH. You do gain in pixel resolution as opposed to source resolution. Instead of wasting pixels on black bars top and bottom, they actually get used. In other words, you would be using 1080 vertical pixels of the projector to display 810 lines of source resolution as opposed to limiting yourself to 810 pixels. That works out to +33% INCREASE IN RESOLUTION. Higher pixel resolution is always an advantage and that's why 720 and 1080 resolution machines were purchased even though most source material remains at 480 (for SD material).


You also take full advantage of the projectrs light output.

 

[Barry928]

I agree with the increase in light output but I do not agree there is an actual resolution improvement. When you factor in the stretch by the scaler and then passing through a second lens to reformat the resolution may actually decrease or stay the same depending on the quality of both the scaler and the lens. I am a fan of the 2:35 screen and love how it looks at the trade shows but the added expense of the high quality second lens and a motorized sled is hard to justify for only a light increase when displaying 1080p native source material.

 

[usualsuspects]



Quote:
Originally Posted by Thunder /forum/post/0


Does a VP deinterlace first, then scale, then vertically stretch? Or is it in a different order?

De-interlacing first, then stretch/scale - note that stretch/scale is one operation, not two operations.

Quote:
When you say 1080p24, is that the same as 1080 24sf?

From a practical standpoint P and PSF (segmented frame) are the same. They are different formats, but the point of both is to preserve the progressive nature of the signal. The only reason to choose one over the other is equipment compatibility reasons.

Quote:
When the VP receives the 1080p24 signal, all it would do is the required stretch and not re-process it in any way, retaining the image enhancements made through the source. Correct?

Yes - that is the point exactly, the VP will re-process the image (stretch it) but you will not loose the image enhancements.

Quote:
When VP receives an interlaced signal, it will need to deiterlace and when it does, any image enhancement from the source will be lost

Not true - I am going to restrict my answer to film sources, video is a different story, but we are not interested in that right now. De-Interlacing film sources that have been enhanced does not degrade the image in any way or reverse any manipulation of the image that occurred before the image was interlaced/de-interlaced. If Interlacing and de-interlacing of progressive sources are done correctly it has zero impact on image quality

Quote:
(this assumes that the source can even provide image enhancements through an interlaced signal - as you point out, this is uncertain).

Yeah, that is the wildcard here, will the Reon do its thing for an interlaced output. There is no reason why it should not work with an interlaced output, but logic and Toshiba HD-DVD players are two different things

The Lumagen will accept 1080p24 on the inputs, so hopefully the worst case scenario is that I will have to wait for a 1080p24 output firmware update for the XA2 (promised).

 

[Thunder]

As "usual" and as I "suspected", a great response. Thanks

 

[usualsuspects]

Got my XA2 last night - Awesome! I mostly concentrated on looking at SD performance vs my HD-A1. SD scaling looks to be better than the A1. I turned on the three SD noise filters (block/mosquito/random noise) and now have the best scaled SD image I have ever seen - easily beats my HD-A1 and Oppo971H/SDI. It works perfectly at 1080i into the Lumagen HDQ.

 

[Thunder]



Quote:
Originally Posted by usualsuspects /forum/post/0


Got my XA2 last night - Awesome! I mostly concentrated on looking at SD performance vs my HD-A1. SD scaling looks to be better than the A1. I turned on the three SD noise filters (block/mosquito/random noise) and now have the best scaled SD image I have ever seen - easily beats my HD-A1 and Oppo971H/SDI. It works perfectly at 1080i into the Lumagen HDQ.

Wow, awesome!


So all of the video procesing is not lost via 1080i output? I am really tempted to pull thr trigger and replace my current source (pioneer 59avi) with one of these. I think you just pushed me over the edge

 

[usualsuspects]



Quote:
Originally Posted by Thunder /forum/post/0


So all of the video procesing is not lost via 1080i output?

You can pause the disk, bring up the "picture" menu, and see the effect as you adjust each filter on-screen at 1080i. One thing, you may have an issue with this player if you object to PC levels vs video levels. I did not try messing with the EDID settings in the Lumagen to see if I could get 422 component video levels out of it. It may or may not work, have not messed around with that yet.