or Connect
New Posts  All Forums:Forum Nav:

Sony VPL-vw1000 - Page 15

post #421 of 9736
Quote:
Originally Posted by d.j. View Post

Millerwill

because there you really can see what the upscalling to 4K do - wow, very impressive .


dj


Yes, the upscaling part starts from the 8th minute in the video:

http://www.youtube.com/watch?v=zStevY7wegk


Best Regards,
Ekie
post #422 of 9736
Yeah. Looks like it is really beneficial for 1080p. I've learned something!
post #423 of 9736
Thread Starter 
Quote:
Originally Posted by AV Science Sales 4 View Post

It seems to me that the JVCs are flashing two times 1920 x 1080 pixels on the screen while the Sony is flashing 4096 x 2160 pixels on the screen. To me this means that the Sony is flashing over twice as many pixels.

I have very little understanding about how JVC's eshift works but JVC claims that their units can do 3840X2160 (8,294,400 pixels) while Sony claims that it can do 4096X2160 (8,847,360 pixels). So the difference is 552,960 pixels.

Is JVC's information wrong or misleading?
post #424 of 9736
Does the vw1000 have the ability to display a 16:9 ratio image inside a 2.35:1 frame? That way lens shift would not be required and all the pixel information would still be in the 16:9 image
post #425 of 9736
I'm not sure if this has been mentioned, but the Cine4Home review mentions the projector having 12k:1 to 20k:1 native on/off CR.

--Darin
post #426 of 9736
Quote:
Originally Posted by darinp2 View Post

I'm not sure if this has been mentioned, but the Cine4Home review mentions the projector having 12k:1 to 20k:1 native on/off CR.

--Darin

I noticed that: it's certainly much better than what HT reported, and one hopes that it continues to improve toward the finished product. It's hard to see how they could get the DI CR up to 1M:1, or even 500K:1 unless the native is 2 or 3 times higher.
post #427 of 9736
Quote:
Originally Posted by millerwill View Post

I noticed that: it's certainly much better than what HT reported, and one hopes that it continues to improve toward the finished product. It's hard to see how they could get the DI CR up to 1M:1, or even 500K:1 unless the native is 2 or 3 times higher.

Or they decided to play some tricks like Panasonic and have a mode for specing where if they notice a blackout scene then after a few seconds they start shutting down more and continue shutting down for close to 20 seconds. In that case it is close to useless imaging (and possibly brings more negative than positive to what it does to the images overall), but it allows for claiming a high number.

I see in the Cine4Home article that they are more aggressive with their dynamic iris, but I don't know if it is one of these slow shutting kind of things.

--Darin
post #428 of 9736
Quote:
Originally Posted by darinp2 View Post

Or they decided to play some tricks like Panasonic and have a mode for specing where if they notice a blackout scene then after a few seconds they start shutting down more and continue shutting down for close to 20 seconds. In that case it is close to useless imaging (and possibly brings more negative than positive to what it does to the images overall), but it allows for claiming a high number.

I see in the Cine4Home article that they are more aggressive with their dynamic iris, but I don't know if it is one of these slow shutting kind of things.

--Darin

We probably won't know until cine4home gets his hands on a production unit!
post #429 of 9736
Quote:
Originally Posted by Lawguy View Post

I have very little understanding about how JVC's eshift works but JVC claims that their units can do 3840X2160 (8,294,400 pixels) while Sony claims that it can do 4096X2160 (8,847,360 pixels). So the difference is 552,960 pixels.

Is JVC's information wrong or misleading?

I am not saying anything other than to create their 4K, it would appear that JVC is flashinbg in sequence two different 1080p frames, one frame shifted one half pixel, horizontally and vertically. Each frame contains 1920 x 1080 pixels. There are two frames flashed and then repeated. This would be a total of about 4 million different pixels flashed on the screen, not 8 million pixels. Tht's all I am saying. I like it. I think it improves things but untils omeone can explain it in other than the pixels overlap (I don't think they do), I just don't see how the JVC can put up 8 million pixels. I know the Sony does because its panel is 4096 x 2160. But the JVC flashes only two 1080p frames, one shifted. Without the shift the two different frames would overlap producing garbage. But some one needs to explain to me how two 1080p flashes can result in 8K pixels.
post #430 of 9736
Quote:
Originally Posted by AV Science Sales 4 View Post

I think it improves things but untils omeone can explain it in other than the pixels overlap (I don't think they do) ...

I believe they do overlap.
Quote:
Originally Posted by AV Science Sales 4 View Post

But some one needs to explain to me how two 1080p flashes can result in 8K pixels.

Someone had a good ASCII example, but I don't recall where it is. Consider that the 2nd set of 1080p is shifted by half a pixel diagonally. You don't end up with 8 million pixels that are each individually addressable, but you do end up with close to 8 million unique squares (or pixels if you want to use that term even though they aren't individually addressable).

Not sure if this will help, but I'll try.

Take a piece of paper and draw 4 squares on it like:
Code:
----------------------
|         |          | 
|         |          | 
|         |          | 
|         |          | 
----------------------
|         |          | 
|         |          | 
|         |          | 
|         |          | 
----------------------
Consider that those squares are red and green on the top row, then blue and magenta (purple) on the bottom row. If you now draw 4 squares of the same size but shifted half a pixel to the right and half a pixel down the very center will have 4 squares that take up the same amount of space as one of the original squares. If the 2nd set of pixels are say green, blue, red and magenta then the 4 small center squares will be:

yellow (from red + green)
green (from green + green)
cyan (from blue + green)
white (from magenta + green)

And the same thing is possible elsewhere. That is, "pixels" that are 1/4th the size of the original pixels, but aren't individually addressable.

--Darin
post #431 of 9736
Quote:
Originally Posted by darinp2 View Post

I believe they do overlap.
Someone had a good ASCII example, but I don't recall where it is. Consider that the 2nd set of 1080p is shifted by half a pixel diagonally. You don't end up with 8 million pixels that are each individually addressable, but you do end up with close to 8 million unique squares (or pixels if you want to use that term even though they aren't individually addressable).

Not sure if this will help, but I'll try.

Take a piece of paper and draw 4 squares on it like:
Code:
----------------------
|         |          | 
|         |          | 
|         |          | 
|         |          | 
----------------------
|         |          | 
|         |          | 
|         |          | 
|         |          | 
----------------------
Consider that those squares are red and green on the top row, then blue and magenta (purple) on the bottom row. If you now draw 4 squares of the same size but shifted half a pixel to the right and half a pixel down the very center will have 4 squares that take up the same amount of space as one of the original squares. If the 2nd set of pixels are say green, blue, red and magenta then the 4 small center squares will be:

yellow (from red + green)
green (from green + green)
cyan (from blue + green)
white (from magenta + green)

And the same thing is possible elsewhere. That is, "pixels" that are 1/4th the size of the original pixels, but aren't individually addressable.

--Darin

So, Darin, assuming that the RS55's e-shifting and the Sony1000's 1080p=>4K upconversion both work as they should, can you predict will have the better pic?
post #432 of 9736
Quote:
Originally Posted by millerwill View Post

So, Darin, assuming that the RS55's e-shifting and the Sony1000's 1080p=>4K upconversion both work as they should, can you predict will have the better pic?


Not Darin but Millerwill

I did see the X70 ( RS55 ) and it looked very good, but the picture was about the same in quality as my 95ES ( the X70 had more light and AMO a little less ghosting ) and black level looked be at about same level for both ( but in ldark scenes, I actully did find the 95 to reveal more details then the X70 ).

I expect and hope ( ) the 1000ES to have more light then both the others and to have a little better native then the 95ES and much higher on/off CR ( probely thanks to a more agressive DI to )
( Not to talk about, I love the motionflow on the low setting )

Then there is the point with the 1000ES, who can take native 4K in ( know we dont have any materiel yet ) and I do think is a very good and importend thing, because if I/you get it, we will probely have to keep it for some years ( !? )and then I guess and think the 4K materiel will be avaible.

So I expect the 1000ES to smoke both the 95ES and the X70/RS55 out off the water and it better do, because it cost a lot more then the others

Just my opion offcause


dj
post #433 of 9736
Thanks guys.

I got lost in the woods because of conceptualizing the shifting over and down one half pixels to put a new scaled set of pixels in the center of each box created by each set of four original pixels. Now I get the overlap bit but I guess I could use an explanation from Darin. With the overlap now none of the subpixels comprising the plus 8 million pixel array would be original and would be 1/4 the size of the original pixels? And if this is the case, what happens re interpixel perceived spacing? Obviously the space between pixels would be less if the chip size in a true 4K panel didn't change much from the size of a 1080p panel and the pixel size would have to be smaller. But what happens with a single panel (actually 3 converged panels) flashing the original and then the scaled eshifted? The overlap essentially elimibnates any interpixel spacing and thus no grid would be visible there being no spacing between pixels.
post #434 of 9736
NM uploading file failed.
Reply
Reply
post #435 of 9736
[quote=AV Science Sales 4;21346201]Thanks guys.

I got lost in the woods because of cconceptualizing the shifting over and down one half pixels to put a new scaled set of pixels in the center of each box created by four original pixels. Now I get the overlap bit but I guess I could use an explanation from Darin. With the overlap now none of the subpixels comprising the plus 8 million pixel array would be original and would be 1/4 the size of the original pixels? And if this is the case, what happens re interpixel perceived spacing? Obviously the space between pixels would be less if the chip size in a true 4K panel didn't change much from the size of a 1080p panel and the pixel size would have o be smaller. But what happens with a single panel (actually 3 converged) flashing eshift?[/QUOTE]


What ?

dj
post #436 of 9736
Quote:
Originally Posted by AV Science Sales 4 View Post

With the overlap now none of the subpixels comprising the plus 8 million pixel array would be original and would be 1/4 the size of the original pixels?

Yep. The original pixels aren't there anymore in their original form once your eye integrates them with the 2nd flashed image (the shifted one).
Quote:
Originally Posted by AV Science Sales 4 View Post

And if this is the case, what happens re interpixel perceived spacing?

My experience with the JVCs is that I can't see any interpixel spacing wit the eShift on. That is, not black lines between the pixels anymore. They should still be there, but because of limitations of my eyes combined with limitations of the lenses, I can't see them, at least not on a white image. This is one reason we have turned off eShift in the service menu before focusing the projector.

--Darin
post #437 of 9736
Quote:
Originally Posted by millerwill View Post

So, Darin, assuming that the RS55's e-shifting and the Sony1000's 1080p=>4K upconversion both work as they should, can you predict will have the better pic?

All else being equal it would seem that the real 4K panel would have advantages, but it may be hard to tell the difference at normal viewing distances.

W.Mayer on the DCI forum is probably the best one to ask.

--Darin
post #438 of 9736
I received a whole bunch of material from Sony today on the VPL-vw1000ES. I will post the new latest spec sheet shortly and I have lens calculators and the owner's manual.

The specs really haven't changed but I do notice the only way a resolution above 1080P in is limited to what can be carried by a single HDMI 1.4a cable. Thus, the highest rate a 4096 x 2160 signal can be fed in is 24p. Lower rates can go to 30. Its all on the spec sheet. Full panel resolution is 17:9 while 3840 x 2160 is 16:9 and can be fed 24p, 25p, or 30p. 1920 x 1080 can be fed at the usual rates up to 60p.
post #439 of 9736
Quote:
Originally Posted by AV Science Sales 4 View Post

I received a whole bunch of material from Sony today on the VPL-vw1000ES. I will post the new latest spec sheetshortly and I have lens calculators and the owner's manual.

The specs really haven't changed but I do notice the only way a resolution above 1080P in is limited to what can be carried by a single HDMI 1.4a cable. Thus, the highest rate a 4096 x 2160 signal can be fed in is 24p. Lower rates can go to 30. Its all on the spec sheet. Full panel resolution is 17:9 while 3840 x 2160 is 16:9 and can be fed 24p, 25p, or 30p. 1920 x 1080 can be fed at the usual rates up to 60p.

Look forward to it, Mark.
post #440 of 9736
Reply
Reply
post #441 of 9736
Quote:
Originally Posted by AV Science Sales 4 View Post

I received a whole bunch of material from Sony today on the VPL-vw1000ES. I will post the new latest spec sheetshortly and I have lens calculators and the owner's manual.

The specs really haven't changed but I do notice the only way a resolution above 1080P in is limited to what can be carried by a single HDMI 1.4a cable. Thus, the highest rate a 4096 x 2160 signal can be fed in is 24p. Lower rates can go to 30. Its all on the spec sheet. Full panel resolution is 17:9 while 3840 x 2160 is 16:9 and can be fed 24p, 25p, or 30p. 1920 x 1080 can be fed at the usual rates up to 60p.

So if an ultra HD blu-ray were to be released...would it be 16:9 3840x2160 or 17:9 4096x2160 ?
post #442 of 9736
I have no idea. If I had to guess, I would guess 3840 x 2160 at 24p. Hell, i have a 50/ 50 chance of being right and I have made a whole lot of guesses with far lesser odds of being correct. What do you think Jon?
post #443 of 9736
I wonder if someone could kindly post an English translation of cine4home's preview of the 1000. When I do the 'google translate' of it, and then try to print it, I get only the first page to print. (My primitive computer skills, I'm afraid!) Yet when I've clicked on the translated version of cine4home articles that others post, I can print the whole thing.

Thanks much!
post #444 of 9736
Quote:
Originally Posted by AV Science Sales 5 View Post

Here are the links that Mark referenced above.
http://avscience.com/2011/12/sony-vp...-4k-projector/
http://avscience.com/wp-content/uplo...Spec-Sheet.pdf

Mark mentioned that the User Manual would also be available, but I didn't see it on these two items. Do you have a listing for it? TIA
post #445 of 9736
Will get it posted soon. Its pretty much the sames as any other Sony projector manual. Nothing in there to really make you say thanks for the info.
post #446 of 9736
How is Sony getting the increased brightness out of this unit? Is the lamp higher wattage, or some other design enhancement?
post #447 of 9736
The Lamp is listed at 330 watts. How they getting the rated lumens out I don't know. Like anything else the actual lumens at d65 will undoubtedly be less. I have no data other than the 2000 lumens unspecified spec. Results will I think have to await field measurements.
post #448 of 9736
Reply
Reply
post #449 of 9736
Quote:
Originally Posted by AV Science Sales 4 View Post

I have no idea. If I had to guess, I would guess 3840 x 2160 at 24p. Hell, i have a 50/ 50 chance of being right and I have made a whole lot of guesses with far lesser odds of being correct. What do you think Jon?

I would have to agree with your guess. I don't know when/if it will happen for blu-ray but it looks like for the majority of use cases, 3840x2160 is the most likely typical "home" use case. I was just trying to figure out where this 17:9 material was going to originate from?!
post #450 of 9736
Quote:
Originally Posted by AV Science Sales 4 View Post

The Lamp is listed at 330 watts. How they getting the rated lumens out I don't know. Like anything else the actual lumens at d65 will undoubtedly be less. I have no dated other than the 2000 lumens unspecified spec. Results will I think have to await field measurements.

Cine4home speculated that production units would have ~ 1500-1600 lumens calibrated to D65. That would be nice!

This would presumably be with high lamp at min throw, with ~20% less in low lamp. 1200 calibrated lumens in low lamp would be very nice!
New Posts  All Forums:Forum Nav:
  Return Home