Return to the CurtPalme.com main site CurtPalme.com Home Theater Forum
A forum with a sense of fun and community for Home Theater enthusiasts!
Products for Sale ] [ FAQ: Hooking it all up ] [ CRT Primer/FAQ ] [ Best/Worst CRT Projectors List ] [ Setup Tips & Manuals ] [ Advanced Procedures ] [ Newsletters ]

 
Forum FAQForum FAQ   SearchSearch   MemberlistMemberlist  Photo AlbumsPhoto Albums  RegisterRegister 
 MembershipClub Membership   ProfileProfile   Private MessagesPrivate Messages   Log inLog in 
Blu-ray disc release list and must-have titles. Buy the latest and best Blu-ray titles to show off in your home theater!

E-shift in software?
Goto page Previous  1, 2, 3
 
Post new topic   Reply to topic   Printer-friendly view    CurtPalme.com Forum Index -> Lumagen Products
View previous topic :: View next topic  
Author Message
cmjohnson



Joined: 03 Apr 2006
Posts: 5180
Location: Buried under G90s


PostLink    Posted: Mon Jan 02, 2017 2:34 pm    Post subject: Reply with quote


        Register to remove this ad. It's free!
Yes, I am quite aware of what MTF is. I saw Scott working with that 8500HR in his lab at VDC years ago, too.

I say fully resolved, and I mean it. All details present in the picture as seen on the RS45 are present and equally as visible on the Ultra, at least in the central area. (Still have to get edge focus handled.)

With a slight vertical size increase, I'm resolving individual scan lines at 1080p. Not one on, one off alternating patterns, but actual scan lines at 1 for 1. And I know that there is room for additional improvement because Greg9518lc has shown what he states is scan lines at 1080p on a white raster.

I know, scan lines do not equal resolution. They equal system focus. They are as much an artifact of CRT as pixel edges are of digital.

The max bandwidth needed for 1080p is 165 MHz and I'm pretty sure I'm getting it. (MP modded VIM and neck cards installed.)
Back to top
kal
Forum Administrator


Joined: 06 Mar 2006
Posts: 15630
Location: Ottawa, Canada

TV/Projector: JVC DLA-RS56


PostLink    Posted: Mon Jan 02, 2017 2:49 pm    Post subject: Reply with quote

cmjohnson wrote:
We are only asking this software e-shift to deliver 1080p at ALL TIMES. But half the frames are shifted in position very slightly and both frames are derived with slightly different details from the same source 4K frame.

That's the part that doesn't make sense. You can't "shift the position very slightly" without the projector being able to display more than 1080p. You can't fit 4 pixels into 2 holes.

Quote:
If you were to tap into the circuits of the JVC unit in e-shift mode and separate out the shifted frames and the non-shifted frames, they could all be displayed on a Marquee or G90 or 909.

E-shift IS NOT SOFTWARE. There are no different "frames". The "frames" are identical. e-shift is purely mechanical. There is NO software, there is no way to 'tap into the circuits'. A frame is displayed normally then the panel is MECHANICALLY moved down and over 1/2 pixel and the SAME frame is displayed again.

Kal

_________________

Support our site by using our affiliate links. We thank you!
My basement/HT/bar/brewery build 2.0


Last edited by kal on Mon Jan 02, 2017 3:17 pm; edited 1 time in total
Back to top
View user's photo album (18 photos)
xmob135lc



Joined: 15 Sep 2012
Posts: 80



PostLink    Posted: Mon Jan 02, 2017 3:07 pm    Post subject: Reply with quote

Originally it is software, 2 new fields / "subframes" supposed to give a "single" high resolution single frame.
http://phys.org/news/2005-10-wobulation-hp-hdtv-resolution.html
Hewlett Packard rear projection tellys have 960x1080 diagonal "diamond" mirror device updated at 120hz and a diagonal shift between every subframe .
So afterall they had both field sequential color and field sequential resolution.

It is something to spend 120hz on for JVC .

Notice the wording:
Quote:

TVs with wobulation technology are able to double the number of pixels in the image without doubling the pixel on the modulator, so there’s no need for a more expensive SLM – and no need to spend nearly twice as much for the same high-quality picture.


Read more at: http://phys.org/news/2005-10-wobulation-hp-hdtv-resolution.html#jCp

it's not double resolution... Also you don't gain high frequencies, but loose it .
Back to top
garyfritz



Joined: 08 Apr 2006
Posts: 10500
Location: Fort Collins, CO


PostLink    Posted: Mon Jan 02, 2017 3:36 pm    Post subject: Reply with quote

Kal, I think what CJ is proposing is basically 2160i. 1080i displays 1080 lines with 540 lines of bandwidth. CJ wants to send a 2160-line 4K signal to the CRT, and use 1080p@60 bandwidth to display alternating frames. Then shift alternating frames 1/2 scanline so you effectively have 2160 distinct lines displayed, albeit with a 30fps effective frame rate. Is that what you're looking for, CJ?
Back to top
kal
Forum Administrator


Joined: 06 Mar 2006
Posts: 15630
Location: Ottawa, Canada

TV/Projector: JVC DLA-RS56


PostLink    Posted: Mon Jan 02, 2017 3:48 pm    Post subject: Reply with quote

That's probably what he means. It's just confusing as he talks about tapping into the JVC circuitry as if JVCs doing something like that, which isn't true. It's mechanical on JVCs.

Kal

_________________

Support our site by using our affiliate links. We thank you!
My basement/HT/bar/brewery build 2.0
Back to top
View user's photo album (18 photos)
justin_f



Joined: 27 Nov 2010
Posts: 49
Location: Australia


PostLink    Posted: Mon Jan 02, 2017 4:14 pm    Post subject: Reply with quote

kal, to quote the link you shared earlier.

"Now we have twice as many pixels (not four times as many), displayed in a checkerboard pattern. Since each pixel is refreshed only 60 times per second, not 120, the pixels are half as bright, but because there are twice as many of them, the overall brightness of the image stays the same."

there is only twice as many pixels, not four.

"Notice that the black pixel grid has become virtually invisible from what it was before. However, the quality of the image has not really improved, because all we did was to copy the same information diagonally.

But — what if instead, we had the projector treat this larger set of pixels (twice as many as 1920x1080) as a single canvas, and we gave each pixel its own color, using some very sophisticated algorithms that analyze the content of the 1920x1080 source image to find object edges and things like that?"

the author of this article is suggesting there IS software involved... there is an algorithim that picks what pixels are to be added into the holes.

cj is suggesting that the mechanical shifting jvc uses to display the "B" (additional algorithim picked pixel) frame can be done on a htpc for the second 60hz of every frame? and i agree... the algorithim the article speaks of will indeed be advanced and the tricky part. I actually scaled the red letter "A" pictures in the article down to a smaller size and flicked between them continously and the shifted picture is brighter than the standard. that alone to me would deem the idea worthy of trying. if all you think would happen is fill a scan line?

as for the resolving of an image. did anyone look at the pictures gjaky posted of the 720p image him and his friend made? why on earth dont we stay within the projectors abilities and push from there first? I am amazed by what they achieved there.

EDIT: gary summarizes this better. thanks!
Back to top
cmjohnson



Joined: 03 Apr 2006
Posts: 5180
Location: Buried under G90s


PostLink    Posted: Mon Jan 02, 2017 6:30 pm    Post subject: Reply with quote

Yes, I'm quite aware of how JVC's e-shift works. One term for it is a "wobulator". It's basically a mirror that shifts the image diagonally, then back to the original position, and it does it in a very precise manner at full frame rate. But later iterations of e-shift also add a lot of additional video processing to the equation.

The whole nature of this topic is to suggest that a suitably powerful scaler could replicate e-shift functionality IN SOFTWARE, specifically to give CRT users the potential benefit of image shifting in functionally the same manner that JVC's E-shift allows a
simulacrum of 4K video to be reproduced with a 2K wobulated image generator.

I am pretty sure that E-shift as implemented by JVC processes the 4K signal and creates, from a single 4K frame, two slightly different 1080p frames, one intended for "base position" and one intended for "shifted position".

There would be precisely zero point to shifting the SAME exact image diagonally. That's called SMEARING and it doesn't simulate higher resolution, it creates lower effective resolution.

The processing of the 4K frame to make two slightly different 1080p frames is critical to making E-shift work.

JVC even indicates that's how e-shift works, and yes, it's not just moving the same 1080p frame diagonally. Far from it.

Here, read this:

http://www.jvc.ca/en/dla-x/feature01.html
Back to top
kabuby77



Joined: 28 Mar 2011
Posts: 147
Location: Italy


PostLink    Posted: Tue Jan 03, 2017 9:22 am    Post subject: Reply with quote

cmjohnson wrote:
Yes, I'm quite aware of how JVC's e-shift works. One term for it is a "wobulator". It's basically a mirror that shifts the image diagonally,


e-shift mean electronic shift, there's no moving parts.
Maximunm image refresh rate of LCD projectors is 120Hz but the panel refresh rate i something linke 480Hz!!!! (or more)!
So when you watch a 60Hz material the projector refresh every frame 8 times alternating the two diagonal positions.
Regarding crt, if you move the 'image with a mirror, the persistence of the phosphors would mix the two images.
That said the vobulated picture is not as pretty as a true 4k , it smears the edges.
Back to top
xmob135lc



Joined: 15 Sep 2012
Posts: 80



PostLink    Posted: Tue Jan 03, 2017 11:15 am    Post subject: Reply with quote

kabuby77 wrote:
cmjohnson wrote:
Yes, I'm quite aware of how JVC's e-shift works. One term for it is a "wobulator". It's basically a mirror that shifts the image diagonally,

Maximunm image refresh rate of LCD projectors is 120Hz but the panel refresh rate i something linke 480Hz!!!! (or more)!

There are no 8 positions , only 2 and a single diagonal shift, because it has binary grayscale like plasma and DLP, and binary grayscale takes time , it has its own subframes .

So in fact JVC has much more persistence, and the active matrix pixels depend on persistence/ temporal smearing for light output.

CRT has better motion resolution , but less light output.

But then again you don't watch HFR video too often or at all, so you don't care about motion resolution, and digital can get away with temporal smearing , while CRT can't with gaussian.
Back to top
kabuby77



Joined: 28 Mar 2011
Posts: 147
Location: Italy


PostLink    Posted: Tue Jan 03, 2017 2:21 pm    Post subject: Reply with quote

I'm not good at writing, but do read at least, where you found 8 positions for shifting?

Anyway, it works for digital because there's a pixel matrix, and it works better with sharp edges (see for example the new benq w11000 )
With crt vobulate is the same as interlacing(except the misalignment of half pixel).
For my taste when two adjacent rows overlap the image loses impact.
IMHO this technique does not make sense
Back to top
xmob135lc



Joined: 15 Sep 2012
Posts: 80



PostLink    Posted: Tue Jan 03, 2017 3:11 pm    Post subject: Reply with quote

kabuby77 wrote:
I'm not good at writing, but do read at least, where you found 8 positions for shifting?

Anyway, it works for digital because there's a pixel matrix, and it works better with sharp edges (see for example the new benq w11000 )
With crt vobulate is the same as interlacing(except the misalignment of half pixel).
For my taste when two adjacent rows overlap the image loses impact.
IMHO this technique does not make sense


misread...

well if there was analog grayscale like Sony's SXRD , it could work 8 times with 2 ms for every position at 60hz,
because no subframes taken away by binary grayscale.

But then, that's no good motion , with 300 lines of motion resolution. So it will be a blurry mess anyway. Besides Sony LCOS has apparently problem with hysteresis , hence the 3d ghosting.
Back to top
Display posts from previous:   
Post new topic   Reply to topic   Printer-friendly view    CurtPalme.com Forum Index -> Lumagen Products All times are GMT
Goto page Previous  1, 2, 3
Page 3 of 3
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum