or Connect
AVS › AVS Forum › Display Devices › Digital Hi-End Projectors - $3,000+ USD MSRP › Official JVC RS55/X70 owners thread.
New Posts  All Forums:Forum Nav:

Official JVC RS55/X70 owners thread. - Page 117

post #3481 of 3669
I'm going to have to adjust that formula after some quick testing. I think at 1.0x seating distance, 1 pixel (uncorrected) off on RED = 79 (C = Average to Barely Fail), whereas at 1.2x seating it might be 85 (B=Keeper).

I think my modifier for seating distance will be 20% for every 0.1 seating distance closer. Hence 1.0x seating distance is 40% more strict than 1.2x seating distance. Hence you might get a barely passed grade for 1 pixel off on RED at 1.2x seating distance, but a barely fail grade for 1 pixel off on RED at 1.0x seating distance.

I also think blue might need to be up'd to -7.5 lost grading per pixel, because 2 pixels off on blue in HTPC text is in some ways close to 1 pixel off on RED as far as perceptible loss in sharpness. It's hard to judge exactly, I will need to mess with this for a couple hours, will do another day.

Gear mentioned in this thread:

post #3482 of 3669
Quote:
Originally Posted by coderguy View Post

I'm going to have to adjust that formula after some quick testing. I think at 1.0x seating distance, 1 pixel (uncorrected) off on RED = 79 (C = Average to Barely Fail), whereas at 1.2x seating it might be 85 (B=Keeper).

I think my modifier for seating distance will be 20% for every 0.1 seating distance closer. Hence 1.0x seating distance is 40% more strict than 1.2x seating distance. Hence you might get a barely passed grade for 1 pixel off on RED at 1.2x seating distance, but a barely fail grade for 1 pixel off on RED at 1.0x seating distance.

I also think blue might need to be up'd to -7.5 lost grading per pixel, because 2 pixels off on blue in HTPC text is in some ways close to 1 pixel off on RED as far as perceptible loss in sharpness. It's hard to judge exactly, I will need to mess with this for a couple hours, will do another day.

I think you are on the right lines and its a great idea to try and put a weighting on this.

Misconvergence is the displacement of TWO colours from each other, and not, as is often misconceived, the movement of one colour from an imaginary correct line. So misconvergence is actually a measurement and/or combination of the following

Red/Green
Green/Blue
Blue/Red

Obviously there is no need to repeat Green/Red as well as Red/Green as they are the same thing. It is important to quote them like this because:-

Red/Green misconvergence is by far the most intrusive and would have the highest weighting. Blue only makes up a fraction of the image definition. I don't think Green/Blue or Blue/Red has more impact than the other so should have the same weighting.

The reason it is important to consider them in pairs is simple. Imagine in a vertical orientation that Red is 0.3 of a pixel above Green which is 0.3 above Blue. This means a horizontal line will show as a white line with red above it and blue below it. The Red/Green misconvergence is 0.3 of a pixel but the Red/Blue misconvergence is 0.6. If we weight red/green as having double impact, we could get a starting formula (not including distance etc), of eC = (0.3*2)+0.6 = 1.2 . If the colours were such that green and blue were swapped, then it would be eC = (0.6*2) +0.3 = 1.5

Then to factor in distance, you need to normalize the effective pixel size. In other words, in an imaginary unit for our calculations, ePS, this is a calculation that takes screensize and viewing distance and produces an effective pixel size. As you correctly point out the effect of misconvergence is perceived differently depending on screensize/seating distance. It is important though to remember that the misconvergence doesn't change, and therefore neither can its true impact on image sharpness. We are talking about a perceived difference. In other words, just because you sit double the distance away, doesn't mean the convergence magically gets better as in absolute terms it is still the same disparity.
post #3483 of 3669
Quote:
Originally Posted by JonStatt View Post


The reason it is important to consider them in pairs is simple. Imagine in a vertical orientation that Red is 0.3 of a pixel above Green which is 0.3 above Blue. This means a horizontal line will show as a white line with red above it and blue below it.

Thanks for the comments.

Right, but the pattern I am using most likely this is already taken into account outside the formula because it is automatically excluded in the viewer's testing of convergence due to the appearance of that being 0.3 off on blue visibly anyhow, and as such will be entered into the calculator in that fashion. I haven't gotten that far into this yet, and I haven't decided what pattern to use. Even if it weren't excluded that way, the math would exclude it because the formula takes into account directions and overlaps.

The grading system would most likely be assumed at 1.2x seating distance for the default grading (this having nothing to do with how close they stand while they are assessing it, but only to do with assumed seating distance "damage"). The user will be able adjust it to get a "personal grade" that is not part of the overall grading system. In the end, whatever I can do to keep improving the calculator.
Edited by coderguy - 5/20/13 at 7:35am
post #3484 of 3669
I thought of what you said further, but I think people might be mixing up analog and digital convergence (not sure).

I think the real datum would optimally be the distinguishable pixel fill border in the non-converged zones of a pixel (but this is hard to test due to the small pixel size), because that is what the CA error is actually offsetting if you think about it in terms of pixel definition color. Hence, even though moving 2 pixels over to the left, might be the same as moving 1 pixel away theoretically, let's try this...

Move 2 pixels over to the left, the distinguishable pixel fill border did not move in the non-converged or converged zone. Essentially, if you moved both those pixels off by 3/4 pixel, they are fringing the pixel definition at a greater rate due to the pixel definition appearing to be the datum and them being out of alignment to the pixel grid itself, even though it might appear to be the same if we were using an analog device that has no pixel grid/fill if we were sitting far enough back. Now is this because green is the datum on our JVC's, I don't know, or is it because when you move 2 colors, you don't move the pixel grid of the panel where the pixel fill border exists. I would think the real datum on a digital device is the location of the pixel fill / pixel grid border, which is close to (but maybe not exactly) the same as the datum color (or green in this case).

My guess is the pixel grid is the datum, the green is the "attempted" datum color, and the other 2 colors are aligned next, red being second, and blue being last. Which might explain why blue is usually off the most.

All this said, I am just going to average it from the datum color that is the least offset on average. Anyhow, just some theoretical stuff, not certain...Back to work....
Edited by coderguy - 5/20/13 at 8:32am
post #3485 of 3669
Quote:
Originally Posted by coderguy View Post

I thought of what you said further, but I think people might be mixing up analog and digital convergence (not sure).

I think the real datum would optimally be the distinguishable pixel fill border in the non-converged zones of a pixel (but this is hard to test due to the small pixel size), because that is what the CA error is actually offsetting if you think about it in terms of pixel definition color. Hence, even though moving 2 pixels over to the left, might be the same as moving 1 pixel away theoretically, let's try this...

Move 2 pixels over to the left, the distinguishable pixel fill border did not move in the non-converged or converged zone. Essentially, if you moved both those pixels off by 3/4 pixel, they are fringing the pixel definition at a greater rate due to the pixel definition appearing to be the datum and them being out of alignment to the pixel grid itself, even though it might appear to be the same if we were using an analog device that has no pixel grid/fill if we were sitting far enough back. Now is this because green is the datum on our JVC's, I don't know, or is it because when you move 2 colors, you don't move the pixel grid of the panel where the pixel fill border exists. I would think the real datum on a digital device is the location of the pixel fill / pixel grid border, which is close to (but maybe not exactly) the same as the datum color (or green in this case).

My guess is the pixel grid is the datum, the green is the "attempted" datum color, and the other 2 colors are aligned next, red being second, and blue being last. Which might explain why blue is usually off the most.

All this said, I am just going to average it from the datum color that is the least offset on average. Anyhow, just some theoretical stuff, not certain...Back to work....

LOL...this is one of those topics where you need to lock a few heads in a room with a whiteboard and properly rationalize this.

Here is one thought for you though. You are talking about datums. This to me implies an imaginary line of perfection, and any colour can deviate from it. But if all three colours were equally shifted so they remained in perfect alignment with each other, then that would still be perfection for a viewer. So we have to talk about relative displacement of colours and not an imaginary line of perfection. If you like to think of green as that datum then that's fine, as long as you are talking about the deviation between colours rather than a fixed point in space. We know that JVC don't do calibration as such in the factory. It is simply tolerances. The reason that blue always ends up being the most out may simply be the choice of mounting design for the panels and the fact that for whatever reason in manufacturing it has the widest deviation. Perhaps it is less clamped than the other two....I would have to look at one carefully opened up to see.

You raise an interesting point about 3/4 pixel blurring pixel boundaries, but a whole pixel maintaining the boundary even though on the example I gave, a 2:35 film on a 16:9 screen, you would see a very obvious fringe top and bottom.
post #3486 of 3669
I think the pixel grid is the datum, in digital panels there would be a fixed point. The perfect point is alignment to the pixel grid, the pixel grid is not projected, it is the electronics blocking the light path of the panels projections, right?

Aligning even all pixels to a location off the grid alignment is going to fringe it and cause a loss of pixel definition, could explain why some projectors even when focused and well converged have poor pixel definition. Moving the colors by adjusting with the digital convergence does not move the panels, and only with a full pixel movement can you "trade pixels" and ignore mis-converged ones due to extra pixels on the panel. However, none of this changes the fact that the projected light of the 3 colors is interrupted by a fixed pixel grid that doesn't move when you modify convergence (unless I am incorrect, but that is how I interpreted it).

Edited:
Also error tolerances often stack in manufacturing, if there is a datum every layer that is next placed has an increased error margin due to the assembly order or alignment of previous color. Hence, green = near perfect, red = good, blue = bad. I doubt it's a coincidence or the way they "set it up", many projectors are the same way, heck most are. You occasionally get red way off on the cheaper projectors, but in general blue is usually the offender. Blue is placed last and the worst offender because it matters the least.
Edited by coderguy - 5/20/13 at 10:33am
post #3487 of 3669
I just received my new (to me) rs55 and the first thing I want to do is verify everything is in good order oerationally. Among other things I want to check convegence. What else should I be looking at? Also I read somewhere i should turn eshift off when performing convergence. If so how do I do that?
Thanks!
post #3488 of 3669
On the remote center pad, hit 'right, left, right, left' then center button. You can then toggle eshift on / off with the right / left action.

You'll see 'demo on / off' as you toggle back and forth. Menu button will exit.

Welcome to the best 2d under 10k. smile.gif
post #3489 of 3669
Here is someone about 003 lamp, since we haven't had many reports, ouch...

"I measured the light output of the X3 with the 003 lamp before returning it to Kenwood. It had declined by 60% from when it was new after only 750 hours use on low setting."
http://www.avsforum.com/t/1296327/official-jvc-rs40-x3-owners-thread/9660#post_23338046
post #3490 of 3669
That guys projector obviously had other issues as well, I wouldn't use that as a reference. Considering how long the 003's have been out, it's been surprisingly quiet here and on AV forum for early dimming.

I saw the recent argument w / RJ on this topic. Folks with the first gen lamps were losing major lumens in the first few hundred hours. surely there's 003's out there with 300-500 hours on them.

besides, there's been some killer deals on the 003 lamps with the flapper to the point where swapping it out every 1000 hours wouldn't be a big deal for most.
post #3491 of 3669
Can you PM me the deal(s) so I can get another spare?

I haven't seen any reports on 003, good or bad to be honest, just that one. Trust me I don't want to be right on the new 003 lamps, just paranoid smile.gif
Edited by coderguy - 5/20/13 at 7:35pm
post #3492 of 3669
It looks like I got a pretty good sample. My one question is this projector supposed to make the kind of loud snapping noises when changing certain picture profiles and grinding when adjusting the aperature. My RS 40 was not this loud when making adjustments. For instance when changing to the stage setting it made a loud snapping sound that made me think something was wrong electrically.
post #3493 of 3669
Yah, I think there is some variance on how loud the different ones are, or maybe it was because I was in different rooms with different acoustics and different distances from the PJ.
Right now on my RS-45, if I change the aperture (I'm 7 feet away from the PJ), then I hear a soft clicking sound, very quiet though. Changing into stage mode is more audible and is a quick "D-ZZZ-TA" sound, but still not loud (easily audible). It was louder on the RS-55 I heard, but I was in a different room. I don't remember on the RS-46, because I wasn't paying attention to the sounds at the time.
post #3494 of 3669
Quote:
Originally Posted by nohjy View Post

It looks like I got a pretty good sample. My one question is this projector supposed to make the kind of loud snapping noises when changing certain picture profiles and grinding when adjusting the aperature. My RS 40 was not this loud when making adjustments. For instance when changing to the stage setting it made a loud snapping sound that made me think something was wrong electrically.

you are hearing a filter move into place in certain modes. The RS40/45/46 doesn't have this filter. it's quite loud and startling when you first hear it.

Also the dual iris will be a bit more noisy during adjustments. everything you're describing is normal.
post #3495 of 3669
Mine makes that racket too and yes it's rather loud.
post #3496 of 3669
Quote:
Originally Posted by zombie10k View Post

...there's been some killer deals on the 003 lamps with the flapper to the point where swapping it out every 1000 hours wouldn't be a big deal for most.

I would appreciate a pm on this.
post #3497 of 3669
Just a reminder that the first 003 bulbs did NOT have the flapper. There are , just for confusion, two kinds of 003 bulbs!
post #3498 of 3669
Quote:
Originally Posted by coderguy View Post

I think the pixel grid is the datum, in digital panels there would be a fixed point. The perfect point is alignment to the pixel grid, the pixel grid is not projected, it is the electronics blocking the light path of the panels projections, right?

Aligning even all pixels to a location off the grid alignment is going to fringe it and cause a loss of pixel definition, could explain why some projectors even when focused and well converged have poor pixel definition. Moving the colors by adjusting with the digital convergence does not move the panels, and only with a full pixel movement can you "trade pixels" and ignore mis-converged ones due to extra pixels on the panel. However, none of this changes the fact that the projected light of the 3 colors is interrupted by a fixed pixel grid that doesn't move when you modify convergence (unless I am incorrect, but that is how I interpreted it).

Edited:
Also error tolerances often stack in manufacturing, if there is a datum every layer that is next placed has an increased error margin due to the assembly order or alignment of previous color. Hence, green = near perfect, red = good, blue = bad. I doubt it's a coincidence or the way they "set it up", many projectors are the same way, heck most are. You occasionally get red way off on the cheaper projectors, but in general blue is usually the offender. Blue is placed last and the worst offender because it matters the least.


I want to make sure we are not misunderstanding each other. Are you referring to the wire grid polarizer? I understood the purpose of this is purely to enhance contrast by polarizing the light and removing any scattered light. I don't believe the grid corresponds to the pixels. My understanding is that in terms of pixels, each panel has its own pixel structure and that the prism simply combines them together. That's why you may find it hard to see the pixel structure clearly with white due to misconvergence, but you can still clearly see it on a single primary.
post #3499 of 3669
Quote:
Originally Posted by bigdogaxis View Post

I would appreciate a pm on this.

Quote:
Originally Posted by zombie10k View Post

That guys projector obviously had other issues as well, I wouldn't use that as a reference. Considering how long the 003's have been out, it's been surprisingly quiet here and on AV forum for early dimming.

I saw the recent argument w / RJ on this topic. Folks with the first gen lamps were losing major lumens in the first few hundred hours. surely there's 003's out there with 300-500 hours on them.

besides, there's been some killer deals on the 003 lamps with the flapper to the point where swapping it out every 1000 hours wouldn't be a big deal for most.
Quote:
Originally Posted by bigdogaxis View Post

I would appreciate a pm on this.

Zombie,

First, thanks again for all of your help. Second, can you tell me where I might go to find a good deal on an 003 lamp and what is a good price? I would like to purchase a backup. I see them for $279 on ebay. Is that a good price? I am absolutely not paying the $400 JVC is asking, that's for sure.

Thanks again!

John
post #3500 of 3669
Quote:
Originally Posted by JonStatt View Post

My understanding is that in terms of pixels, each panel has its own pixel structure and that the prism simply combines them together. That's why you may find it hard to see the pixel structure clearly with white due to misconvergence, but you can still clearly see it on a single primary.
I certainly agree with this. The only element in the chain that has pixel structure is the D-ILA device itself. Perfect convergence would have all three exactly and perfectly aligned. The prism combines all three light paths and has no pixel structure. The wire grid polarizer is nothing more that a "high end" polarizer (a polarizer made with a wire gird has an excellent extinction ratio) and it also has no pixel structure. It is the device that actually forms the image since all it does is pass light in one orientation or block light that is oriented 90 degrees relative to the light that passes thru it. It is a major contributor to the overall CR of the projector.
Edited by Geof - 5/21/13 at 4:39am
post #3501 of 3669
Quote:
Originally Posted by JonStatt View Post

Just a reminder that the first 003 bulbs did NOT have the flapper. There are , just for confusion, two kinds of 003 bulbs!

Thanks for the heads up...caveat emptor. eek.gif
post #3502 of 3669
Quote:
Originally Posted by JonStatt View Post

Just a reminder that the first 003 bulbs did NOT have the flapper. There are , just for confusion, two kinds of 003 bulbs!

I can verify what jon is saying. My RS55 came with a 003 without the flapper. It appears, from what I have read, that the flapper is the key to longevity.
post #3503 of 3669
I still have the original 002 lamp it shipped with. As an experiment, I've been running the RS55 in high altitude mode since it was new. The hours are under 300, but it's held up much better than the RS40 & RS50 at the same hours.

I think it's plausible that airflow was part of the root cause. It explains the attempt of adding the metal fin and the obvious increase in fan size for the new models. The new models in high lamp are slightly louder than the RS55 in HA mode.


The other night I watched 'The Last Stand' on the RS55 and forgot that I turn off e-shift to refocus the projector. I knew in the first scene something wasn't right. This movie looks great with MPC @ 2.
post #3504 of 3669
C'mon Zombie, don't leave us hanging. Where can I get an 003 cheap?
post #3505 of 3669
Quote:
Originally Posted by zombie10k View Post

I still have the original 002 lamp it shipped with. As an experiment, I've been running the RS55 in high altitude mode since it was new. The hours are under 300, but it's held up much better than the RS40 & RS50 at the same hours.

I think it's plausible that airflow was part of the root cause. It explains the attempt of adding the metal fin and the obvious increase in fan size for the new models. The new models in high lamp are slightly louder than the RS55 in HA mode.


The other night I watched 'The Last Stand' on the RS55 and forgot that I turn off e-shift to refocus the projector. I knew in the first scene something wasn't right. This movie looks great with MPC @ 2.

I really don't see how you have less than 300 hours on your RS55/lamp because of all great screenshots you have provided. I guess with all of the projectors you have had at your place this has helped out.

Mike
post #3506 of 3669
Quote:
Originally Posted by JonStatt View Post

I want to make sure we are not misunderstanding each other. Are you referring to the wire grid polarizer? I understood the purpose of this is purely to enhance contrast by polarizing the light and removing any scattered light. I don't believe the grid corresponds to the pixels. My understanding is that in terms of pixels, each panel has its own pixel structure and that the prism simply combines them together. That's why you may find it hard to see the pixel structure clearly with white due to misconvergence, but you can still clearly see it on a single primary.

I never mentioned anything about a wire polarizer, that was added into it by you guys. The alignment in regards to the pixel fill matters because the colors against the BLACK pixel fill itself will cause fringing.

The pixel grid structure is WHAT makes up the pixels as our eyes see it, it defines it from the pixel structure. Yes, each panel also has a pixel it can project, but the goal is to shoot it through the same pixel fill area as the previous pixel. Otherwise if it hits the pixel fill boundary, then you get fringing. How this is addressed digitally is irrelevant to this discussion.

This is why the datum does exist in a digital alignment, but not in analog.
Edited by coderguy - 5/21/13 at 12:01pm
post #3507 of 3669
Quote:
Originally Posted by mbw23air View Post

I really don't see how you have less than 300 hours on your RS55/lamp because of all great screenshots you have provided. I guess with all of the projectors you have had at your place this has helped out.

Mike

That's pretty much it... I only use the RS55 for my 2D BD movies and the other projectors for TV, 3D, etc. You guys have to check out movie I mentioned above, it was made for the JVC's. There's many low APL scenes that aren't as convincing on projectors w/ less native contrast.
post #3508 of 3669
Quote:
Originally Posted by Geof View Post

I certainly agree with this. The only element in the chain that has pixel structure is the D-ILA device itself. Perfect convergence would have all three exactly and perfectly aligned. The prism combines all three light paths and has no pixel structure. The wire grid polarizer is nothing more that a "high end" polarizer (a polarizer made with a wire gird has an excellent extinction ratio) and it also has no pixel structure. It is the device that actually forms the image since all it does is pass light in one orientation or block light that is oriented 90 degrees relative to the light that passes thru it. It is a major contributor to the overall CR of the projector.

This is what I said in the first place, I said the pixel grid is NOT projected. He said the opposite. It is IC's blocking light. I never mentioned anything about a wire polarizer, lol.

What he was arguing was that it doesn't matter where the pixel fill is, of course it does, because the colors against the BLACK pixel fill itself will cause fringing. It wouldn't matter if the convergence was 100% perfect, but it never will be, so it matters because it is always imperfect and determines how defined we see the pixels.

The reason the pixel grid is the datum, is because look at it, where is the most white the MFR tried to align to, exactly. When the convergence gets off, what are we adjusting it back to, we are adjusting it back to be contained to prevent flaring. What is white, white is the spectrum of all colors combined. Where is the green aligned on our JVC's, it's exactly in the middle of the pixel fill.

What he said was there is NO datum, but there is, the pixel grid / green color is the datum for us JVC users (because that is what we have to go by). The goal is for the colors to be fully contained inside the pixel fill. The goal is to NOT fringe the pixel definition, and the pixel definition is made up from pixel fill.The panels have to align in a way that all pixels between the pixel fill do not fringe, that is the point.
Edited by coderguy - 5/21/13 at 12:27pm
post #3509 of 3669
Quote:
Originally Posted by coderguy View Post

This is what I said in the first place, I said the pixel grid is NOT projected. It is IC's blocking light. I never mentioned anything about a wire polarizer, lol.

What he was arguing was that it doesn't matter where the pixel fill is, of course it does, because the colors against the BLACK pixel fill itself will cause fringing. It wouldn't matter if the convergence was 100% perfect, but it never will be, so it matters because it is always imperfect and determines how defined we see the pixels.

The reason the pixel grid is the datum, is because look at it, where is the most white the MFR tried to align to, exactly. When the convergence gets off, what are we adjusting it back to, we are adjusting it back to be contained to prevent flaring. What is white, white is the spectrum of all colors combined. Where is the green aligned on our JVC's, it's exactly in the middle of the pixel fill.

What he said was there is NO datum, but there is, the pixel grid / green color is the datum for us JVC users (because that is what we have to go by). The goal is for the colors to be fully contained inside the pixel fill (even if placing the pixel grid is done after alignment, it doesn't matter). The goal is to NOT fringe the pixel definition, and the pixel definition is made up from pixel fill.The panels have to align in a way that all pixels between the pixel fill do not fringe, that is the point.

LOL - I don't think anyone thinks you are changing the story. We may be saying the same thing, but where I got stuck was that you kept referring to "THE" pixel grid. Where I am getting confused with what you are saying, is that there are in fact 3 pixel grids. Each of the RGB panels each has their own one. And each one floats in space, hopefully aligning with the other two. If you want to take green as the "reference" grid that is okay. But I don't think there is a special reason to assume that green is somehow any more special than the other two panels. We have three pixel grids overlaying each other. As you pointed out, adjusting convergence in whole pixels is not moving any of the grids. The end result, and this is where I "think" we are in agreement is that it would be wrong to say

Red is off by 0.5
Green is off by 0.3
Blue is off by 0.2

If green is always assumed to be the reference, then you can only refer to the deviation of red and blue from green. If you only want to refer to those two deviations, then you have to be careful to define directionality. e.g Upwards is positive, and downwards is negative, such that

Red is off by +0.5
Blue is off by -0.2

Then you can also infer that the delta between red and blue is 0.7.
post #3510 of 3669
Quote:
Originally Posted by JonStatt View Post

Where I am getting confused with what you are saying, is that there are in fact 3 pixel grids. Each of the RGB panels each has their own one. And each one floats in space,

Once they lay the first one, then that can establish the preferential resulting positions of the grid, it becomes a datum. Or a datum can be the tightest possible pixel grid alignment that can possibly exist post-alignment based on the position of the first. The goal then becomes the error margin reduction / tolerance of aligning the others to the first. It does not matter if it is 2 or 20, or by hand or by machine. If the goal was to align each one in correspondence to the previous instead of to the first, then you get stacking errors that grows as you add things. That said, you can use something like BIAS averaging to take into account all the objects, but that's beyond where this needs to go. A datum can in fact be the resulting measurement of multiple objects or coordinates. A datum can also be used as a reference point to handle imperfect shapes and positions. That is why it used geodetic coordinates since the Earth is actually not a perfect shape.

The pixel grid exists well defined, you do not see 3 pixel grids when the convergence is off, nor does the pixel fill change. According to the way you describe it, the pixel fill in LCOS is actually based on convergence, and it is not. The three colors are merged through the light path with no pixel fill initially, they obtain this later is my understanding. I am not exactly sure how the grid is formed because none of the sites I went to explained it clearly other than stating it is formed by IC's, however there appeared to be variances depending on the type of LCD/LCOS. Is it the individual IC lines on the panels, or is somewhere else, well that doesn't really matter. The point is what we can measure by eye and what we see.

Furthermore, keep in mind we are looking at convergence by test patterns, not with precise measuring equipment.
Edited by coderguy - 5/21/13 at 3:15pm
New Posts  All Forums:Forum Nav:
  Return Home

Gear mentioned in this thread:

AVS › AVS Forum › Display Devices › Digital Hi-End Projectors - $3,000+ USD MSRP › Official JVC RS55/X70 owners thread.