or Connect
New Posts  All Forums:Forum Nav:

NEW RANGE JVC 2014 - Page 46

post #1351 of 4022
Quote:
Originally Posted by blee0120 View Post

Yep, that's basically what I'm thinking, they will look the same. But we will find out soon

I don't think they will look the same at all. The point is it is better to start with too much information and throw away the less important bits, than to start with too little and to try and create the missing bits. We know that JVC's e-shift effectively gives approx. 4 million addressable pixels compared to a true 8.x mega pixel UHD display. But if you start with a full UHD image and take the most valuable pixels (through a process of sampling), you should end up with a better image than the other way around.

As for e-shift 3 vs e-shift 2....I get the impression they are fairly similar in terms of concept but that there is now an Auto mode to save you having to select the different presets and of course that it can cope with a 4K input.
post #1352 of 4022
Quote:
Originally Posted by JonStatt View Post

Not doing well today finding links but there was a great little presentation tucked away somewhere which showed this exact calculation.

Hi Jon,

Is this what you're looking for?

post #1353 of 4022
Quote:
Originally Posted by JonStatt View Post

I don't think they will look the same at all. The point is it is better to start with too much information and throw away the less important bits, than to start with too little and to try and create the missing bits. We know that JVC's e-shift effectively gives approx. 4 million addressable pixels compared to a true 8.x mega pixel UHD display. But if you start with a full UHD image and take the most valuable pixels (through a process of sampling), you should end up with a better image than the other way around.

As for e-shift 3 vs e-shift 2....I get the impression they are fairly similar in terms of concept but that there is now an Auto mode to save you having to select the different presets and of course that it can cope with a 4K input.

It will to interesting to see the outcome. Taking the 2K images from the 4K source, then combining them for UHD or having the blu ray player take the 2K image from the 4K source and appying eshift3. I'm curious to see the results. It does seem as the 4K source directly to the jvc will have less processing and produce the better image. I'm sure there will be a way to play 4K blu rays with these new jvc projectors.
Edited by blee0120 - 10/30/13 at 3:56am
post #1354 of 4022
Quote:
Originally Posted by blee0120 View Post

It will to interesting to see the outcome. Taking the 2K images from the 4K source, then combining them for UHD or having the blu ray player take the 2K image from the 4K source and appying eshift3. I'm curious to see the results. It does seem as the 4K source directly to the jvc will have less processing and produce the better image. I'm sure there will be a way to play 4K blu rays with these new jvc projectors.

Unless they have a non HDCP 2.2 protected output for 1080p downscale (which is highly unlikely), you won't see anything on the JVCs when connecting a Bluray 4K playing 4K content as they are not HDCP 2.2 compatible.

The 4K input of the JVCs only supports non-protected content (like demo footage from a Redray player or a 4K camera) or standard HDCP protected content, like standard bluray (whether native 1080p or upscaled to 4K with say a Radiance 204x). That's the only thing we are sure about.
Edited by Manni01 - 10/30/13 at 4:39am
post #1355 of 4022
Quote:
Originally Posted by Manni01 View Post

Hi Jon,

Is this what you're looking for?

Thank you Manni....that was the one I was looking for smile.gif
post #1356 of 4022
Quote:
Originally Posted by JonStatt View Post

I don't think they will look the same at all. The point is it is better to start with too much information and throw away the less important bits, than to start with too little and to try and create the missing bits. We know that JVC's e-shift effectively gives approx. 4 million addressable pixels compared to a true 8.x mega pixel UHD display. But if you start with a full UHD image and take the most valuable pixels (through a process of sampling), you should end up with a better image than the other way around.

As for e-shift 3 vs e-shift 2....I get the impression they are fairly similar in terms of concept but that there is now an Auto mode to save you having to select the different presets and of course that it can cope with a 4K input.

Jon. Obviously he knows all that. Its always better to start with more and throw it away rather than starting with less and creating. What he is saying that e-shift is the great quality equalizer dragging everything down to at least the level possible with the lowest quality 1080p source. I have no clue as to why he is saying this and for what purpose. Why would one even assume e-shift 3 does this?
post #1357 of 4022
Quote:
Originally Posted by JonStatt View Post

4K convergence is an interesting one if the panel sizes are the same because as Geof pointed out, if tolerances don't improve, the convergence error on a pixel level doubles. However if the panel chip increases in size then even without improved tolerances, the convergence improves again. The theory is that JVC's improvement in reducing the pixel gaps this year is in some way towards their eventual aim of a 4K chip. If they can make the pixel gap smaller, then theoretically they can fit more pixels into the same chip size. But while a 40% reduction in pixel gap is promising, it would seem to me that the chip size still needs to increase. If not, other factors come into play including chip yields (the risk of stuck/dead pixels increases although current yields are exceptionally good with only occasional users reporting 1 dead/stuck pixel) and as talked about earlier in the thread the longstanding LCoS issue of pixel bleed.

I had a quick search to see if I could find the chip size on the VW1000 and couldn't find it...I suspect I am missing something obvious....
I pretty much agree with your comments regarding chip size. If the 4K chip is twice as large as the 2K chip then similar convergence tolerances would yield the similar pixel errors in either case. That said, as the chip size increase so does the size of the optics and then prices start escalating quickly...so much like any other design there are always tradeoffs. I suspect the 4K chip will be larger than the 2K chip but not anywhere near double the size.

Darin points out that electronic correction of convergence errors becomes more effective as pixel count increases. Electronic correction techniques probably are more cost effective than spending more in factory labor to achieve tighter convergence tolerances. That said, I would still prefer tighter factory convergence tolerances because if convergence is correct there is nothing to fix (correct).....if mechanical convergence is not correct then some form of electronic manipulation is required to "fix it" and fixing it is okay but I'd rather it not need fixing in the first place.
post #1358 of 4022
Its really not a labor cost issue. Its a question of what tolerance can be achieved by whatever assembly method.
post #1359 of 4022
Quote:
Originally Posted by Geof View Post

I pretty much agree with your comments regarding chip size. If the 4K chip is twice as large as the 2K chip then similar convergence tolerances would yield the similar pixel errors in either case. That said, as the chip size increase so does the size of the optics and then prices start escalating quickly...so much like any other design there are always tradeoffs. I suspect the 4K chip will be larger than the 2K chip but not anywhere near double the size.

Darin points out that electronic correction of convergence errors becomes more effective as pixel count increases. Electronic correction techniques probably are more cost effective than spending more in factory labor to achieve tighter convergence tolerances. That said, I would still prefer tighter factory convergence tolerances because if convergence is correct there is nothing to fix (correct).....if mechanical convergence is not correct then some form of electronic manipulation is required to "fix it" and fixing it is okay but I'd rather it not need fixing in the first place.

Electronic correction is more effective as pixel count increases? If by that point we mean that as we have so many more pixels to start with, if we lose some through resolution loss due to manipulation of fine zonal adjustment, that its effect is unlikely to be noticed....well that is true to an extent. But purists and perfectiontists do not want to lose a single pixel to correct a manufacturing flaw....and just seeing convergence errors of 2 pixels or more is likely to cause heart attacks! smile.gif
post #1360 of 4022
Quote:
Originally Posted by JonStatt View Post

I had a quick search to see if I could find the chip size on the VW1000 and couldn't find it...I suspect I am missing something obvious....

Sorry I missed that one. The chip size of the VW1000ES is 0.74". See the last page of this brochure from Sony Pro Asia.
post #1361 of 4022
Quote:
Originally Posted by Manni01 View Post

Sorry I missed that one. The chip size of the VW1000ES is 0.74". See the last page of this brochure from Sony Pro Asia.
Do you know what their 2K chip size is?

I believe JVC's 2K chip is .7" so Sony managed to cram in 4 times the pixels in a .74" chip. About what I would have expected.....
post #1362 of 4022
Quote:
Originally Posted by JonStatt View Post

Electronic correction is more effective as pixel count increases? If by that point we mean that as we have so many more pixels to start with, if we lose some through resolution loss due to manipulation of fine zonal adjustment, that its effect is unlikely to be noticed....well that is true to an extent. But purists and perfectiontists do not want to lose a single pixel to correct a manufacturing flaw....and just seeing convergence errors of 2 pixels or more is likely to cause heart attacks! smile.gif
I resemble that remark...tongue.gifwink.gifsmile.gif
post #1363 of 4022
Quote:
Originally Posted by blee0120 View Post

It will to interesting to see the outcome. Taking the 2K images from the 4K source, then combining them for UHD or having the blu ray player take the 2K image from the 4K source and appying eshift3. I'm curious to see the results. It does seem as the 4K source directly to the jvc will have less processing and produce the better image. I'm sure there will be a way to play 4K blu rays with these new jvc projectors.

At CEDIA I saw unconverted 1080P on a RS49. It paled in comparison to 4K on the RS67. Granted, I would have liked to have seen 1080P unconverted on the 67, but that was not a choice.
Reply
Reply
post #1364 of 4022
Double post.
Edited by AV Science Sales 5 - 10/30/13 at 8:16am
Reply
Reply
post #1365 of 4022
Quote:
Originally Posted by mark haflich View Post

Its really not a labor cost issue. Its a question of what tolerance can be achieved by whatever assembly method.
Certainly different designs and/or assembly methodfs can affect the costs involved (not just labor) but labor (and/or pure luck) must be involved because some machines come really well converged and some really not well converged. Some may be jarred during shipping but there is still a rather large variance in out of box convergence tolerances and with all else being equal what separates the well converged from the poorly converged except the labor to dial the poorly converged one in better?
post #1366 of 4022
One has to assume, in the case you postulate, that the three chips can be mounted repeatedly in the tolerance you require. This requires precise essentially identical three axis mounting. Not only to the chips have to be in three identical planes, the chips need to perfect planes as to their entry surfaces and the xy coordinates need to be the same.. so its a question of a lot of things including the ability to manufacture the parts essentially perfectly, then to precise locate them and attach at that point and keep everything the same during operation in spite of temperature differentials. Could one cherry pick the best ones out of a batch. Sure. But manufacturing the batch so that a large majority came out that way. Not a chance given current technology. and it gets harder the more pixels.
post #1367 of 4022
Quote:
Originally Posted by mark haflich View Post

One has to assume, in the case you postulate, that the three chips can be mounted repeatedly in the tolerance you require. This requires precise essentially identical three axis mounting. Not only to the chips have to be in three identical planes, the chips need to perfect planes as to their entry surfaces and the xy coordinates need to be the same.. so its a question of a lot of things including the ability to manufacture the parts essentially perfectly, then to precise locate them and attach at that point and keep everything the same during operation in spite of temperature differentials. Could one cherry pick the best ones out of a batch. Sure. But manufacturing the batch so that a large majority came out that way. Not a chance given current technology. and it gets harder the more pixels.

Didn't one of the Sims models have adjustable mounting that can be hand adjusted? (not by you or I but at the factory?)

One of the things that happened with this current chassis, that has never been resolved, is the warm-up changes. On the previous chassis, the convergence error was a constant. But on this chassis it changes as it warms-up and also changes differently if mounted upside down. For whatever reason, normally its the blue panel that shifts the most which is the least harmful. My X75 is like that. But the X70 I had before, it was the red panel doing all the moving and it actually rotated away from green as it warmed up. It looked superb when first switched on, and then not wonderful after it warmed up.
post #1368 of 4022
Quote:
Originally Posted by Manni01 View Post

Sorry I missed that one. The chip size of the VW1000ES is 0.74". See the last page of this brochure from Sony Pro Asia.

Thanks again Manni! I thought it might be nearer 0.9" so that is smaller than I expected
post #1369 of 4022
Quote:
Originally Posted by mark haflich View Post

One has to assume, in the case you postulate, that the three chips can be mounted repeatedly in the tolerance you require. This requires precise essentially identical three axis mounting. Not only to the chips have to be in three identical planes, the chips need to perfect planes as to their entry surfaces and the xy coordinates need to be the same.. so its a question of a lot of things including the ability to manufacture the parts essentially perfectly, then to precise locate them and attach at that point and keep everything the same during operation in spite of temperature differentials. Could one cherry pick the best ones out of a batch. Sure. But manufacturing the batch so that a large majority came out that way. Not a chance given current technology. and it gets harder the more pixels.
Tolerances are a good point and I agree that final convergence accuracy a combination of several factors likely all of which impact cost.
post #1370 of 4022
Just an update to let people know. We only have five open spots for RS57 preorder and five open spots for RS6710 preorder. Several remaining spots on the RS4910 list. The RS57 preorder list will most likely be filled before the ship time for the projectors. The 6710 list may also be filled.
Reply
Reply
post #1371 of 4022
Quote:
Originally Posted by AV Science Sales 5 View Post

Just an update to let people know. We only have five open spots for RS57 preorder and five open spots for RS6710 preorder. Several remaining spots on the RS4910 list. The RS57 preorder list will most likely be filled before the ship time for the projectors. The 6710 list may also be filled.
I'm dense and not understanding "The RS57 preorder list will most likely be filled before the ship time for the projectors." Are you saying we should expect these before the end of November?
post #1372 of 4022
Quote:
Originally Posted by Geof View Post

I'm dense and not understanding "The RS57 preorder list will most likely be filled before the ship time for the projectors." Are you saying we should expect these before the end of November?

The preorder ends when either the projectors start shipping or we sell out. The 57's will most likely sell out and the same could happen with the 6710.
Reply
Reply
post #1373 of 4022
Quote:
Originally Posted by Geof View Post

I'd rather it not need fixing in the first place.
Just to be clear, I agree with you. My points were only about the possible extra capabilities for zone and sub-pixel adjustments to have smaller maximum misconvergence in physical space on the same screen as the resolution is quadrupled, but I agree with you that it would be better to not ever need those.

--Darin
post #1374 of 4022
Quote:
Originally Posted by JonStatt View Post

...it would seem to me that the chip size still needs to increase. If not, other factors come into play including chip yields (the risk of stuck/dead pixels increases...

If you're saying bigger chips improve yields, I don't follow; seems to me that the bigger the piece of substrate, the higher chance for flaws.
post #1375 of 4022
Quote:
Originally Posted by darinp2 View Post

Just to be clear, I agree with you. My points were only about the possible extra capabilities for zone and sub-pixel adjustments to have smaller maximum misconvergence in physical space on the same screen as the resolution is quadrupled, but I agree with you that it would be better to not ever need those.

--Darin
Yes, sorry Darin, I was confident that was your position....I did not mean to imply otherwise. I enjoyed the discussion and learned something in the process....!!
Edited by Geof - 10/30/13 at 10:43am
post #1376 of 4022
Quote:
Originally Posted by AV Science Sales 5 View Post

The preorder ends when either the projectors start shipping or we sell out. The 57's will most likely sell out and the same could happen with the 6710.
Ah, now I understand....thanks.
post #1377 of 4022
Quote:
Originally Posted by noah katz View Post

If you're saying bigger chips improve yields, I don't follow; seems to me that the bigger the piece of substrate, the higher chance for flaws.

Well I could be wrong, but I based that on something I read a few years ago when smaller LCD tv's were stuck at 720p resolution, not because 1080p wasn't desirable, but because the yield was poor resulting in stuck/dead pixels. Nowadays they have overcome this. However, and maybe my assumption was wrong, that with more space to align the tracks/transistors, larger panels did not have these issues. I then extrapolated this to assume that it is probably easier to make bigger LCoS panels more reliably with fewer pixel defects than smaller ones.
post #1378 of 4022
A general question:

1. Why is projection behind in producing consumer-level 4K native projectors? Ya, I know the Sony came out a couple of years ago first, but at this point 4K panels are being churned out by practically every flat panel manufacturer. Why is projection so far behind with only Sony's 2 models? Is it just the old economies of scale, much bigger market for flat panels, that explains it?

2. Flat panel contrast is now far overtaking even JVC-projector-level contrast, with the introduction of OLED panels. Is there any reason we don't hear of similar advances in the pipeline for projection? How long will it be "left behind" in that regard, or is there some fundamental issue that makes the creation of light engines for projection harder to crack than for flat panels? (Taking the projection/room issues out of the equation, for now...)
post #1379 of 4022
Quote:
Originally Posted by JonStatt View Post

Well I could be wrong, but I based that on something I read a few years ago when smaller LCD tv's were stuck at 720p resolution, not because 1080p wasn't desirable, but because the yield was poor resulting in stuck/dead pixels. Nowadays they have overcome this. However, and maybe my assumption was wrong, that with more space to align the tracks/transistors, larger panels did not have these issues. I then extrapolated this to assume that it is probably easier to make bigger LCoS panels more reliably with fewer pixel defects than smaller ones.

I think you're both right (but I could be wrong myself) smile.gif..

If the number of pixels is constant, a larger panel should have better yields as the pixel density is lower, therefore the panel is easier to manufacture.

If the size of the panel is constant, a panel with less pixels will have better yields, for the same reason.

It's the pixel density which matters most regarding yields, like the transistors density for processors.

So Jon would be right to say a larger 4K panel should have better yields than a smaller one.

And Noah would be right to say that a larger panel (as long as he's also implying a higher density of pixels, ie 4K vs 2K) would have more chances to have defective pixels.
post #1380 of 4022
Typically smaller parts have better yields...the bigger the surface area of a chip the better the chances are of having a defect within that area. I don't think pixel size enters into the yield equation.
But bigger chips also cost more. Many semiconductors are made on 12" diameter blanks. The number of parts that fit on one 12" blank depends on the size of the parts...the more parts they can produce on one blank the cheaper they become. So smaller parts are cheaper and yields are likely better.
New Posts  All Forums:Forum Nav:
  Return Home