Jump to content

Sign in to follow this  
Mozitron

benefits of resolution and lens quality

Recommended Posts

I've been digging through all of the discussion regarding projector releases and seeing the endless debate over the fauxK vs native 4K mix blend into a mess with arguments of lens quality.

 

Given that we can calculate the benefit of resolution vs screen size vs seating distance in a simple graph like below (thanks Carlton Bale), is it possible to quantitatively measure the sharpness of a projectors lens and the resolution created form the 2x, 4x fauxK projectors and the 4k or 2x faux8K projectors? This could be represented by a plotted line for each different projector, so you could determine that at a set distance/screen size their would be no benefit in choosing a JVC Nx7 (4K) over a Benq LK990 (4x fauxK) or an Epson 5050 (2x FauxK).

 

 

image.png.8006ce69a37faf3baf18cf0e5cf964f6.png

Share this post


Link to post
Share on other sites

If you want to add another level of detail, the lines for each projector could end at a screen size where the calibrated lumens are too low for HDR or SDR

Share this post


Link to post
Share on other sites
36 minutes ago, Mozitron said:

so you could determine that at a set distance/screen size their would be no benefit in choosing a JVC Nx7 (4K) over a Benq LK990 (4x fauxK) or an Epson 5050 (2x FauxK).

its muddies .. since its highly unlikely you can receive a full 100% mtf 4k uhd source via UHD just not possible.... its been argued at best tis 3-3.5k ? how many films and source are even mastered at over 2k DI ? 

 

even a base benefit is regardless of tech we will likely see 2k and beyond for the first time ? 

 

then there is point that some eg even native 4k sony cant display a 4k pattern unless you go and mess with their internal menus...

 

a much ado about nothing i would suggest overall... rather than get too caught up in 4k and upshifted or what ever 2k base discussion is far more worthwhile in pure PQ terms...? 

 

38 minutes ago, Mozitron said:

If you want to add another level of detail, the lines for each projector could end at a screen size where the calibrated lumens are too low for HDR or SDR

dont really need a chart for this I dont think its simply matter of measuring light output given particular installation and checking whether getting in there abouts of 100 nits 30FL for gains with HDR. there is also the projector central calculator as well. but I do agree too many people likely go for far too large a screen or mount projector too far away and then wondering why its too dim ? 

Share this post


Link to post
Share on other sites

Is there a simple way of measuring sharpness/resolution that is actually portrayed on a screen during a review?

 

So far its comes down to a direct comparison with arbitrary comments about perceived sharpness with qbf patterns.

Share this post


Link to post
Share on other sites


15 minutes ago, Mozitron said:

Is there a simple way of measuring sharpness/resolution that is actually portrayed on a screen during a review?

 

So far its comes down to a direct comparison with arbitrary comments about perceived sharpness with qbf patterns.

just perceived I think...

Share this post


Link to post
Share on other sites

Optics are more important than the amount of pixels on the screen. A 16K projector running out of ordinary optics will look terrible, whereas a 2K projector with sublime optics will look better. 

 

For LCD e-shift you would need to sit a little further back, say 20% or reduce screen size when compared to say sxrd and lcos pixels. At present I'm 1 x image width and have the Epson 9400, I can see a hint of the pixel grid in bright scenes, but that is very close to the screen. most won't be that close so it's not an issue.

Share this post


Link to post
Share on other sites
17 minutes ago, oztheatre said:

Optics are more important than the amount of pixels on the screen. A 16K projector running out of ordinary optics will look terrible, whereas a 2K projector with sublime optics will look better. 

 

For LCD e-shift you would need to sit a little further back, say 20% or reduce screen size when compared to say sxrd and lcos pixels. At present I'm 1 x image width and have the Epson 9400, I can see a hint of the pixel grid in bright scenes, but that is very close to the screen. most won't be that close so it's not an issue.

yep no pixel grid with jvc pixel shift

 

this is superseded JVC from main viewing position captured with full frame camera and a lens to best capture as seen,

 

24812482936_02791aae26_k.jpg

 

this right up at the screen as close as camera can focus.... no pixels no screen door....

24471140189_cd55c39f40_k.jpg

 

while lcd has large inter pixel gap as a prior lcd owner... can say with confidence things like screen door are a non event watching from viewing position... ie you are NOT pixel peeping ! 

Share this post


Link to post
Share on other sites

Yes I agree, had both JVC x5000 and Epson 9200 before that and pixel gap is larger on Epson but at normal viewing distance it was not noticeable at all.

Share this post


Link to post
Share on other sites


I have read that the lens on the 9400 is a very good one, maybe you can confirm, Rich.

Share this post


Link to post
Share on other sites
43 minutes ago, chris2003 said:

I have read that the lens on the 9400 is a very good one, maybe you can confirm, Rich.

It's  very good. super clean, up there with the LK970 and 990 I kid you not. No other 4K projectors I've seen (JVC N7, Sony 570, Sony 760 and everything native 4K before that including the e-shift JVC,s) can best the BenQ lenses only this epson and it's not even native 4K.. which is clearly doesn't need to be.

 

Seems we're getting the 6050 model in the USA which Art claims 'has the better lens' than the 5050 model.. so win win finally for Australia! Though I'll try and confirm this with Epson.

 

Taken with a 6 yr old phone camera, no tripod. 3.8 metres away 165 inch scope screen zoomed so about 180 inch 16:9 image. cinema mode, high lamp, mildly tweaked until I get it calibrated.

 

I finally see what all the fuss with HDR is about too. .it's quite astonishing switching between SDR and HDR on the fly.

 

 

20190514_215929.jpg

20190514_212326.jpg

20190514_122157.jpg

Share this post


Link to post
Share on other sites

Thanks Rich, Could you also confirm if the 9400w when it arrives has the same lens as 9400, I would assume so , but if you could confirm with Epson. I have read HDR on the 9400 is so much better than the 9300 with its curves loaded. Just sold my other house and new owner wanted the JVC to stay so am without projector, so am in the hunt for a new one.

Share this post


Link to post
Share on other sites
9 minutes ago, chris2003 said:

Thanks Rich, Could you also confirm if the 9400w when it arrives has the same lens as 9400, I would assume so , but if you could confirm with Epson. I have read HDR on the 9400 is so much better than the 9300 with its curves loaded. Just sold my other house and new owner wanted the JVC to stay so am without projector, so am in the hunt for a new one.

even if they use the lens they have always been using, I believe going back to start of 8000-9000 series its been an all glass fujinon affair. 

Share this post


Link to post
Share on other sites


forgot to add this one

20190511_203232.jpg

Share this post


Link to post
Share on other sites
23 hours ago, Mozitron said:

Is there a simple way of measuring sharpness/resolution that is actually portrayed on a screen during a review?

No.  Measurements for sharpness/resolution of lenses (two different issues by the way) required special equipment (an MTF test bench) and the lens must be removed from the projector to be tested.

I have never seen any test of a domestic projector lens ever, its the sort of thing only companies who make lenses do. If you buy a lens for a commercial cinema projector (BIG $$$) it may come with an MTF test chart to prove it meets specifications.

 

When viewing images from a domestic projector you are looking at a "system" not just a lens, its a combination of the lens, the optical engine and the drive electronics, plus any enhancements the electronics applies, and there is almost certain to be some even when you think its turned off.

When it comes to image sharpness its not really possible to ascertain if what you see is due to the lens or other parts of the "system", in particular the projectors electronic sharpening which has a much greater influence over the overall end result then the lens, unless the lens is absolute crap, and its not going to be unless its faulty. The lenses in cheap data grade projectors are remarkably sharp, right down to the pixel level, but that does not mean they produce good quality video images, they don't.

To put lens resolving power into perspective, the black gap that surrounds an Sony LCoS or JVC D-ILA pixel is less than 10% the size of the pixel, that represents a spatial frequency equivalent to a “revolution” of about 20K for a 1080 panel and 40K for a 2160 panel, think about that for a while.

 

The requirements for data presentation (PC text and graphics) are VERY different to whats ideal for video. Video captured from any type of camera MUST BE low pass filtered, everything at the pixel level is therefore VERY blurred and no sharp edges exist. When we reconstruct that real world analogue image on screen we don't want to introduce hard pixel edges as sharp edges at the pixel level represent spatial frequencies many, many times higher than anything that was captured by the camera and are nothing but distortion. The real world is not made up of little sharp edged blocks so they have no place in the on screen image.

So, for PC text and graphics sharply defines pixel edges are desirable, however for video content they are not. We want soft pixel edges that blend together seamlessly for the most accurate image. Sharp pixel edges, especially ones with significant and distinct black borders introduce what I call false sharpness, many like this sharper “look” but its not an accurate representation of the image.

 

The true sharpness of video images cannot be ascertained by looking at pixels or their edges so evaluating a projector intended for the display of movies by looking at computer generated test patterns up close is pretty pointless and tells you stuff all. Sharpness for movies is all about relative image contrast over a wide range of spatial frequencies, with low to middling spatial frequencies (resolutions) being BUY FAR the most important for the human visual system and they are heavily affected by digital sharpening. Two projectors that look the same when viewing pixels can be quite different when displaying the lower spatial frequencies critical to video image sharpness, so simply looking at pixels is not a wise move. Look at the projected image as a whole viewing movies, if you start pixel peeping you have lost the plot IMHO.

 

If you pixel peep with a triple chip projector (anything other than single chip DLP) you will be encouraged to start playing with the so called "convergence" correction controls to better align the red, green and blue images.

The imaging chips are fixed in place and cannot be re aligned outside the factory, all the adjustment system is doing is using 2 or more pixels to do the job of one, its a blur function and should NOT be used if you value "resolution". Make sure the sub pixel alignment system is DISABLED for best results, on a Sony that requires service menu settings which not everyone would be wise playing with, just enjoy the projector as it is.

The inexperienced my also be temped to correct chromatic aberrations (purple-blue fringing, a lens problem) using the "convergence" system and doing so with REALLY screw things up, NEVER EVER DO THAT.

If convergence is more than about halve a pixel out for red and green (blue is not critical) or drifts off significantly in different areas of the screen the projector is faulty IMHO. Having said that, at 4K such errors are insignificant and not visible so don't get all worked up about perfect convergence, sit back and forget about it.


 

It should be remembered that video is VERY limited in actual visible resolution. Even under laboratory perfect test conditions images shot at 4K plus with the best cameras and mastered in 4K are nowhere near “4K” visible resolution, even without any video compression. This is due to that requirement to low pass filter to make sure that any spatial frequencies higher than the 4K limit are filtered out as they can irreversibly damage the image. The end result of multi stage low pass filtering is an amplitude response that starts falling off at below 1K, is significantly down at 2K and typicality more than 70% down at 3K. Beyond 3K response is so low its useless and thats “BEST CASE”.

So our perfect 4K video poorly resolves 3K and can only achieve that if the subject detail is very high contrast, most detail is not so the practical limit is lower.

 

Domestic video is chroma sub sampled, so colour is encoded into 4 pixel blocks with only one colour value for all 4 pixels. This means colour resolution is HALVED and for 4K video is limited to less than 1.5K. Some much for “4K”, its at best 3K luma resolution and 1.5K chroma. 
 

There is a lot more to a quality lens than sharpness, many cheap lenses are sharp. Other important aspects of performance sort the good lens from the also rans such as even focus over the entire screen at all zoom settings, freedom from chromatic aberrations (colour fringing), freedom from vignetting (dimmer corners), lack of geometric distortions over a wide range of zoom and lens offset, freedom from internal reflections, etc, etc. 

To do everything well is very difficult and invariably involves compromises or LOTS of money.

 

Lens quality control is also an issue, a review sample may be very good and the one you receive not so great in one area of performance or another, its luck of the draw unfortunately. 


 


 

 

 

 

Share this post


Link to post
Share on other sites
11 hours ago, Owen said:

If convergence is more than about halve a pixel out for red and green (blue is not critical) or drifts off significantly in different areas of the screen the projector is faulty IMHO. Having said that, at 4K such errors are insignificant and not visible so don't get all worked up about perfect convergence, sit back and forget about it.


 

There is a lot more to a quality lens than sharpness, many cheap lenses are sharp. Other important aspects of performance sort the good lens from the also rans such as even focus over the entire screen at all zoom settings, freedom from chromatic aberrations (colour fringing), freedom from vignetting (dimmer corners), lack of geometric distortions over a wide range of zoom and lens offset, freedom from internal reflections, etc, etc. 

To do everything well is very difficult and invariably involves compromises or LOTS of money.

 

Lens quality control is also an issue, a review sample may be very good and the one you receive not so great in one area of performance or another, its luck of the draw unfortunately. 


 


 

 

 

 

Owen, thanks for your detailed input. But I dare say there is a vast difference in image quality when you see a decent quality lens vs something that's full of chromatic aberrations. The difference can be night and day. As you point out there is more than just the lens quality that makes the image acceptable or not but a good lens is really noticeable. The BenQ LK series projector have sublime lenses, they even proudly tout this. but these are 13 and 15K price tags whereas cheaper projector won't have such quality optics. I've seen this time and time again from perhaps 60+ projectors I've had in my showroom over the years.

 

And if the convergence is out by more than half a pixel, wouldn't you simply move the panel 1 full pixel with the other way? I thought you said that's the only time you should use such adjustments? And that the micro 1/16th adjustments were software driven and blur the pixels? It's when you're out by just less than half a pixel width that you couldn't or shouldn't use the alignment feature? Perhaps I see the colour fringing more than the average person.

 

The new Epson lens is seemingly fantastic when compared to other units even with native 4K panels.

Share this post


Link to post
Share on other sites

 

13 hours ago, oztheatre said:

And if the convergence is out by more than half a pixel, wouldn't you simply move the panel 1 full pixel with the other way?  I thought you said that's the only time you should use such adjustments?

Thanks mate, I aim to be informative.

Digital projectors are "fixed pixel" displays and the red, green and blue imaging chips can't be moved/realigned outside the factor setup environment. The only so called "convergence" adjustment that can be done without image degradation is to move the image on one or more of the imaging chips electronicly in full pixel steps horizontally or vertically to better align the colours, but this should have been done in the factory and be the "zero" or all adjustments disabled point. Obviously moving the image by a full pixel affects the entire screen by the same amount and if there was a half pixel error there will still be a halve pixel error after adjustment, the misalignment will just be in the other direction.

If, for example, red is offset to the right of green by say 3/4 of a pixel, moving red to the left by a full pixel will give an error of 1/4 pixel to the left, which is obviously a better solution. One must be very careful to use an appropriate test pattern to check that you have not inadvertently incorrectly diagnosed which pixel-line the red pixel-line belongs to or you will end up moving red in the wrong direction and end up with a 1 and 1/4 pixel error in one direction or a 1 and 3/4 pixel error in the other. This can look just fine on a some test patterns but is obviously NOT good for a real picture.

If the factory didn't do its job by setting the full pixel adjustment correctly and set it as the zero point of the user adjustment there are improvements to be had by full pixel "adjustment".

 

14 hours ago, oztheatre said:

And that the micro 1/16th adjustments were software driven and blur the pixels?

Thats correct. We can't actually move the pixels in a digital projector, we can only use adjoining pixels to counteract a colour fringe caused by mis convergence of less than a full pixel.

I'll give an example. Lets say green and blue are aligned but red is offset to the left by half a pixel causing a red fringe. The "correction" system can't make the red go away, it can only illuminate the green and blue sub pixels belonging to the adjoining pixel on the left to neutralise the red colour of the fringe, however in doing so it has made the pixel-line we where attempting to correct 2 pixels wide. You don't need to be a rocket scientist to figure out what that does to "resolution".

 

So, using anything other than the full pixel adjustment involves using 2 or more pixels to do the job of one and obviously degrades resolution. Correcting a half pixel error can halve the resolution of the projector instantly and yet Sony have been using sub pix "correction" on their 4K projectors from the factory and give the user no way to turn it off. The amusing part to me is that users just don't notice the loss of resolution because 4K video is never actually 4K (other than in pixel count). Manufacturers take advantage of this to hide poor factory convergence which could lead to consumers returning projectors. What they don't know wont hurt them, as long as they don't see significant colour fringing they are happy and don't complain so problem solved. Its cheaper than better quality control of the optical engine manufacturing process and allows a lot of less than best optical engines to be used.

 and never will be so getting all worked up about "resolution" is just plain silly IMHO.

 

On 17/05/2019 at 7:22 AM, oztheatre said:

It's when you're out by just less than half a pixel width that you couldn't or shouldn't use the alignment feature?

Whatever errors remain after the full pixel alignment CANNOT be corrected without causing blur and a notional loss of "resolution". I say "notional" because movies are inherently very blurred at the pixel level to begin with and unless you view digitally created test patterns people aren't going to be aware of it, especially with 4K systems. 

 

Share this post


Link to post
Share on other sites


Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this  

  • Classifieds Statistics


    Currently Active Ads

    Total Sales (Since 2018)

    Total Sales Value (Last 14 Days)

    Total Ads Value (Since March 2020)
  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...