FauxK vs 4K Projectors – A Look At True 4K, 4K UHD and 1080p Pixel Shifters

Faux-K or FauxK: What is it? First, some writers started referring to projectors that could accept 4K content, but didn’t have 8 million discreet pixels to show it, as “FauxK.” I joined that party. I wish I had invented the term, but I commend whoever did. I found an early thread on one of the forums from way back in 2011 with another reviewer “debating” what he felt was a 1080 pixel shifting projector line that might be misconstrued at true 4K. He used the term FauxK. For those who are curious, there’s a link below.

VPL-VZ1000ES_4K-HDR_deepwater-horizon_rig
4K content from The DeepWater Horizon, projected by Sony’s ultra short throw true 4K projector – 8.3 megapixels no overlapping pixels.

FauxK vs 4K – what’s the verdict? I consider any projector that can put at least 3840 x 2160 discreet pixels on the screen – without any overlap, to be “true 4K” (not an official term).

Let’s start with the simple truth that FauxK isn’t any official term that some standard committee has blessed. Consider it slang.The term has been showing up projector industry since at least 2011.

Here’s the thing: Traditionally we measured resolution (still do), by the number of discreet pixels horizontally and vertically, such as 1080p being 1920 x 1080 pixels, that’s roughly 2 megapixels. But, when pixel shifting showed up first in projectors by JVC, and shortly after by Epson, we ended up with a 1080p projector capable of hitting the screen with 4 million pixels, but it required overlapping those pixels. They didn’t create smaller pixels so they could get more on the screen, they just fired them twice, shifting by a half pixel. There are real advantages to that, but there is one truth that holds for all pixel shifting projectors.

They cannot produce as sharp an image as a projector with the same number of pixels hitting the screen, when the other projector has discreet pixels – each one is separate, no overlap. Technically, it’s no contest! It’s this simple:

The area of each 4K (3840 x 2160) pixel on, say a 100″ screen, is exactly one-fourth the area of a 1080p projector’s pixel, whether that 1080p projector is a pixel shifter or not. Those projectors simply cannot be a sharp (unless there are serious problems with the 4K projector. The true 4K projector can do a perfect vertical or horizontal line, that’s 1/3840 of the screen width. For a pixel shifter that’s 1080p, that line has to be at minimum twice as thick as a true 4K line. Got it?

Epson-LS10500_4K-HDR_ghostbusters_manhattan
CGI enhanced image of lower Manhattan, projected by Epson LS10500 1920×1080 with pixel shifting, from 4K Ghostbusters.

 

Now welcome to the world of DLP – TI (Texas Instruments) created a new DLP chip, neither true 1080p, nor true 4K (we’re going to stick to consumer 4K – technically true 4K would be 4096 pixels (as is found in movie theaters), not 3840.

This DLP chip has 2716 x 1528 discreet pixels, not 3840 x 2160. Multiply that out, you get 4.15 megapixels and by using pixel shifting, they get 8.3. That 4.15 megapixels is double the standard 1080p projector and the same as a pixel shifting 1080p projector.

HT8050_ghostbusters-4K_manhattan-fauxk
CGI enhanced image of lower Manhattan, projected by the BenQ HT8050 4K UHD projector with 2716×1528 native resolution plus pixel shifting, from 4K Ghostbusters.

Well, folks, that makes them 4K UHD, not true 4K, by my take. What is 4K UHD? It is a resolution used by Blu-Ray UHD players and YouTube (UHD = Ultra High Definition). That is, it’s part marketing, part reality.

To test maximum resolution, I’m creating a 3840 x 2160 image in photoshop of vertical lines, all one pixel wide, I’ll change colors as frequently as from one pixel to the next.  Something like R,G,B,R,G,B and then others such as: R,G,B,C,Y,M, RRGBB, GGYBB, and wider ones, perhaps R,R,B,B,G,G… Only the true 4K projector will be able to see all those colors.

VPL-VZ1000ES_4K-HDR_ghostbusters_manhattan
CGI enhanced image of lower Manhattan, projected by the Sony VZ1000ES true 4K projector with 3840×2160 native resolution plus pixel shifting, from 4K Ghostbusters.

Click on, and zoom in as close as you can – say, to the area around the tallest tower – and look closely at the lighting. The LS10500 1080p pixel shifter is a little soft – less detailed than the BenQ HT8050 the 4K UHD projector, which in turn is not as detailed and defined as the $25K Sony VPL-VZ1000ES immediately above. The differences aren’t huge, but they make a visible difference. One note, the Sony needed to have its digital pixel alignment done. I hadn’t noticed that I had taken this image before I did the pixel alignment until later. You can see a green edge on one side of pixels and a blue on the other.

I say that using this logic: If that’s real 4K, then so would a “4K UHD” projector based on a standard HD – 720p – of 1280 x 720 (you know, that’s the lower HD resolution, the one ESPN and ABC still broadcast at). Using pixel shifting, a company could start with that native resolution and shift pixels 9 times and you are back at 8.3 megapixels. Of course, each of those individual pixels would be about the size of a bowling ball compared to a baseball (for true 4K), or maybe a baseball compared to a golf ball.

Hey, I didn’t do the math on those, call them guesses.  After all, each of those pixels would be 9 times the size of one true 4K pixel. Good luck getting fine detail out of that, but I believe it would still technically be 4K UHD. (I haven’t read the specs for 4K UHD beyond the 8-megapixel requirement, and the various relationships to tech like BT2020 color space or HDR.)

So, since Faux-K doesn’t have a clear definition, the one I’ve been using the past few years is simply this:

If a projector can’t put up 3840 x 2160 pixels without pixel shifting (call it overlapping pixels), but can input 4K content and display it – that’s my definition of true 4K, and that’s the one we use in our reviews.

If it’s a projector I call Faux-K, it’s probably a 4K UHD projector, or a 1080p pixel shifter, but conceivably there are other ways to put at least 8.3 million pixels on the screen. (Not that they make sense.) Normally, when I am talking 4K-content-capable, I’m always trying to be clear whether it’s a 1080p pixel shifter, or now that we have projectors with the new TI chip, also 4K UHD (aka a 1528p pixel shifter), or a true 4K projector.

OK – regarding that early Faux-K reference. That far back, I recall an article by Andrew at HomeTheaterReviews, from 2011, of him using the term to describe pixel shifting 1080p projectors that accepted 4K content. In light of recent conversations with various manufacturers about 4K resolution of late, I’m reminded of follow up his and JVC’s follow-up comments on one of the forums. There, he and JVC debated whether their projectors were being misrepresented by using terms such as “4K Precision,” and it’s an interesting back and forth. The point being, the term Faux-K (or FauxK) has been batted around for quite some time without a clear definition of what Faux-K is, only what it is not.

As far as the terms go, FauxK vs 4K, I pretty much treat Faux-K as a fun term to describe competent projectors that are capable of accepting 4K, but can’t pass my “true” test. Manufacturers should not be offended. At the same time, makers of 1080p pixel shifters and 2716 x 1528 UHD projectors should be clear in their marketing. They are not true 4K, but they accept 4K content, and use advanced techniques to project them so that they seem more detailed and sharper than standard 1920 x 1080, or 2718 x 1528 projectors that lack pixel shifting, but that they cannot resolve detail as fine as a true native resolution 3840 x 2160 projector can.

Between 4K, FauxK, and the addition of HDR and BT2020 color space, things are still a bit murky as we all scramble to understand the benefits, and limitations, and how to maximize the viewing experience when there are trade-offs. In addition, things will be changing at a healthy pace going forward. Hang on for a bit of a bumpy ride. – art

 

News and Comments

  • aksdad

    Your “not as detailed and defined” statement is a subjective analysis of the picture quality. There are differences in contrast and color quality and accuracy between the BenQ HT8050 and the Sony VPL-VZ1000ES—which costs 3 times as much as the BenQ—but they are both displaying a true 4K picture.

    The BenQ can display 8,294,400 discrete pixels (3840 x 2160) and the Sony can display 8,847,360 discrete pixels (4096 x 2160). Both of these resolutions are considered 4K. To be precise, the BenQ’s 3840×2160 is the television and Ultra HD Blu-ray standard, sometimes called UHDTV, UHD-1 or 2160p, and is the resolution of all the 4K TVs on the market and all the 4K Blu-rays. It is also far more common than the slightly higher 4096×2160 DCI motion picture standard (DCI 4K) which is typically found in expensive digital cinema projectors. The difference, about 6%, is negligible, but both are considered 4K and both have 4 times the resolution of HD (1920×1080).

    There is nothing “faux” about the 4K resolution of the TI 4K DLP chip used in the BenQ. At CEDIA 2015, TI demonstrated with still images that their pixel-shift technology was indeed producing 8.3 million discrete pixels. See the closeup of the spreadsheet in this image:

    https://www.insightmedia.info/wp-content/uploads/2015/10/TI-spreadsheet-image.jpg

    You can clearly see the single-pixel white line between the single-pixel black lines forming the border. The TI chip was dramatically sharper than the LCD pixel-shift 4K on JVC and Epson projectors and was considered (subjectively) sharper than a Sony projector using native 4K SXRD chips because of the common problem of aligning the Sony’s three 4K chips (one for each primary color) which produces lines that aren’t quite as sharp due to slight color shifts.

    By the way, if you’re demonstrating the differences in perceived resolution between 4K projectors, the pictures posted in the review should be at least the same resolution as the projected 4K image, or higher (at least 8 megapixels). Images that are only 1.7 to 2.6 megapixels (about a quarter the resolution of 4K) can show differences in color and contrast but they can’t pick up individual projector pixels to show how detailed and sharp the projected image is. If you want to keep the image sizes small, perhaps you could take a close-up of the same area of the image or just crop the images to 1/4 of the projected image.

    • ProjectorReviews.com

      Hi aksdad, I’m always up to debate what’s faux K.
      First, I concur – I don’t distinguish between 3840 and 4096, to me that’s an aspect ratio difference. My opinions are subjective – always have been. I deal in the practical when viewing a projected image.

      BTW regarding the images sizes we put up, yes most are too small, but the highest resolution ones I do put up, are usually one of a pair, such as a credits shot in the lab from Ghostbusters 2016. First I show about the full frame, but then I zoom in to less than 1/4th of the area of the frame (often a lot less), and shoot those at 2000 pixels wide. As a result, I’m using more pixels for that image, than the data has. But, to do it right, I really should be shooting at a resolution that would be 16K or higher (across the whole screen) so that the affects of camera pixels hitting image piixels don’t create patterns or soften the image.

      I’m not thrilled with the image you linked to. First of all it’s low resolution. and no scale. (BTW, I did see the TI demo, although I think it was 3 months later at CES 2016). At that demo, there was a Sony, and an Epson. I remember commenting that I had never seen a worse aligned Epson. but then one never trusts demos from one company when they show their competition, unless their competition tuned their one product. (Epson had confirmed they hadn’t), but that’s neither here nor there.

      I fully agree that any good DLP 4K UHD is visibly sharper than any good 1080p 3LCD or LCoS pixel shifter. We’re simply debating the the term Faux-K. I also agree that on the right content, and with good optics, a single chip DLP projector that’s 4K UHD can look sharper than a true 4K Sony with although the panel alignment of the Sony can be awefully good, so that misalignment is only a small fraction of the size of their pixels.

      Again, it comes down to the smallest object that can be clearly defined, and a good Sony (or any other true 4K) should rule in that case.

      Re the image you sent, the nice thing about pixel shifting is you don’t have to.

      I haven’t thought through all the implications, but that single pixel wide white line inside of two black lines, I would suspect is 1/2718th of the screen width. How could they possibly make it smaller than that (1/3840)? That’s the big question.

      Look at the vertical and horizontal spreadsheet fine lines (not those drawn as objects fonts and numbers. Those fine lines should be razor sharp, why aren’t they, unless there is overlapping data.

      To me its simple – true 4K can completely resolve a single red line 1/3840th of the width of the image, and fully resolve a second vertical line adjacent to it on it’s right, also with a width of 1/3840 of the screen width that is green. I’m happy to call any projector that can do that “true 4K” Having black next to white masks what would be learned if it was red next to green, since black is “off.” In that case, only every other line is projecting light, not every line.

      “True 4K” With a Sony 4K projector (or the JVC, or some really high end ones), no problem putting that red line on the screen with the green line next to it. (give or take the usual relatively minor misalignment of panels (which wouldn’t be detectable if we are only doing primary colored lines.)

      With any pixel shifting 2716×1528 DLP chip based projector, trying that, at best will get you red fringe on one end, green on the other, and yellow in the middle. (Andof course neighboring vertical lines will also impact the one to the left of the red, will affect part of what’s supposed to be red, while It’s that simple.

      The CTA standard (not exactly a organization that focuses primarily on display issues) is too soft for my taste. One could have an Epson 1080 pixel shifter such as the 5040UB or LS10500, or any of the 1080p JVCs, and using firmware, have them shift 3 times (with design modifications needed I assume), instead of just one:

      That would make those 1080p pixel shifters 1920x1080x4 = 8.3 megapixels. So, even with their pixels being far larger than the DLP chip’s they too would still be 4K UHD.

      Further, one could even design (not that anyone would – except to make a point) – a projector that pixel shifts 8 times using 1280×720 panels, and you still get 8.3 megapixels – 4K UHD. But those 720p pixels would be 9 times the area of a true 4K – basically a basketball compared to a baseball. The final result on the screen would probably be mush, but my point is, technically it would still be 4K UHD.

      Bottom line, I don’t care for that standard relying only on # of pixels. I think it should be number of pixels on the screen and the size of the smallest resolvable dot, or line.

      BTW using white lines and black is beneficial to supporting pixel shifting. It’s when using colors that things start becoming “clear.”

      OK way too long a ramble. Other than that… Nice to meet you! -art

  • TK Goldman

    “If a projector can’t put up 3840 x 2160 pixels without pixel shifting (call it overlapping pixels), but can input 4K content and display it – that’s my definition of true 4K, and that’s the one we use in our reviews.”

    …don’t you mean ‘If a projector CAN put up 3840 x 2160 pixels without pixel shifting”?