DIY Dual-Layer LCD monitor

Could you use just a sheet of cellophane?  Please forgive me if I'm wrong. Something like this: https://amzn.eu/d/dGW4zCG

Wouldn't excluding the first diffusion layer create brighter & darker spots in the first LCD? The depolarising layer will diffuse the light but the second LCD may still show inconsistent contrast levels. Either way, this is a very impressive DIY feat considering the technology has a starting price of around £20,000.

I think the Monitor turned out pretty cool. I never thought you would reach black levels that good with just two LCDs.

For improvement I have some things floating around my head:

I am not sure if this is possible with the LCDs you used but I have seen videos of LCD screens where the second Polarizing filter was removed.

Searching "Private LCD" should yeald some results. This would remove the need for the filter paper and it could also stop the shadowy effect behind the text.

Another idea could be to only take the LCD layer itself from the scond monitor and sandwich it into the other one. Hower I am not sure if this is possible without breaking the LCD as i guess that it may be glued together in most panels.

Fascinating as always. It might be worth mentioning that this is very similar to how Samsungs QD-OLED displays work, as in using multiple layers. While they aren't 2 lcd panels, it does use two layers. The first is the OLED panel that just produces blue light across the whole screen while the second quantum dot panel changes the colors of that blue light. Just note that the QD layer doesn't just filter otherwise you'd only get blue light. Probably a bit hard to replicate.

For this project, what happens if you remove the polarization layer from one of the lcd panels? I've seen a video where you can remove a screens polarization layer and stick it to some glasses to create a privacy screen so perhaps it can be removed to help with the efficiency? 

Why not remove the first LCD polarisation filter?

People sometimes that to make screen white unless you have glasses with matching polarisation filter.

I think if you can write a code where if amount of light colored pixels on screen are more than the dark pixels, then the 2nd panel (which is below the visible the visible one) should turn white. this will increase the brightness of the screen when viewing bright content.
Further, the amount of white of the 2nd panel could be adjusted dynamically. by trial and testing.

PS: I don't know if it will work or not. just a random idea. 😆 

Dude this is AWESOME! That is all. 😉

Could you just add a polarizing sheet at a different angle (I forget if it’s 45 or 90) to let the polarized light through? 

Or is that what he was referring to as the expensive “reverse polarizing sheet”?

Very good experiment! 😉

I'm currently planning to do a similar experiment with a dual layer LCD projector. Hopping to get an insane contrast!

For a dual layer LCD monitor I think a simple way to make it more efficient. For better light transmission, best to use glossy panels instead a matte finish panels. And to match the polarization between panels flipping the first panel upside down, and now you can remove the diffuser film between panels. I think it would be more efficient by far.

Regards!

I know about the BVM-HX310, but I never really thought about doing a DIY version of this. It's pretty insane. 

What kinda brightness does this get (in nits)? I think doing this with HDR and smaller 4K screens to reduce power consumption could be super interesting, but I have no idea how much power that would require. The one in the video does look darker than the seemingly 300 nit Asus monitor though. I also didn't spot any brightness enhancement films in the video, so maybe that could help some more. 

edit: I've been looking for hours and can't find good answers regarding color gamut/volume and EOTF calibration on driver boards, which would be essential for HDR and there are basically no existing solutions that I'm aware of. 

If you could horizontally flip the image displayed on one of the lcds, you could mount them in a back to back configuration. This should theoretically match the light polarization of both lcds where they mate up, and remove the need for the intermediary diffuser while reducing the amount of light blocked from the polarization layers (every time nonpolarized light passes through polarization layer the intensity is cut by roughly half).

However knowing absolutely nothing about how lcd panels are actually driven, flipping the image for one panel is probably a nightmare task. I imagine its also important for the rgb sub pixels to match up as well, which it probably wouldn't anymore.

EDIT: This is one of those things that seems like it would be pretty easy to do from a manufacturing standpoint. Just slap an additional lcd layer and its top/bottom electrodes right behind the one already there, which should have a minimal impact on the maximum brightness of the monitor.

This is all based on some quick googling and could be based on a lack of understanding of lcd principals.

Super cool!

Can the colour filter be removed from the lower panel?  It would hurt the saturation but I wonder if it would be worth it for the increased brightness.

Dealing with polarization between the layers seems like the biggest challenge for the DIY solution.  Finding an alternative to the diffuser would be a huge win.

For years I've wondered why manufacturers haven't launched this tech to compete with OLED in the consumer space.  It seems like manufacturing costs should be manageable.  Maybe it's because of the brightness problem?

Thanks for the extra tips!

Sticky tape like Scotch tape? Laying down strips perfectly flush to each other without overlapping may be tedious, but the results may be worthwhile. I'll try to think of similar products and materials that are available in the form of sheets.

I presume a diffusion layer should still be used somewhere in the assembly, or do two LCD panels combine to diffuse the light sufficiently?

Posted by: @diyperks

↑

 

Some thoughts on making it brighter/more efficient would be to exclude the first diffusion layer (the one I fitted on top of the acrylic) as I actually don't think it's necessary when you consider that the second tracing paper layer (the depolarising layer) will do this diffusion anyway. Needs testing though.

 

Another idea is to rotate the polarised light rather than diffuse it. Surprisingly sticky tape can do this (acting a bit like a half wave plate) but they're in small strips so unless the backing plastic can be found somewhere in a large sheet form I'm not sure how good it would look. It preserves brightness shockingly well, though.

Mr. Engineer.

Is there a functional panel to enlarge the image on the TV.

For example:

To put some material on my 50-inch TV (in front of the panel) that enlarges the image to many more inches.

p.s.

Sorry for the really complicated question.

I wonder if the desired effect could be achieved using a single lcd screen but using the mechanisms of a CRT TV(tube tv)

A CRT works by scanning an electron beam across a phosphorus grill very fast, and like an OLED display illuminate pixels individually producing the desired blackness (the older crt models were terrible at this but the newer models were getting pretty close to almost oled level blackness, sadly the tech was dropped in favour of cost effective led and lcd screens) also CRT tvs can have refresh rate of 75Hz and maybe overclockable.

A crt could be modified replacing the grill with the lcd screen and a uniform florescent paper behind it so each pixel is individually iluminated

Of course the circuitry would have to be modified and getting the signal timings right would be difficult but i assume an older lcd screen with vga input and a similar era crt monitor could be used as proof of concept, Running the same vga signal to both the lcd and the monitor via a split vga might work.

The end result could be a good retro gaming monitor with authentic scanlines while maintaining image sharpness.

Couldn't you get that 90 degree rotation by putting the panels back-to-back? There's nothing saying the light has to go back-to-front, so as long as you can mirror your displays with one of them flipped you can get higher brightness by keeping the light between the panels ideally polarized

Could also make it brighter by having the panel behind be grayscale and a slightly lower resolution. Human eyes have a certain amount of bloom already, and can't detect contrast accurately when extremes are next to each other, so you can get away with some bloom there in exchange for making point lights much brighter (pure white/color for a larger patch behind the outer "detail" layer)

Some really interesting suggestions here, particularly those around back-to-front arrangements. More experimentation needed!

Posted by: @chopped

↑

Could you use just a sheet of cellophane?  Please forgive me if I'm wrong. Something like this: https://amzn.eu/d/dGW4zCG/p >

That's actually not a bad suggestion... I'll order some 🙂

Posted by: @amaraldo

↑

Wouldn't excluding the first diffusion layer create brighter & darker spots in the first LCD? The depolarising layer will diffuse the light but the second LCD may still show inconsistent contrast levels. Either way, this is a very impressive DIY feat considering the technology has a starting price of around £20,000.

Initially I thought yes, but thinking about it more the brightness spots are only visible to my eyes from my perspective. From the LCD pixel's perspective it's just getting light that's hitting it from the back, and it will cast a shadow onto whatever is in front of it (the tracing paper) hence my thought that it's not necessary.

Posted by: @mrpuravida

↑

Could you just add a polarizing sheet at a different angle (I forget if it’s 45 or 90) to let the polarized light through? 

Or is that what he was referring to as the expensive “reverse polarizing sheet”?

That's a really interesting thought. Changing both polarisation sheets on a single LCD panel (there are two, one on the front, one on the back) may yield some interesting results.

Very interesting project! I did have an idea about how to reduce the power consumption by reducing the need for such an increased brightness to selectively black out dark areas. You have two different image signals you can input into the panels, but they don't need to be identical. What if the back LCD was displaying a matted out version of the image with just the areas that need to be very dark, or black,  showing and then everything else would be #FFFFFF white to allow the maximal light through, and have the second LCD signal actually contain the color information for those brighter portions.

Say you were trying to light the screen with a full red image. You're passing the light through two red filters, which is decreasing the brightness each time, even if it's the same color (No filter has perfect transmission), but if you 'removed' one of those filters, you'd have an increased brightness with the same coloration, as double red will not be any redder than a single pass through the same type of red filter. You're just wasting energy.

You could even adjust the threshold of what gets matted and what doesn't during software image processing. Instead of using a splitter, just use a single GPU with two HDMI outputs, and mirror one screen's 'image' onto the other with whatever post-processing you'd like, like a selective pixel matte. The delay should be negligible, especially on a 60hz panel. How much this would actually help, I have no idea, as I don't know the transmission loss from the redundant light filtering, but it's definitely worth a shot!

Back-to-back with one screen mirrored is a good idea but I worry it won't work because the colour filters are no longer aligned?

You are a wonderful engineer. You exposed the companies that produce screens, as the cost of a fairly good screen appears from 500 dollars, and it is also of high quality in terms of color. My point is that the companies do not cost in manufacturing, but they sell you just a name and brand only, no more. I hope that the companies reduce the prices In order for customers to buy high-quality products and wonderful pictures

(87) Circular Polarization - YouTube

see if it helps

Posted by: @fakeinternetnametm

↑

I wonder if the desired effect could be achieved using a single lcd screen but using the mechanisms of a CRT TV(tube tv)

A CRT works by scanning an electron beam across a phosphorus grill very fast, and like an OLED display illuminate pixels individually producing the desired blackness (the older crt models were terrible at this but the newer models were getting pretty close to almost oled level blackness, sadly the tech was dropped in favour of cost effective led and lcd screens) also CRT tvs can have refresh rate of 75Hz and maybe overclockable.

A crt could be modified replacing the grill with the lcd screen and a uniform florescent paper behind it so each pixel is individually iluminated

Of course the circuitry would have to be modified and getting the signal timings right would be difficult but i assume an older lcd screen with vga input and a similar era crt monitor could be used as proof of concept, Running the same vga signal to both the lcd and the monitor via a split vga might work.

The end result could be a good retro gaming monitor with authentic scanlines while maintaining image sharpness.

---

 

This is a great idea, so great in fact that JVC pretty much did this! Their LCCS displays used a black and white CRT to essentially function as a sophisticated back light for an LCD panel placed right in front of it. The color info is sent to the LCD so that when CRT's black and white image gets filtered through it, you end up with a clean full color picture.

These displays were only ever sold and used professionally, and weren't on the market for much more than a brief blip. While they did provide a very good image, today they're not much more than a novelty and a neat relic of societies transition from CRT's to flat panels. That said, they're pretty damn cool! It's wild to see the same idea pop up organically decades later!

Posted by: @fakeinternetnametm

↑

I wonder if the desired effect could be achieved using a single lcd screen but using the mechanisms of a CRT TV(tube tv)

A CRT works by scanning an electron beam across a phosphorus grill very fast, and like an OLED display illuminate pixels individually producing the desired blackness (the older crt models were terrible at this but the newer models were getting pretty close to almost oled level blackness, sadly the tech was dropped in favour of cost effective led and lcd screens) also CRT tvs can have refresh rate of 75Hz and maybe overclockable.

A crt could be modified replacing the grill with the lcd screen and a uniform florescent paper behind it so each pixel is individually iluminated

Of course the circuitry would have to be modified and getting the signal timings right would be difficult but i assume an older lcd screen with vga input and a similar era crt monitor could be used as proof of concept, Running the same vga signal to both the lcd and the monitor via a split vga might work.

The end result could be a good retro gaming monitor with authentic scanlines while maintaining image sharpness.

---

 

This is an interesting idea that I also have thought about once or twice. However there are multiple issues which make this not really effective:
First of all a CRT doesn't have the pixels/lines aligned perfectly. Small magnetic interferences and imperfections usually cause the picture to warp weirdly around the corners. This makes it impossible to align an LCD grid in front of it. The result wouldn't be that great.
The second issue is that the resolution of most black and white tube TVs and Monitors (which would be ideal for a project like this) is pretty low. So in the end if you take the high pixel cound of an LCD into consideration you would end up with an LCD that just has very small dimming zones.
Lastly CRTs aren't that bright anyways so the image would be very dimm.

As @free_microwave has already pointed out that there are some JVC TVs which used an LCD layer (no pixels just one big LCD layer) in front of a B&W CRT which creates the illusion of color by swapping the RGB rapidly. The result however is still just a low resolution color image which now has a randbow effect like DLP projectors.

In conclusion I think its best to just get a nice CRT Monitor. They have great colors and contrast and some Models support high resolutions. Mine can do 2048x1536@80Hz 🙂 and has perfect dark levels

I'd be curious to see if Harcoreta's proposal would work. As other users have mentioned, the color filter array order would probably need to stay aligned between the two panels. So along with flipping one of the panels you would also need to rotate it 180 degrees (this is dependent on how the color filter array is laid out). But now the image from that panel is flipped both vertically and horizontally.

Assuming you have a GPU with multiple outputs, Windows 11 has an option to change the display orientation for each monitor/display. At least on my system setting it to "Landscape (flipped)" flips it both vertically and horizontally. Unfortunately, it doesn't seem to allow you to duplicate the displays while maintaining one flipped vertically and horizontally. If someone knows how to do this that'd be great!

In Matt's video he was using a splitter to feed the video to both displays. You could use a hardware based solution to the the vertical and horizontal flip. AJA makes the ROI-HDMI that can do this. However, it's pricey ($1245 USD) and you're limited to 1920x1080.

I've been researching pretty much the whole day, I have some thoughts.

Replacing the first layer of tracing paper with the diffusion layer that was in the monitor might help a tiny bit. As others pointed out, you can flip the second LCD to have an upside down image to hopefully get rid of the polarity problem (I'm honestly not sure) while still having the sub-pixels align (provided they're in a horizontal layout, which they often are). 

Using RGB LEDs or LCD-specialized ones rather than generic white ones (white ones should also be 6500k) would make sense since their output spectrum is much closer to what we want, which is just the three primary colors. You could also use Quantum Dots for the backlight (or printed onto an LCD). Either way having narrow emission bands for the primary colors would be quite helpful.

Having a wider color gamut also means it would be easier to go monochrome for the first panel, which is what other dual layer LCD displays do. I did find a few monochrome ones on eBay, but only in really awkward sizes.

For calibration and HDR I looked all over the place and found no existing solutions for driver boards or ways to easily create your own. I found a couple of monitors with no local dimming, no ABL and accurate PQ EOTF tracking that hard-clips, which sounds perfect for modding in a different backlight and boosting the brightness. Actually getting that to work with editing the static HDR metadata and dodging weird tonemapping seems pretty complicated though. 

I'm currently looking for ways to input an HDR signal into some separate hardware and then use a massive LUT to find the correct values for both LCD panels. 

Of course that's all assuming that you can overcome the engineering challenges to get enough brightness for decent HDR lol. Sony did it with a 31" screen that can do 1000 nits using 450 watts peak.

Great video!

I did wonder though, instead of adding tracing paper between the panels you could try removing the polarizing layers of the back panel. This should help with the brightness problem as the light will only be polarized once, but im not 100% sure it would work due to how they help form the image.

To keep the tradition of mixing water and electronics, you should consider adding a layer of water with sugar to rotate the polarization of the light 😆

Having the rear LCD black and white could allow for might light to pass through