Tinkerer builds VR headset with CRTs repurposed from Sony Watchman — makeshift VR rig isn't the sharpest, but eliminates screen-door effect
It's also limited to just black and white.
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
You are now subscribed
Your newsletter sign-up was successful
Virtual reality headsets come in all sorts of shapes and sizes, and the technology has matured enough to even become accessible for budget buyers. But what if you wanted something a little more unique that simply doesn't exist on the market? That's the predicament the aptly-named YouTuber 'dooglehead' found himself in when he wanted to enjoy the glory of cathode-ray tubes blasting directly into his eyes.
dooglehead is a fan of the nostalgic look that CRTs provide, and he wanted to chase that dreamy, characteristic glow and apply it to the modern era. He explains in the video that virtual reality headsets as a whole aren't new, and some even employed CRT screens back in the day. Granted, they were often too bulky and awkward to be practical, but even if you look past the heft, those CRT-based VR devices ran on proprietary hardware that isn't easy to source.
This is where the neurons start firing and dooglehead decides to build his own CRT VR headset. He starts with two Sony Watchman portable TVs from the '90s and extracts the 2.7-inch CRTs inside. These are unique in the way they operate because in the Watchman, you view them through a window from the side — the same side as the electron beam — instead of viewing from the opposite end of the electron gun.
Article continues below
This makes the assembly much more compact, which is great for a VR headset, but it's limited to black and white reproduction only due to shadow mask constraints. The Watchman displays also don't have any composite input, so dooglehead whips out an FPGA dev board and devises his own HDMI input. Through some programming, he makes the CRT appear as an external monitor connected inside Windows.
Now comes the fun part, figuring out the tracking for the headset. Our enthusiast follows one of many DIY VR headset projects hosted on GitHub and starts building out his own tracking system. He creates a custom PCB with an IMU for rotational tracking, the aforementioned HDMI port, and two USB-C ports — one for powering the CRTs and one for connecting an HTC Vive tracker.
To enable positional tracking, lighthouse base stations firing infrared laser beams were placed in the corners of the room and sweep it thousands of times per second to detect movement. If the laser hits the photodiodes in a VR headset, those SteamVR lighthouses can precisely triangulate where you are in the space.
Since dooglehead's makeshift contraption obviously didn't have the constellation of tracking points of a Vive or Index, he had to resort to using an external tracker. Everything was put together on a rather dangerous-looking Google Cardboard chassis, with the custom PCB, the FPGA, the CRT screens, and an HTC Vive tracker just strapped together with hopes and dreams. The entire thing weighed only 544 grams, which is lighter than some actual retail headsets.
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
The project itself doesn't go any further with a 3D-printed shell or other refinements because according to dooglehead, the resulting CRT VR headset was not the most impressive thing ever. The YouTuber includes a 3-minute gameplay segment going through various VR titles that include driving, shooting, role playing, and more, while showing the CRT screen in action. It looks exactly like you'd expect — a greyish luminescent image with some character.
Dooglehead's first comment was that he wished the screen was in color because in some games it made it hard to distinguish between certain elements. Secondly, the screens felt blurry: not due to the 640x480 (per eye) resolution, but because the electron beam wasn't perfectly in focus. The fact that it was then hitting the screen at a perpendicular angle just exacerbated this issue.
The biggest positive to using the CRTs was the lack of screen door effect (SDE) since they have a natural anti-aliased look that blends pixels to its advantage. VR screens are especially susceptible to SDE, since they're so close to our eyes.
The YouTuber closed out the video by saying he won't switch to this DIY CRT VR headset mainly because he prefers a color screen. He was glad to have experienced this mashup of retro and modern tech, but it's simply too outdated to use on a daily basis. CRT shaders can likely emulate a similar look, and manufacturers keep shipping insanely high-res displays on VR headsets these days that combat the SDE. But we have to commend dooglehead for seeing this crazy idea all the way through to the end.
Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.

Hassam Nasir is a die-hard hardware enthusiast with years of experience as a tech editor and writer, focusing on detailed CPU comparisons and general hardware news. When he’s not working, you’ll find him bending tubes for his ever-evolving custom water-loop gaming rig or benchmarking the latest CPUs and GPUs just for fun.
-
edzieba ReplyVR screens are especially susceptible to SDE, since they're so close to our eyes.
Utter nonsense. The reason SDE is common for VR displays is because they are spread over a massive field of view, which means pixel density (pixels per degree subtended angle) is extremely low. Since you do not view the display without the optical engine in front of it, the display density of the panel itself, and the physical distance of the panel itself from your eyes, are irrelevant. For example, with a 90° horizontal FoV, a 1920x1080 panel will only have 12 pixels per degree.
For example, the Quest 3 has the display panels less than half the distance from your eyes as the Rift DK1. But the SDE is dramatically reduced on the Quest 3 compared to the DK1.
What really affects SDE is the fill factor of a display: the ratio of area of emitting pixels - the bits that emit light (for OLEDs, or the non-aperture-masked areas of a CRT, or the translucent areas of an LCD - to the dark areas (the support electronics between pixels in an LCD or OLED, or the aperture mask in a CRT, or the gaps between scan lines in a CRT). The lower the fill factor, the worse the SDE. -
w_barath Hello, ionizing radiation is a real thing with CRTs at this range from your head.Reply
While the original product surely had strontium, barium, or lead formula glass, that is still engineered for a standard viewing distance of 30-100cm from the display. Not sub-2.5cm!
That's also assuming the device has proper shielding in the housing, which this clearly does not. -
Draaza Reply
This was literally my thought upon seeing the headline. I know it's just a silly throwaway project but please don't deliver a concentrated beam of highly energetic electrons (which produce x-rays when they hit the phosphor) directly to your retinaw_barath said:Hello, ionizing radiation is a real thing with CRTs at this range from your head.
While the original product surely had strontium, barium, or lead formula glass, that is still engineered for a standard viewing distance of 30-100cm from the display. Not sub-2.5cm!
That's also assuming the device has proper shielding in the housing, which this clearly does not. -
heffeque Reply
This.Draaza said:... but please don't deliver a concentrated beam of highly energetic electrons (which produce x-rays when they hit the phosphor) directly to your retina -
fireaza Reply
This is why the "OLED or no sale!" attitude isn't as cut and dry as many people think. Yes, OLED has superior contrast over LCD. However, there's very few OLED panels that use an RGB stripe sub-pixel arrangement, most use pentile. This means that even at the same resolution, an OLED panel can have noticeable SDE over an LCD panel, simply due to the fact that the sub-pixels are more tightly packed together on the LCD. Considering how much of an impact the SDE can have on a VR headset, a moderate reduction in contrast is a worthy sacrifice.edzieba said:What really affects SDE is the fill factor of a display: the ratio of area of emitting pixels - the bits that emit light (for OLEDs, or the non-aperture-masked areas of a CRT, or the translucent areas of an LCD - to the dark areas (the support electronics between pixels in an LCD or OLED, or the aperture mask in a CRT, or the gaps between scan lines in a CRT). The lower the fill factor, the worse the SDE. -
edzieba Reply
There's an additional gotcha with OLED panels: pixel on switching time. Whilst OLED grey-to-grey times are very good, the time for a pixel to go from completely off (true black) to any level of 'on' (but more pronounced on low levels) is much higher. This is why OLED have the 'black smearing;' phenomena. This can be avoided by never turning the pixels completely off, but this means OLEDs lose that 'true black' advantage.fireaza said:This is why the "OLED or no sale!" attitude isn't as cut and dry as many people think. Yes, OLED has superior contrast over LCD. However, there's very few OLED panels that use an RGB stripe sub-pixel arrangement, most use pentile. This means that even at the same resolution, an OLED panel can have noticeable SDE over an LCD panel, simply due to the fact that the sub-pixels are more tightly packed together on the LCD. Considering how much of an impact the SDE can have on a VR headset, a moderate reduction in contrast is a worthy sacrifice.
Secondly, OLEDs have to deal with 'mura': a difference in the light output response per-pixel for the same input voltage. This is fixed by a pixel-by-pixel LUT of actual brightness vs. commanded brightness, calibrated at the factory. The problem here is that either you cap the maximum brightness of every pixel on the panel the the least responsive pixel (to avoid dark pixels on a bright background) with all pixel responses being negative values and use nonlinear pixel response curves in the LUT, or you use both positive and negative pixel offsets in the LUT with simpler linear lookups but accept that some pixels will never enter the 'off' state in dark scenes (And some will go completely dark even in dim-but-not-black scenes). LCDs also have to deal with Mura, but because they are not self-emissive the undershoot penalty does not exist, and the overshoot penalty can be compensated by changes to the entirely independent backlight.
The brightness cap has an additional issue for VR specifically, where the move to pancake optics (hybrid polarisation catadioptrics) from basic doublet lenses or hybrid-fresnel lenses - a move that has allowed for smaller panels, more compact and lighter HMDs, larger eyeboxes, reduced chromatic aberration, and greatly reduced glare artefacts ('god rays') all at the same FoV - comes at the cost of optical efficiency. For the same apparent brightness to a viewer, a pancake optic HMD needs more than double the brightness from the panel compared to doublet or hybrid-fresnel optics. -
Chokkymalk Reply
Kedzieba said:Utter nonsense. The reason SDE is common for VR displays is because they are spread over a massive field of view, which means pixel density (pixels per degree subtended angle) is extremely low. Since you do not view the display without the optical engine in front of it, the display density of the panel itself, and the physical distance of the panel itself from your eyes, are irrelevant. For example, with a 90° horizontal FoV, a 1920x1080 panel will only have 12 pixels per degree.
For example, the Quest 3 has the display panels less than half the distance from your eyes as the Rift DK1. But the SDE is dramatically reduced on the Quest 3 compared to the DK1.
What really affects SDE is the fill factor of a display: the ratio of area of emitting pixels - the bits that emit light (for OLEDs, or the non-aperture-masked areas of a CRT, or the translucent areas of an LCD - to the dark areas (the support electronics between pixels in an LCD or OLED, or the aperture mask in a CRT, or the gaps between scan lines in a CRT). The lower the fill factor, the worse the SDE. -
roboj1m Reply
I mean, yes all of that is true, except don't forget that these are 2" tubes from a handheld television.w_barath said:Hello, ionizing radiation is a real thing with CRTs at this range from your head.
While the original product surely had strontium, barium, or lead formula glass, that is still engineered for a standard viewing distance of 30-100cm from the display. Not sub-2.5cm!
That's also assuming the device has proper shielding in the housing, which this clearly does not.
So designed to be ~50 cm from your face.
That's not to say it's safe and you're wrong, but probably not as bad.
They might have thought of that and designed these to be much safer because they're always going to be close to the viewers eyeballs.
Personally I'm more freaked out by the idea of strapping a circuit that's running at 25-ish kV to your face! 😂
And then having to take it off blindfolded!!😳
TVs can outright kill you if you poke the wrong bit. -
fireaza Reply
Indeed. And I believe Value said slow pixel response time and the loss of brightness was the reasons they went for LCD panels in the Steam Frame. But if you ask the internet, they'll say the reason is because they're greedy and didn't want to spring for a superior OLED panel.edzieba said:There's an additional gotcha with OLED panels: pixel on switching time. Whilst OLED grey-to-grey times are very good, the time for a pixel to go from completely off (true black) to any level of 'on' (but more pronounced on low levels) is much higher. This is why OLED have the 'black smearing;' phenomena. This can be avoided by never turning the pixels completely off, but this means OLEDs lose that 'true black' advantage.
Secondly, OLEDs have to deal with 'mura': a difference in the light output response per-pixel for the same input voltage. This is fixed by a pixel-by-pixel LUT of actual brightness vs. commanded brightness, calibrated at the factory. The problem here is that either you cap the maximum brightness of every pixel on the panel the the least responsive pixel (to avoid dark pixels on a bright background) with all pixel responses being negative values and use nonlinear pixel response curves in the LUT, or you use both positive and negative pixel offsets in the LUT with simpler linear lookups but accept that some pixels will never enter the 'off' state in dark scenes (And some will go completely dark even in dim-but-not-black scenes). LCDs also have to deal with Mura, but because they are not self-emissive the undershoot penalty does not exist, and the overshoot penalty can be compensated by changes to the entirely independent backlight.
The brightness cap has an additional issue for VR specifically, where the move to pancake optics (hybrid polarisation catadioptrics) from basic doublet lenses or hybrid-fresnel lenses - a move that has allowed for smaller panels, more compact and lighter HMDs, larger eyeboxes, reduced chromatic aberration, and greatly reduced glare artefacts ('god rays') all at the same FoV - comes at the cost of optical efficiency. For the same apparent brightness to a viewer, a pancake optic HMD needs more than double the brightness from the panel compared to doublet or hybrid-fresnel optics.