Hacker News Reader

Comment by plorkyeran | Hacker News Reader

I would guess that up until now the biggest thing preventing anyone from investing in small-scale CRT manufacturing is the simple fact that there was an inventory glut that covered the world demand for CRTs for nearly a decade after the last one was made. Now that we're apparently getting close to the point where there's actually demand for new CRTs is when we can find out if it's something that you can build a business around.

View on HN

> there was an inventory glut that covered the world demand for CRTs for nearly a decade after the last one was made.

I collect and restore analogue oscilloscopes and even today you can still get original NOS scope tubes for many models that have ceased production 30, 40, 50, 60 years ago.

But is there really a small cohort of CRT nostalgics out there? I mean I get the records, the sound is unique and music is absolutely ubiquitous. Arcade gaming is a much smaller market and I find it hard to imagine that there are enough CRT purists out there to warrant a business.

Especially considering a nice new OLED could be fashioned to look like a CRT.

Ah, but it would be less "vibrant" and "nostalgic", according to the article.

vinyl : digital audio as crt : lcd

That's an interesting comparison, particularly now. Until the last few years CRTs certainly had better image quality than the alternatives.

And vinyl forces a better mastering where digital audio allows loud and compressed sound...

Vinyl, nokia 3310, vintage clothing, furniture. The whole Etsy phenomenon... Nostalgia and older stuff is a huge market. Just look at vintage gaming in general. You still cannot buy a NES classic edition for less than double retail recommended pricing. Also CRTs have 'something' that no panel will ever have. The curve, the thick glass edges, the static buildup, the scan error, the whole heavy tactility, the legendary white noise. And as gen X ages, they'll get nostalgic, and willing to pay. I'd invest.

I use CRT as my main screen. Some reasons for it:

1. I can choose any resolution without blurring... problem now worse with 4k screens (lots of angry posts in Stellaris forums due to UI that is hard to read in 4k and become blurry if you lower the resolution...)

2. Ridiculous response speed... playing Crypt of the Necrodancer or Super Hexagon is much, much more fun on CRT, all mistakes feel like it was purely lack of skill, instead of feeling something unfair happened. Ditto for fps... CS community in fact is hogging all CRT, the ones I knew for sale all ended in some CS player home willing to fudge settings to end with crazy refresh rate and response speed (there is even a video of a guy pulling 200+ fps by using interlaced mode!)

3. Better blacks and whites... in fact, this was the main reason, when CRTs died I also abandoned them, but all screens I had after had colour or contrast problems, I had to fiddle with settings constantly, one day my screen broke and a CRT was available nearby, when I plugged it in, I was blown away by the image quality that required only geometry settings, no endless messing with gamma, colour profiles and whatnot.

4. Not valid anymore, but until recently CRTs were much cheaper than flat panels of equivalent quality, at least on my country... but now with "new" stocks empty, used prices around rising, I even own a CRT tv that now on eBay costs more than the price it has when it was new... (this particular model is also the one used as reference for some SNES emulators CRT "filters")

You can get all sorts of Sony Trinitrons off of sites like Craigslist for free... just turn up and haul them away

And CRT are just a giant light bulb, bombarding you with not so good radiation. Add to this the fact that the corner are never going to be perfectly square. I get the attraction for some of the superior quality of the CRT, but once this technology has been overtaken by a superior one, there is little sense to look back.

> And CRT are just a giant light bulb, bombarding you with not so good radiation.

In other words, you do not know how a CRT works.

Educate yerself, it's free: https://en.wikipedia.org/wiki/Cathode_ray_tube

The article is a poor source for your argument, because it does at length argue about the health risks of X-Rays emitted by CRTs.

It's more helpful to put it into perspective with other radiation sources:

https://upload.wikimedia.org/wikipedia/commons/2/20/Radiatio...

IOW, radiation dosage of using a CRT for a year comparable to ten bananas.

I don't see the "does at length argue about the health risks of X-Rays emitted" in the article. It discusses them and concludes that they are widely considered safe.

> CRTs can emit a small amount of X-ray radiation as a result of the electron beam's bombardment of the shadow mask/aperture grille and phosphors. The amount of radiation escaping the front of the monitor is widely considered not to be harmful. The Food and Drug Administration regulations in 21 C.F.R. 1020.10 are used to strictly limit, for instance, television receivers to 0.5 milliroentgens per hour (mR/h) (0.13 µC/(kg·h) or 36 pA/kg) at a distance of 5 cm (2 in) from any external surface; since 2007, most CRTs have emissions that fall well below this limit.

Out of curiosity, color CRTs do have a kind of native resolution. It's the RGB pattern used to show colors.

See: https://www.reddit.com/r/askscience/comments/3a2tx6/do_old_c...

> Especially considering a nice new OLED could be fashioned to look like a CRT.

Except for all the old cabinets that just won't work with it I guess.

If you are seriously considering going into this business, consider also the possibility that somebody will figure out a way to make an LCD screen that is acceptable for this use case once the CRTs are no longer available. Light-guns are likely gone no matter what, or at least going to be inconsistent and hard to use, but an LCD screen explicitly designed to emulate a CRT may be able to fill in. LCD screens not explicitly designed for this use case are getting down to ~25ms latencies, which is getting to within spitting distance of one frame.

It seems to me there's no fundamental reason that an LCD can't be designed nowadays to paint the screen almost identically to the way a CRT did it, complete with an emulation of their "flaws" (with at least a line or two of buffering to pull that off). The fact that older LCDs use an internal framebuffer and are locked to a certain refresh rate seems like an artifact of the older technology rather than a fundamental limitation. Just as G-Sync and friends can start playing with the refresh rate it seems like nothing would prevent you from designing an LCD controller that doesn't wait for an entire frame but just starts painting the NTSC or similar signal.

If I'm wrong, I eagerly await informative corrections. But bear in mind I'm aware that I'm calling for fairly significantly different supporting electronics behind the panel, that work on a fundamentally different architecture than current LCD panels. I do realize this is a non-trivial modification.

You'd have to run numbers but it's possible this is more cost-effective than trying to source all the CRT parts and/or run your own manufacturing line; this approach would involve creating custom electronics (feasible even for a startup) and then sticking them on commodity screens, which is almost certainly much cheaper than trying to make CRTs.

Edit: Apparently I have gored some sort of sacred ox or something? I'm getting downmodded for this? Really? I'm talking about deeply redesigning how LCD screens work, so your current issues with LCD screens don't apply any longer. There are some things about how LCDs work now that are historical accident, not fundamental to the technology.

25ms is still several orders of magnitude too slow for a lightgun. It's important to remember that ~16ms is the approximate time for a CRT to paint an entire frame at 60hz. A lightgun is way, way faster than that. It needs to determine the x and y position of the user's aim. To accomplish this means being faster than a horizontal pulse (the signal that draws one line of pixels) which in the case of NTSC is 4.85µs. I highly doubt you can get an LCD to refresh that quickly since the process involves actual physical twisting of liquid crystals in response to a change in electric current.

This is assuming that no one will just make a new lightgun that works with LCDs.

The discussion is in the context of playing classic arcade games. A new lightgun would not work with these games.

"Light-guns are likely gone no matter what, or at least going to be inconsistent and hard to use,"

If I am correct about how the LCDs could be redesigned to work, I can imagine a world where you're paying about 1ms + the actual liquid crystal latency for the display which can apparently also be in low-single-digit milliseconds, which is probably good enough even for frame-perfect fighting games, but you're still going to have a screen that is fundamentally persistent rather than a scanning flash of light. (Possibly 3 or 4 ms if you're going to do some basic CRT emulation. Such emulation wouldn't be able to do full screen distortions induced by too much brightness, but even in the CRT era that was considered a failure, not a feature. It should be able to do slight persistence and local blurring.)

Modern 144Hz LCDs are getting close to sub 1ms latency and we have 200Hz LCDs coming out.

sub 1ms latency

Which is still several orders of magnitude slower what is needed to simulate a scanning electron beam in order to make lightguns work.

Couldn't light guns be emulated? Use something wii-mote-ish, and then when the trigger is pulled, wait the requisite time after the next frame to trigger.

Rather than actually fabricating new tubes, I think a smarter approach would be to make a convertible circuit with adjustments and connectors that could revive any/most CRT's. So you could find any non-punctured CRT, gut all the electronics, replace with the add-on board, and fire it up again. The new circuit would have HDMI or breakout boards for any interface you want, IR, Wifi, audio, etc.

Reply on news.ycombinator.com