Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

That would be great.

My main issue with smart TVs isn't the network connectivity, its their horrible UI. The menus are universally confusing, slow and poorly thought-out. And because these TVs have to have some complicated OS to run all this stuff, they're slow to start up. My AppleTV goes from standby to on faster than my circa 2008[1] (only kinda smart) TV and receiver are ready to display an image. The AppleTV should not be the fastest device in the chain.

I assume your commercial displays have a time-to-first-image that more closely resembles a computer monitor. That's what I want in a TV. If Apple had a TV-version of CarPlay that TV manufacturers could license, I'd be pretty happy with that.

Given that embedded TvOS isn't coming any time soon, my ideal TV would have:

    * 70" - 80" screen
    * the smallest possible bezel
    * 4k HDR
    * capable of professional color calibration
    * one HDMI 2.1 input with ARC
    * TV tuner with auto-scan
    * any crappy speaker will do (I'm not going to use it)
    * IR/RS-232 control with distinct codes for on and off (i.e. not toggled)
    * excellent CEC support
    * time-to-first-image under 3 seconds, immediate audio
Nice to haves:

    * simulated snow for dead channels/no input (much nicer than blue IMO)
    * power, HDMI, serial connections should be down-facing
    * physical buttons on the back of the TV for power, channel, volume, menu
    * a rear or bottom-mounted red power LED that reflects off the wall/table
I would be willing to pay $3000 - $4000 for this TV, with the expectation that it would last at least 15 years.

[1] In December 2008, I bought a Samsung LN52A750 52" 1080p TV for $1937. That's roughly $2772 inflation-adjusted. It has a huge bezel; a modern TV with the same physical dimensions would be closer to 60" diagonal.



This is a very good and comprehensive list. Thank you! Our industrial displays already have most of your "must-haves," except for three:

1. Professional color calibration - this is doable, but it adds significant cost. When referring to color calibration, it is not the typical marketing gimmick used for consumer displays, but instead something equivalent to what you would get with Sony's BVM-HX310 reference monitor. So, my question would be what level of professional color calibration is acceptable?

2. HDMI 2.1 - HDMI-2.1a to be specific. Our controllers are all compliant up to HDMI-2.0 as of now. This is something in the works, but due to chip shortages, it has been challenging to come by industrial-grade chips/SoCs that handle HDMI 2.1a. However, this is on our radar.

3. TV Tuner - it is unclear if you mean an actual channel tuner or an auto-scanner for the available input. If referring to a channel tuner for COAX-based inputs, this is something that has been phased out, not by us but by our chip suppliers.

Regarding the price point, that sounds very doable. My personal expertise is in overall system integration and sheet metal design. The intent is to create even the consumer-grade display with a fabricated aluminum shell that feels rock-solid and heavy-duty aesthetically.

FYI, all the "nice-haves" are also already present, with the exception of down-facing connections. Can you clarify? Also, did you check out the pictures from my link above?

Thank you again for your valuable feedback!


Color calibration - I'm not entirely sure what would be involved here, TBH. "Professional" in this context would be what a pro home AV installer[1] would look for, not what a Hollywood color grader would need.

TV Tuner - Yes. An ATSC tuner with a COAX antenna input that can auto-scan for receivable channels. If this isn't feasible anymore, then I guess the HDMI input doesn't need ARC.

Connectors - looking at your photos, your input connections are on the side. If you put them on the bottom (down-facing), I think they'd be easier to reach if the TV is wall mounted (maybe not as easy if its on a table-top stand though). As long as they don't stick straight out the back, any orientation would be acceptable.

What's your experience with outdoor displays? 1000+ nits, IP-67 rated?

[1] In my area, that would be someone like https://www.gramophone.com


In that case, I would say that color calibration is very easy to accomplish and having the connectors facing down is something I was already planning to do. Our industrial displays are typically mounted 15-20ft in the air, so customers have an easier time accessing them from the side while on their lift.

As for outdoor TVs, that's a whole other Pandora's box, but yes, we also make those displays, mainly for digital billboards, advertising, and rental purposes. They typically range from 5000 nits to 12000 nits in brightness, and their pixel pitch ranges from 3mm to 9mm. They're all IP67 rated. Displays with a brightness lower than 5000 nits are typically indoor LED displays with much finer pitch, ranging from 0.9mm to 6mm. A 0.9mm pitch would be very similar to Samsung's QN90A series TVs with QLEDs (a marketing gimmick for an actual LED pixel-based display). These displays may cost a pretty penny, but we hope that once semiconductor plants are up and running in the US, costs will start to come down. Here's a project we did with a circular 3mm display back in 2017. This unit is 7ft tall with a 4.5ft diameter: http://bit.ly/43keDD5


Your wraparound display is very cool.

So my use case for outdoor displays is in the recreational marine market. Screen sizes are typically between 7" and 18" with some very high-end "glass bridge" displays coming in around 24". Retail for a 12" chart plotter is $4000. Typical display brightness is 1200 nits, which is OK, but still hard to see in direct sunlight with sunglasses. Conversely, they also don't get dim enough at night.

The glass bridge solutions all use a "black box" chart plotter (essentially a rugged PC) and dumb displays. But the smallest displays are 16" and way more expensive than the integrated solution.

I'd love a 12" to 15" HD display with capacitive touch, optically bonded LCD, wide viewing angle, viewable in direct sunlight (with sunglasses), and dimmable to nearly off (20 - 50 nits maybe?) so that its not blinding at night (not city night, but offshore night). It should have one cable: a USB connection for power, video, and HID output.

In the vein of your pole display, a 14-18" tall, 4" wide screen would be very cool for the sailing market. Most folks mounting mast displays are still stacking a bunch of individual 3" or 4" monochrome LCDs (at $1000/ea).


Thank you! I think I know what you're talking about. I live in a gated subdivision where the gate entrance control panel has a similar 15" screen but with terrible brightness and UI. Your display also sounds a lot like an HMI panel but for outdoors. Do you happen to have a product link for this display with an embedded PC? If the market is big enough, I'd be more than willing to explore some sort of joint venture with you.


What would be the best way to stay up to date with this? Could you add a little newsletter or maybe a social media account somewhere?


That circular display is amazing


Thank you! What you see in there is a WebGL creative put together using ThreeJS running at 60FPS on a RaspberryPi with a custom/frameless Chromium build. This was back in 2017, and I can only imagine what someone can accomplish nowadays!


Ever look into Touchdesigner or Resolume to drive it?


Yes, a few of our permanent install/broadcasting projects use TouchDesigner and Ventuz. For rental applications, Resolume, Modul8, and VDMX are preferred depending on the use case.


FYI, this is our current OSD structure: https://imgur.com/xNnsjwT. If you look at the "User Color", it lets you set independent gain/offset for each color channel. You can also select one of the predefined "Color Balance" options between Warm, Normal, Cold, or sRGB. The sRGB option arranges the panel color to the sRGB gamut. Additionally, you can set the Gamma curve between 1.8 all the way up to 2.2. I hope this information helps.


For the AppleTV crowd, a rf coax tuner isn't necessary, just want a big dumb panel


> So, my question would be what level of professional color calibration is acceptable?

Naively, I'd say… Delta-E <5. Asus ProArt monitors originally shipped with that, and could be calibrated by the user to Delta-E <2; afte a while, Asus simply started shipping them them with the latter.


Personally, I explicitly do not want a tuner in a TV. I would much rather feed that through my existing system and have the TV be as dumb as possible -- I literally just want a display.

Down-facing connections are a must.

I also don't want anything except a power button -- no channel, volume, menu, etc. I don't want crappy TV speakers and I'd really rather not have an OSD. I would love a display where I could fire up some software on my computer and have it issue commands over HDMI CEC to change settings. There are USB CEC adapters out there that would make this possible if computer HDMI ports won't emit CEC commands for whatever reason.

Ideally, strip the display down to the simplest possible thing and focus on the core features: big display, small bezel, high resolution, HDR, color accuracy (configurable through the aforementioned CEC mechanism). I don't even need multiple inputs: just give me one working HDMI input.


Eh, menu for basic colour calibration would be nice, doesn't need CEC for that though. I do agree with throwing out the speakers, many people have at least a sound bar and those that don't are unlikely to want this product (That market wants all in ones). I also agree with the tuner, there would be no DVR, signal filtering, etc on the TV but those may be desirable and would therefore require an external tuner anyways.


Interesting! Personally i don't want it if it doesn't have CEC. I want to be able to control it via something I'm SSH'd into.


Agree on exactly 1 HDMI input honestly. It's much easier to have the rest of my system mux HDMI (because you either have a receiver, or just a simple switcher, or exactly 1 device anyway).


I'm still looking for a monitor that can switch between inputs within 50ms. Should be possible, no?


A video mixer a-la Blackmagic ATEM Mini will do this. Maybe there are cheaper options if it's just being used for input switching.


That will introduce a lot of delay - they make synchronous cuts by frame buffering every input.


Then I'm afraid it's not possible. The reason the input switch takes so long is because of the HDID negotiation.

A video mixer acts as the source and sink for the output and the inputs respectively, where a HDMI switch will just physically disconnect and connect the output port to a different input port, meaning the HDID negotiation has to be re-done every time.


It still seems the right hardware, upstream or downstream depending on which direction you're viewing things from, could keep the connection alive with the HDID pre-negotiated. Like, assume the hardware doesn't change between switching. If you're the TV's firmware and have the CPU power to, why renegotiate when switching inputs. Solve for the default case and have it standing by and running hot. it's not like the panel is going to change in the interim.


FWIW, I don't care if the screen briefly flickers during switching.


You just need to strip the hdcp on the source.


> switch between inputs within 50ms

Why do you need this?

What kinds of inputs?

How many?

Are they all operating at the same resolution and frame rate?

What resolutions?

Do they require scaling? De-interlacing?

Are they synchronized? (genlock)


Why would you want to hang onto a TV for 15 years? Half of the tech you mention will be obsolete.


My TV is 12 years old, so I don't think it's unreasonable? Only down-side to the old TV is I have to strip HDCP to hook it up to my laptop through my receiver (a direct connection works because it negotiates the older version).




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: