Eviction of CRT being a net saver in input

Tech Talk
1

How many people would like to play Street Fighter 30th and say if only I had a real CRT TV or else I won’t have to lead my actions?

Well I decided to do something about it.

I broadcast on Twitch a test broadcast specifically for TV ping. We didn’t introduce the layer of the network ping yet.

I’ll describe the routes from the Xbox one console to the 3 screens respectively.

All 3 screens gobfrom consoles HDMI port to five-way hdmi switcher, then via HDMI cord to a sound extractor so that my computer and any headphones could be fed sound for headphones surround sound.

Then the video goes to a 2 way simultaneous powered splitter.

On path A it goes by HDMI to a device I bought on either Amazon or ebay which converts HDMI into VGA. Then it goes into a VGA input hub where a second future thing could be connected.

Then it outputs to a y splitter this one is unpowered.

Path a-a go straight to my CRT monitor.

Path a - b goes to a VGA to HDMI converter which can work either powered or not.

Then it goes into the computer input of an HDMI to USB 2.0 capture card. Where OBS displays a live preview on the upper left

Back on path a - A you see the TV screen which is captured by my high definition camera, and that gets piped straight into a USB port where it’s picture is shown on the upper right window.

Meanwhile back on path B it goes straight to the Sony PlayStation’s 3D TV HDMI port on input 2.

Then a second HD USB camera films the result of the PlayStation 3D TV when it’s switched to input t who.

Finally the output of the whole TV screen goes via thunderbolt 1 cable to my HDMI connector and then to an HDMI to USB connector and then to that to my Android phone, a Moto G7 where a trial version of steam Labs obs for Android broadcasts what’s captured on the screen if you notice at the end of the video I have to do some swiping to stop my Twitch.

Just being up front and honest about my ping time.

I assume the most direct path is past B if path AA saves over path B, then you have savings switching to a CRT TV. If not, then all this saving is futile.

I noticed two problems with something with this method.

One is anything that is either 480i and possibly 240p does not work right on the CRT TV. GameCube’s Legend of Zelda Wind Waker does work right in 480p modr, and a capture device does accurately capture the 480i menu and the 480p game but the monitor only works in 480p mode, but has a weird glitch. There is some sort of gobbledygook written in the black space of the picture on my PlayStation 3D monitor. It’s like my VGA to HDMI conversion device doesn’t know how to deal with something especially when converted from HDMI back to HDMI.

The second is input lag timed the delay to be 31 milliseconds in a PlayStation 3D TV. So if I save one frame, around 16 milliseconds assuming 60 frames per second, then the three devices in between take about 16 milliseconds of ping. (HDMI switching port audio extractor splitter and HDMI to VGA converter) if more the times it registers two frames then probably we have an efficient everything in between the HMDI monitor. Therefore everything except light Gun Games should be low ping enough to work as good as possible.

I understand light gun games won’t work for other reasons like different Hertz rates.

last thing I want to point out is I found a very weird error in the website speechchat.com. If someone types in the phrase “Am I John Cena?” which is a sentence I typed a couple streams ago unprompted because I was testing to see if they could hear me or see me through my waved hand, with a low hertz webcam, and I wasn’t sure if they could hear me so I had to type it out on speech chat. Then I heard my computer voice say: “Am one John Cena?”

why the heck would it read I as a word one unless it’s going into Ro7man numeral mode? I think I is more commonly used as a first person reference to oneself in the subject of the sentence than the Roman numeral for one. That is a bug that seemed to slip.

How in the heck did that happen? How is it not found that it thinks I is the Roman numeral one and not a first person reference to oneself in the subject of a sentence?

Here is the link of my twitch video on my Twitter site under the handle @tripletopper:

Is it opens on the app and not a link just look on twitter.com for the username @tripletopper

2

Different devices lag at different rates and different methods of measuring lag yield different results, so the only thing this proves, if I am not mistaken, is that the combination of converters you use to play games on a C.R.T. are faster than that L.C.D. sony made to facilitate features that require active shutter glasses for the Playstatiion 3. That is unsurprising given that in order to work, active shutter glasses require 120hz panels so that they can flicker to block visuals from being viewed by the wrong person or eye.

The Playstation television minimally needs to reprocess a 60 F.P.S. video into a 120 F.P.S. video to match the refresh rate of its screen, which by itself causes lag. It is a nice piece of tech to have for its intended purpose, but probably not the ideal display to be using in terms of lag, and hence not the ideal display to use for this test.

Also, I am not sure if I understand the post enitrely, but the mere fact that your C.R.T has a V.G.A. port very strongly implies that it is not a television. There are some very specific exceptions, but for the most-part V.G.A. inputs were only available on devices sold as monitors. The reason 480i would not work on a V.G.A. C.R.T. is because it is a 15khz signal below the display’s synchronization range. 15khz compatible R.G.B. compatible cathode ray tubes are a rarity in North America, which is the main reason why people make a such a fuss over professional video monitors, although it is worth noting that a y/pb/pr colorspace conversion is supposed to be equivalent in theory and 480i compatible C.R.T. televisions generally don’t get even half of the hype for some strange reason.

I am not really sure what the very, very best L.C.D. for this purpose would be, but the Playstation television is not it. Now I have considered using a high speed camera to see how fast the SCPH-131 screen for the Psone is compared to a composite or s-video C.R.T. before, just to see how much faster a C.R.T. is compared to an L.C.D. when we eliminate scaling as a factor altogether, (It’s a 240p L.C.D. made specifically for the console) but that is a very different sort of test and I lack a high speed camera anyway.

3

I was comparing it to my TV and what was considered a low lag TV from 2010 to about 2014.

Then the invention of the TN monitor came.

Inputlag.com rated the lag of the PS3DTV at 31 milliseconds or approximately two 60 Hertz frames versus reality.

If I caught up one frame then I caught up with lag of that TV’s lag, then I caught up 1 frame vs reality. I was hoping that it was exactly two frames behind then that would prove that the HDMI to VGA converters and back combined are less than one millisecond therefore over 95% of the delay is in the display not in the conversion.

And during this test, I had one direct path to the PS3DTV (lower right) one path straight to to CRT (upper right) and one “double converted” HDMI to VGA then. VGA to HDMI that was HDMI captured.

I’m not insisting the Twitch stream was zero ping relative to the game reality. If OBS or Twitch was late, we factored that out by making them the same broadcast.

If left and bottom right are equal, and one route was 0 converters, and the other route was 2 converters, and the left is internal input and the bottom right was internal input from camera, then that shows is that the display delay is more significant than the converter delay.

VHS and Betas, are considered zero ping, but actually throw off a light gun by a couple of frames. A DVD-R in preview mode still makes the light gun work, but is more wildly off. It doesn’t register in half the screen, and is thrown off vertically.

Retro Tink’s SCART to YCbCr converter only throws light guns off as much as a VCR does.

By the way, how does one “frame advance” and “frame retreat” in the Twitch playback on an Android app? I’d like to do an analysis on Twitch.

If I am right this is how a grocery list Style order of operations exists on a CRT TV.

Step one get sync signal to start at top of line.

Step 2 draw line from left to right as one continuous analogy colored gradually changing line.

Step 3 do a carriage return to start the next line.

Step 4 repeat two and three until end of frame.

Step 5. Either send new sync signal if 240p 60 or, on alternate passes, send new field signal if 480i 30 and new sync signal on alternating trips.

Each field trip takes 16.7 milliseconds assuming 60 Hz or i30.

The color divisions within a row are continuous yet the color divisions within a column are discrete.

I don’t know much about digital TV but I assume it gets the whole frame drawn out then puts it in a frame buffer then broadcasts it and most of the data it receives are Delta frames or difference frames between that frame and the last frame and some could be Delta pixels meaning pixels described in terms of the previous pixel not as an absolute analog value. That’s how you save bandwidth.

I know there is a video bandwidth limits to keep within live video. Because even with our pathetic 1.5 Meg in for our DSL we sometimes lead the movie and disconnect well after our internet has gone out because it has buffered lots of movie in between now and later. Of course we have to keep it at 480i in order to assure that.

A movie you could phone in ahead because it’s predetermined and loaded faster than it plays. if a video game would do that, you would question the meaning of life, or at least, the meaning of a game if everything is predetermined.

that’s the reason why there are alternate signals carried on the same digital carrier: because for TV to make sense, the video must either be live or else we get into really big tough existential questions about the true meaning of video games and displays and fate.

Arguably most TV is not used for live stuff. For video games to be of any worth, it must be live.

with smaller data footprints (assuming same time stamp) does that mean a digital signal could be transmitted stored, processesd and actually fully displayed before 16 milliseconds which is the natural, organic time it takes to display one field in a i30 or p60 game?

I am discussing this. I believe what I said is true. If a TN monitor shows a complete p60 picture in 8ms instant complete draw time, then maybe I have to think about this. Maybe data footprints savings that are greater account for faster complete frames vs live CRT draw time.

Does someone have a breakdown of 16 milliseconds of one frame how an analog TV versus a digital TV processes and displays the screen? If the lower data packet is faster than 16 milliseconds of a live spooling then it is possible to display the whole thing faster than 16 milliseconds, the natural length of an analog CRT frame.

4

The PlayStation 3D TV was known for terrible lag.
There are affordable displays without even 1/3rd of the delay.

No that wrong, the picture is still drawn line by line.
Its why Lag testers still test at three spots, at the top, middle and bottom of a screen image.

Also stop confusing Network lag with Video latency.

An NTSC Analog CRT screen image the refresh rate is 29.97 Hz per field or 59.94 Hz per second.
The Reason it’s 59.94 hz and not 60 hz is so that the signal has room for the color subcarrier signal
There an whole hour long documentary on how the math works out and exactly is the color sub carrier so I am not going to go over it in a forum post.

Digital Video refresh is variable, and can be anywhere from as low as 24hz to 288hz depending on the model of display and the video codec used.

5
  1. the PS3DTV was considered lag friendly, FOR A 3D MONITOR IN 2010-2014.

2 and analog VHF/ UHF signal, draws a signal line for line and is uncompressed. When you digitize a signal, you can do a 60 HZ signal in less time, using much less of the wave. If it was pre-written content it can go into the future and continue downloading. but if your broadcasting a live signal you have plenty of spare room so you have other secondary signals carried along the same wave. If you haven’t over the year antenna you can get the “point two” and “point three” channels of every broadcast station. Some frequencies have more channels, some have less, depending on how they use the bandwidth. My dad still uses 1.5 Meg DSL sets the download to 480p and let’s the Blu-ray player upscale and sometimes when the internet goes out we have some room to play with. so I know it is possible to load a movie and finish loading it before you finish actually playing it.

So you’re downloading a frame faster than a frame could be displayed on a mechanical CRT TV, then the question is how fast is the signal draw. Bad modern TVs will overprocess the signal which is good for watching movies but awful for playing games. Most TVs after a certain Year have a game mode which turn off the auto processing and deliver the raw signal.

I heard that TN, whatever that stands for, are the initials for the type of monitor that is considered the quickest on the draw. I don’t know whether a TN draws it dot by dot, or line by line, or frame by frame, but, in the end, the complete frame can in theory be drawn before a CRT.

I noticed on replays in wrestling that ever since the arenas use digital displays, on hyper slow motion the displays alternate between full display and totally black.

But I am right about analog TV and analog displays “shopping list of directions” that it draws it from left to right one line at a time with no distinction between pixels within a row. Then takes a slight break to set the second gun on the next row.

In theory if it could process a full frame in less than 16ms, if the screen pixels change combined with the full frame being loaded is can beat 16 milliseconds. the reason why you measure top left middle and bottom right on those testers is because do it gradually whereas modern technology doesn’t instantly in one step.

However that would be useful for Press Your Luck because the “always big bucks in a spin” square is in all three iterations of it (the 80s, the zeros, and the prime time 2019-20) is always on the top. if you’re a reflexes are quick enough and you have a CRT TV you could “Do the Larson” and get $100,000 Plus when the largest square is $5,000, and not have to memorize patterns. I heard Larson thought exactly like me look and press and he could just time it but then his VCR was always slowing to a pause which means he had to anticipate it where he learned the patterns.

Me and less than 100 other people have done this during the life of zero’s version of the only version of the game. (Gotten 100k+, but since it was a private “for fun” game, we got nothing but the satisfaction.) and they promised they randomized it to prevent another Michael Larson. And I only did it on my CRT iMac and I did it again when I realized my new iMac was a problem and went down a few systems and tried it again on the CRT iMac.

6

First off everything you posed is false, and I don’t know where you getting your bull shit from

No it was never considered Lag Friendly. Lag testing wasn’t really explored in 2010, and Display manufactures don’t lag test (or at least publish their results anywhere).

All TV Images Analog and Digital is drawn Line by Line
And this goes for Progressive and Interlace

STOP CONFUSING NETWORKING WITH VIDEO.

No such thing

The picture is always drawn at the Refresh rate. So Modern displays can draw the image faster.
This has nothing to do with Input latency for the video

Wrong OLED is technically Faster.

Sort of, this is the only thing that kinda right. Image is drawn top to bottom, left to right.
There no BREAK per say, part of the signal is off screen which is the overscan and the instructions to start a new line of video.

What are you even going on about here?

Irrelevant.

7
  1. Here’s a photo of exactly what I said. A PlayStation 3D Display testing at 31 ms.
    Screenshot_20201228-122339
    Screenshot_20201228-1223391080×2270 219 KB

    It’s from the site displaylag.com , on that same website it shows 2 3d monitors from that year that were faster. The quickest one being 11ms.

I’m questioning if CRT is always fastest. I understand that with digital signals, a frame’s picture data can be transmitted in faster than real time. If the show was pre-recorded and streamed on a streaming service then you could finish loading it well before you finish watching it. On VHF/UHF TV, the quickness of digital picture loading manifests itself in having multiple signals coming off the same VHF/UHF Channel. And I know analog CRTs draw the screen as it comes.
It’s like the difference between dumping a cassette and dumping a CD. Most cassettes (unless you have special equipment) can only be dumped in real time, but CDs can be dumped faster than real time.

Probably at a PC, that manifests itself, as either quick or frame rates or more points of articulation in an analog wave, meaning pixels of resolution in a digital sense.

If the HDMI to VGA converter comes to one millisecond of translations I’m or less, then the only way I’m going to beat a CRT monitor is if I get a monitor with faster than 17.7 ms.

Is it possible? If I’m describing things correctly then yes. Is it here? one of those 2011 3D TVs had a TV delay test of 11 ms, so yes.

Is 16.7 ms of analog VGA CRT + up to 1 ms for the HDMI to VGA converter good enough? I guess, I can make some of blocks better by instinct.

If a good 3D gaming monitor is withing the sub $200 range when new, that would be cool.

If there were such things as shutter 3D add on kits, like the SMS Sega Scope 3d in the 80s, meaning any patents would have run out, then that widens my options.

BTW currently I only need it for Macintosh and HD console gaming.