Update: I was actually able to dig up the service manual for this monitor and it is a fully digital design. The schematics does not mention what chip it is using but it says "NTSCPAL ASIC" and is connected to the a/d converters for the RGB input signal from the video decoder chip and two 512k x 16 bit DRAM memories. I bet my money on bucket brigade chips like the Panasonic nm3205 4096 stage analog delay line. Assuming a 640x480 resolution with a chroma resolution of 1:4 of the luminance it would only take about 20 of those chips to delay the chroma signal. The delay line for the luminance signal would have to be 2-4 larger depending on if the quarter rate signals from the chroma is used to recreate the luminance signal.
That's way too many bucket bregade chips! Why not commission a CCD camera sensor manufacturer to give you a semi-custom analogue brigade device? There are other problems. If memory serves the BBD chips had been discontinued during the early 90s, and only supported about 100KHz clock operation, while the pixel clock of this display needs to be on the order of 2-4MHz. Besides, that era already saw plenty of digital framebuffer devices, which are by all reason cheaper. Can i have your money now?
Pretty sure it was mentioned by others on previous vids that it uses a framebuffer. If you think about the timings alone, trying to store a whole frame (or field) would be extremely difficult in the analog domain. It has to store a whole field so it can repeat it for each of the colours. The Macrovision is probably messing with the sync during VBlank, and that can cause a slight "foldover" near the top of the screen. The monitor probably has a pretty good TBC (Time Base Corrector) as well, which is why it's able to keep sync lock. But Macrovision may still be upsetting the colour burst, and hence the weird phase shift in the Chroma during image sampling. (that's assuming it's being fed a Composite NTSC signal, as there's obviously no need for a colour burst with an RGB or Component input.) Also, sometimes to increase the effective sampling rate, they would use multiple ADCs in parallel, then have offset clocks for each, which could produce some strange effects if the video signal isn't quite right. Certainly an interesting display tech.
The NTSCPAL ASIC is a custom chip.. hins ASIC :) an ASIC is a purpous designed chip for a particular product. Genrilly anyway. as for it being digital... you sure the ASIC isnt an analog chip for the most part. seems more likely that the lcd is being timed and controlled digitally but all the video is being handled by analog. i'm betting that ASIC also has the jungle in it for osd and such. i wonder what kind of input lag this thing has.
Hi I know it’s over a year since you posted it but do you still have the manual? I have this JVC (from long before TechConnect did a video) and want to be future proofed if it cacks out on me.
That’s true, it was always called Technology Connextras. I miss our viewers who had to disappear. For some reason they insisted this channel wasn’t always called Technology Connextras. They are in a better place now, being re-educated.
The best part of Technology Connextras is the unexpected Disney geek deep dives, which has a history even longer and prouder than the name. Which is what it’s always been.
Epcot Resorts side note: this area is where my wife and I honeymooned in 2009. It was considerably more relaxed and quieter than other areas at WDW, and would be amazingly convenient for any of the culinary festivals at Epcot. I also would like to second Alec's recommendation of the "Keys to the Kingdom" tour, mentioned in the Utilidors video, and rec the "Backstage Magic" tour. No joke, getting to see the sheer scale of the laundry facilities was one of the highlights of our trip!
The "can this be used to colorize B&W videos" question is terrifying. I imagine those people are still puzzled when their old B&W films don't show in color on their modern LCDs.
They're probably the kind of people who go to photo processing places and ask if they can remove somebody from a photo so they can see what's behind them.
Some variants of macrovision introduce additional features such as wide luma pulses and additional hsync pulses. Somewhere in that monitor there is a PLL that is tripling the video clock, and since the basic unit of time you need to lock to in analog video processing is the hsync, it stands to reason that macrovision's techniques here are probably interfering with it until it gets a good lock. If the display ciruitry is indeed all analog, it's likely all a bunch of delay locked loops and RF mixers. Basically you triple the frequency, split it into three, mix each one with independently filtered chroma to derive RGB then send two colors down independent delay loops, then mix them all when they come back. That's the gist of it anyway. There are a lot of delay lines and PLLs in all analog TVs; pretty normal part of the technology.
As someone suggested I should take a closer look at the boards to see if there's a bunch of RAM chips on there. I've been hesitant to do that, though, as they're really crammed in there and I really don't want to break anything
Having designed my own CRT TV, I can speculate a bit further. Indeed, Macrovision messes with the horizontal PLL. If it was a digital framebuffer though, normally messing with the horizontal PLL wouldn't split colors - it would only shift the signal left and right (and maybe mess up the DC restoration causing the picture to fade in and out). But in this particular case, I bet the memory sampling clock, for both input and output (usually a multiple of 27MHz) is generated directly from the input horizontal PLL, and the output frequency is also directly derived from that. This allows a much cheaper and simpler single port RAM (reads and writes can alternate every clock). The NES and SNES also have famously wonky sync timing, I wonder if they would show similar artifacts. Also, nothing about this design precludes it from working with Macrovision, it's just that they didn't bother (it's a studio monitor after all). The artifacts just look a lot weirder.
No you can't just trippe frequencies and do some mixing to speed up the video signal. Signal processing doesn't work that way. (Will comment in more detail in another comment)
I love how the manual basically says: “We devised this complicated processing chain and it turns out it’s affected by MacroVision. Rather than source other processing ICs and do a bunch of redesign, we decided NOT to do that instead. Sorry.”
@@nickwallette6201 Well I cannot blame them for not giving enough effort. The market for color preview CRT monitors was rather small. Most people were contempt with monochrome ones. Also LCD monitors of that size were already available in decent quality and you can make those nearly arbitrarily bright. So this clearly was more of a prototype than a full thought out device.
To answer the "is it digital?" question I'd _usually_ suggest going all EEVBlog on it and taking it apart to see, but IIRC you prevously mentioned that these were pretty darned expensive; plus it looks like way too neat of a device to risk in any case. (I think it's extremely cool all around.) Thanks for the great follow-up.
Being a professional monitor, it is designed to show all faults in the original video, including copy protection artifacts and "flag waving", which is what we used to call the switching / sync issues at the top of the screen from playing video tape without a Time Base Corrector (like your home VCR). So, while your consumer TV's job is to hide all faults and make the picture look good despite issues with your VCR or TV signal, professional monitors want to show every blemish so engineers can take steps to minimize it. A lesson I learned a while back when trying to use a professional monitor to watch time shifted TV.
This was always the big thing for me during the HD-DVD wars. HD-DVD _supported_ encryption, whereas Blu-ray _requires_ it. After all, you don't want someone pirating your home video recordings!
LOL! Loved the Disney diversion right in the middle! Might've been a bit much of a leap to make that connection in a main channel video, but it seemed to fit here perfectly!
I swear to god your making me rethink my life watching your channel I almost though Connextras was a third channel, until you said this was the same channel as connections 2
Well any time shifting basically requires time discrete operation - either way the picture content is represented on a grid. It's either an analogue BBD chip (time-discrete, but value-continuous or analogue), or a digital buffer (time-discrete and bitwise value-discrete). The only way to manufacture a BBD array, especially of such size, in the late 90s would be to talk a CCD (camera sensor) manufacturer into that, perhaps they can repurpose one of their existing designs, maybe even without changes other than connecting an extra set of bond wires and blocking the light. The shifting of the signal in scanout direction is not at all a common or known CCD artefact. I suspect the framebuffer is purely digital though. However, the rest of the system, other than ADC/framebuffer/DAC, is likely a fully analogue design or a mixed design with severe analogue parasitic traits. Whether there is a grid of pixels? Well you need to catch up on Nyquist theorem - and no, if the filtering is perfect sinc or close enough, both ADC and DAC side, it just wouldn't be visible. Monitors today deliberately eschew this kind of filtering and emphasize the pixels to make things appear sharper, but since this JVC is dealing with analogue sources or DAC sources with unknown resolution, it would be important to them to eliminate sampling artefacts. You also won't easily find grid-like sampling artefacts on a late-90s/early-2000s SHARP SVHS VCR that uses digital video processing/reconstruction, or a same-era SHARP CRT TV that does digital framerate doubling. Furthermore the little grid-shaped artefacting that is present due to filter implementation quality will likely be easily hidden by CRT artefacts - slightly inconsistent line width, luminophore bloom, etc. I'm thinking the analogue frontend to the digitisation circuit or the readout circuit clock is somehow disturbed by the Macrovision signal, possibly in a very indirect manner. It was likely an engineering mishap that was orders of magnitude easier to document than fix, or where a fix would incur quality compromises for its primary use as a PVM.
@@CODMarioWarfare Would you mind completing your thought? What kind of a delay line exactly? And what would be the point? Sure if you had to use a BBD (rather than a fundamentally 2dimensional CCD), they could effectively be in separate banks per colour, and clock each bank in and out sequentially. But then you still have some issues, for one the max clock rate on the BBDs before they died out around early 90s was around 100KHz, but you have 3MHz-ish pixel rate on this CRT, and well the issue of them being discontinued. You could parallel a bunch of BBDs and run offset clocks on them and multiplexing, but then you'd have the issue that the BBD introduce a bit of smearing, the tiniest bit, but it would imprint as vertical lines if you tried to multiplex them like this. I lied a bit about time shifting being necessarily time discrete, you could build a resistor-capacitor chain, that's a true analogue delay line, they even used to make these long bois for PAL televisions back before BBD came around - while such a device is not time discrete on the face of it, the maximum frequency and and length of propagation is limited by the number of stages too; alternatively they experimented with transmitting the signal acoustically through some kind of crystal or solid material. Obviously all of this has been obsolete for a very long time. All of this is pretty pointless really. Game consoles shipped with significantly more RAM than needed for such a monitor's framebuffer in 1994, you could say at that point RAM cost is no longer an issue. And there are no massive inherently painful problems associated with a digital framebuffer. 512x240 framebuffer with 8-bit luma and 4-bit Pb and Pr is quarter of a megabyte. You need two such framebuffers or memory banks, one per field, at any given moment one of them is in write mode and one reading, and they alternate the tasks every field. Alternatively within the same amount of RAM you can use 8-bit Pb and Pr and just drop their horizontal resolution in half. But of course this being a 1999 product, significantly more RAM is found in here, and no quality compromises are necessary. Someone else down here found the service manual and from that is evident that each bank has 1MB capacity.
Instead of a digital frame buffer, it could be using an analog delay line. They're sometimes used in crt tvs, and can act as a frame buffer. Wikipedia has a pretty good article on it.
There was a tram from the EPCOT resorts to I Gateway, yes. I think it stopped running when more resorts were built around the lake and foot traffic got heavy.
I wonder what it would look like if you started one movie playing through the monitor, then detached the LCD and held it up against a different screen playing a B&W movie. Might be trippy.
I think you might get some colour artifacts around movement or other changes, as a result of the frame rate not being synchronized to the shutter speeds. Otherwise, it will look mostly the same (or a bit dimmer). Remember: the luminosities of the red, green, and blue values are produced by the CRT, then colourized by the LCD. You'll get 1/180 of a second with the CRT showing the red channel in white, which is then filtered by the LCD to be only red; 1/180 of a second with the CRT showing the green channel in white, which is then filtered by the LCD to be only green; and 1/180 of a second with the CRT showing the blue channel in white, which is then filtered to be only blue. An area of the screen is only blue if the CRT lights that area when the LCD is set to blue, and does not light it when the LCD is set to the other colours. If the CRT shows the same image during all three parts of the LCD colour cycle, as would be the case when you hold it in front of a different screen playing a black and white movie, it just shows the same black and white image if the frame rates would be in sync, or almost the same with some colours around moving objects when they aren't. I could imagine a bright object moving with a red shadow just in front of it and a blue shadow just behind it, for example.
Reason for the colour split is probably that the ADC is using the blanking interval as a way to reference black level and peak white level during the conversion, so that the input signal can be offset and gain trimmed to make full use of the limited number of bits of the ADC, probably only an 8 bit flash converter ( which still needs a resistor ladder and 256 comparators on the flash converter chip to drive the logic, that provides a binary value equal to the peak value of the ADC input) that is fed a sampled voltage of the black level as bottom reference, and then fed a buffered voltage equal to peak white for that frame. 8 bit flash as the design probably was late 1990's, and faster flash converters with higher resolutions were very expensive, but CMOS flash converters with 8 bits ( CA3308 for example) were almost same price wise, and could sample at 15MHz with 8 bit resolution, just needed a fair bit of work to make maximum use of the limited input voltage range. Likely 3 ADC's all fed after the colour matrixing decodes the signals into luminance and the 2 chroma difference signals, and then the digital values are used to generate the 3 bitmaps of the frame, for sequential read out and conversion back to analogue video for the CRT. If you select RGB input they use the 3 converters direct per colour, so no distortion, but analogue video would need the matrix to get the signals apart, and red is typically the one that suffers the most, as it is always the poor cousin to the dominant green signal. Having only 253 values available you wanted to use the range fully, so had to use the blanking interval sampling, as in a VCR, to set the gains, thus the Macrovision interference, and the same for head switching tearing, as that also corrupts the reference levels in the blanking interval. The red would probably be due to the maths required having underflows or overflows, and ignoring the carry or borrow in the output to the DAC driving the CRT after the frame buffer. As the data is only there for a frame no real worry, but a really interesting technique used there, which I think was pioneered by Tektronix in a series of digital oscilloscopes, that had a colour screen, using a very similar system.
I love the international gateway. I got to stay at beach club once and walking right into world showcase was amazing. I would go over there in the evening just to get some snacks from France 😄
Similarly, here on the other coast there used to be a tram from the Disneyland Hotel to the gate. It went away in the late 90s during the Resort construction. It's not a long walk either, but now it's through Downtown Disney's shops and restaurants instead of across a massive parking lot with nothing to spend money on.
International gateway used to feel like your own personal entrance to EPCOT. With the skyliner, it’s fairly busy. Still worth a jaunt to Ample Hills ice cream at the boardwalk.
Before 1990, a TV with a digital frame buffer would be ludicrous. After 1990, a TV without a digital frame buffer would be ludicrous. The sampling was probably being done at 4SC, so it is completely synchronous with the color. Demodulation then could be done entirely digitally. The frame buffer chips I saw in the late 80s operated at 4SC, 910 samples per horizontal line in NTSC. Fun fact, when I was designing and building my CBS field sequential TV replica, I purchased a Tektronix TDS744A oscilloscope that uses exactly the same color shutter technology with a BW CRT. I was stunned by the coincidence.
Yes, a professional monitor would never be expected to show copyrighted content. It probably has quite small margins of error, as errors is what you look for in a professional setting.
My Dad owned a television from Teac in 1996. I moved out of home in 1997. Anyway, if we played a video with Macrovision on it, the top quarter of the screen would be kinked by up to 45 degrees towards the right. Our VCR was an Akai but I tested it on another TV and it only happened when it was connected to that exact television. As a result, we never rented new release movies. It helped for us that we lived in Australia and Macrovision wasn't in really widespread use here until 1994. Yes, some older tapes had it, but they were in the minority - mainly Fox and Disney.
Such a monitor obviously includes a fairly complex interaction of many components. First of all it needs to create some sort of pixel clock from the incoming signal so that the AD converter can take an equal number of samples for each frame. For that you take what's known as a PLL. In a nutshell it allows you to one frequency to be an exact multiple (or fraction) of another signal by generating the signal you want, dividing the frequency of that to be that of your reference signal, and then comparing the phase of both signals to adjust your frequency generator. From that frequency you derive the clock for your AD converter and your DA converters which run at 3 times the frequency. For television that is fairly simple as you have a rock steady signal. For signals with VCRs that is somewhat more complex. First of all the signal has strong variations in time (jitter), secondly during head switching you can have a jump in time of up to half a line. Building a system that can reliably track a VCR yet be steady enough not to ignore noisy signals is hard, and that's probably where the problem comes from. Since the deflection circuits have similar problems it could very well be that the first scan (red) has some timing problems at the start. Converting a 525/60 signal into a 525/120 one requires some sort of frame store. There actually were analogue ways to do this... most involved a CRT and a TV camera running in sync. For 625/50 to 525/60 (and vice versa) conversion you can actually use switched acoustic delay lines as explained in the BBC research department report 1972/22 as it's enough to just shift the active parts of a line in time. However for this it wouldn't be possible as you'd have to continuously vary the "delay" over a huge span of time. (you can delay an analogue signal by a voltage determined delay via discrete delay lines with varicap diodes, it sounds inside) In any case a field full of delay lines would fill multiple racks. There is another way to store and delay analogue signals and that's Charge Coupled Devices. Those deposit charges on elements of a microchip and you can shift them around. It doesn't matter if you use them for analogue or digital signals in fact there are even special models to be used as image sensors. Some people have used those to convert 625/50 to the UK 405?/50 standard to continue to use their monochrome TV sets. www.nostalgiatech.co.uk/Vintagetech.htm (scroll down to the standard converter s European TV sets have been using those to do field doubling to eliminate flicker since the early 1990s (though they stored the data on CCD chips digitally). Given the time period the most likely explanation for this is that they used an analogue NTSC decoder, then a chain of 3 AD-converters with field stores to then convert them to analogue. BTW, no you would not see any kind of "pixel structure" just like you don't see staircases at the "analogue" output of a CD-Player. The so-called "reconstruction filter" creates a smooth waveform no matter what you input. You can try that with a CD-Player by playing a CD with a single sample width pulse. (1 sample with a high value, the rest zero) What you will get is some sort of hill as a low-pass filter will smooth out your waveform. With better CD-players this will more and more approximate a sin(x)/x function. (with f(0)=1) I know this sounds incredibly counter intuitive, but if your signal was properly band-limited when it entered sampling, the reconstruction filter will re-create it perfectly well.
Bring back Dreamfinder! Also, I remember seeing somewhere that at one time there was a tram that went around the Seven Seas Lagoon directly to the Magic Kingdom from parking/TTC, though it would mean missing a major Disney ride - the Monorail
I'm sure there was talking going on (and that it was likely interesting too), but I got distracted reminiscing about "A Day At Epcot Center" (aka "my childhood ") and promptly gazed longingly at Imagination, Horizons and well, poor clothing chooses. Ah, the early 90s!
"Quantization noise" You are right. We the engineers treat the limited bit depth as "noise." You can directly translate from the number of bits to an effective noise floor for a given signal. It isn't clear in that sentence it it refers to an internal framebuffer or the framebuffer of the source producing a black signal.
I would suspect that this is one complex circuit. The real interesting question is whether the image brightness is controlled more by the CRT or the LCD display, and whether the CRT portion is all analog directly off the NTSC or RGB inputs, or whether it is running off the digital signal that drives the LCD using A/D and D/A converters to maintain sync. I would suspect the LCD is likely 720x480 or perhaps 640x480, or maybe 800x600 based on the age of the unit. First, you have to convert the signal to the LCD using analog to digital conversion, and to field sequential or perhaps the memory buffer allows multiple colors displayed simultaneously and changes the values for one color at a time. This likely adds a delay to the video. Then you have to drive the monochrome CRT, either field sequentially three times faster or by displaying the entire image at once, based on the grid pixel pattern of the LCD panel at a higher resolution. Somehow this circuit must derive how much light information goes to the CRT versus using the LCD light to dark of each color pixel. The real question if the CRT light intensity is also controlled to allow for more contrast than just what the LCD panel can provide. My varying the CRT light intensity behind the pixels you can send varying levels of light to the LCD panel filters, adding a major increase in the number of colors that the LCD can display without expanding the color depth of the panel. The sound will also have to be delayed to be in sync with the delay related to the complex signal processing of that unit.
They were worried people were going to copy this tape...? Pro NTSC/PAL gear was often pretty unforgiving about signal levels and timing, so it's not surprising that a monitor like this would have trouble with Macrovision. Anyway, it was cool to see another video about it!
The signal is almost certainly being buffered and read out digitally, it's just that it isn't using any sort of visually "lossy" compression like JPEG or MPEG, but rather something uncompressed like bitmap . A "bitmap" image or whatever lossless format this thing is digitizing each frame (excuuuse me, field) of video into is giving an exact value to each red, green and blue pixel, therefore is unnafected by randomness of stuff like snow that would wreak havoc on lossy compression schemes like MPEG. We're talking about 1 field of standard definition here, so a couple megabytes of memory (if that) would be all that was needed, which would not at all have been a big deal in 2000.
There probly using an analog delay circuit. Lots of tv's use them to verieng degrees mid 90's on up to keep audio in sync or for other video trickery PIP freezframe excetra. i once had an analog tv or studio monitor *cant remember* that had a full 5 second delay from it's inputs. this thing didn't even have an OSD all knobs and chunky pop buttons. kind of a channel selector for which input you were using. Never could turn off the delay and i went through every single button knob and gubbens on the thing. Delay circuit was something to behold tho. took up a large portion of the internal space. huge jackpack on it tho. lots and lots of input. MMmmmm innnnpuuut.
Interestingly the BBC was able to recover the color from black & white footage that was originally broadcast in color due to the unique interference that was created from the color signal leaking into the luminance signal . Interesting stuff. There was another way used for old Dr. Who episodes where they had lower quality color copies from North America and took the chroma signal and essentially overlayed it over the surviving black and white broadcast copy and with a little work, had a color show.
I have one idea that might demonstrate whether the display is using a frame buffer or not. It's my understanding that a frame buffer will add some small amount of input lag to a video game. For a number of old game consoles, there's a piece of homebrew software called the 240p Test Suite that, among other things, contains a couple of input lag tests. Ideally, you would split the signal from the console to both this monitor and a CRT that you know is fully analog, then go into the Lag Test (not the Manual Lag Test) and film both screens simultaneously with a high frame-rate camera. The on-screen animation basically counts frames, so you can go through the footage and compare whether the two displays are on the same frame of the animation at any given moment. If they're out of sync, you can even tell by how much. That would probably take a decent amount of effort to set up, of course, but if you want to figure out whether this monitor does any digital processing of its own, that's one approach you could take. Edit: I perused the comments more and found one where someone mentioned they dug up the service manual and confirmed that the monitor is digital. So yeah, my suggestion probably isn't at all necessary.
EPCOT is wonderful and deserves more tangents The Carousel of Progress is my favorite thing ever, which for some reason my brain keeps trying to associate with EPCOT I haven't been to Disney in far too long
I think that the the digital frame buffer is so good that, quantization noise apart, the reconstructed signal is identical to the analog one (Nyquist).
The image might be digitised but the pixel clock may be derived from the original signal, and that process could be upset by the macrovision blocks. The wobble is probably it locking back on.
The CRT scans at 180 frames and if the lccs screen part was put in front of another b/w CRT TV and the video signals into both TVs are the same source would result be a 20 fps color image? I want to try that on a vintage projection tv from 1951 i have.
Its definitely digitizing the video. You can't see the horizontal pixels because the CRT hides the edges (it literally isn't sharp enough to show them). Being broadcast quality, it should be digitizing each line at D1 resolution (720 pixels across). What model JVC VCR is being used for playback? Some of the high end Digipure/TBC equipped SVHS models will show the curl at the top of the screen depending on the tape (no surprise its a Disney tape as they have been known to use very aggressive variants of Macrovision). The reality is an external full frame TBC can squash this problem and one can enjoy viewing all their VHS videos on this monitor.
1. it appears to me that they also didn't do a great job of NTSC color decoding, say using a process like the Faroudja decoder. There are color subcarrier artifacts visible. Does this monitor have a Y/C or component input? 2. did you WORK for DISNEY at one point in your life? 3. awesome video I wish I had one of these.
It could be using analogue delay lines, as used in PAL receivers, to store the lines as they come in and then play the line out three times with different colour decoder matrices each time. That would imply that the shutter would operate at 3 times the line frequency rather than 3 times the frame rate. You need to probe with your scope the interface to the LCD - if it’s switching at a multiple
Found this video that shows it does actually have a field store. th-cam.com/video/GTDpVZWQaFE/w-d-xo.html If you play the video at 0.25 playback speed you can clearly see that it is field sequential.
It probably is digitizing the video; but at such a high resolution that the small monitor cannot resolve the individual pixels together and smoothly blends them together.
I believe macro vision worked largely off of messing up the video timing and sync...a device called a time base corrector was used widely in tv stations back in the analog days...though digital magic did much of this in the digital world tv now works in largely. A TBC matched the input signal to a reference signal and made the timing synchronized...thereby stopping the macro vision. We used to copy VHS movies in master control for our own personal use and macro vision never presented an issue. You could also alter the brightness, contrast, black level, hue, saturation and other signal components with a TBC. I'd be curious if the back of this monitor has a reference signal input or how it would be affected with analog macro vision through a TBC.
I was going to say that maybe people with mobility problems would like the trams, but . . . then you're in Epcot at the world showcase, a MILE LONG LOOP, so I'm not sure. Might still be useful, though--people do ride the friendship boats often enough and they serve the same purpose. My primary "EPCOT has changed" complaint is that I will forever lament the loss of the beaver tails stand at Canada. Sweet cinnamony fried goodness. I also miss the old Motion but not as much as the beaver tails.
it could be that, rather than truely pixelified, the signals waveform is digitized, stored in the framebuffer, but then ultimately DAC'd back to a typical TV signal.
Kind of odd that I don’t remember that old video you were talking about. Because I do have perfect memory, like remembering this channels name Technology Connextras. Weird what you “forget” in just a couple months.
Presumably if it is an analogue device it would use signal delay lines to split the video into three and then delay the signal by 1 or two frames? What’s cool is this is super similar to DLP projection.
I laid in bed and thought about why Macrovision would interfere with the color and why just red and blue. The best idea I can come up with would be related to how the Chroma signal is sent. It is sent in an I/Q format, so the chroma signals dont represent what we'd call real color. It's sadly not as easy as YPbPr where one is difference blue and the other is difference red. But if my memory is correct the I channel carries the red to blue information and Q carries purple to green. These signals are quadrature mixed...plus more bandwidth is devoted to the Q channel. So this would explain why the red and blue gets wonky, apparently Macrovision is causing the thing to lose sync on the Q, probably I too but due to the insensitivity of human vision its not noticable. Why would the thing get screwed up...thats a really good question. It shouldn't be screwing with the colorburst, but its possible as I had a lot of later TVs with digital comb filters that would fail entirely with Macrovision. A lot of TVs didnt process true IQ either, they found cheaper ways of doing it. All my TVs that had true IQ processing always had dark splotches with Macrovision. If I can find a service manual and reverse engineer the decoding circuitry I might be able to get to the bottom of it.
If Technology Connections is so great, how come there's never been a Technology Connections 2?
@Idiot Online Wondering Aloud YES!
I like you :-D
Technology Connections 2 sounds like something I would subscribe to.
@@MrBblhed Or a channel called Alec Connectify.
@Idiot Online Wondering Aloud yes, we already have LGR foods, now we need Technology Foodnections.
"The channel that's always been called 'Technology ConnExtras.'"
"Also, we have always been at war with Eastasia and allied with Eurasia." ;-)
@scaper8 Comment double-plus good!
What a doubleplus good member of MiniTruth! Big Brother is pleased.
I love that reference
“Always been called Technology Connextras.”
*Mandela Effect intensifies*
"Forked with the vertical blanking interval."
I see someone's been watching The Good Place.
it is such a good show
@@TechnologyConnextras Definitely a great one, that's for sure! Also, what happened to "fudge"? I guess The Good Place now beats A Christmas Story!
Why can’t I say ‘fork’? Haha - what a creative device to get around broadcast censorship. :-)
Some people say this channel is forking shirt. I disagree.... 😁
I'm sad that it is going away soon.
Update: I was actually able to dig up the service manual for this monitor and it is a fully digital design. The schematics does not mention what chip it is using but it says "NTSCPAL ASIC" and is connected to the a/d converters for the RGB input signal from the video decoder chip and two 512k x 16 bit DRAM memories.
I bet my money on bucket brigade chips like the Panasonic nm3205 4096 stage analog delay line. Assuming a 640x480 resolution with a chroma resolution of 1:4 of the luminance it would only take about 20 of those chips to delay the chroma signal. The delay line for the luminance signal would have to be 2-4 larger depending on if the quarter rate signals from the chroma is used to recreate the luminance signal.
That's way too many bucket bregade chips! Why not commission a CCD camera sensor manufacturer to give you a semi-custom analogue brigade device?
There are other problems. If memory serves the BBD chips had been discontinued during the early 90s, and only supported about 100KHz clock operation, while the pixel clock of this display needs to be on the order of 2-4MHz.
Besides, that era already saw plenty of digital framebuffer devices, which are by all reason cheaper.
Can i have your money now?
Yea I posted that on Twitter as well. 6-bit AD converters too at that!
Pretty sure it was mentioned by others on previous vids that it uses a framebuffer.
If you think about the timings alone, trying to store a whole frame (or field) would be extremely difficult in the analog domain. It has to store a whole field so it can repeat it for each of the colours.
The Macrovision is probably messing with the sync during VBlank, and that can cause a slight "foldover" near the top of the screen.
The monitor probably has a pretty good TBC (Time Base Corrector) as well, which is why it's able to keep sync lock.
But Macrovision may still be upsetting the colour burst, and hence the weird phase shift in the Chroma during image sampling.
(that's assuming it's being fed a Composite NTSC signal, as there's obviously no need for a colour burst with an RGB or Component input.)
Also, sometimes to increase the effective sampling rate, they would use multiple ADCs in parallel, then have offset clocks for each, which could produce some strange effects if the video signal isn't quite right.
Certainly an interesting display tech.
The NTSCPAL ASIC is a custom chip.. hins ASIC :) an ASIC is a purpous designed chip for a particular product. Genrilly anyway.
as for it being digital... you sure the ASIC isnt an analog chip for the most part. seems more likely that the lcd is being timed and controlled digitally but all the video is being handled by analog.
i'm betting that ASIC also has the jungle in it for osd and such. i wonder what kind of input lag this thing has.
Hi I know it’s over a year since you posted it but do you still have the manual? I have this JVC (from long before TechConnect did a video) and want to be future proofed if it cacks out on me.
That’s true, it was always called Technology Connextras. I miss our viewers who had to disappear. For some reason they insisted this channel wasn’t always called Technology Connextras. They are in a better place now, being re-educated.
I'm glad to hear that we take such good care of our dissidents ;)
We have no dissidents. We love everyone. We have those who love us now, and those who soon will...
"They are in a better place now, being re-educated."
Ah, yes. There is no war in Ba Sing Se.
The best part of Technology Connextras is the unexpected Disney geek deep dives, which has a history even longer and prouder than the name. Which is what it’s always been.
Offhand Disneyland collab when 😂
Ah yes, im glad i subscribed to "Technology Connextras" (as it was always called) some time ago in order to not miss new videos like this one.
Epcot Resorts side note: this area is where my wife and I honeymooned in 2009. It was considerably more relaxed and quieter than other areas at WDW, and would be amazingly convenient for any of the culinary festivals at Epcot. I also would like to second Alec's recommendation of the "Keys to the Kingdom" tour, mentioned in the Utilidors video, and rec the "Backstage Magic" tour. No joke, getting to see the sheer scale of the laundry facilities was one of the highlights of our trip!
The "can this be used to colorize B&W videos" question is terrifying. I imagine those people are still puzzled when their old B&W films don't show in color on their modern LCDs.
Don't be silly, these people don't have LCDs, they've got 8mm projectors.
They're probably the kind of people who go to photo processing places and ask if they can remove somebody from a photo so they can see what's behind them.
Well, it can make old square movies into widescreen ones. Why CAN'T it make b&w films into color films?
I hope at least some of them were joking.
Imagine if those people watched silent films! Yeesh.
Technology Connextras, where the "extras" part is better than many channels primary content...
Some variants of macrovision introduce additional features such as wide luma pulses and additional hsync pulses. Somewhere in that monitor there is a PLL that is tripling the video clock, and since the basic unit of time you need to lock to in analog video processing is the hsync, it stands to reason that macrovision's techniques here are probably interfering with it until it gets a good lock. If the display ciruitry is indeed all analog, it's likely all a bunch of delay locked loops and RF mixers. Basically you triple the frequency, split it into three, mix each one with independently filtered chroma to derive RGB then send two colors down independent delay loops, then mix them all when they come back. That's the gist of it anyway. There are a lot of delay lines and PLLs in all analog TVs; pretty normal part of the technology.
As someone suggested I should take a closer look at the boards to see if there's a bunch of RAM chips on there. I've been hesitant to do that, though, as they're really crammed in there and I really don't want to break anything
Having designed my own CRT TV, I can speculate a bit further. Indeed, Macrovision messes with the horizontal PLL. If it was a digital framebuffer though, normally messing with the horizontal PLL wouldn't split colors - it would only shift the signal left and right (and maybe mess up the DC restoration causing the picture to fade in and out). But in this particular case, I bet the memory sampling clock, for both input and output (usually a multiple of 27MHz) is generated directly from the input horizontal PLL, and the output frequency is also directly derived from that. This allows a much cheaper and simpler single port RAM (reads and writes can alternate every clock). The NES and SNES also have famously wonky sync timing, I wonder if they would show similar artifacts.
Also, nothing about this design precludes it from working with Macrovision, it's just that they didn't bother (it's a studio monitor after all). The artifacts just look a lot weirder.
No you can't just trippe frequencies and do some mixing to speed up the video signal. Signal processing doesn't work that way. (Will comment in more detail in another comment)
I love how the manual basically says: “We devised this complicated processing chain and it turns out it’s affected by MacroVision. Rather than source other processing ICs and do a bunch of redesign, we decided NOT to do that instead. Sorry.”
@@nickwallette6201 Well I cannot blame them for not giving enough effort. The market for color preview CRT monitors was rather small. Most people were contempt with monochrome ones. Also LCD monitors of that size were already available in decent quality and you can make those nearly arbitrarily bright. So this clearly was more of a prototype than a full thought out device.
Engagement
thank you for your cooperation
You always have the strangest stock footage. Love the video bro.
"I didn't record trivially-recorded snow, BUT OH HEY THERE IS DISNEY! HERE IS ALL THE VIDEO YOU NEED!" -- love you dude!
I wonder if this technology could be used to blackandwhitize color films.
Yes. It's called a monochrome crt
😆👍
Monochrominize?
Maybe those fancy 3-strip technicolor films
@Idiot Online Wondering Aloud oh of course. Silly me
To answer the "is it digital?" question I'd _usually_ suggest going all EEVBlog on it and taking it apart to see, but IIRC you prevously mentioned that these were pretty darned expensive; plus it looks like way too neat of a device to risk in any case. (I think it's extremely cool all around.) Thanks for the great follow-up.
Being a professional monitor, it is designed to show all faults in the original video, including copy protection artifacts and "flag waving", which is what we used to call the switching / sync issues at the top of the screen from playing video tape without a Time Base Corrector (like your home VCR). So, while your consumer TV's job is to hide all faults and make the picture look good despite issues with your VCR or TV signal, professional monitors want to show every blemish so engineers can take steps to minimize it. A lesson I learned a while back when trying to use a professional monitor to watch time shifted TV.
I think the biggest issue here is that Disney actually put copy-protection on a freaking ad. ¬_¬
I was going to say. Then again, maybe it was just a built-in feature of the tape manufacturing plant they used.
This was always the big thing for me during the HD-DVD wars. HD-DVD _supported_ encryption, whereas Blu-ray _requires_ it. After all, you don't want someone pirating your home video recordings!
LOL! Loved the Disney diversion right in the middle! Might've been a bit much of a leap to make that connection in a main channel video, but it seemed to fit here perfectly!
I swear to god your making me rethink my life watching your channel
I almost though Connextras was a third channel, until you said this was the same channel as connections 2
"That's not how this works, that's not how any of this works"
a quality meme sir. it checks out.
Alright you "conned" me into watching this "extra" video you uploaded.
I see you know the ConnTexxt!
Well any time shifting basically requires time discrete operation - either way the picture content is represented on a grid. It's either an analogue BBD chip (time-discrete, but value-continuous or analogue), or a digital buffer (time-discrete and bitwise value-discrete). The only way to manufacture a BBD array, especially of such size, in the late 90s would be to talk a CCD (camera sensor) manufacturer into that, perhaps they can repurpose one of their existing designs, maybe even without changes other than connecting an extra set of bond wires and blocking the light. The shifting of the signal in scanout direction is not at all a common or known CCD artefact. I suspect the framebuffer is purely digital though. However, the rest of the system, other than ADC/framebuffer/DAC, is likely a fully analogue design or a mixed design with severe analogue parasitic traits.
Whether there is a grid of pixels? Well you need to catch up on Nyquist theorem - and no, if the filtering is perfect sinc or close enough, both ADC and DAC side, it just wouldn't be visible. Monitors today deliberately eschew this kind of filtering and emphasize the pixels to make things appear sharper, but since this JVC is dealing with analogue sources or DAC sources with unknown resolution, it would be important to them to eliminate sampling artefacts. You also won't easily find grid-like sampling artefacts on a late-90s/early-2000s SHARP SVHS VCR that uses digital video processing/reconstruction, or a same-era SHARP CRT TV that does digital framerate doubling. Furthermore the little grid-shaped artefacting that is present due to filter implementation quality will likely be easily hidden by CRT artefacts - slightly inconsistent line width, luminophore bloom, etc.
I'm thinking the analogue frontend to the digitisation circuit or the readout circuit clock is somehow disturbed by the Macrovision signal, possibly in a very indirect manner. It was likely an engineering mishap that was orders of magnitude easier to document than fix, or where a fix would incur quality compromises for its primary use as a PVM.
What if you amplified the signal and sent it down three delay lines, multiplexed to the CRT?
@@CODMarioWarfare Would you mind completing your thought? What kind of a delay line exactly? And what would be the point?
Sure if you had to use a BBD (rather than a fundamentally 2dimensional CCD), they could effectively be in separate banks per colour, and clock each bank in and out sequentially. But then you still have some issues, for one the max clock rate on the BBDs before they died out around early 90s was around 100KHz, but you have 3MHz-ish pixel rate on this CRT, and well the issue of them being discontinued. You could parallel a bunch of BBDs and run offset clocks on them and multiplexing, but then you'd have the issue that the BBD introduce a bit of smearing, the tiniest bit, but it would imprint as vertical lines if you tried to multiplex them like this.
I lied a bit about time shifting being necessarily time discrete, you could build a resistor-capacitor chain, that's a true analogue delay line, they even used to make these long bois for PAL televisions back before BBD came around - while such a device is not time discrete on the face of it, the maximum frequency and and length of propagation is limited by the number of stages too; alternatively they experimented with transmitting the signal acoustically through some kind of crystal or solid material. Obviously all of this has been obsolete for a very long time.
All of this is pretty pointless really. Game consoles shipped with significantly more RAM than needed for such a monitor's framebuffer in 1994, you could say at that point RAM cost is no longer an issue. And there are no massive inherently painful problems associated with a digital framebuffer.
512x240 framebuffer with 8-bit luma and 4-bit Pb and Pr is quarter of a megabyte. You need two such framebuffers or memory banks, one per field, at any given moment one of them is in write mode and one reading, and they alternate the tasks every field. Alternatively within the same amount of RAM you can use 8-bit Pb and Pr and just drop their horizontal resolution in half. But of course this being a 1999 product, significantly more RAM is found in here, and no quality compromises are necessary. Someone else down here found the service manual and from that is evident that each bank has 1MB capacity.
Instead of a digital frame buffer, it could be using an analog delay line. They're sometimes used in crt tvs, and can act as a frame buffer. Wikipedia has a pretty good article on it.
The trams went between the Yacht and Beach Club resorts and the International Gateway for a brief time in the early 90's. I rode one.
There was a tram from the EPCOT resorts to I Gateway, yes. I think it stopped running when more resorts were built around the lake and foot traffic got heavy.
I wonder what it would look like if you started one movie playing through the monitor, then detached the LCD and held it up against a different screen playing a B&W movie. Might be trippy.
I think you might get some colour artifacts around movement or other changes, as a result of the frame rate not being synchronized to the shutter speeds. Otherwise, it will look mostly the same (or a bit dimmer). Remember: the luminosities of the red, green, and blue values are produced by the CRT, then colourized by the LCD. You'll get 1/180 of a second with the CRT showing the red channel in white, which is then filtered by the LCD to be only red; 1/180 of a second with the CRT showing the green channel in white, which is then filtered by the LCD to be only green; and 1/180 of a second with the CRT showing the blue channel in white, which is then filtered to be only blue. An area of the screen is only blue if the CRT lights that area when the LCD is set to blue, and does not light it when the LCD is set to the other colours. If the CRT shows the same image during all three parts of the LCD colour cycle, as would be the case when you hold it in front of a different screen playing a black and white movie, it just shows the same black and white image if the frame rates would be in sync, or almost the same with some colours around moving objects when they aren't. I could imagine a bright object moving with a red shadow just in front of it and a blue shadow just behind it, for example.
I'm always in the mood for your Disney world trivia
Reason for the colour split is probably that the ADC is using the blanking interval as a way to reference black level and peak white level during the conversion, so that the input signal can be offset and gain trimmed to make full use of the limited number of bits of the ADC, probably only an 8 bit flash converter ( which still needs a resistor ladder and 256 comparators on the flash converter chip to drive the logic, that provides a binary value equal to the peak value of the ADC input) that is fed a sampled voltage of the black level as bottom reference, and then fed a buffered voltage equal to peak white for that frame. 8 bit flash as the design probably was late 1990's, and faster flash converters with higher resolutions were very expensive, but CMOS flash converters with 8 bits ( CA3308 for example) were almost same price wise, and could sample at 15MHz with 8 bit resolution, just needed a fair bit of work to make maximum use of the limited input voltage range.
Likely 3 ADC's all fed after the colour matrixing decodes the signals into luminance and the 2 chroma difference signals, and then the digital values are used to generate the 3 bitmaps of the frame, for sequential read out and conversion back to analogue video for the CRT. If you select RGB input they use the 3 converters direct per colour, so no distortion, but analogue video would need the matrix to get the signals apart, and red is typically the one that suffers the most, as it is always the poor cousin to the dominant green signal. Having only 253 values available you wanted to use the range fully, so had to use the blanking interval sampling, as in a VCR, to set the gains, thus the Macrovision interference, and the same for head switching tearing, as that also corrupts the reference levels in the blanking interval.
The red would probably be due to the maths required having underflows or overflows, and ignoring the carry or borrow in the output to the DAC driving the CRT after the frame buffer. As the data is only there for a frame no real worry, but a really interesting technique used there, which I think was pioneered by Tektronix in a series of digital oscilloscopes, that had a colour screen, using a very similar system.
I love the international gateway. I got to stay at beach club once and walking right into world showcase was amazing. I would go over there in the evening just to get some snacks from France 😄
CRT, shutter, syncing, cool! Intermezzo: Epcot, Disney, resort, tram, gateway, Hollywood, uhm.... I *think* you're rambling about a theme park, not really sure though. CRT, shutter, LCD, yay!
This is my favorite spinoff channel that has always been named what it is currently named, along with D!NG.
Similarly, here on the other coast there used to be a tram from the Disneyland Hotel to the gate. It went away in the late 90s during the Resort construction. It's not a long walk either, but now it's through Downtown Disney's shops and restaurants instead of across a massive parking lot with nothing to spend money on.
I really enjoyed the extra bit of WDW content there!
Can it decode the secret message from the side of my 70's cereal box?!
Or my 80s Transformers tech specs card?
Maybe it will work with red/blue 3D glasses?
International gateway used to feel like your own personal entrance to EPCOT. With the skyliner, it’s fairly busy. Still worth a jaunt to Ample Hills ice cream at the boardwalk.
Connextras?! This is main channel material
This channels name may appear different at some times. This is due to an indecisive host and is NOT a malfunction.
Before 1990, a TV with a digital frame buffer would be ludicrous. After 1990, a TV without a digital frame buffer would be ludicrous. The sampling was probably being done at 4SC, so it is completely synchronous with the color. Demodulation then could be done entirely digitally. The frame buffer chips I saw in the late 80s operated at 4SC, 910 samples per horizontal line in NTSC.
Fun fact, when I was designing and building my CBS field sequential TV replica, I purchased a Tektronix TDS744A oscilloscope that uses exactly the same color shutter technology with a BW CRT. I was stunned by the coincidence.
The Slow mo guys have to shoot 10:24 at 1000fps or more, I think it would look very interesting to see those bubbles affected by a shutter
I love the tangent
That’s not a bug in the display. That’s Macrovision deliberately outputting a non-NTSC signal.
Yes, a professional monitor would never be expected to show copyrighted content. It probably has quite small margins of error, as errors is what you look for in a professional setting.
My Dad owned a television from Teac in 1996. I moved out of home in 1997. Anyway, if we played a video with Macrovision on it, the top quarter of the screen would be kinked by up to 45 degrees towards the right. Our VCR was an Akai but I tested it on another TV and it only happened when it was connected to that exact television. As a result, we never rented new release movies. It helped for us that we lived in Australia and Macrovision wasn't in really widespread use here until 1994. Yes, some older tapes had it, but they were in the minority - mainly Fox and Disney.
0:51 It's Buzzy! It's our boy! Someone get Jenny!
Defunctland is attacking Alec's second channel! Quick! Send help!
In all seriousness, Kevin and Alec doing a feature *together* would probably implode TH-cam. And I'd pay good money to watch that happen.
Such a monitor obviously includes a fairly complex interaction of many components. First of all it needs to create some sort of pixel clock from the incoming signal so that the AD converter can take an equal number of samples for each frame. For that you take what's known as a PLL. In a nutshell it allows you to one frequency to be an exact multiple (or fraction) of another signal by generating the signal you want, dividing the frequency of that to be that of your reference signal, and then comparing the phase of both signals to adjust your frequency generator. From that frequency you derive the clock for your AD converter and your DA converters which run at 3 times the frequency.
For television that is fairly simple as you have a rock steady signal. For signals with VCRs that is somewhat more complex. First of all the signal has strong variations in time (jitter), secondly during head switching you can have a jump in time of up to half a line. Building a system that can reliably track a VCR yet be steady enough not to ignore noisy signals is hard, and that's probably where the problem comes from. Since the deflection circuits have similar problems it could very well be that the first scan (red) has some timing problems at the start.
Converting a 525/60 signal into a 525/120 one requires some sort of frame store. There actually were analogue ways to do this... most involved a CRT and a TV camera running in sync. For 625/50 to 525/60 (and vice versa) conversion you can actually use switched acoustic delay lines as explained in the BBC research department report 1972/22 as it's enough to just shift the active parts of a line in time. However for this it wouldn't be possible as you'd have to continuously vary the "delay" over a huge span of time. (you can delay an analogue signal by a voltage determined delay via discrete delay lines with varicap diodes, it sounds inside) In any case a field full of delay lines would fill multiple racks.
There is another way to store and delay analogue signals and that's Charge Coupled Devices. Those deposit charges on elements of a microchip and you can shift them around. It doesn't matter if you use them for analogue or digital signals in fact there are even special models to be used as image sensors. Some people have used those to convert 625/50 to the UK 405?/50 standard to continue to use their monochrome TV sets. www.nostalgiatech.co.uk/Vintagetech.htm (scroll down to the standard converter s
European TV sets have been using those to do field doubling to eliminate flicker since the early 1990s (though they stored the data on CCD chips digitally).
Given the time period the most likely explanation for this is that they used an analogue NTSC decoder, then a chain of 3 AD-converters with field stores to then convert them to analogue.
BTW, no you would not see any kind of "pixel structure" just like you don't see staircases at the "analogue" output of a CD-Player. The so-called "reconstruction filter" creates a smooth waveform no matter what you input. You can try that with a CD-Player by playing a CD with a single sample width pulse. (1 sample with a high value, the rest zero) What you will get is some sort of hill as a low-pass filter will smooth out your waveform. With better CD-players this will more and more approximate a sin(x)/x function. (with f(0)=1) I know this sounds incredibly counter intuitive, but if your signal was properly band-limited when it entered sampling, the reconstruction filter will re-create it perfectly well.
TIL that those black and white static thingie is called 'Snow"
You're a kid?
Some quirks and features, you say?
Doug DeMuro crossover when?
@Idiot Online Wondering Aloud omg ...next hoovies garage where he buys hooptie equipment
it could breaks it out to RGB signals and then uses some sort lf a delay line to draw the fields. SECAM used a delay line as part of its approach.
Always love Disney deviations in my Connextras
Bring back Dreamfinder!
Also, I remember seeing somewhere that at one time there was a tram that went around the Seven Seas Lagoon directly to the Magic Kingdom from parking/TTC, though it would mean missing a major Disney ride - the Monorail
I'm sure there was talking going on (and that it was likely interesting too), but I got distracted reminiscing about "A Day At Epcot Center" (aka "my childhood ") and promptly gazed longingly at Imagination, Horizons and well, poor clothing chooses. Ah, the early 90s!
I loved the ending :D I get questions like that all the time and try to avoid an instant facepalm.
"Quantization noise" You are right. We the engineers treat the limited bit depth as "noise." You can directly translate from the number of bits to an effective noise floor for a given signal. It isn't clear in that sentence it it refers to an internal framebuffer or the framebuffer of the source producing a black signal.
I would suspect that this is one complex circuit. The real interesting question is whether the image brightness is controlled more by the CRT or the LCD display, and whether the CRT portion is all analog directly off the NTSC or RGB inputs, or whether it is running off the digital signal that drives the LCD using A/D and D/A converters to maintain sync. I would suspect the LCD is likely 720x480 or perhaps 640x480, or maybe 800x600 based on the age of the unit. First, you have to convert the signal to the LCD using analog to digital conversion, and to field sequential or perhaps the memory buffer allows multiple colors displayed simultaneously and changes the values for one color at a time. This likely adds a delay to the video. Then you have to drive the monochrome CRT, either field sequentially three times faster or by displaying the entire image at once, based on the grid pixel pattern of the LCD panel at a higher resolution. Somehow this circuit must derive how much light information goes to the CRT versus using the LCD light to dark of each color pixel. The real question if the CRT light intensity is also controlled to allow for more contrast than just what the LCD panel can provide. My varying the CRT light intensity behind the pixels you can send varying levels of light to the LCD panel filters, adding a major increase in the number of colors that the LCD can display without expanding the color depth of the panel. The sound will also have to be delayed to be in sync with the delay related to the complex signal processing of that unit.
funny, the Pruett video about this device was recommended to me this morning....
TH-cam knows of your plans.
They were worried people were going to copy this tape...?
Pro NTSC/PAL gear was often pretty unforgiving about signal levels and timing, so it's not surprising that a monitor like this would have trouble with Macrovision. Anyway, it was cool to see another video about it!
The SECRET meaning of the word EPCOT is: Every Person Comes Out Tired -_-
During Food & Wine, *_Every Person Comes Out Tipsy_*
@@TechnologyConnextras You don't mean to suggest that I'm not supposed to have one of every drink from every nation, are you? ^_^
Cracking up at the end. People are funny.
The signal is almost certainly being buffered and read out digitally, it's just that it isn't using any sort of visually "lossy" compression like JPEG or MPEG, but rather something uncompressed like bitmap . A "bitmap" image or whatever lossless format this thing is digitizing each frame (excuuuse me, field) of video into is giving an exact value to each red, green and blue pixel, therefore is unnafected by randomness of stuff like snow that would wreak havoc on lossy compression schemes like MPEG.
We're talking about 1 field of standard definition here, so a couple megabytes of memory (if that) would be all that was needed, which would not at all have been a big deal in 2000.
I noticed I had stopped listening and was drawn in to the 1980's video and it's....80's things. Okay, I went back and listened this time.
There probly using an analog delay circuit. Lots of tv's use them to verieng degrees mid 90's on up to keep audio in sync
or for other video trickery PIP freezframe excetra.
i once had an analog tv or studio monitor *cant remember* that had a full 5 second delay from it's inputs.
this thing didn't even have an OSD all knobs and chunky pop buttons. kind of a channel selector for which input you were using. Never could turn off the delay and i went through every single button knob and gubbens on the thing.
Delay circuit was something to behold tho. took up a large portion of the internal space. huge jackpack on it tho. lots and lots of input. MMmmmm innnnpuuut.
You can use a 45 degree mirror to test the color filter upside down, with the filter 90° from the screen
"Technology Connextras"... ah... yes... this ol' chestnut... of course... no mandela effect ref whatsoever....
Interestingly the BBC was able to recover the color from black & white footage that was originally broadcast in color due to the unique interference that was created from the color signal leaking into the luminance signal . Interesting stuff. There was another way used for old Dr. Who episodes where they had lower quality color copies from North America and took the chroma signal and essentially overlayed it over the surviving black and white broadcast copy and with a little work, had a color show.
Wow! We even get connextras 2!
I have one idea that might demonstrate whether the display is using a frame buffer or not. It's my understanding that a frame buffer will add some small amount of input lag to a video game. For a number of old game consoles, there's a piece of homebrew software called the 240p Test Suite that, among other things, contains a couple of input lag tests. Ideally, you would split the signal from the console to both this monitor and a CRT that you know is fully analog, then go into the Lag Test (not the Manual Lag Test) and film both screens simultaneously with a high frame-rate camera. The on-screen animation basically counts frames, so you can go through the footage and compare whether the two displays are on the same frame of the animation at any given moment. If they're out of sync, you can even tell by how much. That would probably take a decent amount of effort to set up, of course, but if you want to figure out whether this monitor does any digital processing of its own, that's one approach you could take.
Edit: I perused the comments more and found one where someone mentioned they dug up the service manual and confirmed that the monitor is digital. So yeah, my suggestion probably isn't at all necessary.
This video loaded with no suggestions
EPCOT is wonderful and deserves more tangents
The Carousel of Progress is my favorite thing ever, which for some reason my brain keeps trying to associate with EPCOT
I haven't been to Disney in far too long
It always seemed to belong in EPCOT more than Magic Kingdom, even Tomorrowland.
As usual, anti-piracy practices come and ruin the day!
Newspeak at it again. Damn INGSOC
Just like my second favorite second channel, D!NG, which has always been called that.
10:45 Hey, that is a fair enough question
I think that the the digital frame buffer is so good that, quantization noise apart, the reconstructed signal is identical to the analog one (Nyquist).
This is infor-medy (Informatics & Comedy) at its finest.
Hmmm Im kinda disturbed by something. Its not quite right but i cant lay my finger on the connection.
The image might be digitised but the pixel clock may be derived from the original signal, and that process could be upset by the macrovision blocks. The wobble is probably it locking back on.
I didn't watch the whole thing, but could you use this technology to colorize black and white films?
Obviously, it adds the colors
@@albino_gringo1912 No, it actually just adds the colors
Can it colourize black and white audio?
@@l9day Can it monochromize color footage?
Holy shit, is this a Techmoan reference?!
The CRT scans at 180 frames and if the lccs screen part was put in front of another b/w CRT TV and the video signals into both TVs are the same source would result be a 20 fps color image? I want to try that on a vintage projection tv from 1951 i have.
Were here!
Its definitely digitizing the video. You can't see the horizontal pixels because the CRT hides the edges (it literally isn't sharp enough to show them). Being broadcast quality, it should be digitizing each line at D1 resolution (720 pixels across). What model JVC VCR is being used for playback? Some of the high end Digipure/TBC equipped SVHS models will show the curl at the top of the screen depending on the tape (no surprise its a Disney tape as they have been known to use very aggressive variants of Macrovision). The reality is an external full frame TBC can squash this problem and one can enjoy viewing all their VHS videos on this monitor.
I was hoping he would go on a tangent about color recovery at the end. Alec, if you're reading this... I'd love to see a video about color recovery!
Remembers EPCOT with one entrance. Feels old. :-)
10:33 ...the teacher failed! 😂
H-hey, I'm subscribed to this, who knew?
1. it appears to me that they also didn't do a great job of NTSC color decoding, say using a process like the Faroudja decoder. There are color subcarrier artifacts visible. Does this monitor have a Y/C or component input?
2. did you WORK for DISNEY at one point in your life?
3. awesome video I wish I had one of these.
Cool little device
It could be using analogue delay lines, as used in PAL receivers, to store the lines as they come in and then play the line out three times with different colour decoder matrices each time. That would imply that the shutter would operate at 3 times the line frequency rather than 3 times the frame rate.
You need to probe with your scope the interface to the LCD - if it’s switching at a multiple
Found this video that shows it does actually have a field store. th-cam.com/video/GTDpVZWQaFE/w-d-xo.html If you play the video at 0.25 playback speed you can clearly see that it is field sequential.
Wait wait wait... wait. How'd we get to Epcot, sir? lol Love it.
epcot is life
It probably is digitizing the video; but at such a high resolution that the small monitor cannot resolve the individual pixels together and smoothly blends them together.
I believe macro vision worked largely off of messing up the video timing and sync...a device called a time base corrector was used widely in tv stations back in the analog days...though digital magic did much of this in the digital world tv now works in largely.
A TBC matched the input signal to a reference signal and made the timing synchronized...thereby stopping the macro vision. We used to copy VHS movies in master control for our own personal use and macro vision never presented an issue. You could also alter the brightness, contrast, black level, hue, saturation and other signal components with a TBC.
I'd be curious if the back of this monitor has a reference signal input or how it would be affected with analog macro vision through a TBC.
he has done a video all about macrovision on his main channel, i'd recommend you watch it before complaining here.
I was going to say that maybe people with mobility problems would like the trams, but . . . then you're in Epcot at the world showcase, a MILE LONG LOOP, so I'm not sure. Might still be useful, though--people do ride the friendship boats often enough and they serve the same purpose.
My primary "EPCOT has changed" complaint is that I will forever lament the loss of the beaver tails stand at Canada. Sweet cinnamony fried goodness. I also miss the old Motion but not as much as the beaver tails.
When are you and Jenny Nichols doing a colab video about Disney theme parks?
it could be that, rather than truely pixelified, the signals waveform is digitized, stored in the framebuffer, but then ultimately DAC'd back to a typical TV signal.
I think maybe those people meant to ask if you could get a originally black and white tv to display a color image if you put this filter over it.
I would guess, unless you could change the TV's circuitry to output at 180Hz, you'd be limited to 1/3 frame rate.
an* originally
Without a sync between the TV and the filter it'd probably look rather interesting.
That's not how it works.
That's not how any of this works.
@@K-o-R I don't think persistence of vision would hold up at that frequency.
I went to Epcot this week, I’m kinda sad because there was so much more exciting stuff(to me) when I last went in 2003... no more Body Wars :(
Kind of odd that I don’t remember that old video you were talking about. Because I do have perfect memory, like remembering this channels name Technology Connextras. Weird what you “forget” in just a couple months.
Presumably if it is an analogue device it would use signal delay lines to split the video into three and then delay the signal by 1 or two frames?
What’s cool is this is super similar to DLP projection.
I laid in bed and thought about why Macrovision would interfere with the color and why just red and blue.
The best idea I can come up with would be related to how the Chroma signal is sent. It is sent in an I/Q format, so the chroma signals dont represent what we'd call real color. It's sadly not as easy as YPbPr where one is difference blue and the other is difference red. But if my memory is correct the I channel carries the red to blue information and Q carries purple to green. These signals are quadrature mixed...plus more bandwidth is devoted to the Q channel.
So this would explain why the red and blue gets wonky, apparently Macrovision is causing the thing to lose sync on the Q, probably I too but due to the insensitivity of human vision its not noticable.
Why would the thing get screwed up...thats a really good question. It shouldn't be screwing with the colorburst, but its possible as I had a lot of later TVs with digital comb filters that would fail entirely with Macrovision.
A lot of TVs didnt process true IQ either, they found cheaper ways of doing it. All my TVs that had true IQ processing always had dark splotches with Macrovision.
If I can find a service manual and reverse engineer the decoding circuitry I might be able to get to the bottom of it.