It's not rare it's a currently manufactured product. It requires a very fast processor so yes the tech being used really is appreciable. There's really no competition though so they can charge whatever they want, which is probably a matter of setting the price at a point where their production all gets bought up. The fact that they have "budget" models like xrgb mini tells you it's not just BS markup though.
My trintron died and after hunting for a PVM or BVM for months in my are i ende up caving and getting one of these. Im pretty happy with it and it obviously takes up less space which is a HUGE plus. I even found an LG 1080p tv that gives me like 5ms so i seriously lucked out.
That said, I do miss my old RGB monitor and wish it was immortal but sadly they are all doomed a pitiable death. Those of you with working ones are fortunate, but the next best thing is easily one of these Framemeisters. Quite possibly my favorite purchase in the last few years.
>>2885525 >>2885558 In addition to pro applications, these Faroudjas were used in high end home theaters with $40k 9 inch CRT projectors like Electrohome Marquee 9500. The peak of 90s video tech. I have a Faroudja LD-100 doubler and it's a solid piece: those knobs are solid metal, all BNC connectors, and the inside is packed with circuit boards and components.
>>2885735 The xrgb-mini is fine if you're just playing games for casual fun but if you play games with strict timings, speed running, arcade cabinets etc it's not the best option. It's also nice if you can't really afford tons of RGB mods for your consoles and don't have the money or space for high-quality CRT to play them on.
The unit adds almost 1.5-2 frames of lag to the picture, not counting the lag if your LCD which is probably going to be another 1-5 frames of lag depending on how good your TV/monitor is (unless you use one of those low-latency gaming monitors). Games that have strict timing will be frustrating and almost feel like they play different tot he original.
I've beaten Mega Man X in like 38 minutes on a XRGB, which is pretty much identical to my time off of it. I can't recall, but my time was always only a few minutes off the best speed runs, and that's just one of my games that I speed run. So depending on the game, its fine.
That said unless you're lucky like I was, you're going to be hunting for a good HDTV to pair with it. I found a LG that gives me practically nothing I only have the base lag the unit gives. I don't recommend fighters or SHMUPs, but anything else is actually fine provided you find a good HDTV.
>>2885792 >Even the fastest crt with a 60hz refresh has at best 8ms of lag. Since they are top down refresh. So are LCDs (in the best case, in the worst case they buffer the whole frame first). The GPU/PPU/whatever renders a pixel at a time, the cable transmits a pixel at a time, the only way to speed it up is with 120Hz + emulator BFI. The point is CRTs have zero additional latency beyond this unavoidable latency.
>>2885832 >>2885798 No experience with that LG but it's looking like last year was the last year for great 1080p sets. Oleds are the only thing that may come about and change that. We'll know soon with ces coming up.
>>2885843 >Some assholes will claim so. Or some smash kid from /v/ Lol you ain't lying. I'm surprised one hasn't chimed in here yet. I've played through many fast, precision based games lately and can't tell anything.
I'm still toying with finding the optimum framemeister settings for each console. I wish they would fix is issue with scanlines on 1080p.
>>2885936 >It is functionally identical to the scart to HDMI scalers that are like 50 dollars. People say it looks "way better" but there's no side by side difference. There is very much a difference. Maybe not enough of a difference for you to spend the extra money, but the XRGB scales a much better picture.
I had the $50 upscaler everyone has, and it did look decent on a still screen, but any background or sprite movement caused a blurry mess.
>>2885425 >Why exactly is this thing so expensive? Very niche product and literally no competition as of right now. If you're willing to wait, there's a guy on the Shmups forums who's currently working on an alternative that's gonna be cheaper and apparently superior to Micomsoft's offerings as far as features go. http://shmups.system11.org/viewtopic.php?f=6&t=52158
I have a basic understanding of how CRT's and HDTV's handle resolutions.
From what I've been told, CRT's could technically handle any native resolution you could throw at them, up to a certain point. Making sure devices being shown properly at their native resolution don't have any problems and look the most optimal. Output resolution of the console syncing up 1:1 with the output resolution of the tv
With modern displays like HDTV's and Monitors, they only have 1 Native resolution that is absolutely assured to handle the content and display it correctly. Because that's how just how the video scalers in modern displays are. Any oddball resolution causes a blurry output and high latency, depending on the quality of the scaler in the tv.
Two things. How wrong am I? And what's stopping TV manufacturers from putting in better scalers into their tv's? shouldn't this problem have been solved by now? Is it simply neligence on the manufacturers part or is there a technical reason?
Why would they waste time and money on a feature that barely anyone is going to use? For every schmuck like me that wants to retro game with a great picture and little lag, you're gonna have like 1000 normies who just want to watch netflix in HD and even more normies that just want to play CoD/Madden or watch sports in HD.
Theres no reason for them to actually put time into that when designing and manufacturing a television.
Honestly, I prefer RGB monitors and the XRGB to most CRTs.
>>2885425 In short, nothing. Nothing. does. It's a cash in designed to take advantage of consumer whores who are gullible into believing they are some how missing out on some deep connection with their old games because how they play isn't authentic enough, probably because they never grew up and connected with these systems and games as only a kid can, but rather because they were told as adults what was to be revered and coached into having taste.
If you have RCA cables, play your retro system on the TV 'the way it was meant to'. If not, just hook up your computer to your TV and emulate what you can, add filters if you want, or maybe hack a game. Just have fun with the game, don't be caught up in minor details. The people designing these games surely weren't. Try to keep in mind what makes retro special to us, and while there is a physical aspect to it that is for the realm of diehard collectors or people who never tossed their systems away, just being able to freely talk about games that impact us for one reason another is way more important than how you project it on your Guy-Montag sized LCD screen display.
Just enjoy the game, especially because they are free for the taking, and you are doing them a service IMHO by keeping them alive. Share with your friends some Soul Calibur, not you fancy gizmo that adds fake rips in your jeans for $400.
If you really think that your post has merit when you imply I lost sight of what makes retro special just because I use RGB cables and a XRGB mini then shit friend, you've got to start learning how to not generalize just because you've seen some picture quality elitists in the CRT/RGB monitor general.
>>2888196 I've been feeling cynical lately because of all the bad threads on this board. I don't contest that you get something out of your device, how can I? If that's the case, this thread has merit. But I am always suspicious of shills for obvious reasons.
What I am cautioning about is the over worry of trying to recreate an experience rather than let the experience naturally talk to you. If you're cool with that, I think we can see eye-to-eye.
>>2885593 >those knobs are solid metal Placebo and muh tech benis >inside is packed with circuit boards and components. Yeah all that tech is replicated with a single processor chip these days. You as retarded as audiophiles and muh golden ears.
>>2888202 >Shill b8 Fine.I'll bite. If you're gonna insist on using a modern TV instead of a cheapass CRT that people are literally giving away, you might as well emulate on a computer with cheap but decent monitor. Any nVidia or AMD videocard can output to almost any resolution. If not, just add a 3X Raw filter and stop being autistic with muh scanlines.
I swear shillfags are innit with the CRTfags to trigger me.
There are a bunch of people who insist that anything that is not a high quality CRT is garbage, yet there are people who say a HDTV with a framemeister is just about the best you can get. Sections of the CRT crowd go into detail about not just why CRTs are great, but the many problems with LCDs, plasmas and every other display technology. Yet somehow this $400 memebox supposedly bypasses all the flaws of a LCD.
The upscaler crowd is just weird. The only conclusion that made sense was a mix of crowd mentality and buyer's denial on a large scale.
I just started ignoring everyone after I kept asking for research done on human latency perception. All the people arguing couldn't provide anything more than their own feelings. The whole conversation is riddled with bad science and confirmation bias.
I appreciate the concern. I dont really think there is anything wrong with gsetting the best quality picture from your old systems. And that is using RGB.
Playing Mega Man X on an emulator with filters like that other anon is not really the way to go, but is likely a better option than a really shitty CRT or playing on a LCD or HDTV without a scaler like the XRGB
>>2888647 >There are accuracy tests that prove certain emulators are more accurate than the real thing.
I know the emulation crowd is disillusion about "emulation is 100% perfect", but I didn't believe anyone thought that they could be "more accurate than what they emulate". That's such a logic fail, I'm surprised typing it didn't cripple your brain.
>>2888459 So much butthurt. Some of us appreciate things that aren't made out of plastic. Good luck repairing the latest junk from China which has everything integrated into a single ASIC. Those Faroudja processors won Emmy awards for advancing NTSC broadcast technology, and I'm sure the analog stages in them are still better than what's around today given the required stability and accuracy for pro use.
Bullshit about any audiophile analogy, stuff that's built well is simply better. I have an Ikegami pro monitor that has a solid metal faceplate around the CRT and it's fucking sweet.
Even audiophiles have a point: you would probably prefer a Durabrand boombox from Walmart over a Mark Levinson amp because the boombox is highly integrated while the Levinson has all discrete output transistors...kek.
>>2889079 In the case of machines with mixed signal parts (eg. the SID in the Commodore 64), emulation can be more accurate than the original, with accurate meaning "more like a typical example of real hardware". Your real hardware SID chip might have weird filters because they are analog with poor tolerances.
>>2889203 The only reason to emulate at transistor level is if you want to emulate broken or misused (eg. cartridge tilting) hardware. For normal use, it is a digital system, so RTL is sufficient. https://en.wikipedia.org/wiki/Register-transfer_level
And theoretically it is possibly to formally prove that a conventional emulator is equivalent to an RTL hardware description, so if you managed to do that a conventional emulation with such proof would be equally perfect.
>>2891036 Objectively wrong. The vast majority of TVs are high persistence, and those that aren't have unacceptably high latency. There are literally only 2 ways to get low persistence + low latency: Original hardware + CRT Emulator + 120Hz LCD + BFI
>>2891041 If you care are about blur, then the only two acceptable ways to get rid of it is high refresh rates. 120 or 144hz. Not really applicable to retro games.
The other way is a strobe.
BFI is trash. It's distracting and still causes blur. It's essentially a very long strobe.
CRTs have their own quirks and anyone that isn't a fanboy has to admit they are terrible at displaying an image properly. You can write paragraphs on that. All a CRT really offers in this application and in this modern time is a nice feeling of playing on old hardware. Even that is really pushing it with the popularity of using CRTs that were released years and decades after the game came out.
The other way is a strobe.
BFI is trash. It's distracting and still causes blur. It's essentially a very long strobe.
CRTs have their own quirks and anyone that isn't a fanboy has to admit they are terrible at displaying an image properly.
>>2891069 >CRTs have their own quirks and anyone that isn't a fanboy has to admit they are terrible at displaying an image properly. What, an SD CRT doesn't display old SD content well? Or did you mean a PC CRT doesn't display SD content well?
>>2891662 I was born in the early 80's, and love CRT with a passion, but this anon is right: >>2891684
With some of the new very low latency LCD's and tools like the one in this thread, it's a hard sell to stay on a CRT. For me, the biggest thing is that I am pretty tired of the screaming high pitched noise from the flyback transformer.
>>2891684 >Geometry issues Never solved, this is the only real problem with CRTs >convergence issues No problem on high end sets >bloom Solved by playing in a dark room and not setting the brightness stupidly high >Linearity As in gamma? That's intentional >terrible contrast Competitive with LCDs even in the crappiest sets, close to OLED level in high end >-color “decay” >too sensitive to ambient light See bloom >limited service menus Depends on the CRT
>>2891842 >For me, the biggest thing is that I am pretty tired of the screaming high pitched noise from the flyback transformer. That's only a problem if you insist on using a SD CRT. They have ugly scanlines and are generally worse in every way than a good PC monitor CRT. Use a high end PC monitor CRT + emulation, and the whine will be 100kHz+, so not even your dog will hear it.
>>2891871 Alright, not him, but you clearly know a fair bit. What about persistence vs lag in LCD? Why do people say CRT's have better black levels. And why do station engineers say that OLED has only been better than a (new?) CRT for master monitor use for about two years, and that most places can't afford to use it because it fades like nobody's business? Surely that's way faster than a CRT ages?
How are CRTs sensitive to ambient light?
What about halation and the simple glow of the phosphors? That alone makes SD content look far better than any other way, IMO. Can LCD do anything like that?
>>2892113 >What about persistence vs lag in LCD? It’s over blown since a strobe and modern LCD tech solve both. Strobe gives you no blur just like a CRT. Modern LCD tech has near no lag these days. On LCDs that take advantage of that of course. Most still have well above 3 frames of lag with a 60hz game.
There are tricks to get past pixel persistence blur too. Just being smart with the strobe. That enables zero blur. It is equal or better to a CRT.
>Why do people say CRT's have better black levels They do for the most part. This is confused with contrast very often. Contrast on these CRTs is measured by a black screen vs a gray screen. The ANSI test uses a more realistic test with a checkerboard pattern. CRTs can’t handle that and the black levels plummet. You can easily see this with white text on a black background.
>And why do station engineers say that OLED has only been better than a (new?) CRT for master monitor use for about two years, and that most places can't afford to use it because it fades like nobody's business? I don’t know the tech that well but I do see people saying the half-life with the brightness is a huge issue. I assume it’s an issue like burn in with plasmas that will get fixed down the road.
>Surely that's way faster than a CRT ages? Probably. CRTs have a long usable life but a short life if you’re really crazy about having accurate colors. At the very least they have to be recalibrated often if you’re that crazy about it.
>How are CRTs sensitive to ambient light? A couple ways. You have the glare from the glass. Ambient light causes black levels and colors to be off.
>>2892871 >What about halation and the simple glow of the phosphors? That comes down to resolution and manufacture more than anything. The CRTs we’re talking about aren’t very sharp. If you like that look then a LCD won’t give it to you.
This is where I differ because while I do like to play on a CRT every now and then I prefer the sharpness a LCD and scaler like an XRGB gives me.
You also have scanlines. Some games I like to turn them on and some games I leave it off. Most the time just leave them off. It’s something pretty specific to retro games.
To me these are more flaws as far as display quality goes. But I grew up with them and do admit they are nice to look at sometimes.
>>2892951 You don't seem dumb enough to be the previous anonymous who thought strobing backlights were powered by time travel, so maybe you are somebody else. Yes, there is added latency with a simple strobe (approx half a frame average), because the backlight cannot be lit up until the entire frame data is transfered. CRT starts drawing immediately. If you want to match CRT latency you have to strobe a large number of vertical sections immediately as the pixels in that section update.
>>2892958 I thought you were the same asshole. The guy who ignores fact and videos supporting it.
There is no added lag. >Yes, there is added latency with a simple strobe (approx half a frame average), because the backlight cannot be lit up until the entire frame data is transfered. CRT starts drawing immediately. I was never and am still not talking about simple strobes or whatever the hell you’re talking about. You’re comparing a top down refresh to full screen refresh and saying that somehow adds more lag. You have zero support for this claim. None. You never did.
Regardless if it’s a full strobe or a scanning strobe it’s only on for .5ms or less.
I posted two real world tests last time. One done with a high speed camera the other in comparison with a CRT to fully support this. Both had well under a frame of lag. Here are the links again. http://www.blurbusters.com/gsync/preview2/ http://shmups.system11.org/viewtopic.php?f=6&t=48662
This image explains the chain. Hopefully even you can understand. Once the display has the frame information it needs then it strobes. There is no added lag here. Please tell me where the added lag is? Actually before you start typing. Find me a single test that shows a modern strobe having added lag because of the strobe.
What I don’t think you understand is the display can strobe multiple milliseconds before the entire frame is “done”. Completely dependent on the hz of the strobe and framerate.
Strobe with slow motion video. https://www.youtube.com/watch?v=hD5gjAs1A2s And before you be a dumbass again like last time. The first part is a comparison without strobe. Please don’t go off on another misunderstand rant like last time about that.
The simplified breakdown with a modern strobe is -start refresh -wait for pixel persistence/ghosting to be done -wait for enough frame data -strobe -end for next refresh
With that there is no added lag. Unlike what you’re confusing -start refresh -wait fucking forever because you say so and apparently manufactures are dumb or something. I really don’t know. Best I can make out of your logic -wait for the whole fucking frame -strobe for like 5 mins -end and start again
With a proper strobe the whole thing should done just before or at the end of the frame transmission.
The strobe adds no additional lag. As supported by multiple tests with strobe on and off. No additional lag between them.
>>2893076 >You’re comparing a top down refresh to full screen refresh and saying that somehow adds more lag. Of course it does, because you console renders at the same speed regardless of your display, and your cable transmits a single pixel at a time. Pre-5th gen. consoles don't even have a frame-buffer, so it's obviously impossible to speed up the frame transmission.
>>2893081 >-wait for enough frame data And there is where the added lag come from.
>>2893076 Frame transmission time is even the biggest contribution to latency in the image you posted! Your own image contradicts you. You can cut that latency by half if you use a CRT (reducing it to zero at the top of the screen, and not changing it at the bottom, so halved on average).
>>2893121 >Of course it does, because you console renders at the same speed regardless of your display, and your cable transmits a single pixel at a time. Pre-5th gen. consoles don't even have a frame-buffer, so it's obviously impossible to speed up the frame transmission. >And there is where the added lag come from. There’s no added lag. The panel is refreshing itself while the strobe is off. It then strobes for a full frame refresh. Again as supported by the tests there is no added lag on these panels compared to the strobe being off for this very reason.
You don’t know what you’re talking about and still provide zero evidence for your claims.
>>2893130 >Frame transmission time is even the biggest contribution to latency in the image you posted! Your own image contradicts you. You can cut that latency by half if you use a CRT (reducing it to zero at the top of the screen, and not changing it at the bottom, so halved on average). What the fuck are you saying? A CRT speed up frame transmission now? A CRT is a top down refresh display. It ain’t speeding anything up.
A CRT doesn’t instantly begin drawing lines either.
>>2893135 >There’s no added lag. The panel is refreshing itself while the strobe is off. It then strobes for a full frame refresh. The CRT refreshes each line sequentially (like an unstrobed LCD). You see it without having to wait for the whole frame to be transmitted before the strobe.
>>2893139 >A CRT doesn’t instantly begin drawing lines either. There's no frame buffer in a CRT either. You're claiming data storage by literal magic. When a CRT receives the data for a pixel it displays it immediately. With a strobed LCD it also updates the display immediately, but you can't see it because the backlight is turned off. You have to wait for it to finish updating *all* the pixels before it can turn the backlight on.
>>2893173 >but you can't see it because the backlight is turned off. You have to wait for it to finish updating *all* the pixels before it can turn the backlight on. False. You still don't read any evidence I posted.
The NES (and similar) has no framebuffer: http://www.google.com/patents/US4824106
The CRT has no framebuffer: (google image search "CRT schematic")
Therefore the only possible way to transfer the data is a pixel at a time. There is literally no place to store it.
The CRT moves the electron beam in a raster scan, and lights up the pixels in order as they are received. It has no choice but to do this. If it does not display the data immediately it would be lost, because there is nowhere to store it.
The LCD can have a frame buffer, but more commonly it has a line buffer, because that's simpler and cheaper. Watch the high speed videos of non-strobed LCDs and you will see it update line by line. The time taken to see the new image data is approximately the same as the CRT + the pixel response time.
The strobed LCD does the exact same thing, but you cannot see the new image data immediately because the backlight is turned off. The backlight is only turned on after the complete frame transfer, otherwise you would see tearing artifacts. That is the source of the additional latency.
In the case of the scanning backlight, you do not have to wait for the whole frame transfer, because you can strobe vertical sections as soon as they are updated. With small enough vertical sections the latency could match the non-strobed case.
Sure, if it's a scanning strobe. But if it's a single strobe like it usually is then you *must* wait for the whole frame transfer or you get tearing artifacts. You are claiming your strobe is capable of time travel. You can't show pixels before they are transferred to the display.
>Add the clever trick of the the pixel response time being taken care of in the off cycle of the strobe That makes no difference to latency, you wait for the response time whether you watch it or not.
>>2893217 >Sure, if it's a scanning strobe. But if it's a single strobe like it usually is then you *must* wait for the whole frame transfer or you get tearing artifacts. You are claiming your strobe is capable of time travel. You can't show pixels before they are transferred to the display. See already posted evidence. >That makes no difference to latency, you wait for the response time whether you watch it or not. Same as above.
Or continue to just spout your opinion with zero evidence.
Time-travel anonymous - I have a money making scheme for you: 1. Get access to a high speed stock trading center 2. Set up a DAC outputting stock prices, add blanking generator so it's a valid video signal. 3. Hook it up to your strobed LCD. 4. Put a photodiode pointing at the bottom of the screen, connect to an ADC 5. Strobe before the complete frame transfer 6. By the power of magic, you can read the ADC and figure out what the stock prices will be a few milliseconds into the future 7. Profit!
>>2893238 You're measuring LCD vs LCD, not LCD vs CRT. And your measurement methodology is highly dubious - you don't even report what sensors you used or where they were positioned, which is critically important as added latency depends entirely on vertical position. That's why I keep saying "average" -- the added latency varies from zero to approx one frame depending on position.
>>2893245 That's what you're claiming is possible. You say that you can strobe the backlight before a complete frame is transfered and yet still see the complete frame. If you can see into the future, why aren't you rich? A few milliseconds of precognition will result in great wealth in high-speed trading.
>>2893260 The better the strobe, the lower the duty cycle. Obviously a sufficiently bad strobe is indistinguishable from the unstrobed case with 100% duty cycle. This degenerate case of "strobing" adds no additional latency because there is no real strobing. Your argument is now "strobing doesn't add additional latency if you turn it off and pretend it's still there", which is correct, and also retarded.
Note that even if you're willing to accept tearing artifacts (which nobody else is, and I've never seen on a strobed LCD), you *still* have a latency penalty, because the strobe on time is always shorter than the frame transfer time. No matter how you time the strobe there will always be some pixels that you have to wait to see because the strobe happened to be turned off when they were updated.
The solution is the scanning backlight. Here the "strobe on time" can be equal to the frame transfer time without causing sample-and-hold blur, because it's spread out spatially.
>>2893298 The strobe isn't the frame. How retarded are you?
I'm honestly curious. How fucking brain dead retarded are you?
Your claim of >>2892946 >"Strobing gives half a frame of extra lag." has been completely disproved.
All that's left is to laugh at your dumbass really.
The blur buster site has multiple tests showing anywhere from no added lag in best case scenario to about 3-4ms with lighboost which is not a scanning backlight. The other source showed no lag as well.
>>2893326 >The strobe isn't the frame. But the frame is invisible when the strobe is turned off.
None of your tests actually measure LCD vs CRT.
You are claiming something literally impossible. No pixels are visible unless the strobe is turned on. Pixels are updated on the display, visible or not, for the entire frame transfer time. This is an unavoidable consequence of serial pixel-by-pixel transfer. The frame transfer time cannot be sped up because it's a fixed property on the console's PPU.
Therefore if the strobe is turned on for less time than the whole frame transfer time (which it has to be or it is completely pointless), there will always be some pixels which are not immediately visible. They will not become visible until the next time the strobe is turned on. No other technical details matter, it's simple fact that something that takes 15ms to finish cannot be completely shown in 5ms. You can't get around this with any kind of timing tricks. 15 is bigger than 5. Some of those pixels will not be shown immediately.
The only solution is to increase the strobe time, and the only way to do that without adding sample-and-hold blur is with a scanning backlight. CRT raster scan is effectively a scanning backlight.
>>2893359 >You keep posting shit that is talked about in the links. There's absolutely nothing in your links supporting what you claim.
>Post a single piece of evidence to support your claim. 15ms > 5ms. It's that simple. If the frame transfer time is greater than the time your non-scanned strobe is turned on, then you cannot see the whole frame until the next strobe. There is no trick for making 15ms equal 5ms. Time travel is not real.
>>2893373 No test will show that because that's not what I claimed. The additional lag caused by non-scanned strobing is (frame time - vertical blanking time) / 2, averaged over the whole screen. It varies from 0 to (frame time - vertical blanking time) depending on where you measure it vertically. In no case can it be over a frame.
For your tests to actually prove it can be less than this, they would have to specify where the measurement was taken, and they would have to actually show less lag than (frame time - vertical blanking time)/2 at center of the screen.
That's just over 3ms for a 144Hz display as compared to a CRT. No such LCD exists.
>>2893405 I read your articles last time you claimed this magical bullshit, vertical position was not reported, and no latencies low enough to disprove my claim were reported. I'm not the one claiming literal magic. If you have a precognitive monitor, go claim your $1M from the JREF.
>>2893486 Who claimed that a strobe adds that much lag? I have a cheap LCD and with the framemeister, I have 0.03 to 0.05 frames of lag. That is near perfection. A human literally cannot notice that minuscule amount of lag.
>>2893486 If you can start watching something that takes 15ms and see the whole thing in only 5ms, then you are a genuine magical wizard and you should be out taking over the world instead of shitposting on 4chan.
>>2893554 >I have 0.03 to 0.05 frames of lag Bullshit. You're claiming 0.5ms to 0.8ms. Not even the fastest gaming LCDs can do this, and that's even before you take into account strobing. Your measurements are wrong.
>>2893647 >Though it's measured in ms. It's still limited to 1 frame resolution because input latch is strobed once a frame and the output is by definition only updated once a frame. And even if really did have millisecond resolution, that anonymous is claiming sub-millisecond resolution, so he would still be lying.
>>2893657 I just checked out the latest SNES version and it's only 1 frame resolution.
>>2893627 Here's how to do it properly: Get a CRT, an upscaler, an LCD, a 4th gen. or older console, a signal splitter, a dual channel oscilloscope, and two light sensors (eg. phototransistor + resistor + power supply). Put the light sensors dead center on each screen. Connect one to each scope channel. Set up some software on a flash cart to show a black screen, wait for input, and then display a white screen. Trigger on the CRT's sensor. Measure the time taken for the LCD's sensor to reach 90% of peak (or whatever threshold you consider sufficient, 90% is generous to LCDs).
Assuming your LCD is non-scanned strobed, the time difference will vary according to the position of the sensors.
>>2893669 >>2893669 Lag test starts at 1:43. Single frame resolution, same as the SNES version. Manual lag test is also single frame resolution (it even says "offset 1: 1 frame", so even retarded people like you should be able to figure it out). You appear to be the blind one.
>>2893676 >implying averaging the manual lag test readings means anything I can hit single frame windows, I can give you whatever result you want. The manual lag test is completely invalid for argument purposes, because it's trivial to bias it. Only objective measurements, preferably with an oscilloscope are valid.
>>2893684 Anybody who's not completely incompetent at video games can force the manual lag test to output whatever number they like. The fact that it averages its single frame precision readings is irrelevant when all those readings are potentially meaningless. And somebody who's spent a lot of money on an inferior system has a strong incentive to bias the result, even if only subconsciously.
>>2893696 I never said it didn't display ms, I said it didn't measure fractional frames, which it doesn't.
It's very clear that the claim of 0.5ms to 0.8ms is bullshit, and an obvious result of manipulating the manual lag tester (if it's not completely made up, which is more likely). Even getting that level of consistency isn't humanly possible without deliberately biasing the timing to correct for the missed frame windows.
>>2893698 If lower scores and less wins counts as "manageable", sure.
I honestly don't think there's anything inherent to it that makes it so pricey. It's just because it's a niche thing and it basically has no competition.
From what I undestand, it's not so much what the framemeister does, but what it doesn't do. Other scalers make the same mistakes as TVs as interpret 240p as 480i and then use line doubling and other shit, whereas the FM just straight-up takes the 240p signal and upscales it. This results in a better picture. Again, this is from what I understand, and I'm not talking about the extra features the FM has.
It doesn't seem that complicated to do a straight-up upscaling without doing any interlacing or deinterlacing. Only issue if you ignore that stuff is the fact that 1080p isn't an even multiple of 240p (it's 4.5x as big). What I would do is just scale it to 4x (ie, 960p) and then have a set amount of junk pixels around the border of the screen to bring it technically up to 1080p without compromising on pixels and not needing to worry about that sort of shit. Not sure if that's trickier to do than it sounds (I'm sure it is) but I think it would be possible to make a cheap no-frills upscaler just as cheap as the shitty ones that interlace and deinterlace everything
>>2885452 Not OP, I've been dreaming about doing this for a while now. I'm a burnt-out electrical engineer looking for a way out of the hamster wheel, and I've been wanting to start a project at home. I've really wanted to redesign a lot of retro products using all-American manufacturing and superior quality to the Hong Kong specials I see all over eBay and Amazon, but I don't know if it will even be profitable if it costs more per unit.
>>2894256 Reason leads me to believe you might be right; it's fashionable to make cheap shit that can't quite color within the lines. But is no one willing to pay a few dollars more for a quality product? I paid a few dollars more for those SNES and N64 security bits that were made in the USA, just so I know I can use them years from now. I mean, I'm willing to pay one hundred dollars for a Metal Warriors instruction manual, so maybe I have a few extra bucks, right?
I've bought cheap, foreign made USB cables that are sold in bunches just because they fail often enough that you need a hundred or so to get a functioning ten. Most of us have had to deal with garbage audio connectors that fail within a month or two after use. Is this acceptable to you all? I want to change this shit.
Please forgive the soapbox, I'm drunk on gin and feeling a little megalomaniacal.
All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.
This is a 4chan archive - all of the shown content originated from that site. This means that 4Archive shows their content, archived. If you need information for a Poster - contact them.
If a post contains personal/copyrighted/illegal content, then use the post's [Report] link! If a post is not removed within 24h contact me at firstname.lastname@example.org with the post's information.