RetroTINK 5x-Pro

[ldeveraux]

So yeah, sounds like it's the Carby then.

[bobrocks95]

Everyone throw your Carby's out! Couldn't possibly be the Portta doing it with multiple sources!

[ldeveraux]

Everyone throw your brains out, as they won't allow you to comprehend new information nor admit you were wrong! Couldn't possibly be you, everyone else is out to get you.

[SuperSpongo]

lol, arguing on the internet.
We've all been there. We say we don't care, should have let it go many posts ago, yet we keep coming back to it. New information has come up, I could be right!!

[bobrocks95]

lol, arguing on the internet.
We've all been there. We say we don't care, should have let it go many posts ago, yet we keep coming back to it. New information has come up, I could be right!!

I could test some other HDMI sources with the Portta, but then I really would be wasting my time, at least posting is cheap.

[maxtherabbit]

lol, arguing on the internet.
We've all been there. We say we don't care, should have let it go many posts ago, yet we keep coming back to it. New information has come up, I could be right!!

I could test some other HDMI sources with the Portta, but then I really would be wasting my time, at least posting is cheap.

I'd actually really like to know. I'd always assumed it was just a defect in the device itself, and that it would do it with any source, but in truth I've never actually confirmed that hypothesis.

[orange808]

This sounds like a 'discussion' we had earlier where orange808 suggested (if I remember right) that flat panels can scan out pixels faster than a CRT - which isn't true, but it's a belief that can lead to making a lot of other mistakes about display performance

You're spreading FUD.

When testing with a Time Sleuth and evaluating the actual panel "scan out", that can and definitely does happen on many displays. I'm talking about only the process of "drawing" the image and that's what we were talking about in the OSSC thread.

At 60Hz refresh, a measured difference between the "bottom number" and the "top number" that is less than 16 2/3 ms indicates an absolutely unambiguous phenomenon: the physical panel managed to complete the "scanout" process (from the timed start of "drawing" to finish) faster than a CRT electron gun would (physically) scan the entire image at 60Hz. I've owned multiple LCD displays that did exactly that.

It's not untrue or even controversial. The reason the LCD will still finish the scanout behind a zero latency mechanical CRT is because of video buffering/processing lag represented as the top number. Of course, it's limited by how fast the information arrives. However, HDMI provides a fast frame transfer feature and a theoretical video processor in the chain would already have video information on hand. So, yes, it would be possible to pass it on without incurring huge latency penalties.

This all makes perfect sense, because I have random access to every "pixel" of the screen (limited only by response) the moment I have buffered and processed video information to "send". So, I can write to them as soon as I have the video information ready.

"Butz, butz, Umm... Butz, the entirez imagez isn't knownz until it's all bufferesdz!!! So, it can'tz workz!" (drool, drool, and knuckle drag)'

"Well, a DVDO must be witchcraft." lmao

The display certainly does not need a full frame buffer and (as I said previously) we have random access to the entire panel's pixels. The only restriction on the physical image panel is the response time, but even that wouldn't stop me from changing the "value" of pixels. I believe we call that ghosting when a moving "object" is unable to reach full "saturation" (completely change") before it is changed again.

Want a simple explanation? :
Obviously, when we know a panel can display higher refresh rates than 60Hz, we already know it can "scan out" faster than 60Hz, anyway. It's obvious. So, it depends on the video processing. Kinda makes you feel silly, huh? (However, this explanation is incomplete and doesn't reflect the actual limits of the panel, because the maximum refresh allowed by a manufacturer depends on things like cable bandwidth, image quality, and the video processing hardware design.)

Have a good one. I guess that was a belief that could lead to (what was it you said?) : mistakes.

[Guspaz]

A CRT scans out the image at the same speed that the input provides it. Digital transmissions work the same way, the image is sent gradually such that at 60 Hz one frame takes 1/60th of a second to send. Thus a digital display can't scan out a 60Hz frame any faster than it receives it, and it receives it at the same speed as a CRT would.

If the LCD/OLED appears to be scanning the frame out in less than 1/60th of a second, then it buffered part of the frame before it started scanning it out. You can also cheat by sending a 60Hz signal at 120Hz and duplicating frames, such that you can scan out each 60Hz frame in 1/120th of a second, though you've then got a 1/120th delay before the next frame gets scanned out.

The exception to this is HDMI's Quick Frame Transport feature, which sends the frames as fast as it can and then waits for the next frame. This was added as part of the HDMI 2.1 spec and won't be supported by any retro gaming related hardware or scalers. I'm not even sure if the current-gen consoles support it.

[Unseen]

At 60Hz refresh, a measured difference between the "bottom number" and the "top number" that is less than 16 2/3 ms indicates an absolutely unambiguous phenomenon: the physical panel managed to complete the "scanout" process (from the timed start of "drawing" to finish) faster than a CRT electron gun would (physically) scan the entire image at 60Hz.

It doesn't - even if you manage to perfectly measure the very first and last rows of the picture, the time difference between them in a standard 1080p60 signal is just 16.0 milliseconds. The remaining time is the vertical blanking.

[Fudoh]

On a 480p signal it's only 15.24ms, right ? 525 lines for the full 16.67ms, so 480 lines in 15.24....

[orange808]

At 60Hz refresh, a measured difference between the "bottom number" and the "top number" that is less than 16 2/3 ms indicates an absolutely unambiguous phenomenon: the physical panel managed to complete the "scanout" process (from the timed start of "drawing" to finish) faster than a CRT electron gun would (physically) scan the entire image at 60Hz.

It doesn't - even if you manage to perfectly measure the very first and last rows of the picture, the time difference between them in a standard 1080p60 signal is just 16.0 milliseconds. The remaining time is the vertical blanking.

Granted, but you're splitting hairs.

Does that invalidate my point? The number being slightly smaller doesn't matter much, does it? For instance, a 120Hz capable display will be plenty fast enough to "start late" and "catch up" to a CRT by the end of the scanout--and I believe that was the context of the discussion that Eddie was talking about (I believe that was the OSSC thread).

Given that we have **random access** to all the "pixels" (instead of mechanically scanning), does it prevent the physical panel from actually putting the image on the screen (from the moment it begins to the moment it stops) faster than a CRT would perform the same operation--when we isolate the task and consider only the "drawing" from start to end? And, of course, because we're splitting hairs, we will acknowledge that the CRT just shoots the phosphors on the glass and it glows for an instant. The entire image isn't actually lit all at once--but human beings see it as a complete image. Just so I don't get some shitpost reminding me of that.

Blanking time could vary depending on the source, because we still have two "standards" floating about for 1080p. I believe there's a thread about frustrations surrounding reduced blanking timing interfering with CRT usage and unpredictable timings.

[Guspaz]

As I said, you can only refresh a digital display as fast as you get the input signal, which is the same speed as an analog video hits a CRT. The "data burst" digital video functionality you described is HDMI 2.1 only. The digital signal arrives top-to-bottom, just like a CRT physically draws it. This is why very low latency digital panels (like LG OLEDs) update the image top-to-bottom like a CRT.

[Josh128]

At the risk of pissing somebody off (which I dont care), a digital panel will never update faster than a CRT if its displaying an image from an analogue outputting retro-console. To do so it would have to defy physics. If said console has been modded to provide a digital output, depending on how that is done, I could see how it might be possible. OK, Im going eat my chicken tendies that I just cooked.

[orange808]

As I said, you can only refresh a digital display as fast as you get the input signal, which is the same speed as an analog video hits a CRT. The "data burst" digital video functionality you described is HDMI 2.1 only. The digital signal arrives top-to-bottom, just like a CRT physically draws it. This is why very low latency digital panels (like LG OLEDs) update the image top-to-bottom like a CRT.

I never even acknowledged your existence. Just noticed you had replied in the first place, as I regularly ignore you. Still don't give a shit to reply, because you're just in for an argument.

Most displays probably scan from top to bottom. So?

I know how a CRT scans. Thanks.

I specifically made my point and you chose to ignore the specifics.

Hey, let's argue. Fun, fun.
Going to post another picture of a gross blurry LCD cheap beamer with two frames of lag projected directly onto a wall (with a hotspot, too) and tell us how great it is? Never replied to that, but don't think I didn't notice. It was horrid, mate. Never was fast. Not even for a projector--as there were simulation projectors that could perform with a single frame of latency. And, the image quality was... bad.

[Kez]

I really don't think this claim is so outrageous as to derail another thread. Seems to me like people are just talking at cross purposes.

The claims aren't breaking the laws of physics or anything, they can be illustrated by this:



The purple part is smaller in the LCD. The complete image still arrives later than the CRT, simply the time spent actually putting the image on the screen is shorter because it buffers some information before it starts drawing. I haven't personally observed this but I don't see any problem with it in theory.

[thchardcore]

Was 24khz support ever implemented?

[Strider77]

I usually won't get a picture at all, but if it does it's unstable horizontally.

This was my result with ANY source and I went through the effort of exchanging the unit and it gave the same (shitty) results.

#Trash

[parodius]

Was 24khz support ever implemented?

It did work when I tried last year.
viewtopic.php?f=6&t=67984&p=1457376#p1457376

[Guspaz]

Going to post another picture of a gross blurry LCD cheap beamer with two frames of lag projected directly onto a wall (with a hotspot, too) and tell us how great it is? Never replied to that, but don't think I didn't notice. It was horrid, mate. Never was fast. Not even for a projector--as there were simulation projectors that could perform with a single frame of latency. And, the image quality was... bad.

For somebody who is so aggressively negative towards me, you seem to have confused me with somebody else, because I have no idea what you're talking about: I've never posted photos of an LCD projector pointing at a wall...?

I'm sorry if I murdered your... cat? Dog? Firstborn child? I assume I must have done something terrible to you to evoke this kind of response. I'm afraid I don't remember what it might have been.

Kez's illustration makes my point quite well in visual form. No matter what form of display, be it analog or digital, it can only display the image as fast as its being sent, each frame takes 1/60th of a second to send, and it is sent top to bottom.

[orange808]

Going to post another picture of a gross blurry LCD cheap beamer with two frames of lag projected directly onto a wall (with a hotspot, too) and tell us how great it is? Never replied to that, but don't think I didn't notice. It was horrid, mate. Never was fast. Not even for a projector--as there were simulation projectors that could perform with a single frame of latency. And, the image quality was... bad.

For somebody who is so aggressively negative towards me, you seem to have confused me with somebody else, because I have no idea what you're talking about: I've never posted photos of an LCD projector pointing at a wall...?

I'm sorry if I murdered your... cat? Dog? Firstborn child? I assume I must have done something terrible to you to evoke this kind of response. I'm afraid I don't remember what it might have been.

Kez's illustration makes my point quite well in visual form. No matter what form of display, be it analog or digital, it can only display the image as fast as its being sent, each frame takes 1/60th of a second to send, and it is sent top to bottom.

I got started programming using a little kit for the Atari 2600 as kid. Stop telling me how a television works. I literally learned that before high school.

[Josh128]

Was 24khz support ever implemented?

It did work when I tried last year.
viewtopic.php?f=6&t=67984&p=1457376#p1457376

Oh snap, I must have missed that. Interesting. How did it present in a 480p output though? Centered with black borders? Or did you scale it to the 1024x768p that was once implemented?

[Josh128]

...

[ldeveraux]

and you thought I shit on this thread...

[thchardcore]

We can all be a little more civil and kind...enough ugly to go around at the moment.

[Josh128]

Mike, my buddy has still not come through with the damned slice file of the speaker covers, I will call him again.

[thchardcore]

No worries Josh. Thanks for the update. Sorry for going OT everyone.

[jsteel]

I don't have a RetroTINK 5x (yet) but I was just reading the RetroTINK 5x manual and it says that the triple buffer can cause "occasional judder due to the need to repeat/drop frames". I wonder how noticeable this judder is; if I play a side-scroller for example, am I going to notice anything and how often might it judder? Is anyone able to comment on their experience with this?

Thanks

[vol.2]

I don't have a RetroTINK 5x (yet) but I was just reading the RetroTINK 5x manual and it says that the triple buffer can cause "occasional judder due to the need to repeat/drop frames". I wonder how noticeable this judder is; if I play a side-scroller for example, am I going to notice anything and how often might it judder? Is anyone able to comment on their experience with this?

Thanks

I think it's entirely subjective, but I wouldn't notice the "up to 1.25 frames" of latency myself at all. Some people say they are more sensitive to this.

In any case, it's important to remember that triple buffering is optional, and is offered to help with games that have resolution switches and possibly weird compatibility issues. It's not even needed for the vast majority of use-cases.

[jsteel]

I don't mean the lag/latency but the "judder" that it mentions (but yes I guess this is caused by the variable latency). On a smooth scrolling game is there really a noticeable judder; something that you notice as an annoyance? My TV is not compatible with for eg the OSSC 5x so I wonder if that means I'll need this triple buffer, rather than frame lock.

[vol.2]

I don't mean the lag/latency but the "judder" that it mentions (but yes I guess this is caused by the variable latency). On a smooth scrolling game is there really a noticeable judder; something that you notice as an annoyance? My TV is not compatible with for eg the OSSC 5x so I wonder if that means I'll need this triple buffer, rather than frame lock.

I couldn't say, but you might find someone here that has experience with your TV in this thread if you give the details.