I’ve wanted a Time Sleuth for such a long time, thanks so much for the in depth video. Got my parts ordered to make one, cheers mate!

Definitely gonna make one of these. Thanks for sharing!

Thanks for doing a video on this. Was intrigued when I found out about it. 🥂

Interesting video, I didn’t know about these devices. I am glad that the PCB has the option to use through-hole components.

I'm using this lag video tester, to calculate the lag in, Mist, ZX-Uno, UnAMIGA, ATLAS, Poseidon, Neptuno, FleaOhm, Terasic DE2, Alhambra, Mister, Misterynano, Tang nano 20k, All of these are Fpgas, and I am taking data of all of them, and others FPGAs in the Future, so many thanks TheRetroChannel!

@Rautz.

3 ай бұрын

Will you be uploading your data anywhere?

I love this, great job on breaking down the whole thing in this video and wonderful job from pthalin and zwenergy on porting the original timesleuth project! This is so beneficial to the community.

Thanks for the great tutorial! 👍

You thoroughly covered all the aspects one needs to know to build and use the device in an excellent way! Thanks for sharing my affiliate link!

@TheRetroChannel

Жыл бұрын

Thanks for porting the code to the Tang Nano. Brilliant idea

I don't know if I have a need for this, but it would be fun to bring it into a store and test out displays.... 😈 It's also nice to see the Tang Nano being used ubiquitously like this. I've seen so many projects pop up lately supporting this small FPGA board. It is a very useful, capable, and flexible board.

Excellent stuff! Definitely making me one of these. You can bet I'll be making the janky version too 🤣

@TheRetroChannel

Жыл бұрын

Order the Tang Nano and PCBs now and they should arrive around the same time. Trust me, the janky version even with only 4 components is easy to screw up 😄

@MoreFunMakingIt

Жыл бұрын

@@TheRetroChannel that only sounds a bit like me!

Pretty cool invention. I'm not a skilled enough gamer anymore to notice monitor input lag. Years ago when I used to game online Is where I noticed serious lag. Like 200ms or even worse. One minute you would be alive and the next dead, and you didn't even notice who or what killed you.

@TheRetroChannel

Жыл бұрын

Oh yeah, I remember the dark ages of online multiplayer, especially in Australia where our internet connections were junk (and in a lot of places still are)

@patriksretrotech

Жыл бұрын

Even if you are not a skilled gamer and don't "feel" the lag, it may completely ruin the game experience in a subtle way. I casually play Mario 3D on Wii U with the kids, and believe me, Im not skilled. We haven't played for some time and I didn't even think about that we had a got new tv, but suddenly I sucked bad at the game. First, I started to think it was me getting old. But after building the device, I realized what the problem was. So don't underestimate the importance of lag.

@gcolombelli

6 ай бұрын

Lag can totally ruin your gaming experience even if you're not a terribly experienced player. Display lag usually isn't even the worst part of it though, unless you're unlucky to have a very unusual display. Using a bluetooth controller on a "noisy" (in the RF sense) environment and running games on a software based emulator can add a ton of variable lag. While most people can adapt to a fixed constant lag, variable lag will make the gameplay very inconsistent and will make it nearly impossible and very frustrating to try to adapt how you play, it's just not worth the effort as it will screw with your ability to play on a more sane hardware setup.

@gcolombelli

6 ай бұрын

@@patriksretrotech May I ask what TV model was it, just so I can avoid it? 🤣 Seriously, thanks for your contribution.

Game mode on the plasma probably just turns off the overscan, so there's a bit more pre-processing, so more lag.

@TheRetroChannel

Жыл бұрын

The overscan didn't seem to change, and the tv has a separate option for turning overscan on and off. I was curious so I looked up the manual for this set, apparently it just displays a timer when connecting a game equipment 😄 "Game playing time is displayed only when connecting a game equipment"

@KAPTKipper

Жыл бұрын

@@TheRetroChannel That has got to the lamest "game mode" ever. But does basically answer that the game mode uses a bit more processing. Still 30ms is not much lag

Amazing project! Will testing 720p output on a 1080p display show accurate numbers? I assume the display will have to upscale to native res, and that could add latency.

@TheRetroChannel

Жыл бұрын

Possibly. The plasma display that I showed has a native resolution of 720p, but the numbers really didn't change between any of the resolutions. I believe the Tang Nano 4K isn't capable of doing 1080p60

@junmendero5396

Жыл бұрын

@@TheRetroChannel thank you. I ordered a nano 4k for this!

I like your desk mat, especially the color which is good for the eyes and less tiring. Where can I get it?

@TheRetroChannel

11 ай бұрын

It's an ESD safe mat, but I don't remember where I bought it. Might have been Mouser or Digi-Key

Thanks for the video on this. I got the pcb and everything set up but I’m not getting any readings. I can see the photo transistor is getting 3+ volts but no measurements.

@TheRetroChannel

4 ай бұрын

It could be the phototransistor receiving too much or too little light. I've found the through hole version requires shielding from external light sources and the SMD version requires a good amount of light to get into it

@ShawneeUnion

4 ай бұрын

Thanks @@TheRetroChannel . I'm going to print out the case and see if that helps. Also this tutorial was perfect. I'm not sure I could of made one without your help.

I'm having issues with my phototransistor requiring too much light to turn on, tried a few different models and doesn't seem to work. Anyone else have this problem?

@TheRetroChannel

3 ай бұрын

Yeah I've had some issues with the SMD phototransistor needing more light when it's inside the 3d printed case. I went back to using the through hole phototransistor, which seems to need much less light

Analog CRT lag should be constant across the entire display. There is simply nowhere to buffer 15ms of video. Input lag is measured from when the signal changes on the wire to when photons are emitted from the screen. It appears that the device you are using is measuring from the start of the frame interval to photon emission. Games can update the output between the start of the frame and when the updated area is drawn on the screen (eg. Atari 2600, or anytime screen tearing is observed). This is not reflected in the device's measurement. The measurements should be verified with an oscilloscope displaying the video signal against the photodiode signal. Given a constant scan rate, the measurement can be corrected by subtacting the product of the line number (accounting for progressive/interlaced scan) and the line interval.

@TheRetroChannel

Жыл бұрын

Yes, as I mentioned the time is calculated from when the frame starts. The Time Sleuth and photo transistor have no way of knowing what part of the screen they are seeing, only that it has taken X amount of time from the first scanline to light being emitted

@shanent5793

Жыл бұрын

the position could be encoded in the video signal (eg. flashing the lines at different rates) or made selectable as a user option. It's confusing to call it an input or display lag test or measurement and it gives the impression that CRT's have any perceptible lag. Latencies of even 10ms have been shown to affect targeting and tracking, so I think it's worthwhile to be precise and make measurements that can be compared

@gcolombelli

6 ай бұрын

@@shanent5793 Note that some very late high end CRT based displays do have a framebuffer, some even take HDMI input and even the analog inputs will likely get digitized and stored digitally on such models. Those can have a considerable amount of lag depending on how much video processing is being done and how many frames they're holding in memory. In all likelihood, this is one such model. This device, like most other display lag testers, measure the delay from the start of the frame, not from the start of the line, and it's up to you to interpret what is showing up depending on where on the screen you're positioning the sensor. The project is open source and if you want, you can adapt it for your specific needs. It's not meant to be a super accurate lag tester, but it's enough to tell if there's a framebuffer inside the display and if there's anything fishy going on, like the display being mounted upside down. Really, as ridiculous as that sounds, I've seen it with my own eyes. I kid you not, I have a LCD TV that does exactly that, it's a LG 32LH570B. I was having some fun comparing the lag between different displays when I decided to put a CRT next to it and compare it using the 240p Test Suite "Lag Test" feature and recorded at 960fps with my phone. I'll try to find the video to upload it to KZread, I almost fell off the chair laughing when I checked the video. I was ready to accept almost any wacky thing that this cheap TV could be doing to the video, but that still caught me off guard.

@shanent5793

6 ай бұрын

@@gcolombelli I suppose I'd have to see it too, because measurements that suggest that the display is upside down can often be explained by very slowly responding LCDs or excessively long delays so that the tool and TV aren't synchronized. I think analog SECAM would also be an interesting one to try, sets had an analog delay line to recover the color data that was spread over consecutive lines. It was usually the CRT HDTVs that had that kind of processing. The display is fixed at 1080i so it has to line double or resample anything else. The best CRTs were in projectors, they could sync to any analog input and digital signals were always handled in separate (expensive) boxes. Line doubling in principle adds very little delay but there were all sorts of video processors available to upscale your laserdiscs with various other algorithms. What I don't get is PC gamers attempting to measure latency using mouse clicks. Why not record the screen and mouse at high speed while tapping the mouse with a hammer? The movement delay is what matters most when aiming. LG really does make a full range. I think the cheapest ones I've seen were in hotels, 40 inches, blurry as heck, but at least they only weigh about five lbs!

@gcolombelli

6 ай бұрын

​@@shanent5793 If you don't mind the lack of side by site comparison with a CRT, I can record my cheap LG TV again at 960fps so you can see it updating the image. Some tests on 240p Test Suite are really well suited to making obvious issues with progressive video being treated as interlaced or panels being mounted upside down. I'd rather not move those heavy CRTs again for another shot. If that TV wasn't wall mounted, I would rather move it to sit next to a CRT. Yeah, most measurements people make with a camera and having their controller and display in frame are pretty much meaningless. Usually they won't tell you relevant details about what controller they're using, if it's wired or wireless, what display they're using and so on. MiSTer has tools for measuring both input lag and diplay lag now and there's a huge database of lag for different controllers on different connections when multiple options are available (wired, 2.4GHz dongle, bluetooth). I'd love to see such a database* for display lag over various TVs and monitors, and showing if it varies depending if it's using HDMI, VGA, SCART, YPbPr, S-Video or Composite Video. But that can have a few complications on it's own... like some consoles that seem to have a framerate slightly below or above 50/60Hz. On the default settings, MiSTer will try to output a steady 60Hz signal and will buffer and repeat (and maybe drop?) frames to keep a steady 60Hz output over HDMI, but you can override that, and what goes out over the VGA port is really whatever the core is outputing, unless you turned on VGA scaler. * I actually found one such list, but there's no indication of how it was measured, and considering it only had integer numbers, I wouldn't be surprised if they just took whatever they found in the spec sheets of those products and I wouldn't be surprised if they mixed "response time" with "input lag" either.