Blurriness in Overwatch

[masneb]

Personal preference. When younger thought it was better to play at the highest DPI a mouse offered (grew up in 90s), slowly ratcheting up as each new mouse offered higher DPI, then stopped as mice seemed to get janky when using higher DPIs around 8200 (which was the sensor limitation at the time). 8600 is where I ended up after trying to make my new mouse feel like a older one I was using. Since I haven't increased DPI as I've learned that a lot of sensors don't work well at high DPI, manufacturer implementations of said sensor aren't all good, introduce a load of error, or manufacturers introduce interpolation/'fake' DPI.

Most mice aren't tested well or at all above 2k DPI, meaning it's an absolute crapshoot regardless of what the box says or the sensor is capable of. Even Rtings only tests up to 1600 and that's about the only website that tests things well or with any sort of meaningful scientific methodology. The worst thing about this is the manufacturer wont even tell you if the mouse has been tested at higher DPI or where interpolation starts, for a lot of mice it's at about half the mouse DPI. If sensors/mice worked properly, I'd play at the highest DPI offered, then reduce sens in game. Even today with mice offering 32k DPI, most don't work well, or even acceptably at 8k~.

No reason to play at a lower DPI as I'm not sacrificing accuracy (within reason). I don't play with my hand floating, my palm is sitting on my desk and use the base of it as well as finger tips to move my mouse, which is why I can maintain accuracy while playing. It's the complete opposite end of the spectrum to people who play like a printer.


Still figuring out a way to capture this. The VODs don't show the behavior, in the recording it looks no different when I use my mouse as opposed to keyboard. This might be something where you need a highspeed camera, which I don't have.

I tried Apex and wasn't nearly as prevalent, but that may be due to slower game speed as well as lower contrast and no character names (in training area).

[Chief Blur Buster]

Still figuring out a way to capture this. The VODs don't show the behavior, in the recording it looks no different when I use my mouse as opposed to keyboard. This might be something where you need a highspeed camera, which I don't have.

You may want to pursuit your camera (handwave) along the screen motion while you're turning.

Stationary camera = eyes stationary
Moving camera = eyes tracking moving objects

Basically pan the camera at the object that your eyeball is tracking, to keep the camera-gaze same as eye-gaze.

Similar to
viewtopic.php?f=23&t=4782

Although games don't have the sync track built into them, it *may* partially capture some of what you're seeing. Try 4K60fps filming.

However, the filming may not be useful without a reference (e.g. video of what expected, versus video of problem). So even if you nail the pursuit technique, it is quite possible that without a reference (of what it should look like), it may not capture the essence of the problem.

[masneb]

Rather interesting, so I spent some more time with this tonight as my return window is closing on my monitor, it looks like I'm starting to see the deviation and inconsistencies in my mouse polling.

Somethings worth noting:
-I think anyone can see this themselves in albeit artificial scenario by setting their polling rate to 125hz. It was extremely noticeable when I did this and assume it would be visually present on older monitors. Not sure how new of a monitor this would work down to (240 or 360s? People would need to try).
-I tried 800 DPI, again, this didn't change anything. It appears to be completely unrelated to DPI.
-Overwatch HPMI does not do anything. It should, but it doesn't. I think this has something to do with a technical limitation in some way of HPMI and it might only really help lower end PCs struggling with FPS. I do not, especially in training, I can maintain capped 1%s and close to capped .1%s at 460FPS.
-Another mouse did not change what I was seeing, although I'm sure it plays into things.

The interesting things are when I started messing with polling rate.

On 8000hz I could see jitters and vibrations, almost like it was 1k only in a different way. I'm very sure this is USB struggling to handle the polling, and tripping on itself. 4k seems as though it looked better then 1k and 8k. 2k seemed like 1k. The main thing I'm looking for is clarity of text and outlines on training dummies while doing very smooth circles on them.

This is regardless of feel or consistency of the input. This is PURELY visual. 1K still feels the best, which is why I use it and will keep using it.

I swapped around to multiple USB hubs in my system and some were downright nauseating in how they presented the vibration/blurriness/hazy image. Being able to see a lot of microstuttering, when a poll is off by a lot (nothing to do with FPS, this is polling microstuttering). Just like MouseTester, this almost visually matched the deviation in the input graphs for each port. MouseTester itself being very prone to error and questionable validity at times.

I mentioned a couple years ago that there really needs to be a dedicated USB port for mice or some sort of other standard for peripheral input and this really does look like this is the start of visually 'seeing' polling rate and all the fluctuations dealing with it. It's everything - the polling rate of the mouse, the USB standard, the manufacturer of the mouse, the model, the USB hub that's being used, what chip it's using, it's all of this put together. The best port on my system was in fact a stand alone USB card I bought last year to figure out some weird nuances I was having, but I doubt it's the best solution. I'm still curious about Thunderbolt to USB...

This has nothing to do with click to photon latency and another indicator of why that's only a small piece of the puzzle. Something that would be much more useful then Reflex Analyzer is something that verifies and validates your USB sensor input. FCAT for sensor polling. I spent a bunch of time using Reflex Analyzer thinking that was part of this, it's not. Input latency seems to have absolutely nothing to do with this. I honestly don't even know if click to photo latency matters beyond getting in the ballpark. I think devs for Reflex were chasing the wrong thing, they were chasing the lowest possible latency, instead of the error and consistency of the data it was receiving.

If it looks this way right now, 480hz OLEDs are going to be absolutely wild. People need BFI because the whole peripheral chain is broken and they don't even know it. Just hide the garbage.

[Tell]

The deathadder v3 wired is also 8k polling capable but it still has a bug with the polling rate forcing you to re-plug it every boot to get it to work properly. There was a firmware update for this mouse last month that didn't actually solve the problem.

https://insider.razer.com/mice-and-surf ... ight-44271

In that thread someone mentions the polling bug is also present on your mouse the viper 8k.

Have you updated the mouse firmware? it mentions "Fixed an issue where the cursor will jitter/shutter when the mouse is set to high DPI."

Synapse should find it when you check for updates but if not you can download it here:

https://mysupport.razer.com/app/answers ... MjElMjE%3D

[styhk]

Do not use extreme overdrive

[masneb]

The deathadder v3 wired is also 8k polling capable but it still has a bug with the polling rate forcing you to re-plug it every boot to get it to work properly. There was a firmware update for this mouse last month that didn't actually solve the problem.

https://insider.razer.com/mice-and-surf ... ight-44271

In that thread someone mentions the polling bug is also present on your mouse the viper 8k.

Have you updated the mouse firmware? it mentions "Fixed an issue where the cursor will jitter/shutter when the mouse is set to high DPI."

Synapse should find it when you check for updates but if not you can download it here:

https://mysupport.razer.com/app/answers ... MjElMjE%3D

No, I'm on base firmware as I found out updating the first time caused jitter. You can find my responses in the original firmware update on these forums. The only thing upgrading off the original firmware does is help with a plugging issue, which I've never had across two system upgrades. The cursor jitter was extremely noticeable.

Not sure why this is being suggested as I explicitly mention testing two mice.

Not using extreme OD, that causes all sorts of inverse ghosting, even if it makes it more responsive. That's different and not what I'm noticing.

[babooz99]

I know what u mean. Nobody else will understand it. Most people here are like low skilled players they dont know what theyre talking about

[MatrixQW]

I don't know if this relates to other games as I mainly play OW2 and there isn't really anything as fast paced as OW2.

That's because you never tried QuakeWorld.
I play it. There is a small but strong community.
There is no game that is so low latency and smooth as QW, you can play at 1000fps easily and frametime is stable at 1.3ms
If you watch the entire 10m there are some incredible fast paced scenes.

phpBB [video]

[masneb]

I played Quake 3 at lan club in highschool in 99-03, I've played plenty of Quake over the years. Quake isn't anything new. And at home at the time I was playing Tribes 2.

[MatrixQW]

Quake 3 isn't QuakeWorld, the speed and movement are not comparable.
Of course it isn't new but it's unknown for younger people.
This was just to say QW is the fastest game ever and there are no issues with blurriness or slowness