Blur Busters Forums

lyrill wrote: ↑

02 Mar 2021, 02:19

I'm actually amazed how much the jitter issue is largely alleviated by brute 20kdpi coupling with a lower sens. it's not completely gone obviously, but previously it's unthinkable to confidently venture into 8k or even 5k+ dpi realm. Of course all the low latency benefits help with accuracy too.


This is again done on 2/11 os and isn't sensitive enough to pick up smallest movement(even in desktop, so no it's not strictly an Areson bug or whatever) but slower feeling edpi in csgo does pick up all controls. Basically I don't trust this program to tell the whole picture..

There should definitely be an accurate hz read coupled with the plot drawing or Mouse Paint.exe to have context of perf vs hz vs speed.

As you can notice that you can't really discern if at 2k it's simply because the cross stitch marking from "track" is too thick and de-resolutioned the result or that there's actually still dotting. There should definitely be an adjustable thickness of the marking.

What's that tool you're using?

Chroma wrote: ↑

04 Mar 2021, 01:04

lyrill wrote: ↑

02 Mar 2021, 02:19

I'm actually amazed how much the jitter issue is largely alleviated by brute 20kdpi coupling with a lower sens. it's not completely gone obviously, but previously it's unthinkable to confidently venture into 8k or even 5k+ dpi realm. Of course all the low latency benefits help with accuracy too.


This is again done on 2/11 os and isn't sensitive enough to pick up smallest movement(even in desktop, so no it's not strictly an Areson bug or whatever) but slower feeling edpi in csgo does pick up all controls. Basically I don't trust this program to tell the whole picture..

There should definitely be an accurate hz read coupled with the plot drawing or Mouse Paint.exe to have context of perf vs hz vs speed.

As you can notice that you can't really discern if at 2k it's simply because the cross stitch marking from "track" is too thick and de-resolutioned the result or that there's actually still dotting. There should definitely be an adjustable thickness of the marking.

What's that tool you're using?

mouse test 2007 mfc 1.0.0.1, aka areson. taiwan program.

Chief Blur Buster wrote: ↑

09 Feb 2021, 22:35

hkngo007 wrote: ↑

03 Feb 2021, 19:09

https://mouseinsider.com/mouse-polling-rate-checker/
I get max 5.6Khz and average 5Khz (doing circles).

https://zowie.benq.com/zh-tw/support/mo ... ecker.html
I get average 5Khz (doing circles).

Browser-based mouse testers are currently unable to keep up. I tried with my 8KHz, and I can confirm 8KHz in native apps, but only 3-to-5KHz in browser apps.

Chief Blur Buster wrote: ↑

02 Mar 2021, 01:57

Ah, that. My browsers tend to only work to 3000Hz-5000Hz,

It's a confirmed browser limitation that needs to be fixed by Google/FireFox/Edge/etc. The HTML5 PointerEvents array overflows because the HTML5/browser standards were never designed for 8KHz. But it should be easy to extend with some code commits (to Chromium / FireFox). However, native software works fine, but that unfortunately requires a software download.

Peeps, don't use mouse-Hz-test websites for 8KHz testing yet

It's part of why TestUFO mouse tester has not yet been released.

I deliberately lowered the processor frequency to 4.7 and left a lot of applications in the background so as not to mislead anyone.
https://youtu.be/aA1pTUsKKKE

Any issues using this through a USB Extender? How about high speed monitor port (add latency? I believe BattleNonSense showed it does not but prior to 8k) or Keyboard High speed port?

Alpha wrote: ↑

07 Mar 2021, 10:50

Any issues using this through a USB Extender? How about high speed monitor port (add latency? I believe BattleNonSense showed it does not but prior to 8k) or Keyboard High speed port?

hate to be a broken record here, but it's common knowledge since decade ago NOT to use extender for anything serious. i believe this doctrine has not been debunked. nor close downloading when gaming etc

lyrill wrote: ↑

07 Mar 2021, 12:26

Alpha wrote: ↑

07 Mar 2021, 10:50

Any issues using this through a USB Extender? How about high speed monitor port (add latency? I believe BattleNonSense showed it does not but prior to 8k) or Keyboard High speed port?

hate to be a broken record here, but it's common knowledge since decade ago NOT to use extender for anything serious. i believe this doctrine has not been debunked. nor close downloading when gaming etc

Where was it proved? https://www.youtube.com/watch?v=nbu3ySrRNVc Just checking due to some of the old regurgitated BS being just that.

Edit: no issues with extender. The manual specifically states plug directly into PC.

Alpha wrote: ↑

07 Mar 2021, 14:58

lyrill wrote: ↑

07 Mar 2021, 12:26

Alpha wrote: ↑

07 Mar 2021, 10:50

Any issues using this through a USB Extender? How about high speed monitor port (add latency? I believe BattleNonSense showed it does not but prior to 8k) or Keyboard High speed port?

hate to be a broken record here, but it's common knowledge since decade ago NOT to use extender for anything serious. i believe this doctrine has not been debunked. nor close downloading when gaming etc

Where was it proved? https://www.youtube.com/watch?v=nbu3ySrRNVc Just checking due to some of the old regurgitated BS being just that.

Edit: no issues with extender. The manual specifically states plug directly into PC.

could you not link random youtube links without context and just use words? do whatever you want what works for you m8

lyrill wrote: ↑

08 Mar 2021, 10:02

Alpha wrote: ↑

07 Mar 2021, 14:58

lyrill wrote: ↑

07 Mar 2021, 12:26

Alpha wrote: ↑

07 Mar 2021, 10:50

Any issues using this through a USB Extender? How about high speed monitor port (add latency? I believe BattleNonSense showed it does not but prior to 8k) or Keyboard High speed port?

hate to be a broken record here, but it's common knowledge since decade ago NOT to use extender for anything serious. i believe this doctrine has not been debunked. nor close downloading when gaming etc

Where was it proved? https://www.youtube.com/watch?v=nbu3ySrRNVc Just checking due to some of the old regurgitated BS being just that.

Edit: no issues with extender. The manual specifically states plug directly into PC.

could you not link random youtube links without context and just use words? do whatever you want what works for you m8

Terribly sorry.

Where was it proved? Here is a link https://www.youtube.com/watch?v=nbu3ySrRNVc from a guest author on BlurBusters (this amazing website) debunking someone (lyrill) who may be regurgitating bs they heard that makes it seem like they just don't know what they are talking about ultimately sounding like a broken record becoming the very thing they sought to destroy (Star Wars reference maybe? Not sure honestly). To be fair, please continue if your time warrants such leisure.

In case someone is looking for validation on what could be an old wives tale vs a respected and knowledgeable person's technical video that seems to disprove the decade old "common knowledge " mantra of USB extenders causing latency\issues for reasons not being mentioned though its common knowledge, indicates other wise.

In an effort to save others time that may stumble across this in the future, I edited this yesterday with my experience so feel free to try. No issues, everything works as expected currently but the manual does state, "To optimize the polling rate of your Razer Viper 8K, it's highly recommended you connect the device directly to a USB 2.0 or USB 3.0 port of your computer. Connecting the device to a USB hub, switch, or extender may affect its performance.". I bolded "may" as it did not impact mine.

Opinion piece so thinned skinned folks are not impacted (thats passive aggressive nerds ). Razer is just balling out lately and major kudos. I've enjoyed my ultimate, my viper mini, though I tend to use my XM1 more and lately the GPro Superlight. Certainly feels good having a Razer mouse back on my desk and I appreciate some of the changes to the software which I've always thought of as overly intrusive. I do believe higher refresh rates matter here and believe its felt but would need some more time to play around. Siege has had an on going bug regarding RAW Mouse input. Leaving this ON is actually broke. Siege was janky @ 8k as I have that option Off however enabling it fixed that though I doubt its truly raw. I don't know. I'll have to try the test servers but last I tested Raw mouse option was still broken. Have a great day all!

The image clearly shows a latency increase with a 2 meter extender. Considering it's most likely an absolute value unaffected by the rest of the pipeline it would end up being roughly 5% increased total latency on an optimized high end system.

The odd thing about this screenshot is that using the monitor hub results in LOWER latency than directly connecting it to the back panel which just further reinforces my belief of 20 sample click to photon tests not being conclusive evidence.

lyrill wrote: ↑

02 Mar 2021, 07:10

oh and btw is there a link to the 3080+8k article cus I can't find it in Area and title post said it was due last Nov.

One of my computer failed in December postponed the article to later -- it was such a busy month and I had to cherrypick what tasks to postpone til new year. Sorry about that. It's still coming, as I hurry through my overflowing plate -- there is a Blur Busters Approved monitor I am currently focussing on getting launched first. THEN, I'll definitely hit the 8KHz mouse.

As you all already know, Blur Busters essentially runs on Valve Time for article deadlines. Guilty as charged.

Brainlet wrote: ↑

08 Mar 2021, 16:16

The odd thing about this screenshot is that using the monitor hub results in LOWER latency than directly connecting it to the back panel which just further reinforces my belief of 20 sample click to photon tests not being conclusive evidence.

I think both he and you are right.

20 samples are too few, but there are some interesting insights from background knowledge:

There are rare cases where a hub unexpectedly reduces latency. It doesn't happen often, but it happens -- for mudane reasons like improved USB processing (e.g. better USB chip in hub than in the computer) and/or repeatering for a stronger USB signal that doesn't require error correction by a crappy motherboard USB chip on a poor S/N connection running too close to noisefloor. USB2 packets being re-encapsulated (within USB3 hub) into USB3 if the chip in the motherboard is better and less buggy at processing USB3 than USB2. Or it could also be coincidential (crappy connector mating, worn USB ports, cables, etc). All kinds of rare edge-case reasons that a hub may unexpectedly reduce latency.

But almost always, it's better to connect directly.

USB extension cables occasionally add a bit of latency from USB error correction, since longer cables can reduce signal strength to the point where the motherboard's USB chip has to do more error correction by asking the mouses' USB chip to retry the USB packet that failed. Like a network connection with a 5% packet loss, it happens on USB cables. It's exactly why in high-EMI environments, wireless mice sometimes have less lag than wired mouse (e.g. if you live under a hydro transmission corridor).

If you use good cables and a good USB chip on the computer, then there should be no noticeable lag increases. But in real world, we have crap USB cables and crap EMI environments, with crap USB motherboard chips overloaded with many USB devices (like a simultaneous 4KHz USB keyboard), presenting lots of opportunity to add unexpected lag from USB error correction.

My current favourite recommendation for the 8000 Hz mouse is a direct USB plug to a PCIe USB card with its own PCIe lane directly to the CPU, with no interference by other USB devices (only USB device plugged into PCIe USB card). Single USB chip processing the 8KHz mouse directly.