Blur Busters Forums

Hello BlurBusters,

I have tough question that I wasn't able to find answer yet, maybe you will be able to help me.

I know we have 1 place with very accurate response time measurements and it's TFTCentral reviews. Example here: https://www.tftcentral.co.uk/reviews/as ... d_response

What is interesting to me is Response Time difference between units of same model. I mean can we be sure that all units for ^^^ model for example will have approximately same response times or maybe their responces will be radically different for different units. Like in unit from tftcentral we had 4.8-6.5ms responces. Maybe another unit can have like 3ms-5ms, but other like 5.5-8ms etc.I know about overdrive, but let's omit this.

Please understand me correctly. There are thing in monitors that differs from unit to unit, f.e. backlight bleed (it can be almost 0 visible and it can be very visible), we have units with 0 dead pixels and units with f.e. 4 of them. But we know this because it's obvious - we see it in reviews from different people, they tell us their opinions. But what can we know about response times? People can't measure this at home and post us results. How can we be sure that measurements from tftcentral are approximately same for all units or if it's not a truth - in this case how big this difference can be?

I saw units of same model that had no artifacts with overdrive and had severe artifacts with overdrive enabled, so I believe there is some randomness in overdrive specific stuff (some voltage tunes on factory during manufacturing or smth like that) But like I said above if we exclude big overdrive from this question how can we be sure that response time don't differes much from unit to unit.

From one point of view to clearly answer this question TFTCentral should take like 10 monitors with same model (and even different revisions) and measure each of them and compare results, but this will never be done probably. So maybe you know smth about this, maybe you dealt with big amount of same-model monitors and clearly saw big difference at least on ghosting-test here or you know some manufacturing details that can answer this question or maybe I simply didn't find info even that it's already available.

Thank you in advance.

I've seen this happen, from many cases.

-- Temperature differences between different tests.
-- Different lengths of power-up between different units.
-- Different testing locations for different monitors.
-- Noise in oscilloscope measurements
-- Panel lottery (not always)
-- Firmware differences (not always)
-- Differences in temperature throughout the panel surface (e.g. hot corner from hot power supply)

Often when I unbox 2 units and let them warm up in the same room at the same time, in the same location of room (neither monitor closer to a fan or heater), and power up at the same time, both monitors has darn near exactly the same GtG. But then deviated when I turned off 1 monitor, let it cool down, and then turned back on. It was very different! Sometimes 10ms different, and I've seen 100ms differences -- e.g. a frozen VA panel shipped in the winter and arriving on my desk cold. It ghosts like a mofo, like any frozen LCD. But what many don't realize it affects gaming monitors too! Even a 1 degree temperature differences can create millisecond GtG differences on some panels. It depends on the panel tech.

I've seen GtG faster on one end of the panel than the other, and I found out because it was because of a hot power supply at a corner of the panel.

You know those frozen smartphones? That you forget in your car in the middle of the winter. They ghost a lot and slowly. Or those LCD screens on parking meters in the middle of winter. Or any frozen LCD. They are really, really, really slow -- sometimes GtG measured in seconds. Even a 5 degree temperature rise can speed up GtG by several milliseconds. I've seen 10ms GtG become 2000ms GtG when I parked the monitor outside in the winter -- the typical LCD behaviour when frozen. But smaller temp differences (15C vs 18C vs 20C vs 21C vs 23C) means 10ms GtG may become 11ms or 15ms or 20ms or 30ms.

90% of the time -- Temperature, baby, temperature! Even when it's at the end of the room that is 0.5 degrees warmer.

That's why VESA specifies a controlled temperature when measuring GtG, e.g. measure at a precise 20C.

-- Always unbox the monitor and let it warm up (shipments through freezing weather)
-- Warm up the monitor for 30-60 minutes in the winter if you play in a cold room.
-- Heat up the room in winter if you're playing competitively. You don't want to underperform with an LCD at 55 degrees F or 12 degrees C slowing down your response time.

Few realize it, but it makes a difference.

Yes, firmware differences (messups on the firmware) and panel lottery play a role (one panel may reach GtG spec at 21C and another panel may reach GtG spec at 19C, creating tiny millisecond-scale GtG differences when played at the same temp). But 90% of the time it's plain old fashioned temperature that explains the differences between two identical model monitors with identical firmwares in the same room.

Although you're talking about unit per unit variance, along the same line of thought, I personally wonder if something happened with the XF252Q. They sent out models either to reviewers or suspect reviews will be done in the first month or two so they produce the 'premium' or just models that match the specs for an initial run and then all the models they make cheaper to raise their baseline.

I went through four different XF252Qs and one thing I noticed on the last two models after they went out of stock on Amazon for like three weeks is they seem slower. This is based on playing mobas and seeing name tags scroll across the screen. Where as before, especially with the first monitor I had they were almost completely readable (while scrolling the screen and trying to read nametags, that's a big deal) and the last two they're fuzzy and illegible. It's still better then my old monitor it replaced, however I don't think it's the same specs of the original monitor, using the new AUO panel.

However you can't test this without ripping it apart and looking at part numbers (to which I have no expertise) or having a in depth testing setup.

Most smaller hardware review sites (like TFT), and monitor reviews is a small segment, probably have an agreement setup with a company to test hardware and then return it to them. That being said they probably don't make enough to buy and test a large swath of monitors, however you could definitely do this and return the products to something like Amazon.