World first QD-OLED monitor from Dell and Samsung (34 inch Ultrawide 175hz)
- Thread starter Grim5
- Start date
Associate
- Joined
- 28 Sep 2018
- Posts
- 2,257
Some good testing here: https://www.reddit.com/r/ultrawidem...be10/aw3423dw_full_0100_pixel_response_times/
His numbers make a lot more sense. I was confused by the HuB numbers since had the display transition times were as they measured you’d see massive chrominance overshoot issues. Instead, we have none. Similar to spectros, you need high end equipment for accuracy in this as well.
@Baddass just fyi.
His numbers make a lot more sense. I was confused by the HuB numbers since had the display transition times were as they measured you’d see massive chrominance overshoot issues. Instead, we have none. Similar to spectros, you need high end equipment for accuracy in this as well.
@Baddass just fyi.
Interesting, although a flawed approach really in my opinion. The reason there are different figures floating around is because different reviewers are using different measurement methods, something that i've been working on with TechTeamGB (who have recently produced their OSRTT response time tool many reviewers are now receiving - including Techtesters who i know reviewed this screen) and with Hardware Unboxed. We've been trying to provide guidance to reviewers who are receiving these new devices on an explanation of all the options and possibilities, recommended settings to use, or at the very least a consistent approach to how things are named and labelled to avoid reader/viewer confusion. What we don't need is everything beind bundled in to "response times" when there's many different ways you could interpret the measurement data. Ideally you'd need to understand with any set of results what method was being used.
Using full 0 - 100% response times as the guy on reddit has done is probably fine for OLED as the transitions are near instant anyway, and it won't make much difference to the overall figures. However, for any LCD screen it is flawed and massively too strict to be of use. On LCD screens you see very slow tapering off for the transition rise time and fall time which would make a full 0 - 100% measurement excessively long. Those measurements are included in the measurements I take, but on LCD they have very little relevance when you are considering actual visible response times / pixel changes. After all, capturing "response time" is about capturing something that is relative to what you see in practice and allowing you to compare screens. Adding in a long tapering off time when the shade has already reached a shade that is entirely inperceptable to the user is pointless. That's why having some kind of tolerance level in there is optimal.
Gamma correction is not "bull" as he claims, it helps ensure that when you set a tolerance level you account more closely to what you'd perceive in real life, not the light level that the device is measuring from the screen which won't account for gamma. It allows you to consider visually how close a transition gets to its final state, and at what point that shade is "close enough" to be visually identical or at least very similar.
After a lot of testing and consideration I personally settled on using an RGB 10 tolerance level, which gets you extremely close to the desired shade to be visually indistinguishable in practice (esp in motion content) but avoids being too strict or capturing parts of the measurement curve that are unncessary and not representative of what you'd see in real use. HUB instead opted for a 3% tolerance level which is basically capturing nearly all of the curve and very close to 0 - 100% full transition time, but just removing bits that could be noisy in the data. That's why their results are always slower than mine (and other sites). It's not clear to your average viewer that this is only different because they're capturing more of the curve than we are. In my personal opinion i think that 3% tolerance is too strict, for the same reasons as explained above when considering the full response time (0 - 100%). It also has challenges with being variable depending on the range of the transition (explained more in my article below). I've discussed this at length with Tim at HUB before
but each to their own.For what it's worth i would trust Tim's measurements of the AW3423DW far more than some random guy on Reddit
Techtesters were using another different method as well, which caused some confusion with the AW3423DW in their review. They were using the new OSRTT tool but the main confusion came from their transitions in the top row of the response time table from black (0) to grey. There were some transitions recorded at ~5ms with everything else showing <1ms. I spoke to them about this at the time as well to figure out what they'd done. Basically those few transitions were showing some overshoot as well (also captured in Tim's measurements) but unlike Tim and every other response time tester, rather than capturing the "intiial response time" before the overshoot peak as has been the practice for 20+ years, they also included the overshoot peak in that measurement. So the 5ms was how long it took to change to the desired shade, go up and over the overshoot peak, and settle back down again afterwards. Thats in the OSRTT software and another possible measurement approach but at the time they hadn't been given the guide we'd worked on explaining the different options, and hadn't explained what they were capturing in the review. I've suggested they move to "initial response time" to standardise that with everyone else to avoid that kind of confusion. So in their results any transition that had any overshoot will give you an overly long G2G figure too because of that different method.
Hope that helps explain some of the differences
ps i am hoping to review the AW3423DW soon as well!
Far more detail on it all here if you're interested: https://tftcentral.co.uk/articles/response_time_testing
@mrk Do you have any side by side comparisons of huawei gt 34 and aw3423dw? I have the huawei but the qd oled is on order until august. Would be amazing to see side by side comparison. I am aware VA is known for contrats compared to IPS but I am preety certain the QD OLED is in a league of its own with true blacks.
Caporegime
Have been working from home today and the fan has come on. It's noticeable but appears to be shifting very little air feeling the air vents.
Associate
- Joined
- 28 Sep 2018
- Posts
- 2,257
@Baddass ultimately Tim’s data doesn’t correlate to panel behavior. The tech itself excels at near black handling and the has a distinct advantage there. Vincent can save me a lot of typing: https://youtu.be/iF7Ieqxj4K0 I documented this well before Vincent's video so it's not some piggybacking going on.
If I was to take Tim’s readings, correlate that with his woled readings and then apply that what we know about near black/coming out of black issues on WOLED, it'd be easy to assume the issue will be much worse on QDO.
You're absolutely right. It's new tech and to me that means extra caution and peer review of findings should be done to make sure it doesn't have some weird quirks or requirements. The onus is on the reviewer confirm their tools and methods are still applicable. There's no reason why OLED/QDO/LCD based tech can't have their own process and even individual datasets. You might end up being limited in only being able to compare within the same technology and that's fine also. I understand the urge to have a blanket dataset that covers all tech but that's not always viable.
Thankfully for Tim and others, people are less ocd in the monitor space. Because those who rushed out info on the s95b with insufficient methods and tools are still doing wound licking. We all want data and want it yesterday but accuracy should always be a driving force. Have you noticed how Vincent doesn't have a full review out yet? Or some respected people on AVS are only confirming things as they're able to correlate that data first with others?
For this particular data point, I trust this guys findings for the simple fact the data is inline with panel and tech behavior. I'm sure his data could be refined better as he's not claiming to be anything but an amateur but it gets the point across. So yes, for this use case random redditor > Tim for me.
If I was to take Tim’s readings, correlate that with his woled readings and then apply that what we know about near black/coming out of black issues on WOLED, it'd be easy to assume the issue will be much worse on QDO.
You're absolutely right. It's new tech and to me that means extra caution and peer review of findings should be done to make sure it doesn't have some weird quirks or requirements. The onus is on the reviewer confirm their tools and methods are still applicable. There's no reason why OLED/QDO/LCD based tech can't have their own process and even individual datasets. You might end up being limited in only being able to compare within the same technology and that's fine also. I understand the urge to have a blanket dataset that covers all tech but that's not always viable.
Thankfully for Tim and others, people are less ocd in the monitor space. Because those who rushed out info on the s95b with insufficient methods and tools are still doing wound licking. We all want data and want it yesterday but accuracy should always be a driving force. Have you noticed how Vincent doesn't have a full review out yet? Or some respected people on AVS are only confirming things as they're able to correlate that data first with others?
For this particular data point, I trust this guys findings for the simple fact the data is inline with panel and tech behavior. I'm sure his data could be refined better as he's not claiming to be anything but an amateur but it gets the point across. So yes, for this use case random redditor > Tim for me.
Man of Honour
Ah the GT is in another room and all fixed in (as is my AW really on the arm etc) but I can assure you that even combining both pros of IPS (my old LG) and the GT's VA panel that the QD-OLED wipes the floor with it and leaves room to spare. There simply is zero contest between the three and QD-OLED wins at everything. Well maybe not fully on text rendering because of the subpixel layout but in Windows you can work around that as mentioned previously. But certainly for everything else!
Thanks for respoding! Yea QD OLED is an amazing, the only major downside I see is above average input lag compared to other gaming monitors. I cannot wait to receive mine. Will be close to 700 after Cashback. Bargain if you ask me. Can almost buy 2 x LG C2 OLED for that
Man of Honour
Hmm personally I have not noticed any input lag - Is where you've seen it mentioned/shown via HDMI input perhaps? Via DP it's seemingly instant and generally feels more instant than teh Mateview GT due to how quick the QD-OLED motion and everything else is overall.
@Robert896r1 , ah ok i see what you are referring to now. Yes the longer response times up to 2.58ms don't inintially feel right to me for any OLED tech and something could be wrong with the way they were captured of the measurement device. Although having said that Techtesters got very similar results for the transitions to black here, and thats using a completely different device and setup. I will try and measure with several devices and approachs when i can test the screen next week. Although i don't think we need to get too worried about differences of like 1ms or so to be honest
May not be noticable especially to untrained eye but measurements definitely show the input lag is relatively high compared to more "competitive" gaming monitors.
Here are few sources:
ALMOST Perfect - Alienware AW3423DW QD OLED Review
QD OLED Monitors are here! Can the Alienware AW3423DW live up to the hype?Shop here (soon!): https://geni.us/COcBnF1 or look at your local Dell page.Sponsor:...
www.youtube.com
The Best Monitor Ever? - Alienware AW3423DW QD-OLED Review
MSI SPATIUM SSD’s: https://msi.gm/MeetSPATIUMANZ Promotion: https://msi.gm/GetPoweredbyMSISupport us on Patreon: https://www.patreon.com/hardwareunboxedJoin ...
www.youtube.com
Man of Honour
Hmm interesting indeed, I guess against those gaming monitors it is measured higher, but in reality is it noticed? The pros alone far far far outweigh the measurements for input lag - And as mentioned I see no difference for input lag vs the Mateview GT - But the AW feels quicker because of the other specs/advancements.
I fully suspect that for non competetive gamers, this (input lag) will be a 100% non issue though as nobody will notice any difference.
I fully suspect that for non competetive gamers, this (input lag) will be a 100% non issue though as nobody will notice any difference.
Definitely agree that the pros outweigh any cons. 1st iteration of monitor will always be imperfect. The huawei input/response time is not great from what I can see so I wouldn't use that as a benchmark Like I said the only major drawback for me was the input lag but quicker response times reduces this negative effect. I am sure I will be hapy with the purchase especially at the discounted price. Just getting impatient
Man of Honour
Fair!
I just played some Res Evil The Village on this and man it's awesome, the perfect game to showcase what OLED is all about. Pitch black rooms only lit up with your torch is something else. Coupled with HDR and max settings, it is a great experience. Shame you have to faff with settings initially as it's all set up by default from a console gaming perspective.
I just played some Res Evil The Village on this and man it's awesome, the perfect game to showcase what OLED is all about. Pitch black rooms only lit up with your torch is something else. Coupled with HDR and max settings, it is a great experience. Shame you have to faff with settings initially as it's all set up by default from a console gaming perspective.
Ye I get you. PC is never a plug and play experience like consoles. I guess a pro of that is we get to have cutting edge technology first
OLED is great but QD OLED is the next level, I think QD OLED is better for pc monitors IIRC. Overall QD OLED is superior to OLED. BTW is HDR a good experience if the game does not natively support HDR?
Man of Honour
The Auto HDR feature in Windows that converts SDR games to HDR? I've not used the feature to be honest as have yet to see any evidence to support it being any good so don't want to bother with faffing around etc really.
Edit*
Wow so just noticed that if you enable the HDR mode in Res Evil The Village then it triggers the Windows HDR mode automatically when you launch the game, so there's no need to manually enable HDR in Windows for this game and then turn it off after gaming. It uses the Smart HDR feature of the display like how TVs work and how MPC-BE works with the MPC Video Renderer set to pass through HDR signal to display. This is the first game I have played that auto triggers HDR on launch. Excellent stuff!
Edit*
Wow so just noticed that if you enable the HDR mode in Res Evil The Village then it triggers the Windows HDR mode automatically when you launch the game, so there's no need to manually enable HDR in Windows for this game and then turn it off after gaming. It uses the Smart HDR feature of the display like how TVs work and how MPC-BE works with the MPC Video Renderer set to pass through HDR signal to display. This is the first game I have played that auto triggers HDR on launch. Excellent stuff!
Last edited:
Associate
- Joined
- 28 Sep 2018
- Posts
- 2,257
The input lag is something that's confuses me also. The monitor isn't doing any processing and we know the response times aren't holding it back. So I'm not sure what dell is doing to getting such a high figure. I certainly feel it's an area that could be improved via a FW update but Dell isn't known for those.
Last edited:
Soldato
- Joined
- 1 Jun 2014
- Posts
- 5,066
This is silly, but I'm very much looking forward to playing the Dead Space remake on this monitor. Should look super (and will scare the crap out of me).
Man of Honour
There's a remake??
Soldato
- Joined
- 1 Jun 2014
- Posts
- 5,066