Dell U2713HM

With 14/16 bit internal display LUT (EIZO/NEC) and 8 bit output, what you can get is 100% grayscale on variety of gamma curves
Click to expand...
Yes, there is no loss of tonal values and the possibility of precise gradation in combination with excellent grey balance.

What screen uniformity issues?
Click to expand...
Without active homogeneity compensations the panel area may suffer from unpleasing chroma or luminance differences. That underlies individual variability of course.

Best regards

Denis
 
Not really suprised that this budget segment display have the old LG LM270WQ1 panel in a slightly new revision. I guess we have to await a direct replacement for the U2711 before we get the new LM270WQ3 panel. Maybe early spring 2013 if we are lucky.
 
Peat Moss said:
What screen uniformity issues?

http://www.tftcentral.co.uk/reviews/dell_u2713hm.htm
Click to expand...

Review up already. Yay!

Sounds very good. Slightly better in most ways that will matter to consumers (better responsiveness, better contrast, more accurate sRGB, lighter AG coating).

The tradeoff is less connectivity for inputs most people don't use, loss of wide gamut/10 bit input, again unused by consumers.

Quote on the AG:

The screen coating on the U2713HM is a normal anti-glare (AG) offering. This is contrary to a lot of other screens using variants of the LM270WQ1 panel which offer a glossy screen coating. Readers will be pleased to hear though that the AG coating is actually nice and light and is not the usual grainy and aggressive solution you would normally find on an IPS panel. In fact in practice it is almost what you might call a semi-gloss coating being quite similar to AU Optronics AMVA offerings. Dell seem to have toned down the AG coating which is great news. It retains its anti-glare properties to avoid unwanted reflections, but does not produce an overly grainy or dirty image that some AG coatings can.
Click to expand...


IMO this sounds like a clear winner for most people.
 
Honestly - judging by reviews this is my dream monitor come true. It's like if Dell read all my posts about u2412m and created a monitor specially for me... Can't wait to see it personally.
 
I still don't get it, why can't you do that with an 8 bit LUT?
Click to expand...
Because many input values would be mapped to the same output value; just think of gradation or whitepoint adjustments – apart from that calculations (that are written into the LUT) wouldn't be as precise.

Best regards

Denis
 
Sailor_Moon said:
Because many input values would be mapped to the same output value; just think of gradation or whitepoint adjustments – apart from that calculations (that are written into the LUT) wouldn't be as precise.
Click to expand...
The calculations could and should be more precise without the table being wider than 8 bit. Could you give an example of what exactly would go wrong with an 8 bit table?
 
The calculations could and should be more precise without the table being wider than 8 bit
Click to expand...
No. The calculations are far more precise in a wider space. This is best practice not only regarding display LUTs of high end LCDs. An other example in the same context would be CMM transformations in the PCS.

Could you give an example of what exactly would go wrong with an 8 bit table?
Click to expand...
A display with 100% tonal range @Gamma 2.2-gradation would loose about 25% of the 8bit input signal when calibrated via a 8bit LUT to achieve the L* characteristic.

Best regards

Denis
 
Murzilka said:
Honestly - judging by reviews this is my dream monitor come true. It's like if Dell read all my posts about u2412m and created a monitor specially for me... Can't wait to see it personally.
Click to expand...

Unfortunately they missed the posts about input lag. The HP is still quite a bit faster.


Sailor_Moon said:
No. The calculations are far more precise in a wider space. This is best practice not only regarding display LUTs of high end LCDs. A example in the same context would be CMM transformations in the PCS.


A display with 100% tonal range @Gamma 2.2-gradation would loose about 25% of the 8bit input signal when calibrated via a 8bit LUT to achieve the L* characteristic.

Best regards

Denis
Click to expand...

I think we're talking about different things here.
 
I think we're talking about different things here.
Click to expand...
I don't think so but to clear it up: It makes sense to implement a LUT far more precise than the input signal to avoid a loss of tonal values and to raise the precision of the tonal transformations. Of course you don't raise the tonal range above the input signal, there is no interpolation of new colors which aren't in the input signal.

Best regards

Denis
 
So how does that work at the tech / math level?
Input and output are both an 8 bit value, right?
Then a single table with 256 8-bit entries can provide any mapping you want.
 
I have described it above. Think of the Gamma 2.2 - L* adjustment. Many tonal values of the input signal would be transformed to the same output value in this case because the quantisation steps are way too coarse. Little example for the highly simplified scenario I mentioned @10bit precision:

RGB(127/127/127) => RGB (118/118/118)
RGB(128/128/128) => RGB (118/118/118)
RGB(129/129/129) => RGB (120/120/120)

RGB(127/127/127) => RGB (474/474/474)
RGB(128/128/128) => RGB (476/476/476)
RGB(129/129/129) => RGB (478/478/478)

With a >8bit LUT you can map to different values between one 8bit quantisation step. A dithering stage at the end (scaler or panelinternal) helps to "rescue" the tonal range.

Best regards

Denis
 
Last edited:
Sailor_Moon said:
A dithering stage at the end (scaler or panelinternal) helps to "rescue" the tonal range.
Click to expand...

Ah! Output is not 8 bit.
Is dithering really used for this? I thought it degraded other aspects of quality.

If dithering is indeed being used, wouldn't the videocard be able to take care of it?
 
Looks like a pretty good monitor overall. I'm disappointing with the lag - it's not awful but near 1 frame puts it in territory that I will notice. HP or others will hopefully make good use of this light AG panel and make a better gamers display with less lag.

Best news so far, it has popped up on a few German retailers website at a full £120 less than PCBuyIT. I expected U2713HM to be ~£450 and was hoping it might be obtainable for <£400. It just might be in the run up Christmas and in the New Year when sales and promos are aggressive.

NCX said:
Circuit delay=input lag
Click to expand...
I don't agree. Circuit delay is unambiguous and a term used in industry, whereas input lag is a term coined by gamers posting on forums, it isn't described or measured consistently.

Snowdog said:
A 30" U3013, with sRGB native...
Click to expand...
That's a nice thought and it could happen. <8ms latency would make it a winner for me, but the scaler and image processing is probably more useful rom Dell's perspective.
 
If dithering is indeed being used, wouldn't the videocard be able to take care of it?
Click to expand...
One could think of such solutions but when calibrating via the 8bit videocard LUT you will experience a "hard cut" with increasing banding the more corrections are calculated. My external videoprocessor from Lumagen with extensive CMS uses at least spatial dithering for 8bit output.

Ich spreche keine deutsch.
Click to expand...
I haven't found the whitepaper in english but the diagram is quite helpful.

Best regards

Denis
 
too slow, fuck it.

@Sailor_Moon: i know this is off-topic but you really have to write "Best regards, Denis" on every single post? put it in your sig then, this is annoying.

Best regards

D4rkn3ss
 
i know this is off-topic but you really have to write "Best regards, Denis" on every single post? put it in your sig then, this is annoying.
Click to expand...
Consuetudo est altera natura. I will take this into consideration.
 
XTF said:
Unfortunately they missed the posts about input lag. The HP is still quite a bit faster.
Click to expand...
Um, yeah. But the samsung s27b970 was measured as having 29ms of input lag, and it seemed to be very responsive in practice. This one is measured to have 22ms of input lag which is even faster so that's very OK by me.

Best regards

Murzilka
 
Im glad they got rid of the PWM. Looks like its either U2713HM or the Samsung S27B970 for potential customers with sensitive eyes shopping for the latest display.
 
Murzilka said:
Um, yeah. But the samsung s27b970 was measured as having 29ms of input lag, and it seemed to be very responsive in practice. This one is measured to have 22ms of input lag which is even faster so that's very OK by me.
Click to expand...

Not only that but the HP has much more ghosting, almost as if it doesn't really have active overdrive.

It often looks like there is a tradeoff between good overdrive circuitry and lag. Almost all PVA/MVA lag a fair bit, but there seems to be a few that lag minimally, but they tend to have fairly horrible ghosting.

I had a Dell 2405 that must have had horrendous lag because that I could feel it was like everything was moving through molasses(though it also had tons of ghosting too). Then I got my NEC which has about 33ms of lag. I don't feel that at all.

Here we get 22ms lag, but with very good overdrive. Only the 120Hz TN panels beat it in the ghosting/blur tests. This is an awesome overall result for a good all round IPS screen.

Reposting the review link as it became buried in the LUT discussion.
http://www.tftcentral.co.uk/reviews/dell_u2713hm.htm
 
Last edited:
Murzilka said:
s27b970 was measured as having 29ms of input lag
Click to expand...
Murzilka said:
This one is measured to have 22ms of input lag
Click to expand...

TFT Central's lag tests say:
S27B970 - 28ms (faster RTC mode) & 24.4ms (fastest RTC moded)
U2713HM - 22.2ms
​
Prad's (I'm using 2nd hand info from forum members here)
S27B970 - 13.6ms circuit delay, 19.3ms overall latency
U2713HM - 15ms circuit delay, 22ms overall latency
​
So Prad test Samsung as being faster by an insignificant amount.


Murzilka said:
Best regards

Murzilka
Click to expand...
off topic to talk about signatures, but since it has come up already perhaps it's worth pointing out that aside from good etiquette, the forum rules ask for post signing to be confined to the signature box:

The [H]ard|FORUM Rules

(25) You are allowed one signature per post, do not put your signature in the body of your post. A signature section is supplied for that usage. No one wants to see your name on your posts multiple times as it is annoying.
Click to expand...
 
Snowdog said:
Not only that but the HP has much more ghosting, almost as if it doesn't really have active overdrive.

It often looks like there is a tradeoff between good overdrive circuitry and lag. Almost all PVA/MVA lag a fair bit, but there seems to be a few that lag minimally, but they tend to have fairly horrible ghosting.

I had a Dell 2405 that must have had horrendous lag because that I could feel it was like everything was moving through molasses(though it also had tons of ghosting too). Then I got my NEC which has about 33ms of lag. I don't feel that at all.

Here we get 22ms lag, but with very good overdrive. Only the 120Hz TN panels beat it in the ghosting/blur tests. This is an awesome overall result for a good all round IPS screen.
Click to expand...

mmm.. thanks, i guess i'll wait a little longer to decide
 
Has anyone looked into the marketing/price point behind this? They release this model a few months (I assume) before the replacement for the 2711 but what is sinister is that they price this one ABOVE the 2711 knowing people are looking for a new screen from dell. This means, when the superior, successor to the 2711 does arrive, they will pitch that at a much higher price than its predecessor and the shift in price will appear acceptable or normal due to the 2713's raised price.

And all this at a time where prices, generally are going down.. Bad move Dell. It's not as if you have made an amazing leap forward here.

I'm still interested somewhat, but I dislike the price and strategy behind it.

Still using my 2407 happily after 5 years. No denial on the quality.
 
Murzilka said:
Most unlikely. -)
Click to expand...

Any reason for that? HDMI 1.4 supports 4k resolutions, and I know some of the Korean panels (the microcenter one in particular) can be driven at 2560 by HDMI. Considering this is a fairly modern monitor, I think it's plausible.
 
TerraPhantm said:
Any reason for that? HDMI 1.4 supports 4k resolutions
Click to expand...
At 24Hz, yes...

TerraPhantm said:
and I know some of the Korean panels (the microcenter one in particular) can be driven at 2560 by HDMI. Considering this is a fairly modern monitor, I think it's plausible.
Click to expand...
Is it actually outputting 2560x1440 over the HDMI or is it being scaled by the monitor? I'm not saying it isn't doing it, I'm just curious if you know for sure that it is properly running at 1440p over HDMI. I have had... limited success using HDMI for almost anything :(
 
TerraPhantm said:
Any reason for that? HDMI 1.4 supports 4k resolutions, and I know some of the Korean panels (the microcenter one in particular) can be driven at 2560 by HDMI. Considering this is a fairly modern monitor, I think it's plausible.
Click to expand...

I'm not sure but from what I understand the hdmi supports 4k resolution at 24hz refresh rate, but it will not not support the 2560x1440 @ 60hz.
Sn0_Man beat me to it:D
 
Sn0_Man said:
At 24Hz, yes...


Is it actually outputting 2560x1440 over the HDMI or is it being scaled by the monitor? I'm not saying it isn't doing it, I'm just curious if you know for sure that it is properly running at 1440p over HDMI. I have had... limited success using HDMI for almost anything :(
Click to expand...



Murzilka said:
I'm not sure but from what I understand the hdmi supports 4k resolution at 24hz refresh rate, but it will not not support the 2560x1440 @ 60hz.
Sn0_Man beat me to it:D
Click to expand...

I was definitely getting 2560x1440 - no scaling whatsoever. Picture was identical to what I got over displayport. HDMI definitely supports it - the signal is 340MHz - should allow for slightly higher resolutions than dual links at 165MHz. Admittedly there aren't many devices that support it. But the $400 micro center monitor definitely does. So I was hoping someone tried on this dell

Some simple math would show 4k at 24hz needs about the same bandwidth than 2560x1440 at 60hz. Under HDMI's limit either way at 8-bit color.

Edit:
Found this - U2711 running at 2560x1440@60Hz over HDMI by using a hacked driver. So it is plausible that the U2713 natively supports it. Maybe

XH85lLBaCeh.jpg
 
Last edited:
TerraPhantm said:
I was definitely getting 2560x1440 - no scaling whatsoever. Picture was identical to what I got over displayport. HDMI definitely supports it - the signal is 340MHz - should allow for slightly higher resolutions than dual links at 165MHz. Admittedly there aren't many devices that support it. But the $400 micro center monitor definitely does. So I was hoping someone tried on this dell

Some simple math would show 4k at 24hz needs about the same bandwidth than 2560x1440 at 60hz. Under HDMI's limit either way at 8-bit color.

Edit:Found this - U2711 running at 2560x1440@60Hz over HDMI by using a hacked driver. So it is plausible that the U2713 natively supports it. Maybe

XH85lLBaCeh.jpg
Click to expand...

Scalers between U2711 and U2713 are not exactly the same.
 
TerraPhantm said:
Any reason for that? HDMI 1.4 supports 4k resolutions, and I know some of the Korean panels (the microcenter one in particular) can be driven at 2560 by HDMI. Considering this is a fairly modern monitor, I think it's plausible.
Click to expand...

Because video cards generally use HDMI 1.2 and not 1.4. Seriously, WHY would you want to use HDMI on a freaking PC monitor? Thats like using VGA on a pc monitor. Its for bigscreens and consoles - not good displays. I cringe every time I hear of someone wanting to use that crap on a 2560x screen.

Displayport is the best connectivity option for montiors anyway - has more bandwidth / features than any HDMI spec.
 
Last edited:
You must log in or register to reply here.
Back
Top