About a XL2720Z Review [and XL2720Z strobe timing bug]

[lol37]

Hello everyone !

There is a french XL2720Z review here : http://www.lesnumeriques.com/moniteur-e ... /test.html, it claims the motion blur technology is based upon a scanning method, is that right ? i thought it were a strobing one
i also got another question about PWM, why it's used on strobing ? one wants as much brightness as possible to compensate brightness loss, PWM is used for dimming right ? that's stupid !
oh, and they also claim there's much crosstalk on 3D mode, will it affect the motion blur reduction ? or it only accounts for lightboost hack ?

(I made it quote because i wrote it already to Mark, but he doesnt like my selfish behavior so here is it )

A partial answer

Yes, they are using the buggy firmware.
PWM is not only used for dimming (Wikipedia). It's Pulse-Width Modulation.
Strobing is the same thing as PWM, except we don't call it PWM because it is a nicer low-frequency one-flash-per-refresh PWM that eliminates motion blur -- including LightBoost.

Strobe crosstalk is exactly the XL2720Z bug I'm talking about.

[Chief Blur Buster]

For other readers, see:
Strobe Timing Bug in XL2720Z

(I made it quote because i wrote it already to Mark, but he doesnt like my selfish behavior so here is it )

Oh, it's simply because I get too many emails -- I prefer people asking questions publicly.
It's more effective to educate 1000 people doing 10 detailed public replies rather than 1000 private messages/emails. Y'know what I mean?

Strobing is the same thing as PWM, except we don't call it PWM because it is a nicer low-frequency one-flash-per-refresh PWM that eliminates motion blur -- including LightBoost.

I've recorded this as part of the LightBoost FAQ at http://www.blurbusters.com/lightboost/faq/ -- "Q: How is LightBoost different from PWM?"

Strobing, by its nature, is PWM. Simple black frame insertion (via backlight) is a very simple form of PWM that is synchronized to the refresh rate, not used for dimming, but used for motion blur elimination. PWM on Wikipedia).

[Chief Blur Buster]

There is a french XL2720Z review here : http://www.lesnumeriques.com/moniteur-e ... /test.html, it claims the motion blur technology is based upon a scanning method, is that right ? i thought it were a strobing one

Sometimes some people use "scanning" and "strobing" interchangeably. However, it is all-at-once strobing, confirmed by my high speed camera.

[lol37]

i see, if one choose in the new firmware 1ms strobing length, luminance would be 125/4 cd/m² = 31.25 cd/m² ( according to the review )
that's really low, what are the main cause of brigthness limit (i heard about liquid cristal + color filter light filtering) ? is that actual edge-lit LED actually reached their limit ?

Haha, it isn't so smart replying back in english at a french website !
i'm french so i could help you convert it in french without problems

[Chief Blur Buster]

i see, if one choose in the new firmware 1ms strobing length, luminance would be 125/4 cd/m² = 31.25 cd/m² ( according to the review )

The default persistence of most strobe backlights seem to go around 2ms, for this reason, which results in ~60cd/m2. Sometimes voltage boost is used during strobing (it's done on some models of LightBoost monitors, such as the XL2411T).

Haha, it isn't so smart replying back in english at a french website !
i'm french so i could help you convert it in french without problems

Please do. You can either post French replies, or I can edit the posts.

[lol37]

i can't believe you made that mistake coming for you
60cd/m² isnt an half bright than 120cd/m²
i know you did it on purpose, just to make my current post be!

[Chief Blur Buster]

i can't believe you made that mistake coming for you
60cd/m² isnt an half bright than 120cd/m²
i know you did it on purpose, just to make my current post be!

2ms persistence is actually roughly 60cd/m2 on most 27" LightBoost displays.
A similiar display, TFTCentral LightBoost brightness measurements for VG278HE:

LightBoost 10% = 51.52cd/m2
LightBoost 20% = 55.22cd/m2
LightBoost 30% = 59.00cd/m2
LightBoost 40% = 62.64cd/m2
LightBoost 50% = 66.18cd/m2
LightBoost 60% = 69.48cd/m2
LightBoost 70% = 73.13cd/m2
LightBoost 80% = 76.66cd/m2
LightBoost 90% = 80.23cd/m2
LightBoost 100% = 87.14cd/m2

However, someone made a modification to make LightBoost=10% (1.4ms) much brighter; see chapter within Electronics Hacking: Creating Strobe Backlight on modifying the boost voltages for brighter strobing.

It varies quite a bit from strobe backlight to strobe backlight, EIZO Turbo240 manages to be about 250cd/m2 in strobed mode, which shows a healthy big voltage boost is being used in EIZO's strobing.

[lol37]

obviously one thing is wrong, either the strobe length or the luminance display
2 times brighter than 50cd/m² is 2²*50cd/m² = 4*50cd/m² (since of the square and not linear luminance, it is squared proportional), that's how maths and physics work !!!
if we apply this on the lightboost thing, then the proportional strobe scaling between 10% and 100% is sqrt(87.14/51.52)
i highly doubt it's equal to 2.4/1.4
i might guess the strobe voltage isn't as sharp as a square, PWM based signal

PS : i've translated both of your messages !

[Chief Blur Buster]

obviously one thing is wrong, either the strobe length or the luminance display

Sometimes LED voltage can change during persistence (strobe length) change, which can make LED brighter.
This is a new, additional, extra variable.

Example of voltage boost (overclock voltage, raise voltage, bigger voltage, boost voltage):
CREE: Pulsed Over-Current Driving of Cree® XLamp® LEDs: Information and Cautions.

Persistence is not a synonym of brightness.
With LED, you can increase voltage when shortening strobe length (shorten persistence), to make strobe brighter to compensate.
Not all strobe backlights can do this, but some do.

The XL2720Z, however, does not seem to use any voltage boost, so it is dimmer than the 24" LightBoost monitors at similiar persistence.

If voltage remains constant, it is correct.
Assuming constant voltage, brightness linearly scales with persistence adjustment.

[Chief Blur Buster]

PS : i've translated both of your messages !

Merci!!