HTC One screen reviews: One comes with different display types

puremind · Mar 23, 2013

HTC One Display Brightness

The HTC One uses an adaptative contrast mechanism for the screen brightness that is in operation even when auto-brightness is off. This mechanism means that brightness settles after each picture change only after a 4-5 seconds, which means that measurements that are taken too quickly can overestimate brightness by 15%, which is why there is a wide fluctuation across all reviews.
At which level brightness settles depends on the average picture level. You can sometimes observe this while browsing: when the content displayed varies between dark and bright content, an adjustment is visible.
In my measurements below, I have indicated the brightness range as well as the typical settled brightness for the standard 100% Voodoo test pattern.

HTC One Black Levels

Like with brightness, black levels also fluctuate due to the adaptative contrast mechanism:

Content with darker average picture level will be dimmed
Content with brighter content will will brightened
Overall black levels will fluctuate between 0.29cd/m² and 0.55cd/m², so dynamic contrast will be around 1700:1 whereas ANSI (or intra-picture contrast) will be around 1000:1.

Color Temperature: Three phones, three different display calibrations.

Unit 1 (silver)

Fabricated on 12th March 2013 at factory Tierra del Fuego Factory in Argentina (Serial: FA33CWxxxxxx, Panel ID JDI C2_2)
Display: color temperature slightly on the cool side

White color temperature: 6950K (some meters will incorrectly measure this up to 350K higher, especially if not using the right spectral file for the type of display)
Black Level: 0.37cd/m² to 0.50 cd/m²
Black Level (typical voodoo test pattern): 0.43 cd/m²
White Point: 370cd/m² to 490cd/m²
White Point (typical voodoo test pattern): 435cd/m²
Typical Contrast (without adaptative mechanism): 1000:1
ANSI Contrast: 1100:1

Unit 2 (silver)

Fabricated on 3rd April 2013 at Hsinchu, Taiwan factory (Serial: HT343Wxxxxxx, Panel ID JDI C2_2)
Display: neutral color temperature, slightly lower contrat:

White color temperature: 6550K (some meters will incorrectly measure this up to 350K higher, especially if not using the right spectral file for the type of display)
Black Level: 0.50 cd/m² to 0.64cd/m²
Black Level (typical voodoo test pattern): 0.52 cd/m²
White Point: 400cd/m² to 530cd/m²
White Point (typical voodoo test pattern): 466cd/m²
Typical Contrast (with adaptative mechanism): 900:1
Typical ANSI Contrast: 970:1

Unit 3 (black)

Fabricated on 12th April 2013 at factory Tierra del Fuego Factory in Argentina (Serial: FA34CWxxxxxx, Panel ID JDI C2_2)
Display: Even cooler color temperature unit 1 but same brightness and contrast

White color temperature: 7200K (some meters will incorrectly measure this up to 350K higher, especially if not using the right spectral file for the type of display)
Black Level: 0.37cd/m² to 0.50 cd/m²
Black Level (typical voodoo test pattern): 0.43 cd/m²
White Point: 370cd/m² to 490cd/m²
White Point (typical voodoo test pattern): 435cd/m²
Typical Contrast (without adaptative mechanism): 1000:1
ANSI Contrast: 1100:1

Comparison of warmest color temperature on Unit 2 (left) vs. coolest on Unit 3 (right):

Based on the collective reviews of the HTC One including my own measurements, it is now clear that HTC One displays can have different display characteristics in terms or color temperature, contrast and brightness. Even displays with the same Display ID have can have widely fluctuating brightness and color temperature depending on factory calibration.

Settled brightness between 390cd/m² and 475cd/m²
Color Temperature between 6550K to 7350K
ANSI contrast between 930:1 and 1100:1

Currently available reviews for the HTC One Display

Neutral Color Temperature

20.03.2013 GBR pcpro: 481cd/m² brightness | 1202:1 contrast (measured with i1 Display Pro)
20.03.2013 DEU Computerbase.de: 483cd/m² brightness | 1100:1 contrast | 6,600K white temperature (measured with DTP94)
26.03.2013 FRA 01net.fr: 485 cd/m² | 0,29 cd/m² | 1679:1 contrast | 6638K white temperature (measured with Minolta CA-210 and DTP94)
26.03.2013 FRA Puremind@xda (see below): 466 cd/m² | 0,52 cd/m² | 939:1 contrast | 6650K white temperature (measured with i1 Pro 2)

Medium-High Color Temperature

26.03.2013 DEU puremind@xda 463cd/m² brightness | 1073:1 contrast | 6934K white temperature (measured with i1 Pro)
02.04.2013 DEU notebookcheck.com: 488.9cd /m² brightness | 0.23 cd / m² black level | 2117:1 contrast | 7205K white temperature
26.02.2013 RUS 3dnews.ru: 382cd/m² and 1,258:1 contrast | 6,938K white temperature (measured with Spyder 4)

High Color Temperature

14.03.2013 GBR uk.hardware.info: 426cd/m² brightness | 977:1 contrast | 7,820K color temperature (measured with i1 Display Pro)
26.02.2013 RUS 3dnews.ru: 453cd/m² and 1,328:1 contrast | 7,598K white temperature (measured with Spyder 4)
14.03.2013 NLD Tweakers.net: 489cd/m² | 1159:1 contrast | 8,166 color temperature (measured with i1 Display Pro)
26.03.2013 DEU puremind@xda 420cd/m² brightness | 1060:1 contrast | 8,130K color temperature (measured with Chroma 5)
28.03.2013 RUS high-tech@mail.ru 492cd/m² brightness | 1189:1 contrast | 7980K white temperature

Unknown color temperature:

14.03.2013 USA laptopmag: 463cd/m² brightness
19.03.2013 DEU Notebookjournal: 385cd/m² and 1100:1 contrast
19.03.2013 DEU PC Welt 447cd/m² brightness | 1711:1 contrast
20.03.2013 DEU Chip.de: 479cd/m² brightness | 1020:1 contrast
21.03.2013 FRA Les Numériques 460cd/m² | 1568:1 (measured with i1 Pro or i1 Pro 2 tbc.)
21.03.2013 FRA 01.net 480cd/m² brightness | 1,655:1 contrast (measured with Konica Minolta CA-210)
23.03.2013 BGR GSMArena: 647cd/m² brightness | 1541:1 contrast

Having said that, there are still different display calibrations out there, I myself measured "settled" brightness
I will try and keep this updated with new tests to confirm where each screen type can be found.

For reference, here are the luminance ranges of the color testing devices listed above.

i1 pro__________________________ 0.20 cd/m² to 300 cd/m²
i1 pro 2_________________________0.20 cd/m² to 1200 cd/m²
Chroma 5/Sencore Color Pro V_______0.01 cd/m² 1000 cd/m²
Spider 3/4_______________________0.02 cd/m² 5000 cd/m²
i1 Display 2______________._____._.__0.02 cd/m² 3000 cd/m²
i1 Display 3/i1 Display Pro/C6___._____0.003 cd/m² 1200 cd/m²
Konica Minolta CA-210______________0.01cd/m² 1000 cd/m²

puremind · Mar 23, 2013

Own screen measurements

Here are measurements I conducted today at a Vodafone shop.

White Balance
First of all, let me say that the readings were conducted by i1 Pro, which is certified until 300cd/m². Overall the readings will be very accurate for color temperature, black level, gamma and chromaticty, but maximum brightness may be off by a small margin (to be confimed tomorrow). Unfortunately Calman did not respond to me regarding my license upgrade to Calman5, so I could not use the Chroma5 for this measurement.

I will conduct new readings with the Chroma 5 likely tomorrow. The results are quite close to the PCpro review in terms of brightness and very close to the Russian review in terms of white color temperature (in fact nearly identical). Overall contrast falls in line with most reviews.

White point / maximum brightness: 445cd/m²
Back Level: 0.43cd/m²
Contrast (dynamic): 1034:1
Average gamma: 2.1
White color temperature (read the color tmperature section below): 6934K (same as Russian Review)
Black color temperature (read the color tmperature section below): 8882K
Average color temperature: 7353K

Overall the white balance is good but not perfect.

Color Temperature

The observed white balance translates into color temperature slightly on the cold side but it is colder at low stimulus and warmer towards the white point, which is important for a pleasing browsing experience:

I think displaying color temperature across the whole spectrum is important. Some reviews only mention one value, which can be the average or the white point temperature, but it is hard to interpret taken on its own.

Color decoding / Chromaticity
As far as color accuracy is concerned, the HTC One's display I tested had fairly accurate colors. Only the most critical viewers will detect thes imperfections in daily use.

Based on this measurement, this is more of a type 2 display, however I am not sure about the white point

BarryH_GEG · Mar 23, 2013

Sites and reviewers test differently. I think if you look at the One X's performance you'll see it varied from site to site too.

Here's GSMArena's testing methodology. You can see how other sites might test differently and get different results because of it. It's more accurate to compare multiple devices on a single site than it is a single device across multiple sites. Assuming each site uses the same standardized tests a One X to SGS3 comparison on tweakers.net would be more relevant than a One X's performance off tweakers.net compared to a SGS3's peformance off GSMArena.

An important note about AMOLEDs is due here. As AMOLED units have the ability to completely switch off individual pixels, their black level readings are 0, which gives them an infinite contrast ratio under the testing conditions.

When we measure we take two readings off each device - first with the display brightness set to 50% and then with the brightness setting pushed all the way up. We test the handsets in complete darkness, because when ambient light is present, the luminance levels of the blacks displayed goes up, and affects the perceived contrast ratio.

Contrast ratio is very heavily influenced by the black levels of a display. While a brighter display would normally have an advantage, it will usually be unable to compensate for insufficiently deep blacks.

Our sunlight legibility test aims to show you how legible each screen remains in bright environments, where screen reflectivity matters as much as its natural contrast and brightness. We use fixed studio lighting to simulate sunlight falling on the phone screens and measure the contrast ratio of each of them, when faced with this powerful light source. We measure each display with brightness turned up to 100%.

jasahu · Mar 23, 2013

Re: HTC screen reviews: HTC One comes with different display types

Gsmarena is Bulgarian, not American

BarryH_GEG · Mar 23, 2013

Here's a way to find the display type that doesn't require root.

1. Reboot the phone twice to get a clean "last_kmsg" file
2. Use any file manager that gets you to the root directory
3. Navigate to the /proc folder
4. Find "last_kmsg"

\

5. Open it on the device or send it to a PC and open it as a text file. Your looking for "panel_id" and "panel_vendor." It's a long ass file so you're better off using an editor with a search feature to search on "panel."

Since there aren't many vendors making 1080P panels I'll bet all One's use the same panel type and it's probably from Renasis (Sharp) like the XZ and DNA.

The AUO (Acer) display on the Teg3 One X is "0x4940014"
The Sharp display on the Teg3 One X is "0x294000f"
The Sony display on the One XL is"0x18103" (The One XL only uses the Sony panel)
The AUO (Acer) display on the One X+ is 0x944a03 (The One X+ only uses the AUO panel)
The DNA uses a Renasis (Sharp) panel (only) but I can't find the ID

puremind · Mar 23, 2013

jasahu said:
Gsmarena is Bulgarian, not American

Thanks I was looking for that information yesterday but could not find it. Do you know if this is the parent or did the testers also receive a Bulgarian unit?

puremind · Mar 23, 2013

BarryH_GEG said:
Sites and reviewers test differently. I think if you look at the One X's performance you'll see it varied from site to site too.

Here's GSMArena's testing methodology. You can see how other sites might test differently and get different results because of it. It's more accurate to compare multiple devices on a single site than it is a single device across multiple sites. Assuming each site uses the same standardized tests a One X to SGS3 comparison on tweakers.net would be more relevant than a One X's performance off tweakers.net compared to a SGS3's peformance off GSMArena.

I agree with you with regards to black levels but not for brightness.

Brightness cannot move by more than 10% depending on measurement device or testing protocol
Calibration usually never reduces brightness either by more than 15%, except of course if you are calibrating an AMOLED to neutral color temperature.

The fact that there are several display types makes no doubt to me, as HTC like many others used that approach in the past and the brightness results are too different.

puremind · Mar 23, 2013

BarryH_GEG said:
Here's a way to find the display type that doesn't require root.

1. Reboot the phone twice to get a clean "last_kmsg" file
2. Use any file manager that gets you to the root directory
3. Navigate to the /proc folder
4. Find "last_kmsg"

\

5. Open it on the device or send it to a PC and open it as a text file. Your looking for "panel_id" and "panel_vendor." It's a long ass file so you're better off using an editor with a search feature to search on "panel."

Since there aren't many vendors making 1080P panels I'll bet all One's use the same panel type and it's probably from Renasis (Sharp) like the XZ and DNA.

The AUO (Acer) display on the Teg3 One X is "0x4940014"
The Sharp display on the Teg3 One X is "0x294000f"
The Sony display on the One XL is"0x18103" (The One XL only uses the Sony panel)
The AUO (Acer) display on the One X+ is 0x944a03 (The One X+ only uses the AUO panel)
The DNA uses a Renasis (Sharp) panel (only) but I can't find the ID

Great stuff I will go to the shops and check it out. Why can't you find the ID on the DNA?

Maedhros · Mar 23, 2013

I made a thread about this earlier, but I only got troll responses.

Hope this thread receives actual input, this is kind of an important thing for me.

jonstatt · Mar 23, 2013

Hate to throw a spanner in the works here, but it may also be that different firmwares are producing different default colour temperatures. With the One X there were indeed three manufacturers of screen. For the European models it was Acer and Sharp. Acer had been colour fidelity, but less uniform, and not as bright. The sharp was a light cannon, but very cool colour temperature, and better uniformity. However, HTC reduced the disparity through firmware updates!

Therefore, the only true way is if we can find a reliable way of finding the vendor code such as in the log file.

Also I can tell you that I had a Sony Xperia last year. Even though all the screens are made by Sony, there was a huge variance between two supposedly identical phones, as I found out when I compared to a friends.

---------- Post added at 01:35 PM ---------- Previous post was at 01:30 PM ----------

BarryH_GEG said:
The AUO (Acer) display on the Teg3 One X is "0x4940014"
The Sharp display on the Teg3 One X is "0x294000f"
The Sony display on the One XL is"0x18103" (The One XL only uses the Sony panel)
The AUO (Acer) display on the One X+ is 0x944a03 (The One X+ only uses the AUO panel)
The DNA uses a Renasis (Sharp) panel (only) but I can't find the ID

Absolutely spot on accuracy here. I confirmed the One X+ my wife bought had an identical characteristic to the One X Acer screen I had before. However they seem to have improved the uniformity issues from my original One X.

I also have a suspicion that at least my One uses a Sharp panel and has a number of characteristics similar to the Sharp screen used on the One X. It has a colour temperature that is a bit on the cool side for sure. Although I haven't measured yet, I have a keen eye, and I would have said it was a little over 7000K. I also compared it to my wife's One X+ and it is a bit cooler, but not rediculous (nothing like the S3 which was wildly off). We may find a future firmware update corrects this if the variance between One's with sharp panels is not too severe.

puremind · Mar 23, 2013

jonstatt said:
Hate to throw a spanner in the works here, but it may also be that different firmwares are producing different default colour temperatures.

Even if it were true, battery life will still fluctuate depending on the screen type you have, and it is something end users have no control over once the device has been purchased, so may as well test his before purchasing...

BarryH_GEG · Mar 23, 2013

puremind said:
Even if it were true, battery life will still fluctuate depending on the screen type you have, and it is something end users have no control over once the device has been purchased, so may as well test his before purchasing...

You're sort of on a witch hunt. HTC hasn't used two panel vendors on a single device since the Teg3 One X. When they did, none of the characteristics of the phone (EG: power draw, benchmarks) were any different. You're basing all your assumptions on tests conducted by eight different reviewers, using eight different methodologies, and using eight different sets of conditions.

Here's how the Teg3 One X fared in some of those same reviewer's tests. They were all testing a One X with the AUO panel because HTC didn't provide review samples that had the warmer Sharp panel (for good reason). And from being on the One X forum I assure you, as much as that phone was picked apart with a microscope, it there were any impacts to anything but color temperature because of the two different displays it would have been on the front page.

And if you performed the same analysis you did to reach your conclusion on any phone you'd see similar deviations. The deviations are there but it's differences in the reviewer's testing, not the devices themselves. By all means carry on but I think when people begin posting their panel ID’s it’ll most likely show a single display type.

puremind · Mar 23, 2013

BarryH_GEG said:
You're sort of on a witch hunt. HTC hasn't used two panel vendors on a single device since the Teg3 One X. When they did, none of the characteristics of the phone (EG: power draw, benchmarks) were any different. You're basing all your assumptions on tests conducted by eight different reviewers, using eight different methodologies, and using eight different sets of conditions.

Here's how the Teg3 One X fared in some of those same reviewer's tests. They were all testing a One X with the AUO panel because HTC didn't provide review samples that had the warmer Sharp panel (for good reason). And from being on the One X forum I assure you, as much as that phone was picked apart with a microscope, it there were any impacts to anything but color temperature because of the two different displays it would have been on the front page.

And if you performed the same analysis you did to reach your conclusion on any phone you'd see similar deviations. The deviations are there but it's differences in the reviewer's testing, not the devices themselves. By all means carry on but I think when people begin posting their panel ID’s it’ll most likely show a single display type.

Please be honest about this. It is not even debatable that maximum brightness has an impact on battery life. Just to prove my point. Take this test:
http://www.computerbase.de/artikel/handys/2013/htc-one-im-test/5/

Battery time at maximal brightness (383cd/m²): 277 min (4:37)
Battery time at 200cd/m²: 322 min (5:22)

That's 16% more battery for a 58% decrease in brightness. Therefore, if you do the maths, this means an 11% difference in battery performance for web browsing/HD Video between the 650cd/m² model and the 385cd/m² model.

If you apply this to the current GSM Arena battery charts, which were created based on the top brightness model (650cd/m²), you can see that the HTC One climbs further up on the chart.

And this factor is even an underestimation for the GSM charts, because they measure standard definition video as opposed to HD, so the impact of screen brightness in their testing environment should be even more, probably around 20% between the two devices.

BarryH_GEG · Mar 23, 2013

puremind said:
Please be honest about this. It is not even debatable that maximum brightness has an impact on battery life. Just to prove my point.

Do you know what brightness settings each of the reviewer’s used when they tested battery life? I'm willing to bet it's not maximum brightness and that all the phones each of them test are set to some equivalent level that the reviewer applies consistently.

And what you're saying is debatable. Each of the reviewer's data you're quoting is correct; based on their testing methodology. What I'm saying is that the same display is performing within 10+/-% (normal deviation) and that 382cd/m2 from 3Dnews is the same as 647cd/m2 from GSMArena. The display's not performing any differently, it's how their testing that's causing variations. The swings in the One X comparative I provided show that. Do you honestly believe if the delta on One X displays at max brightness was 386 vs. 550 that the owners of the less bright display(s) wouldn't notice? There are 100 pages discussing the difference between the warm Sharp panel vs. the cool AUO panel so it’s not as if owners aren’t conscious of their display’s performance.

So you’re taking something that’s meant to be viewed vertically (EG: comparing multiple devices performance against each other on GSMArena where the same methodology was used) and trying to make it horizontal by taking a single device's results from eight different reviewers and comparing them. It can't work because, again, they all use different methodologies for each of their tests. No analysis based on an undefined or inconsistent baseline can ever be compared. There are just too many variables that can (and do) skew the results. And here’s how to prove this. If you ran the same analysis across the same eight sites for the XZ you’d get a similar result to what you got for the One. Same for any other phone. If what you’re representing were true every device produced would have three display types and wild swings in battery life based on huge variations in brightness. We know that’s not the case because it would be all over each devices respective XDA forum.

Here's one of GSMArena's battery tests.

The web browsing test is performed using an automated script which reloads a webpage every ten seconds. There are no flash elements on the web pages, so the playing field is even. The brightness of the phones' displays is set to 50% and we use a Wireless N router placed a few meters away to get full connectivity bars.

puremind · Mar 23, 2013

BarryH_GEG said:
You're sort of on a witch hunt. HTC hasn't used two panel vendors on a single device since the Teg3 One X. When they did, none of the characteristics of the phone (EG: power draw, benchmarks) were any different. You're basing all your assumptions on tests conducted by eight different reviewers, using eight different methodologies, and using eight different sets of conditions.

No disrespect, but you don't seem to be intimately familiar with screen calibration. The difference between the top level of observed brightness and the lowest one is huge. Not difference is testing conditions, protocol or testing devices can explain such a difference (650cd/m² to 385cd/m²).

If testing devices had such error margins, it would be a well known fact amongst professional calibrators, and no one would hire their services because essentially eye calibration would not be any less accurate than measured calibration...Please be serious.

As someone who had calibrated many displays, and repeated measurements with different measurement devices and under different conditions, please trust me when I say that brightness measurements are very stable and reliable, using even the cheapest colorimeters (chromaticity and black levels are more sensitive to the device used but brightness is easy).

LadFromWales85 · Mar 23, 2013

This thread makes me nervous. When the One X come out, I had access to several handsets at once, all of which displayed different tints when on boot screen or recovery screen. Some white some yellowed.

Is this the same with the One? Are there cool and warm screens?

BarryH_GEG · Mar 23, 2013

puremind said:
No disrespect, but you don't seem to be intimately familiar with screen calibration.

No disrespect, but you don't seem very familiar with comparative analytics which are the crux of your conclusions.

This is from the site you posted with the lowest cd/m2 score for the One. The One's performance against the One X+ matches up to GSMArena's comparison of the same two devices even though GSMArena's relative cd/m2 score for both devices was almost double. If the One's they each had didn't have screens that were performing similarly the deviation against the One X+ would have been pretty obvious to both of them.

Maximum brightness has remained virtually unchanged from the HTC One X +: was 373 cd / m 2, was 10 candelas more. But seriously change the contrast: on the already considerable 862:1 she grew up in half and is now 1258:1.

puremind · Mar 23, 2013

BarryH_GEG said:
Same for any other phone. If what you’re representing were true every device produced would have three display types and wild swings in battery life based on huge variations in brightness. We know that’s not the case because it would be all over each devices respective XDA forum.
[/I]

Who spoke about wild swings in battery life? I gave you a figure (11% for a 58% difference in brightness) based on 1 testing environment, 1 testing device, 1 reviewer. The difference is there. In fact, I even have no doubt that the difference could be more when testing SD videos instead of HD videos like in the test I quoted or that this percentage could be higher when starting from a much higher brightness point.

I read the GSMArena testing setup. Based on this, the difference is probably even larger than in my estimate for their own setting. As you said, not all settings will produce the same differences. The GSM set up for video testing was less demanding in terms of video processing, therefore the impact of screen brightness should be even greater!

The fact that battery life is impacted by maximum screen brightness is just not debatable, especially if what you say is true and it is only one and the same display. This actually precludes any other factors such as different display energy efficiency between different display models from making the brighter display brightness somehow equally energy efficient than the lower brightness display.

If what you say is true, then the battery difference is just not questionable. if screens had been different, I could have agreed that the brighter screen could have been as energy efficient due to different technology, but as it stands, light must be produced with energy and more light with more energy..

puremind · Mar 23, 2013

BarryH_GEG said:
No disrespect, but you don't seem very familiar with comparative analytics which are the crux of your conclusions.

Comparative analytics reinforce my point - the computerbase setup is actually more forgiving for screen brightness than the GSMArena one (focus on HD Video means the proportion of battery loss explained by brightess vs. that explained by video processing is less vs the GSM Arena setting), so based on comparative analytics, the difference would be even greater than 11%. That will teach me for understating battery improvements in my table just to not appear to arbitrarily inflate my results!

BarryH_GEG · Mar 23, 2013

puremind said:
If what you say is true, then the battery difference is just not questionable.

Of course it is. Your assumption, based on a flawed comparative resulting in a flawed conclusion, is that each One built could have a display that varies from 382-647cd/m2 and therefore battery performance could vary accordingly. I'm saying there's most likely a single panel used on the One and the difference in brightness between any two panels is no more than 10+/-% due to manufacturing tolerances and that all One's will have similar battery life when used the same way. If you're right this thread will balloon in to 100+ pages and end up a sticky because the lucky winners of the "screen lottery" you're describing that draw 382cd/m2 displays will be up in arms. If I'm right this thread will die out fairly quickly as the deviation you're hypothesizing only exists on paper.

HTC One screen reviews: One comes with different display types

Which screen do you think you have?

385cd/m² | 1100:1 contrast (low)

485cd/m² | 1200:1 contrast (typical)

650cd/m² | 1500:1 contrast (best)

Senior Member

Senior Member

Senior Member

Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Attachments

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Senior Member

Similar threads

Top Liked Posts