Also, I propose another, more realistic (at least for some) test to add - using Google Maps. Open up Google Maps on all phones, set it to navigate you to some address somewhere 20 - 30 miles away, add some music streaming app playing music in background (Pandora or Spotify), perhaps add some social media app with someone else sending you messages through it every 10 minutes or so, then just drive for an hour or so and record the battery capacity drop for each phone. Should be much more intensive than just "browsing" or "watching video" tests.
Constantly charging your phone reduces the batteries health and charging it while doing all of those tasks is going to heat your phone up like crazy! This is even worse for the battery. I never understood how people can put their phones on a clip near the A/C vent with the heat on, device baking in the sunlight, plugged in, with the screen always on displaying a map. No wonder why so many people's battery on their iPhones go to shit so fast.
This is the time when the S6 becomes unusable so it would have been an interest test. But the S7 is handling this scenario fine - though the battery doesn't charge by much.
You leave fee too many variables unaccounted for with a test like that tho... Not unaccounted for but out of direct control, which makes repeatability tough. What if tower A is knocked out or swamped on one run but not the next? Are they supposed to rerun the tests with all phones every time? What if tower streaming server happens to cough up a glitch or be losing more packets on any given run? So on and so forth...
Honestly that sounds like an amazing test. Some people browse the web all day, some just text, some abuse their phone like it owes them money. This has lead to many people saying their phone gets 9 hours of SOT while another only gets 2 hours. Sure these standardized tests gauge phones all the same but what you described is the most demanding situation a phone can be in and I think its a true testament to a phones battery life.
I'm still waiting till just sending some messages and pictures in Snapchat or using GPS doesn't heat up your phone anymore. So far my S7 Edge has stayed the coolest but I still expect better.
Well, those tests are bit unright. Lets say, when really using the phone - you wouldnt do those operations for such time. You usually do several operations which eat more battery and you dont care a lot about closing apps each time. So if to be logic - real users battery life will be much less than tests from this article showed. But as user of htc i can say that batteries are really weak and not able to be very effective nowadays. Battery is the biggest problem in each phone. However, seems soon we will get normal batteries. I heard that german company Neutrino INC made a battery which charges itself in some seconds using a neutrino energy and i heard they are going to release their products into market soon, so we will see what happens
That "neutrino" battery sounds TOTALLY legit. Where can I invest?...
The amazing thing is that Lia0009 is NOT confused. There really is a company claiming this: http://www.neutrino-energy.com Truly a sucker is born every minute. Let me just tell everyone out there, who is not a physicist, from someone who IS a physicist, that this is one-hundred percent grade-A scam. There is NOTHING of value in their claims, there is NOTHING that will lead to a new battery, to a better battery, to a battery alternative, or to anything of any value whatsoever. But there may well be a path from the pockets of suckers to a better life in Rio for the hucksters behind the web site...
When a cpu is tested it will have a slight variation of power efficiency, maximum clock speed etc. - the cpu is "binned" according to these differences - so some cpus may operate at a lower voltage and be more power efficient for example.
I haven't come to any definitive conclusions about the audio performance of the HTC 10. I'll have to spend some time doing relative comparisons between the M8/M9 and the 10 before I'm willing to pass any judgment.
I know you've got loads of things to test, and your exams, but gotta ask. Will your review include any sort of "advanced" audio testing, like what Chris Heinonen did couple of years ago?
HTC's been drumming about the HiFi aspects of the 10 so it would be _nice_ if those claims could be put to the test. And pretty much nobody does proper smartphone audio testing (with headphones), GSMArena does some stuff but their methods are a bit questionable and the results rather useless. Same applies to PhoneArena, yey, output vrms numbers, how freaking useful.
Also, would you happen to know: is that audio upsampling/upconversion BS something you can disable? I'll rather have as few conversion steps and the least amount of processing as possible thank you very much.
Could you please verify the bus speeds of the microSD slot in the full review? I'm mainly interested in UHS-I U3 compliance seeing as adoptable storage seems to be a thing with the HTC 10. Thank you in advance.
I asked that from HTC UK over on Twitter, they couldn't give me an answer (doh, a bit technical) so they escalated it so it can be received and answered by someone who actually knows the answer. I'll probably hear back from them monday the earliest. When I have their answer, I'll try and remember to come back here and tell you :D
I’m so glad HTC stayed with LCD screens. AMOLED screens have oversaturated colors and are harder to see outdoors. Also, I will NEVER buy a phone with a pentile display, ever.
Hooray for HTC moving the audio jack back to the top. it’s so much easier to use credit card readers when the audio jack is on top.
I just hope it’s possible to disable the off-screen buttons. Every time I borrow my friend’s Galaxy S5 I always hit those off-screen buttons by accident and it really annoys me.
This may have been true in the past however at the modern density the pentile has no downside and actually increases display efficiency. The maximum brightness of amoleds is actually comparable if not higher than modern lcd units meaning sunlight readability is class leading, furthermore they are technically the most accurate phone displays you can buy once set to 'basic' (S7 and S7 edge). I'd give the S7 and S7 edge displaymate article a read.
Technically AMOLED displays are NOT the most accurate. According to Anandtech, the Nexus 5X with LCD beats the accuracy of every AMOLED screen if that's what you're after. But Samsung's AMOLED screens do beat the Nexus 5X in brightness and of course contrast and black level.
By the way, I know DisplayMate always claims Samsung AMOLEDs are the best, but considering their reviews often come out way before the phones are released or even normal tech blogs get review units, I think it's safe to say Samsung is giving DisplayMate cherry picked units that may have been specially calibrated after the factory calibration. If you look at measurements of Samsung's retail units from Anandtech or Erica Griffin, they are never as good DisplayMate's results.
Good point. Although I enjoy reading the analysis of their display panels, it seems like Samsung goes for the extra mile and calibrate even further for displaymate to go and give high praises.
AMOLED and LCD technology has nothing to do with the manufacturer calibration. AMOLED can be calibrate to be as accurate as LCD as it has been proven many times (Samsung S6 and S6, Lumia 950...). AMOLED can inherently have higher gamut than sRGB which allow them to display color more saturated than most LCD is choosen so (like the first Samsung phone) but this is a user / manufacturer choice. It has nothing to do with the technology itself.
Anyway, once Apple will have AMOLEd screen, it will be the best thing ever.
Maybe I am just more perceptive, but the S7 screen which is suppose to be the best thing ever, well it still has a hint of that puke green tint that's been with Samsung AMOLEDs since day one. It's noticeably better than the S6, sure. But it's still there. On this defect alone, I can never use any Samsung phone (or AMOLED since they're pretty much the biggest panel supplier).
Quite opposite. Before I get interested in the Mi5 or the HTC 10, I want an AMOLED screen. It's just better, 10% to 20% better, depending on your what you value on a screen. I like all the hardware choices now in the S7 except the glass back. I just wish Samsung make a Nexus soon.
Also, do you know the size of the HTC10's SLC write cache? The write performance will drop if the data to be written exceeds the size of the SLC at some point, and I think it would be interesting to see where the performance drop point is.
Why are my Sequential Reads and Writes much higher than yours on my Exynos S7: Sequential Reads - 488.79 MB/s, Sequential Writes - 140.88 MB/s. Random Read - 83.7 MB/s (21428 IOPS, 4KB). Random Write 13 MB/s (3329.06 IOPS, 4KB)?
Because Exynos maybe? As I understand, the Exynos S7 also has better battery life as well. No getting around the S7 being a pretty awesome phone, but the M7 was my favourite phone ever, and this looks close enough for me to jump ship back to HTC maybe. It's kinda like VW and Alfa Romeo, everyone accepts the VW is a better car technically, but the Alfa just makes you feel more special!
I wonder how much of the differences are actually due to binning of the SOC. There are many many more variables at play. They obviously use different motherboards, different voltage regulation, diodes, resistors, dacs, hell possibly different composition of alloy in the traces on the motherboards. In reviews of several motherboards (idk we'll say 5 brands of z97) here on Anandtech with the same cpu and gpu swapped between them you may see +/- 5% performance. Those differences are solely motherboard based differences. I'm sure the SOC in the HTC is 99.9% as good as Samsung.
Emmc though... I'm not convinced it will keep pace for too much longer. There are significantly better interfaces and nand implementations available now. It's only a matter of time before Samsung more or less drops a sata ssd in a galaxy.
SATA SSD? Sure, if they're willing to implement their own and put up with horrid battery life. Performance differences between motherboards, outside of variance, are usually due to subtiming and frequency tweaks. A megahertz here, a couple ticks shaved off timings there, you have your few p.p.
An interesting thing would be to eliminate the SoC bin as a consideration. But AT might need to finagle some for that to happen.
A pity that you should tested the One 10 versus the US version of the Galaxy S7. I would have liked to know how it performs against Samsung's own Exynos 8890 SoC. Maybe you have the possibility to add those devices?
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
50 Comments
Back to Article
Demi9OD - Thursday, April 14, 2016 - link
"while read-intensive operations like loading apps may be slightly faster if storage reads are the critical path."Should that read "while read-intensive operations like loading apps may be slightly faster (on the S7) if storage reads are the critical path."?
JoshHo - Thursday, April 14, 2016 - link
That's correct. I've updated the article to make this clear.KiretoX - Thursday, April 14, 2016 - link
What happened with Galaxy S7 review Part 2? It's been over a month unless I'm missing something.Shadowmaster625 - Thursday, April 14, 2016 - link
The always do that.phoenix_rizzen - Thursday, April 14, 2016 - link
As mentioned in the comments on another article, he's in the middle of exams, so hasn't had time to finish it.Alex J. - Thursday, April 14, 2016 - link
You really seem to like the word "While" ;-)Also, I propose another, more realistic (at least for some) test to add - using Google Maps. Open up Google Maps on all phones, set it to navigate you to some address somewhere 20 - 30 miles away, add some music streaming app playing music in background (Pandora or Spotify), perhaps add some social media app with someone else sending you messages through it every 10 minutes or so, then just drive for an hour or so and record the battery capacity drop for each phone. Should be much more intensive than just "browsing" or "watching video" tests.
tekeffect - Thursday, April 14, 2016 - link
That is the only time I am sure to have my phone plugged into a charger. For me that would be the least useful test you could do.supraman21 - Friday, April 15, 2016 - link
Constantly charging your phone reduces the batteries health and charging it while doing all of those tasks is going to heat your phone up like crazy! This is even worse for the battery. I never understood how people can put their phones on a clip near the A/C vent with the heat on, device baking in the sunlight, plugged in, with the screen always on displaying a map. No wonder why so many people's battery on their iPhones go to shit so fast.OCedHrt - Saturday, April 16, 2016 - link
This is the time when the S6 becomes unusable so it would have been an interest test. But the S7 is handling this scenario fine - though the battery doesn't charge by much.ChronoReverse - Thursday, April 14, 2016 - link
Interestingly enough, Google Maps scrolling performance is often a sore spot for Galaxy phones even with AOSP ROMs.I noticed this since the Galaxy S4, which would have stuttery scrolling even though it had a more powerful GPU than the M7 did.
Flunk - Thursday, April 14, 2016 - link
If you're going 20-30 miles then wouldn't you be in your car... where there are power outlets?kurahk7 - Thursday, April 14, 2016 - link
Or you can record throttling while charging and navigating.Impulses - Friday, April 15, 2016 - link
You leave fee too many variables unaccounted for with a test like that tho... Not unaccounted for but out of direct control, which makes repeatability tough. What if tower A is knocked out or swamped on one run but not the next? Are they supposed to rerun the tests with all phones every time? What if tower streaming server happens to cough up a glitch or be losing more packets on any given run? So on and so forth...supraman21 - Friday, April 15, 2016 - link
Honestly that sounds like an amazing test. Some people browse the web all day, some just text, some abuse their phone like it owes them money. This has lead to many people saying their phone gets 9 hours of SOT while another only gets 2 hours. Sure these standardized tests gauge phones all the same but what you described is the most demanding situation a phone can be in and I think its a true testament to a phones battery life.I'm still waiting till just sending some messages and pictures in Snapchat or using GPS doesn't heat up your phone anymore. So far my S7 Edge has stayed the coolest but I still expect better.
Lia0009 - Thursday, April 14, 2016 - link
Well, those tests are bit unright. Lets say, when really using the phone - you wouldnt do those operations for such time. You usually do several operations which eat more battery and you dont care a lot about closing apps each time. So if to be logic - real users battery life will be much less than tests from this article showed. But as user of htc i can say that batteries are really weak and not able to be very effective nowadays. Battery is the biggest problem in each phone. However, seems soon we will get normal batteries. I heard that german company Neutrino INC made a battery which charges itself in some seconds using a neutrino energy and i heard they are going to release their products into market soon, so we will see what happensname99 - Thursday, April 14, 2016 - link
That "neutrino" battery sounds TOTALLY legit. Where can I invest?...The amazing thing is that Lia0009 is NOT confused. There really is a company claiming this:
http://www.neutrino-energy.com
Truly a sucker is born every minute.
Let me just tell everyone out there, who is not a physicist, from someone who IS a physicist, that this is one-hundred percent grade-A scam.
There is NOTHING of value in their claims, there is NOTHING that will lead to a new battery, to a better battery, to a battery alternative, or to anything of any value whatsoever. But there may well be a path from the pockets of suckers to a better life in Rio for the hucksters behind the web site...
Fergy - Saturday, April 16, 2016 - link
If you know enough it might be good to write an article for anandtech?mortimerr - Thursday, April 14, 2016 - link
So does he mean that the M10 has a different version of the 820 installed? I don't know what is meant by 'bin'.brookheather - Thursday, April 14, 2016 - link
When a cpu is tested it will have a slight variation of power efficiency, maximum clock speed etc. - the cpu is "binned" according to these differences - so some cpus may operate at a lower voltage and be more power efficient for example.RiotSloth - Friday, April 15, 2016 - link
Thanks for that, was wondering the same thing. Same as RAM, right?RiotSloth - Friday, April 15, 2016 - link
ETA: I'm also guessing Samsung gets dibs on the best bin and HTC fights over the scraps...Impulses - Friday, April 15, 2016 - link
If it were a desktop CPU he basically got one that will likely OC worse... :pPhil7C - Thursday, April 14, 2016 - link
Hi Joshua - thanks for the update - could you elaborate on "I’m still not quite sure how I feel about the audio"? Cheers! PhilJoshHo - Thursday, April 14, 2016 - link
I haven't come to any definitive conclusions about the audio performance of the HTC 10. I'll have to spend some time doing relative comparisons between the M8/M9 and the 10 before I'm willing to pass any judgment.lagittaja - Friday, April 15, 2016 - link
I know you've got loads of things to test, and your exams, but gotta ask. Will your review include any sort of "advanced" audio testing, like what Chris Heinonen did couple of years ago?HTC's been drumming about the HiFi aspects of the 10 so it would be _nice_ if those claims could be put to the test. And pretty much nobody does proper smartphone audio testing (with headphones), GSMArena does some stuff but their methods are a bit questionable and the results rather useless. Same applies to PhoneArena, yey, output vrms numbers, how freaking useful.
Also, would you happen to know: is that audio upsampling/upconversion BS something you can disable? I'll rather have as few conversion steps and the least amount of processing as possible thank you very much.
0siris - Thursday, April 14, 2016 - link
Could you please verify the bus speeds of the microSD slot in the full review? I'm mainly interested in UHS-I U3 compliance seeing as adoptable storage seems to be a thing with the HTC 10. Thank you in advance.lagittaja - Friday, April 15, 2016 - link
I asked that from HTC UK over on Twitter, they couldn't give me an answer (doh, a bit technical) so they escalated it so it can be received and answered by someone who actually knows the answer.I'll probably hear back from them monday the earliest.
When I have their answer, I'll try and remember to come back here and tell you :D
lagittaja - Sunday, April 17, 2016 - link
I'm feeling so dumb right now.HTC 10 will support UHS-I U3 (because A9 supported it as well), but not UHS-II unfortunately because this:
https://www.qualcomm.com/products/snapdragon/proce...
0siris - Monday, April 18, 2016 - link
Well, we're both dumb then; I didn't think of checking the 820 spec sheet either. Thank you for remembering to reply, I appreciate it a lot.lilmoe - Thursday, April 14, 2016 - link
Do you mind sharing your exact settings in AndroBench? File size, buffer size, threads etc...LordConrad - Thursday, April 14, 2016 - link
I’m so glad HTC stayed with LCD screens. AMOLED screens have oversaturated colors and are harder to see outdoors. Also, I will NEVER buy a phone with a pentile display, ever.Hooray for HTC moving the audio jack back to the top. it’s so much easier to use credit card readers when the audio jack is on top.
I just hope it’s possible to disable the off-screen buttons. Every time I borrow my friend’s Galaxy S5 I always hit those off-screen buttons by accident and it really annoys me.
randomhkkid - Thursday, April 14, 2016 - link
This may have been true in the past however at the modern density the pentile has no downside and actually increases display efficiency. The maximum brightness of amoleds is actually comparable if not higher than modern lcd units meaning sunlight readability is class leading, furthermore they are technically the most accurate phone displays you can buy once set to 'basic' (S7 and S7 edge). I'd give the S7 and S7 edge displaymate article a read.grayson_carr - Friday, April 15, 2016 - link
Technically AMOLED displays are NOT the most accurate. According to Anandtech, the Nexus 5X with LCD beats the accuracy of every AMOLED screen if that's what you're after. But Samsung's AMOLED screens do beat the Nexus 5X in brightness and of course contrast and black level.grayson_carr - Friday, April 15, 2016 - link
By the way, I know DisplayMate always claims Samsung AMOLEDs are the best, but considering their reviews often come out way before the phones are released or even normal tech blogs get review units, I think it's safe to say Samsung is giving DisplayMate cherry picked units that may have been specially calibrated after the factory calibration. If you look at measurements of Samsung's retail units from Anandtech or Erica Griffin, they are never as good DisplayMate's results.zeeBomb - Sunday, April 17, 2016 - link
Good point. Although I enjoy reading the analysis of their display panels, it seems like Samsung goes for the extra mile and calibrate even further for displaymate to go and give high praises.jlabelle - Monday, April 18, 2016 - link
AMOLED and LCD technology has nothing to do with the manufacturer calibration.AMOLED can be calibrate to be as accurate as LCD as it has been proven many times (Samsung S6 and S6, Lumia 950...).
AMOLED can inherently have higher gamut than sRGB which allow them to display color more saturated than most LCD is choosen so (like the first Samsung phone) but this is a user / manufacturer choice. It has nothing to do with the technology itself.
Anyway, once Apple will have AMOLEd screen, it will be the best thing ever.
deskjob - Sunday, April 17, 2016 - link
Maybe I am just more perceptive, but the S7 screen which is suppose to be the best thing ever, well it still has a hint of that puke green tint that's been with Samsung AMOLEDs since day one. It's noticeably better than the S6, sure. But it's still there. On this defect alone, I can never use any Samsung phone (or AMOLED since they're pretty much the biggest panel supplier).zodiacfml - Friday, April 15, 2016 - link
Quite opposite. Before I get interested in the Mi5 or the HTC 10, I want an AMOLED screen. It's just better, 10% to 20% better, depending on your what you value on a screen. I like all the hardware choices now in the S7 except the glass back. I just wish Samsung make a Nexus soon.heartinpiece - Friday, April 15, 2016 - link
Are both the read/write rates of the HTC 10 and S7 encrypted read/writes?heartinpiece - Friday, April 15, 2016 - link
Also, do you know the size of the HTC10's SLC write cache? The write performance will drop if the data to be written exceeds the size of the SLC at some point, and I think it would be interesting to see where the performance drop point is.Toss3 - Friday, April 15, 2016 - link
Why are my Sequential Reads and Writes much higher than yours on my Exynos S7: Sequential Reads - 488.79 MB/s, Sequential Writes - 140.88 MB/s. Random Read - 83.7 MB/s (21428 IOPS, 4KB). Random Write 13 MB/s (3329.06 IOPS, 4KB)?RiotSloth - Friday, April 15, 2016 - link
Because Exynos maybe? As I understand, the Exynos S7 also has better battery life as well. No getting around the S7 being a pretty awesome phone, but the M7 was my favourite phone ever, and this looks close enough for me to jump ship back to HTC maybe. It's kinda like VW and Alfa Romeo, everyone accepts the VW is a better car technically, but the Alfa just makes you feel more special!Badelhas - Wednesday, April 27, 2016 - link
Exactly my thoughs. And dont forget touchwiz, which sucks big time, imo.Ej24 - Saturday, April 16, 2016 - link
I wonder how much of the differences are actually due to binning of the SOC. There are many many more variables at play. They obviously use different motherboards, different voltage regulation, diodes, resistors, dacs, hell possibly different composition of alloy in the traces on the motherboards. In reviews of several motherboards (idk we'll say 5 brands of z97) here on Anandtech with the same cpu and gpu swapped between them you may see +/- 5% performance. Those differences are solely motherboard based differences. I'm sure the SOC in the HTC is 99.9% as good as Samsung.Emmc though... I'm not convinced it will keep pace for too much longer. There are significantly better interfaces and nand implementations available now. It's only a matter of time before Samsung more or less drops a sata ssd in a galaxy.
Zoomer - Sunday, April 17, 2016 - link
SATA SSD? Sure, if they're willing to implement their own and put up with horrid battery life. Performance differences between motherboards, outside of variance, are usually due to subtiming and frequency tweaks. A megahertz here, a couple ticks shaved off timings there, you have your few p.p.An interesting thing would be to eliminate the SoC bin as a consideration. But AT might need to finagle some for that to happen.
npaladin2000 - Monday, April 18, 2016 - link
They already did. UFS is essentially a small mobile-oriented SSD.quantumbells - Sunday, May 1, 2016 - link
When can we expect a full review?Groomsh - Tuesday, May 17, 2016 - link
A pity that you should tested the One 10 versus the US version of the Galaxy S7. I would have liked to know how it performs against Samsung's own Exynos 8890 SoC. Maybe you have the possibility to add those devices?Greyarc - Tuesday, June 28, 2016 - link
Still waiting on the full review............any updates?ACM.1899 - Monday, September 19, 2016 - link
still waiting.... :(i think HTC 10 worth a full review.