1
00:00:00,480 --> 00:00:05,120
so honor is sponsoring an early look

2
00:00:03,679 --> 00:00:09,679
actually i don't know if by the time this comes out it's going to be early at

3
00:00:07,279 --> 00:00:13,519
all but uh that's not really the point the point is

4
00:00:11,440 --> 00:00:19,359
we are checking out the honor 10 and i absolutely love doing stuff for honor

5
00:00:16,240 --> 00:00:22,400
because i get to make this joke over and

6
00:00:19,359 --> 00:00:25,050
over and over again

7
00:00:22,400 --> 00:00:27,100
are you ready to do the shoot for honor

8
00:00:29,439 --> 00:00:33,039
hey which video are you editing right now

9
00:00:33,600 --> 00:00:39,520
is it the one for honor no

10
00:00:37,440 --> 00:00:55,760
hey tyler did that product arrive yet for honor

11
00:00:55,760 --> 00:01:05,439
so the highlight features of this thing include an octocore processor a 19x9

12
00:01:02,160 --> 00:01:07,680
aspect ratio screen with notch and an

13
00:01:05,439 --> 00:01:13,200
npu or neural processing unit

14
00:01:10,000 --> 00:01:15,040
and it's very

15
00:01:13,200 --> 00:01:17,759
shiny does it look as good on camera as it

16
00:01:16,400 --> 00:01:22,240
does in person dang so we'll start with the obligatory

17
00:01:20,159 --> 00:01:27,520
unboxing you get a sim removal tool you get a bunch of documentation you get

18
00:01:24,840 --> 00:01:30,960
oh that's actually kind of a nice touch they include

19
00:01:29,520 --> 00:01:36,640
a case so that at least while you wait for your

20
00:01:33,759 --> 00:01:40,880
like you know nicer one or whatever you're not going to get that stupid

21
00:01:38,799 --> 00:01:44,320
first scratch on the phone that you're you're otherwise likely to it actually

22
00:01:42,400 --> 00:01:46,640
includes a screen protector as well so those kinds of things are always a nice

23
00:01:45,759 --> 00:01:52,399
touch they've got a supercharged power supply right here with oops

24
00:01:50,399 --> 00:01:57,280
well uh i'm sure you'll get one that's compatible with your region and a usb

25
00:01:54,560 --> 00:02:00,159
type-c cable with an adorable purple inset

26
00:01:58,240 --> 00:02:04,320
of course it's no big secret but i think unboxings are boring so

27
00:02:02,159 --> 00:02:10,720
let's have a look at the phone itself i've actually got a couple of them i've

28
00:02:06,399 --> 00:02:12,560
got that sick purple and blue iridescent

29
00:02:10,720 --> 00:02:16,400
they call it aurora gradient or something like that finish and then i've

30
00:02:14,400 --> 00:02:20,239
got a more plain blue one that quite frankly uh you should be able to tell by

31
00:02:18,319 --> 00:02:24,480
now but i think this one looks so much better looking at the physical

32
00:02:23,040 --> 00:02:29,040
layout of the phone there are a couple of things that i think some folks are

33
00:02:26,480 --> 00:02:34,640
going to find polarizing like it's really the age of phablets these days

34
00:02:31,760 --> 00:02:37,599
but this is a comfortably one-handable phone

35
00:02:36,400 --> 00:02:42,800
like you can actually type

36
00:02:41,200 --> 00:02:47,280
at least if you got swiftkey on anyway that comes pre-loaded which is nice with

37
00:02:44,800 --> 00:02:52,000
one hand on it but i don't know if everyone is going to like the duplicated

38
00:02:49,760 --> 00:02:55,280
home button so you've got a software one as well as a built-in one in hardware

39
00:02:54,400 --> 00:03:01,680
and that's actually where the fingerprint sensor is so it bucks the trend in terms of fingerprint sensor it

40
00:02:59,599 --> 00:03:06,720
bucks the trend in terms of headphone jack which for me anyway is a really

41
00:03:03,920 --> 00:03:11,200
nice inclusion and then it sticks with the trend for the display notch i'm

42
00:03:09,120 --> 00:03:16,159
happy to see a more handleable option though and personally i'm a really big

43
00:03:13,519 --> 00:03:19,360
fan of glass backs because even though actually this one i think i've already

44
00:03:17,440 --> 00:03:23,040
got some fingerprints on even though they do tend to get fingerprinted up

45
00:03:21,280 --> 00:03:27,920
pretty easily what's nice about the glass

46
00:03:24,959 --> 00:03:31,519
is that at least they're easy to clean then they're perfect another choice they

47
00:03:30,000 --> 00:03:37,280
made that's going to be controversial no matter how you slice it is their use of

48
00:03:34,080 --> 00:03:39,840
an ips display so right now the main two

49
00:03:37,280 --> 00:03:44,640
choices are OLED or lcd ips

50
00:03:42,400 --> 00:03:49,280
and they've both got pros and cons with OLED having perfect blacks and the

51
00:03:47,200 --> 00:03:54,879
ability to you know show the time even when the screen is off for example with

52
00:03:51,519 --> 00:03:57,599
ips having the advantage of basically no

53
00:03:54,879 --> 00:04:00,480
chance of running into burning issues so

54
00:03:58,799 --> 00:04:05,840
it's a good ips they're boasting 96 coverage of ntsc and

55
00:04:04,000 --> 00:04:10,959
it looks after i turned off the vivid profile that was loaded by default

56
00:04:07,680 --> 00:04:12,000
subjectively really freaking good

57
00:04:10,959 --> 00:04:18,639
but everyone's going to have their own feelings on a decision like this

58
00:04:17,040 --> 00:04:22,720
now this is something really interesting that i picked up on the way that they've

59
00:04:20,239 --> 00:04:27,040
implemented the notch is not like pure notch or pure knot notch so you can see

60
00:04:25,280 --> 00:04:31,360
that the display stretches up into the notch area when you're vertical but

61
00:04:28,720 --> 00:04:37,280
check this out it actually does a little trickery thing so you watch that change

62
00:04:33,600 --> 00:04:39,440
to curved and then it cuts the notch out

63
00:04:37,280 --> 00:04:43,600
when you put video in landscape mode which personally i think is a really

64
00:04:41,680 --> 00:04:46,960
nice touch another thing that you can actually do while i'm on the subject of

65
00:04:45,520 --> 00:04:52,600
cool little uh software tricks is you can turn the

66
00:04:48,880 --> 00:04:52,600
notch off entirely

67
00:04:54,160 --> 00:04:59,759
and there are actually some nice touches as well wireless charging isn't

68
00:04:57,840 --> 00:05:05,120
supported but it does support four and a half amp five volt fast charging and

69
00:05:02,639 --> 00:05:09,120
their facial recognition it is shockingly fast so i've got it open here

70
00:05:07,440 --> 00:05:15,199
it's looking for my face bam that's really impressive and this

71
00:05:12,400 --> 00:05:19,360
one's cool there's a feature on here called smart rotate

72
00:05:17,039 --> 00:05:24,680
that lets the screen rotation follow the direction of your face can your phone do

73
00:05:21,680 --> 00:05:24,680
this

74
00:05:26,400 --> 00:05:32,720
and it stays in portrait mode cool right anyway the big feature on the

75
00:05:30,800 --> 00:05:38,479
phone is really on the rear shooter so it's a dual rear shooter it's got a 16

76
00:05:34,720 --> 00:05:40,880
megapixel f 1.8 color sensor and then a

77
00:05:38,479 --> 00:05:47,199
24 megapixel black and white and the whole thing is powered by their neural

78
00:05:43,680 --> 00:05:49,759
processing unit or npu so using their

79
00:05:47,199 --> 00:05:54,240
npu they can actually recognize the subject of a photo and then adjust the

80
00:05:52,800 --> 00:05:58,720
way that it looks in order to make it more pleasing

81
00:05:57,440 --> 00:06:02,720
let's go try it out on a few things shall we

82
00:06:01,039 --> 00:06:07,759
all right so let's do some test shots with the ai camera first let's do with

83
00:06:05,600 --> 00:06:11,600
no ai hi james

84
00:06:09,600 --> 00:06:16,880
i'm trying to do you a service here by taking a good picture of you not easy

85
00:06:14,319 --> 00:06:24,000
now let's turn on ai photography and it has identified that this is a person

86
00:06:21,360 --> 00:06:27,759
and it looks like it went and applied a background blur effect

87
00:06:26,160 --> 00:06:33,840
so that's an interesting thing for it to do on its own let's try some food

88
00:06:31,120 --> 00:06:38,240
that is not very photogenic food but i'm sure it tastes good look at that

89
00:06:36,160 --> 00:06:43,199
it identified food thought it was person for a second there all right a car

90
00:06:39,919 --> 00:06:43,199
should be a pretty easy one

91
00:06:43,680 --> 00:06:49,600
yep wow that did not take long the instant i turned on ai it was like yup

92
00:06:48,479 --> 00:06:53,680
that's a car let's find some nature

93
00:06:51,600 --> 00:06:58,560
dang nature you pretty yep i get a little flower icon

94
00:06:56,160 --> 00:07:03,680
so i want to talk about what an npu actually does while we have a look at

95
00:07:00,880 --> 00:07:09,440
the shots from the ai versus non-ai photo comparison here subjectively i

96
00:07:06,880 --> 00:07:13,520
like a more natural look to my photos so i would probably have it off in many

97
00:07:12,479 --> 00:07:18,720
cases but the underlying technology here is

98
00:07:16,319 --> 00:07:24,319
really the big story to me we've been at the point for a while now where a phone

99
00:07:21,599 --> 00:07:29,680
CPU can be expected to do all of the normal stuff that you'd want from it

100
00:07:26,319 --> 00:07:32,000
really well you know run the ui smoothly

101
00:07:29,680 --> 00:07:38,160
bring up your apps list open apps whatever the case may be without

102
00:07:34,800 --> 00:07:40,880
a ton of perceptible leg so what an npu

103
00:07:38,160 --> 00:07:46,080
does is it is a specialized processor it actually gets dedicated silicon that is

104
00:07:43,360 --> 00:07:50,400
designed for convoluted neural networks so this doesn't mean that all of a

105
00:07:47,599 --> 00:07:54,720
sudden you can just turn on ai mode and run around in the world with you know

106
00:07:52,400 --> 00:07:58,879
pointing your phone's camera at things and it's just gonna start to

107
00:07:56,560 --> 00:08:03,919
have a broad understanding of the world but what it does mean is that a neural

108
00:08:01,599 --> 00:08:09,039
network model one that would have to be trained by a very powerful computer

109
00:08:06,639 --> 00:08:15,280
probably in a data center somewhere can be more easily understood using this

110
00:08:12,400 --> 00:08:20,879
kind of client-side neural networking technology so the powerful server trains

111
00:08:18,400 --> 00:08:23,759
the model creates a simplified version of it that

112
00:08:22,000 --> 00:08:27,360
honor will push to your phone with an update and then your npu does what's

113
00:08:26,479 --> 00:08:32,240
called inferencing in order to feed that model

114
00:08:30,639 --> 00:08:37,200
some new piece of data like you're pointing your phone at a cat

115
00:08:34,000 --> 00:08:40,080
and then infer what it's supposed to be

116
00:08:37,200 --> 00:08:44,560
from the model that it has downloaded so at this point the npu is being used

117
00:08:42,080 --> 00:08:48,959
primarily to monitor usage patterns and then optimize the processor's use of

118
00:08:47,120 --> 00:08:53,519
power so that you'll get as much battery life as possible out of the phone

119
00:08:50,959 --> 00:08:57,440
and to recognize photography subjects with the camera and then adjust the

120
00:08:55,360 --> 00:09:02,800
camera's settings accordingly so

121
00:08:58,880 --> 00:09:04,800
calling it ai is a little bit optimistic

122
00:09:02,800 --> 00:09:09,040
it's not like it's got like jarvis built into your phone or anything like that so

123
00:09:06,720 --> 00:09:14,480
right now it has the ability to uh recognize an animal or a pet but new

124
00:09:12,399 --> 00:09:18,240
models that honor could push via updates could give it the capability of being

125
00:09:16,160 --> 00:09:23,279
able to tell the difference between a cat and a dog and a a a possum that

126
00:09:21,519 --> 00:09:28,480
one's unlikely but like a possum for example meaning that it can in fact get

127
00:09:26,240 --> 00:09:33,440
smarter over time even if the actual training doesn't take place on the

128
00:09:30,880 --> 00:09:36,880
device itself so that's pretty much my rundown of the honor 10. you guys can

129
00:09:34,959 --> 00:09:41,680
find a link with more information in the video description so thanks to honor for

130
00:09:39,440 --> 00:09:44,720
sponsoring this video thanks to you guys for watching it if you guys disliked it

131
00:09:43,360 --> 00:09:49,200
you can hit that button but if you liked it hit like get subscribed or maybe consider checking out where to buy the

132
00:09:47,440 --> 00:09:52,560
stuff we featured at the link in the video description also down there is our

133
00:09:51,200 --> 00:09:58,440
merch store which has cool shirts like this one and our community forum which

134
00:09:54,720 --> 00:09:58,440
you should totally join