Unless you’re extremely familiar with security
cameras it’s pretty difficult to read the specs for a camera online and predict whether
its field of view will be right for the area you need to cover. The easy solution for that is to buy a varifocal
camera that lets you dial in the exact field of view and zoom that you need for your mounting
location. Today on the hookup we’re going to test
these 10 Varifocal PoE security cameras ranging from $110 to $1100 to see if 10 times the
price gives you 10 times the performance. Lets start with a quick look at the camera
lineup. The least expensive for $110 is the Reolink
RLC811A, then the $159 Annke C800 Zoom, the $189 Lorex LNB-92-82-B, the $199 Amcrest VB-27-96-EW,
and the $237 Empire tech T-58-42. Then jumping significantly in price we’ve
got the $449 UniFi G4 Pro, the $709 UniView 36-38-SB, the $779 Bosch Flexidome 5000, the
$899 Avigilon H4A-B01, and last, the most expensive camera we’ll be testing today
is the Axis P-32-48-LVE. As I mentioned before, all these models are
motorized varifocal or what some people call zoom cameras, and all of them are 4k, 8 megapixel
resolution except the Bosch Flexidome which is 5 megapixels because I added it to the
video last minute and couldn’t find a source for the 8 megapixel version since they are
primarily sold by system integrators and not direct to consumer. For testing, to be able to compare apples
to apples I took a significant amount of time standardizing the zoom and aim of each camera
and I tested them in 3 different daytime lighting conditions and 2 different nighttime conditions. Starting our tests with daytime, full sun,
no shadows, which should be the easiest lighting condition and should result in the best images. And starting that with the easiest of those
tests, I held up a sign at 50 feet which with this level of zoom and 8 megapixel resolution
corresponds to 200 pixels per meter, which is the lower limit for pixel density that
should be able to identify a person based on camera footage. Ranking their image quality from worst to
best we’ve got this generally fuzzy image from the Avigilon, then the Amcrest was really
pixelated and over sharpened, next the Lorex was less pixelated but still distorted, the
UniFi G4 Pro just overall lacked detail on my face and the sign, next the empire tech
was also extremely pixelated but it did produce legible text. After that the 5 megapixel Bosch also produced
legible test but was too blurry to be able to positively identify me, while the Uniview
captured details in my face and also produced legible text, and so did the Reolink but I
thought the Reolink was a slightly more accurate representation of my face, the Annke did a
great job preserving details without over sharpening, and finally the Axis produced
a pretty much perfect image from 50ft and not only captured a ton of detail, but it
didn’t introduce any strange distortions making it the clear winner for this first
category. The next test was me running toward the camera
which tends to be a little harder for the cameras to capture a good image since most
cameras use sharpening algorithms to make non-moving and slow moving objects more clear. In this test the UniFi G4 Pro was the least
detailed, followed by the Avigilon which still looked fuzzy to me. Next was the bosch, then the slightly distorted
lorex, the extremely oversharpened empire tech, the strangely smooth uniview, the pixelated
Amcrest, a pretty good image from the Reolink, then the Axis did very well, but I thought
my face looked more defined and accurate on the Annke C800 Zoom, which takes first place
for this test. Last is the most difficult test I did in each
lighting condition which was another running test, but this time running across the image
at 25ft. At this range the images have nearly 400 pixels
per meter, which should be plenty to identity a person, so lets see how they did. The worst image came from the UniFi G4 Pro
which looks like it tried to stitch two frames together, but did a really bad job of it. Next was the avigilon, which performed exactly
like it has been producing an image with no other problems other it being generally fuzzy,
and yes, I did clean the lens and make sure the focus was as dialed in, but this is just
how it looks I guess. After that was the bosch which again wasn’t
terrible but lacked detail in the sign and my face. Then the Amcrest was artificially sharpened,
but still a pretty decent image, the Annke was very accurate, but a little blurry, the
empire tech was way too sharp, but I do have to give it credit for producing a legible
image of the 72 point font, and so did the lorex, except without the crazy oversharpening. The reolink produced a very HDR image with
high contrast and legible 72 point font, and the axis image was also very good, but the
sharpening algorithm made some errors in the sign and also produced some strange artifacts
around my legs. I thought the best image came from the UniView
which was clear with no artifacts, although you can tell it’s been run through a smoothing
algorithm to reduce noise. I repeated those same three tests at dusk,
with significantly lower light, which should increase the difficulty for the cameras. Starting with the identification test at 50
feet, here are the results in order from worst to best. … and the top 3 were the Uniview 36-38-SB,
the Reolink RLC-811A, and first place in this category goes to the Axis P-32-48 which produce
a very natural looking image. Next for the test running towards the camera
here are the results again in order from worst to best. … and again the top 3 were the Reolink RLC-811A,
the Uniview 36-38-SB, and the winner overall was the Axis P32-48 again which produed a
pretty much perfect image with legible 72 point font and great detail throughout. In these slight lower light conditions all
the cameras struggled on the cross running test because they increase their exposure
time which causes fast moving objects to blur. … this time the top 3 were the UniFi G4
Pro, the Uniview 36-38-SB, and without a doubt the best image came from the Axis P-32-48
which not only produced legible 72 point font, but also captured the most facial detail. The last daytime category is to determine
how well the cameras can deal with shadows, which is a feature called wide dynamic range
which lowers the brightness of the highlights and raises the brightness of the shadows. Starting with the 50 Feet identification test
here are the bottom 7 performers. … and the top 3 were the Uniview 36-38-SB,
the Annke C800 Zoom, and as usual the Axis P-32-48 produced the best best image with
no distortion and the most preserved details. For the test running towards the camera here
are the bottom 7. … and the top 3 were the Annke C800 Zoom,
the Uniview 36-38-SB, and surprise surprise the Axis P32-48 takes first place again with
legible 72 point font and nearly perfect facial detail. Running across the camera’s view in these
shadowy conditions here are the bottom 7. … and for the first time the Axis didn’t
make the top 3, which went to the Reolink RLC-811A, the Empire tech T-58-42, and the
best image came from the Uniview 36-38-SB which produced a legible image of the sign
without introducing a bunch of sharpening artifacts. I also ran one additional test in these shadowy
conditions driving my car past at 25 miles per hour, which all of the cameras handled
well, but I tried to use the detail in the rims and the emblem on the side of the car
to determine their rankings. Starting with the worst, the bosch was pretty
blurry, followed by the unifi, the axis did surprisingly bad here introducing some weird
artifacts in the rims and failing to produce a legible emblem, the Amcrest was a little
bit pixelated, while the avigilon was surprisingly good but still a little blurry, the annke
was nice and clear but the emblem lacked detail, and the lorex was slightly better than that,
the empire tech used it edge sharpening algorithm to really make the details pop, while the
uniview had that same detail without the oversharpening. The winner in this category was the reolink
which not only captured all the detail of the rims and emblem in a more natural way,
but it also handled the contrast better and was even able to see me through the passenger
side window. So lets look at some overall results from
all of our daytime testing. For the scoring I awarded 10 points to rank
1, 9 points to rank 2, etc. Averaged across all three lighting conditions
the best performer for the 50 foot non moving identification was the Axis with an overall
score of 10.0 and the surprising finishers after that were the Annke Zoom with an overall
score of 8.3 and the Reolink RLC811A with a 8.0. For movement going towards the camera the
Axis finished first with an overall score of 9.7 followed by the UniView with and overall
score of 8.0 and then the Annke Zoom with a 7.3. For movement going across the camera the Uniview
came out on top with an overall score of 9.7 followed by the Axis with a score of 8.7 and
the Reolink with a 7.3. Combining all those individual scores we can
see the total for daytime performance which puts the Axis on top with a total of 28.3,
the Uniview in second with 25.3, and the Reolink in third with 21.7. If we throw those on a price vs performance
chart and we consider the top performing Axis as “you get what you pay for” then the
Uniview performs slightly above its price point, but not nearly as high value as the
Reolink RLC811A or the Annke Zoom. And having tested a lot of Reolink cameras
this comes as no surprise to me since their daytime performance is basically unmatched
for the money, but I know they tend to struggle at night, so lets take a look at those results. First, I turned off all of the infrared lighting
on these cameras and kept my porch and garage lights on for a low ambient light test. Most of the cameras actually preferred to
stay in color mode with this amount of light, but I did have to force a few of them into
this night color mode. Starting with the still image test, this time
at 25 feet instead of 50, which remember is roughly 400 pixels per meter the avigilon
really struggled and clearly wasn’t getting enough light to produce a quality image. The Unifi did slightly better, but still lacked
facial detail. The empire tech’s sharpening algorithm went
crazy here and introduced a bunch of white dots throughout the image. While the bosch image was overexposed on the
sign, but overall not terrible. The lorex had the same amount of facial detail
as the bosch, but also captured the sign, and the Amcrest had a legible sign, but totally
misrepresented my face. The uniview did a good job with the sign,
but lacked facial detail, while the annke got my face, and not the sign. The reolink did a great job on both the sign
and my face but not quite as good as the axis which did about as well as you could ask it
to. For movement going towards the camera I also
decreased the capture distance to 25 feet give the cameras a fighting chance, but pretty
much all of them struggled, in fact, I don’t think any of the bottom 9 cameras produced
a good enough image to be used for identification. In contrast the Axis produced a great image
with legible text and plenty of facial detail that could easily identify me. Unfortunately things were even more grim for
the cross movement test at 25 feet where none of the cameras captured any facial detail,
but here they are anyways from worst to best. In my testing in the past, color night vision
just isn’t as good for capturing movement as infrared night vision can be, so let’s
see if those tests went any better. These are also the only tests that I did separately
since I needed to only use the infrared LEDs from the camera I was testing. All other lights on my house were turned off
and the infrared LEDs of all the other security cameras were disabled. Starting with the 25 foot 400 pixel per meter
identification test here are the bottom 7 from worst to best. … and the top 3 performers were the Reolink
RLC-811A, the Axis P32-48, and the top spot went to the empire tech T-58-42 which while
a little underexposed still captured the most detail and with a little brightness boost
in post processing you can see how good the image actually is. For the test running towards the camera you
might notice that the stills are from slightly different locations and distances, and that’s
because I took the highest quality frame from a 10 foot area for each camera, which as you
can see for the bottom 7 performers, still wasn’t that great. The top 3 performers here were the Lorex which
didn’t perfectly captures the sign, but did a good job with facial detail, the Empire
tech which again captured a lot of detail after a little brightening, but in this case
the Axis was the best performer even without any post processing. Unfortunately just like the color night test
results of the cross running test weren’t nearly as promising and almost all of the
cameras produced a blurry mess, barely capturing my shape. …the best results came from the avigilon
which at least captured my shape, and the Axis which was blurry, but has hints of detail
that could be useful for things like matching my clothes to a description. I guess moral of the story is no matter how
much you spend on your camera, it’s hard to get good footage of fast moving objects
at night, so if you are a criminal watching this make sure to run really fast while committing
your crimes. Anyways, scoring the color night vision again
awarding 10 points for first place, 9 for second, and so on the Axis P-32-48 had a perfect
score of 30 points, 2nd place went to the Amcrest with a considerably lower 21 points,
and third was the lorex with 19. For infrared night vision the Axis also won
that category with a total of 29 points, the Empire tech had 23, and the lorex had 21. And for nighttime performance, it’s not
really important for a camera to perform well in both infrared and color mode since you’re
just going to set it for one or the other, so I took their best overall score from either
color or infrared performance, which you can see here. And on the price vs performance graph here
again using axis as a “you get what you pay for” solution here’s how the rest
of the competition stacks up for value based on their best nighttime score. So then if we combine all that data, we get
the most important chart, which is combined score for daytime and nighttime performance
which I can use to give some final recommendations. First, if you are going to spare no expense,
the Axis P-32-48-LVE is by far the nicest camera I’ve ever tested in terms of build
quality, interface, options, and performance. It has an excellent companion app if you want
to use the camera in standalone mode, it’s ONVIF compatible so you can use it with blue
iris or other NVRs that support ONVIF, it’s got extremely advanced on device person and
vehicle detection, it supports MQTT and it’s fully NDAA and SEA compliant. The only negatives of the Axis P-32-48 are
that it doesn’t have a built in microphone and obviously its price which comes in at
$1100 each which is probably more than you want to spend on your entire system let alone
a single camera. So moving down the chart. The next highest performer was the UniView
36-38-SB, but with a price tag of $709 that’s probably still too high for most people, and
honestly if you’re going to go all in and spare no expense, you’re better off with
the Axis instead. Coming down significantly in price the Empire
Tech T-58-42 is $237 per camera and had a total score of 39.0 which was mostly the result
of its high performance in nighttime infrared mode catagory which is why it is my recommendation
for anyone looking for a reasonably priced camera where nighttime performance is a priority. The empire tech also has a built in microphone,
has the widest field of view when zoomed all the way out, and is fully ONVIF compatible
including using on device person and vehicle detection which can be used to trigger ONVIF
motion events on your compatible NVR. The Empire Tech’s daytime performance was
just okay, and was generally over sharpened and pixelated, which was unfortunately something
I wasn’t able to correct by changing camera settings or firmware. The other downside of the Empire Tech is that
like the Lorex and Amcrest, the camera is manufactured by Dahua which makes it NDAA
and SEA non-compliant, which personally doesn’t matter to me since I block all my cameras
from the internet regardless of manufacturer, but it’s worth noting. The Empire Tech can technically be used in
standalone mode recording to a removable SD card, but the app interface is clunky and
your experience will be much better if you use it with a compatible NVR instead. After that for just $159 per camera is the
Annke C800 Zoom with a score of 38.3 which was the result of good daytime performance,
but it took its nighttime score from the color night vision tests where in my opinion it
didn’t produce great results, and if you’re looking for a high performing daytime camera
and are willing to compromise on nighttime performance then I think the Reolink RLC-811A
is a better choice. Not only did the Reolink have the third highest
daytime score only behind the Axis and Uniview, it also has a lot of other features that make
it an incredible deal for just $109. The Reolink RLC-811A has a built in microphone,
on device person and vehicle detection, a great companion app if you want to use your
camera in standalone mode, it has the option of using infrared or white LEDs, and because
Reolink owns their own OEM factory their cameras are NDAA and SEA compliant. In the past I’ve reported that Reolink 8
megapixel cameras don’t work with blue iris, but I’m happy to say that as of the January
2022 reolink firmware update and blue iris versions 5.5.5.0 and above the incompatibility
issues seem to be solved, and after setting up the ONVIF service in the camera’s web
interface Blue Iris can automatically populate the correct settings for Reolink cameras. The downside of Reolink cameras is always
the same, while its nighttime IR performance is fine for slow moving objects, things that
move fast tend to get disappearing leg syndrome, which is something I unfortunately haven’t
been able to fix by changing camera settings or by providing additional infrared light. Still for 1/10th the cost of the Axis and
less than half the cost of the Empire Tech the Reolink RLC-811A is an absolutely insane
value as long as you understand the limitations of its nighttime performance. Down in the description I’ve linked an unlisted
YouTube video that has all the raw footage from my testing from each camera if you want
to examine it yourself. Also down in the description are links to
all the cameras I tested in this video, which are affiliate links so if you appreciate the
time, effort, and money that it takes me to make a video like this then using those links
gives me a small commission at no cost to you. Thank you to all of my patrons over at patreon
for your continued support of my channel, and if you’re interested in supporting my
channel please check out the links in the description. If you enjoyed this video please hit that
thumbs up button and consider subscribing and as always, thanks for watching the hookup.
