Deconstruction of a dangerous misleading youtube review "Finding the BEST 4K Security Camera NVR Package (Reolink vs Amcrest vs Swann)"

mat200

IPCT Contributor
Joined
Jan 17, 2017
Messages
7,128
Reaction score
6,638
Executive Summary:
DO NOT be a sheep.
Ask what a reviewer can be missing, ask if a reviewer is using hype words to sell the cheapest product.
  • Test your cameras and kits right away.
  • Just do a bench test.
  • Test real world conditions.
  • Test moving suspects.
  • Test at night, test at day.
  • Use a test rig to place cameras at positions which you plan to place the cameras.



Finding the BEST 4K Security Camera NVR Package (Reolink vs Amcrest vs Swann)
The Hook Up

4:30 - Mentions Blue Iris "NVR" interface is much better but probably not $600 better
Response: Many members here go with a good used i7 Business Class PC and save a lot on that. ( some getting a good PC for $200+/- and also getting a decent PoE switch ) - so it may cost a bit more than an NVR in a kit, but for many here they really like Blue Iris and feel it is worth it. ( note, if buying an NVR separately - it can cost more than going Blue Iris + used business class machine )

11:10 - "let's check out the daytime clarity of each camera"
this is the section which I feel is very poorly laid out.

Reviewer tests a image capture at 25ft of text.

Fact: Wider FOV = less pixel density at 25 ft.
Fact: Tuning parameters also matter.

Reolink B800
Image Sensor: 1/2.5" CMOS Sensor
Effective Pixels: 3840x2160 (8.0 Megapixel)
Lens: f=4.0mm F=2.0
Angle of View: Horizontal: 97° ; Vertical: 51°
ref:

Amcrest Camera IP8M-T2499EW-28MM
Image Sensor 1/2.5” 8Megapixel progressive scan CMOS
Focal Length 2.8mm
Max. Aperture F1.6
Angle of View H: 112°, V: 69°
ref:
and



The reviewer FAILS completely to mention FOV and Lens ( 4mm vs 2.8mm ) in his comparisons - which in my book is completely misleading.

11:49 - "night 25 ft."
".. results were very similar..." again testing an image capture at 25 ft of text.

claims the other cameras have very poor night time performance.

Again major FAILURE on the reviewer.
Failed to test moving subject.
Failed to even mention lens differences.
Failed to even mention FOV differences.

12:12 "..reolink seems to have really perfected their algorithms check out this image from 10 feet away the Reolink imagine looks almost fake but it's just a result of their specific post processing .. "

"... Reolink the results are amazing..."


This is complete BullSh-t. So you're telling me Reolink which uses cheaper processing, has less engineers doing R&D is gonna beat Hikvision and Dahua?

This review is a dangerous one, full of misinformation and omissions. It is not a honest review by any measure, and all the thumbs up on the video are from people who do not know they've just been had by omissions of facts and failure to do a proper review.

This reviewer did such a frame up of the competition and failed to test any moving objects at night ( Reolinks major downfall ), that I highly suspect he is on Reolink's payroll.


1598654267932.png


Update:
At first the review looks legit, until you notice he compares are 4mm lens Reolink vs a 2.8mm Amcrest ( Dahua OEM ) and claims the Reolink is amazing "..reolink seems to have really perfected their algorithms check out this image from 10 feet away the Reolink imagine looks almost fake but it's just a result of their specific post processing .. ... Reolink the results are amazing..."
when the simple truth is he is comparing a 4mm lens vs a 2.8mm lens.. nothing amazing, just reducing pixel density for a wider FOV.. simple math stuff...
 
Last edited:

mat200

IPCT Contributor
Joined
Jan 17, 2017
Messages
7,128
Reaction score
6,638
Feel free to point out anything I missed..
 

mat200

IPCT Contributor
Joined
Jan 17, 2017
Messages
7,128
Reaction score
6,638
Reolink has been discussed and there are some videos captures with motion shared here at ipcamtalk.

To find the Reolink threads, use the search feature and search on titles:

 

mat200

IPCT Contributor
Joined
Jan 17, 2017
Messages
7,128
Reaction score
6,638
A fairer review would be the Reolink B800 (4mm lens ) vs an Amcrest 4mm lens like the IP8M-T2499EW-40MM or IP8M-2496EW-40MM ( both 1/2.5" sensor cameras )

Turret

Bullet
 
Joined
May 1, 2019
Messages
1,147
Reaction score
1,493
Location
Reno, NV
To be fair... that gentleman is more so geared towards Home Automation (which he knows much of) rather than security cameras. Tis like myself (low voltage wiring guy) trying to talk about what is the best dual handled bathroom sink. And he does mention Blue Iris as a starting reference point many of times, so that's good right? :)
 
Last edited:
Joined
May 1, 2019
Messages
1,147
Reaction score
1,493
Location
Reno, NV
If i were interested in debating in a open public forum, I would use science and facts. Example: image sensor size, h.264/h.265, cloud/local, dudes in Russia gawking at my camera footage/me gawking locally at my camera footage, IR distance...bleh bleh bleh.
 

mat200

IPCT Contributor
Joined
Jan 17, 2017
Messages
7,128
Reaction score
6,638
@looney2ns it looks like he responded to you. Getting my popcorn ready...
View attachment 69492
wow... The Hook Up has tested 50 different makes and models of cameras and failed to discuss the most basic math principles, that of pixel density ( and thus quality of image capture ) which in this particular test case varies significantly between the cameras compared at 25 ft due to H FOV and H Resolution, and instead claimed
" ..reolink seems to have really perfected their algorithms check out this image from 10 feet away the Reolink imagine looks almost fake but it's just a result of their specific post processing .. "

"... Reolink the results are amazing..."


No the results are in no way amazing, on the most basic level it is simply the result of comparing a 4mm lens( with a tighter H FOV ) vs a 2.8mm lens ( with a wider FOV ).

From the Cliff Notes:

The equation to calculate the Radius (the distance in feet) for identification for 100 ppf as discussed above: Radius = (( Horizontal Res / 100 ppf ) * ( 360 / Angle ))/2*Pi

Just looking at the lens differences based on the H FOV specs we can see a significant difference.

8MP = 3840 H resolution

(( 3840 / 100 ) * ( 360 / Angle )) / 2*Pi

H: 97° => 22.7 feet to the 100ppf density
H: 112° => 19.65 feet to the 100ppf density
 
Last edited:

Mark_M

Pulling my weight
Joined
Jul 2, 2019
Messages
211
Reaction score
166
Location
NZ
I was surprised when I saw an IPVM article about where Swann, etc OEM's the cameras. Apparently they've gone off Dahua/Hik and gone to this brand known as 'RaySharp'.
This 'RaySharp' probably buy the cheapest scraps from the bin and attempt to make it work. Nothing special like IVS we're known to see, just simple motion detection and it's 'amazing technology'.
 
Joined
Feb 26, 2017
Messages
4,576
Reaction score
3,105
Location
USA
Actually, I believe RaySharp is another predominant manufacturer in China after Hikvision and Dahua, but it doesn't get much attention here on the forum, because RaySharp doesn't seem to brand itself much outside of China; it goes through channels (i.e. rebranding)
I'd rather get directly from Hikvision / Dahua

I was surprised when I saw an IPVM article about where Swann, etc OEM's the cameras. Apparently they've gone off Dahua/Hik and gone to this brand known as 'RaySharp'.
This 'RaySharp' probably buy the cheapest scraps from the bin and attempt to make it work. Nothing special like IVS we're known to see, just simple motion detection and it's 'amazing technology'.
 
Joined
Feb 26, 2017
Messages
4,576
Reaction score
3,105
Location
USA
He's not the only YouTube channel with a misunderstanding in cameras, there's intermit.tech, and many more channels out there with a limited scope of understanding on these cameras. You should not be promoting 4K unless you are going to mention when, where, and how to use them. For most people, 4mp is the way to go.

Executive Summary:
DO NOT be a sheep.
Ask what a reviewer can be missing, ask if a reviewer is using hype words to sell the cheapest product.
  • Test your cameras and kits right away.
  • Just do a bench test.
  • Test real world conditions.
  • Test moving suspects.
  • Test at night, test at day.
  • Use a test rig to place cameras at positions which you plan to place the cameras.



Finding the BEST 4K Security Camera NVR Package (Reolink vs Amcrest vs Swann)
The Hook Up

4:30 - Mentions Blue Iris "NVR" interface is much better but probably not $600 better
Response: Many members here go with a good used i7 Business Class PC and save a lot on that. ( some getting a good PC for $200+/- and also getting a decent PoE switch ) - so it may cost a bit more than an NVR in a kit, but for many here they really like Blue Iris and feel it is worth it. ( note, if buying an NVR separately - it can cost more than going Blue Iris + used business class machine )

11:10 - "let's check out the daytime clarity of each camera"
this is the section which I feel is very poorly laid out.

Reviewer tests a image capture at 25ft of text.

Fact: Wider FOV = less pixel density at 25 ft.
Fact: Tuning parameters also matter.

Reolink B800
Image Sensor: 1/2.5" CMOS Sensor
Effective Pixels: 3840x2160 (8.0 Megapixel)
Lens: f=4.0mm F=2.0
Angle of View: Horizontal: 97° ; Vertical: 51°
ref:

Amcrest Camera IP8M-T2499EW-28MM
Image Sensor 1/2.5” 8Megapixel progressive scan CMOS
Focal Length 2.8mm
Max. Aperture F1.6
Angle of View H: 112°, V: 69°
ref:
and



The reviewer FAILS completely to mention FOV and Lens in his comparisons - which in my book is completely misleading.

11:49 - "night 25 ft."
".. results were very similar..." again testing an image capture at 25 ft of text.

claims the other cameras have very poor night time performance.

Again major FAILURE on the reviewer.
Failed to test moving subject.
Failed to even mention lens differences.
Failed to even mention FOV differences.

12:12 "..reolink seems to have really perfected their algorithms check out this image from 10 feet away the Reolink imagine looks almost fake but it's just a result of their specific post processing .. "

"... Reolink the results are amazing..."


This is complete BullSh-t. So you're telling me Reolink which uses cheaper processing, has less engineers doing R&D is gonna beat Hikvision and Dahua?

This review is a dangerous one, full of misinformation and omissions. It is not a honest review by any measure, and all the thumbs up on the video are from people who do not know they've just been had by omissions of facts and failure to do a proper review.

This reviewer did such a frame up of the competition and failed to test any moving objects at night ( Reolinks major downfall ), that I highly suspect he is on Reolink's payroll.


View attachment 69478
 

mat200

IPCT Contributor
Joined
Jan 17, 2017
Messages
7,128
Reaction score
6,638
Just to be clear... it is not a fair review when the reviewer fails to address pixel density, FOV, 4mm vs 2.8mm, etc.. and then claims the Reolink cameras are amazing ( when it is really just a simple 4mm lens vs a 2.8mm lens that makes for the "amazing" results at the expense of a wider FOV ). In fact it is an extremely deceptive review to those new to this subject.



1598761032999.png

Ref:
@The Hook Up, was this video and others you made sponsored by Reolink? I noticed the Reolink coupon code in the descriptions of the videos.
Hide 2 replies
The Hook Up
No, I generally do that after the fact. Inform a company that I chose their product and ask for a coupon code. All 3 NVRs systems were provided to me by their manufacturer and no other money changed hands.
Chewyfood
@The Hook Up Thanks for the clarification - appreciate you keeping up to date on replies, even on 10 month old videos. People at IPCamTalk have essentially blackballed your reviews claiming they are Reolink spam (especially since you don't test night vision movement, which, according to them, causes Reolink to have serious motion blur problems because of their "night vision cheating"). Side note, can you inform people in your video description that if you purchase on Reolink Website it will be recognized as a Foreign Transaction? 8% off is great, until it is countered by 3% foreign transaction fee - i.e. make sure to use a CC without those fees. Reolink site looks like it is US based, but the purchase gets sent to HK.
 
Last edited:

mat200

IPCT Contributor
Joined
Jan 17, 2017
Messages
7,128
Reaction score
6,638
To be fair... that gentleman is more so geared towards Home Automation (which he knows much of) rather than security cameras. Tis like myself (low voltage wiring guy) trying to talk about what is the best dual handled bathroom sink. And he does mention Blue Iris as a starting reference point many of times, so that's good right? :)
Good point Holbs,

Still his review is being used to sell Reolinks because of his expertise and seemingly "fair review", a review which is so deeply flawed due to the conclusion that Reolink is amazing with the 4mm lens vs the 2.8mm lens.

Until his review of the imaging, the Amcrest kit was ahead of the Reolink for other evaluated reasons. Reolink pulled ahead according to the review once he compared the images between the 4mm Reolink at 25 feet and the Swann ( ? I think also 2.8mm ) and Amcrest cameras ( 2.8mm )

Unsuspecting new purchasers will make poor decisions based on a significantly flawed review.
 

mat200

IPCT Contributor
Joined
Jan 17, 2017
Messages
7,128
Reaction score
6,638
Note:
Even users whom have experience with cameras are noting and referencing this video on other forums as a positive for Reolink ignoring the FOV / pixel density math.

"I watched a video comparison on YouTube between amcrest, Reolink, and swan. Reolink actually does the best at reading from a distance. Mind you this was a comparison of 4k cameras.
..
"

It is critical to the basics of security cameras to have an understanding of the most basic FOV relationship with pixel density and distance.

remember see the DORI section in the cliff notes on more info on this topic ( Wider FOV = less pixel density, and thus short ID distances, shorter DORI ranges.. )

fyi full quote:
I watched a video comparison on YouTube between amcrest, Reolink, and swan. Reolink actually does the best at reading from a distance. Mind you this was a comparison of 4k cameras.

I have the Reolink cameras, model D800, turret 4k. Excellent quality. The max distance that I can make out license plates is probably 75 feet, and that requires some digital zoom and squinting to try to make it out. Without too much manipulation I can see fine at 50 feet. It's not a limitation of the camera, its just that the size of license plate text is actually pretty small when looking from a distance. Beyond 75 feet, making out any details just becomes a challenge for any camera... unless the text is very big. When identifying neighbors who are walking in my area, I know who they are based on their body shape and walking style. Can't make out faces beyond 75 feet.
 
Last edited:

wittaj

Known around here
Joined
Apr 28, 2019
Messages
3,941
Reaction score
4,768
Location
USA
No way they are reading plates at 75 feet with a 4mm lens regardless of the camera, and certainly not a moving plate. Unless the plates are huge or they took great liberty at guessing plates.

So this person can read a plate at 75 feet but not make out a face - how does that make any sense?
 

mat200

IPCT Contributor
Joined
Jan 17, 2017
Messages
7,128
Reaction score
6,638
quick check on ipvm

Not a very clear license capture imho, remember this would be straight on - a slight angle would be a real challenge imho
( fyi for those new to this subject, please see the LPR section - license plates at night due to the reflective nature of the plates are a real challenge )


1607560384838.png
 

wittaj

Known around here
Joined
Apr 28, 2019
Messages
3,941
Reaction score
4,768
Location
USA
Yeah, but the plate would not be that big on the image either and once you zoom in it would be worse.
 

wittaj

Known around here
Joined
Apr 28, 2019
Messages
3,941
Reaction score
4,768
Location
USA
So for kicks, I had tried a 4mm lens set up for LPR (1/2000 shutter) temporarily next to the mailbox to see if it would be a good location. You couldn't read it much past 14 feet out.

Here is a plate at 75 feet out zoomed in to roughly the same size as the IPVM above:

1607563169970.png1607563183157.png
 
Top