Customers often want wider FOV, and yet when they compare image captures they often end up with the impression that a tighter FOV camera is a better camera not realizing that the major part of that difference they see is the FOV / pixels on target. ( see the DORI section of the
cliff notes )
The fact is that
FOV matters .. and you should pick a good camera for the purpose of that view .. if you want a camera for those at your front door, I like a 2.8mm mini-dome wedge at face level.
For a camera covering the front of my driveway I want a varifocal camera that I can adjust just to cover that area ( or perhaps a 6mm lens. )
For most camera positions, a 3.6mm lens is a typical choice. ( it gives about a 90° FOV for many models )
Comparing a 2.8mm lens with one that is closer to 4mm leaves poor perceptions by those new to this topic. Many people new to this subject, and some unclear of the nuances .. will misunderstand the reality of the situation.
Yes, there is more nuance to this than just FOV .. the sensor size matters, the f-stop, lens quality, DSP chips, algorithm, .. and even the tuning of the parameters. ( hint, most testers will just use the factory defaults as it makes testing easier.. in the real world case, you will want to tune the parameters to get a better image capture .. )
When looking at online reviews, and many review on Youtube:
many Reolink cameras have a tighter FOV giving you a longer DORI range .. thus the text may appear sharper compared to other cameras which have wider FOVs .. this is why Reolink seems to produce sharper text than 2.8mm cameras.
It is NOT because Reolink is a better camera, it is simply because Reolink is compared to a camera with a significantly wider FOV.
Ideally, the reviewers SHOULD be comparing similar lens models. ( example
Amcrest 5MP 3.6mm in this case vs the Reolink 5MP .. more on that below )
FOV matters, Math matters, Physics matters ..
From Rob's TheHookUp video referenced in the first post:
He lists:
Top Picks:
Reolink RLC-510A:
https://amzn.to/3qdBy0s
Amcrest 5MP-T1179EW:
https://amzn.to/3KGRfHc
Amcrest
Amcrest 5MP Turret POE Camera, UltraHD Outdoor IP Camera POE with Mic/Audio, 5-Megapixel Security Surveillance Cameras, 98ft NightVision,
2.8mm Lens (103° FOV), IP67, MicroSD (256GB), (IP5M-T1179EW)
Field of View: H 103° V 71° D 132°
Reolink
REOLINK Security IP Camera Outdoor, 5MP Home Surveillance Outdoor Indoor PoE Camera, Human/Vehicle Detection, 100Ft IR Night Vision, Work with Smart Home, Lens
4mm, Up to 256GB Micro SD Card, RLC-510A
Field of View Horizontal: 80° Vertical: 42°
Just the FOV difference will result in the Reolink having more pixels on target at the distance Rob is holding the text .. thus creating the false impression that the Reolink does a better job without covering the issues of FOV differences.
" .. the Amcrest 5MP that had consistently decent performance in all conditions, between the two the Amcrest also has a significantly greater field of view at 105 degrees vs the Reolink's 80 degrees which could definitely account for some of the differences in the image clarity at 25 feet but it shouldn't completely discount the Reolink's excellent daytime performance .. " - TheHookUp *
Let's do the math .. as Math Matters .. ( Physics .. )
From the cliff notes, DORI section:
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
Amcrest ProHD Outdoor Security IP Turret PoE Camera, 5-Megapixel, 98ft NightVision, 2.8mm Lens, IP67 Weatherproof, MicroSD Recording (256GB), White (IP5M-T1179EW-28MM)
2.8mm Lens with a 103° viewing angle
Image Sensor 1/2.7” 5Megapixel progressive CMOS
Effective Pixels 2592 (H) × 1944 (V)
Angle of View H: 103° / V: 71°/ D: 132°
Thus
Radius = (( 2592 / 100 )* ( 360 / 103 )) / 2*Pi =
14.4 feet for the "100 ppf ID distance"
IP5M-T1179EW-36MM ( which currently is not available in the Amcrest store ) specs report:
H: 84° / V: 58°/ D: 110
Radius = (( 2592 / 100 )* ( 360 / 84 )) / 2*Pi = 17.688 feet for the "100 ppf ID distance"
Reolink RLC-510A specs listed as ( from Reolink website )
2560x1920 (5.0 Megapixels)
Field of View
• Horizontal: 80°
• Vertical: 42°
Lens f=
4.0mm fixed; F=2.0, with IR-cut filter
pixels listed a bit less ...
Radius =((2560 / 100 )*(360 / 80 )) / 2*Pi =
18.34 feet for the "100 ppf ID distance" ... ( if the same resolution as Amcrest .. then 18.57 feet )
So, to really compare the quality of the image capture between the Reolink 5MP RLC-510 and Amcrest 5MP 2.8mm cameras and reduce the FOV / lens differences as much as possible you should compare the image captures from the following distances: ( "ID distance" )
Amcrest 5MP 2.8mm - 14.4 feet
Reolink 5MP RLC-510A - 18.3 feet
Since most of the video comparisons compare a Wider FOV Amcrest 5MP cameras vs the Reolink 5MP .. the quick view seems to show the Reolink is better quality in image capture ( example of the TEST text ).
Ideally you want the same lens strength in the cameras you are comparing .. in this case Amcrest did have a 3.6mm version, and that would be better to compare to this Reolink model.
Such it is, I believe most just looking at the image capture will over look the math and physics .. and think the Reolink is somehow superior even tho Rob only has one leg in the low light image capture with the Reolink and has 2 legs with the low light image capture with the Amcrest 5mp
Also remember, best to tune the parameters for your setup to get better results, as each location has different lighting ..
Ok .. so at 25 feet ... find X .. ( ppf )
Radius = 25 feet = (( Horizontal Res /
X ppf ) * ( 360 / Angle ))/2*Pi
X ppf = ( Horizontal Res * ( 360 / Angle ) ) / ( 25 * 2*pi )
So at 25 feet .. ( i'll use the same H Res for both cameras )
Amcrest 5MP 2.8mm 103° :
57 ppf = ( 2560 * 360 / 103 ) / 50*Pi
Reolink 5MP 4mm 80°:
73.4 ppf = ( 2560*360 / 80 ) / 50*Pi
Comparing the 2 .. Reolink / Amcrest ppf at 25 feet = 73.4 / 57 = 1.28
SO the Reolink has 28% more pixels per foot at 25 feet ..
Do you think 73.4 ppf vs 57 ppf makes a difference?
Do you think this makes a significant difference?
imho YES .. absolutely. ( remember TheHookUp does note this, but I do not think he's done to math to really see how significant this difference is .. )
Now, had we had a Amcrest 3.6mm which has a 84° H FOV the ppf calculation ( again, I'll use the same H Res )
69.9 ppf = ( 2560 * 360 / 84 ) / 50*Pi
The Amcrest 5MP 3.6mm is the model you really want to test vs the Reolink 5MP ..
It really is important to understand the DORI principles and FOV when planning your setup.
ipcamtalk.com
