Megapixel myth
Does more megapixels guarantee performance?
Short answer: Quality before quantity.
Sales people and marketing departments often use the megapixel count as a numerical value that performance can be judged. Unfortunately, this is not a guarantee, and the megapixel count can have very little to do with quality. There are many other factors that can influence the end result. As an extreme example if you were to smear vasoline on an otherwise perfectly good lens of an 8mp camera, the end result would be a very soft 8mp image. So, if a lens is of inferior quality this will compromise the performance, and what of the lens’s ability to handle low light. What is its widest aperture (F-stop)?. There’s also a processor involved that needs to process the image. Can it handle wide dynamic applications? How well does it perform this role?
So why are we not talking about these important aspects of the camera? Because it’s easier to sell with a singular number. Megapixel. For some, more is meant to imply better.
Pixels are light sensitive elements on a CCD chip. Typically for most CCTV cameras this is a 1/3” CCD chip. And for the most part the more pixels on a 1/3” chip means the smaller the pixels. (Keep this last sentence in mind, as we will visit this shortly)
Megapixel values is how many pixels a CCD chip has counted horizontally and vertically. Typical CCTV “resolutions” are listed below:
- 2MP = 1920×1080
- 3MP = 2048×1536
- 4MP = 2560×1440
- 5MP = 2592×1920
- 6MP = 3072×2048
See below illustration of what this would actually equate to;
Like calculating an area, megapixel count is the multiplication of horizontal by vertical. It only takes a 40% increase in linear dimensions to double the pixel count. Ie a doubling in Pixel count actually relates to an approximately 40% increase in linear resolution
Notice in the bullet points above how the 4MP resolution is actually only 640 pixels wider and 360 pixels higher than 2MP resolution. 4MP therefore is not actually 2 x 2MP
To view these images on a typical 1080P HD screen they would all look the same. It is if/when electronic zoom on recorded images is performed that the difference in resolution becomes apparent.
As mentioned above one trade off for increase in pixel count is smaller light sensitive elements (pixels) on the same 1/3” landscape. Colour reproduction, low light performance and the ability to handle high contrast scenes can then become compromised with the more pixels you have. Why is this? It’s because the more pixels that populate the same 1/3” area means that each pixel needs to be smaller. The smaller the pixel, the smaller the area it has to capture light. It effectively becomes a smaller bucket to capture light. Meaning it needs to work harder to produce the same result as a natively larger pixel.
Below is an example of the same brand and model camera with the same lens. Images were taken at the same location, same conditions and on the same day/time. One camera is a 2MP the other, 4MP. Both are Tiandy with fixed F1.2, 4mm lenses
Day time comparison
2MP camera
4MP camera
Electronic zoom on both images.
2MP camera
4MP camera
Night time comparison
No illumination other than the cameras own IR illuminators are being utilised. The nearest street light is over 60 meters away and out of field of view. The light in the background is actually from a home
2MP camera
4MP camera
Notice the individual in the driveway next to the hedge on the left in the below images?
2MP camera
4MP camera*
*Individual is in the same spot in this image, however can barely be seen
The 4MP camera does show some difference in resolution upon electronic zoom, however it does not provide twice the resolution one would be lead to believe. Notice the difference in low light performance? The smaller pixels have a harder time in lower light. What some cameras may do to compensate for this is to extend the length of time the shutter stays open. Trade off with this however is what is often called comet trails, or simply softer images on moving objects. So, if your requirement includes surveillance in low light conditions less would be more for this instance, and 2MP would be advisable.
And what about a 6MP camera?
For comparison, below is another manufacturer’s 6mp camera. Focal length on this camera is 2.8mm
6MP camera from a “leading” manufacturer
What if I want colour at night?
If night time surveillance is a criteria, a camera utilising Starlight technology can render colour images even at night. Below is a sample image from the Tiandy Starlight camera. Some ambient illumination is required to achieve this performance. (note: These are 2mp cameras!!)
Good, high megapixel cameras have their use, and camera manufacturers are continuing advancements to develop and improve resolution.
The purpose of this article is to inform. The higher the number of pixels does not guarantee performance, nor should it be an end unto itself. Pixel count should be one factor in determining what product is suitable for your application. High megapixel cameras are best suited for well lit conditions where specific identification such as number plates are required when performing electronic zoom on recorded images. (Having said this a well placed 2mp camera is more than capable of identifying number plates if this is actually required).
With each increase in pixels there is a corresponding increase and load on network bandwidth. Larger files are also recorded, meaning more hard drive is required to archive footage as the megapixel count increases.
In short, choose the product that best suits your application, and not simply off the promise of performance based on a singular number. Best way to determine quality is to view actual footage and images. Ask for example footage, see a demonstration, be mindful though that most things look great in isolation, so make comparisons or choose a manufacturer that not only sells on numbers, but is happy to demonstrate performance.
* cameras used for demonstrations were Tiandy TC-NC214 and Tiandy TC-NC414(2&4MP). Starlight ccamera images were taken with Tiandy TC-NC23MS