Svard Cameras

Looking to Upgrade Your Cameras from an iPhone to Stream?

A lower resolution three-chip camera can actually render more accurate images to that of a higher resolution single-chip camera when you compare on a pixel-by-pixel resolution.

Let's face it camera technology has come a long way over the last two decades. Does anyone remember the first-generation digital cameras from the 1990s that used floppy disks?

The first thing is get the lighting right, as no camera will make up for poor quality lighting.

The portable camera has for the most part been replaced with the latest generation of smartphone. The current image sensors are able to record HD—and even 4K resolution—with pretty surprising results.

In order to really dig into the what, how and why of camera technology and what makes one better than another, I need to dig into a bit of the technical details. That way you can understand what sets a $5,000 broadcast camera (or one even more expensive) apart from a $500 smartphone or DSLR camera.

For those of you who have followed me on laser projector discussions you know that I talk about the single chip DLP projector versus the three-chip projectors, and how they offer much better color brightness and accuracy. The reason is that the one-chip projector only shows one color at a time. To display the full range of color, it flashes a red image, then green, and blue sequentially. This happens so fast that most of the human population renders that as a full-color image. The three-chip projectors are showing all three colors continually, so you always see a full-color image.  

The simple answer here on the camera side is also true to an extent. There are single chip cameras and three-chip cameras.

The primary way an imaging sensor works is by light hitting it and the photons energizing the photo diode. Unfortunately, I don't have the bandwidth in this article to get into what the P-N junction is (in short, it is a boundary or interface between different types of semiconductor materials), and how the phosphors or boron doping affect that, so let's rather look to how these things work.

The Basics Behind Image Sensors

All image sensors are "dumb" to color they generally sense light from about 190 to 1,100 nanometers. In the middle of that range is visible light which for most is around 380-700nm. For those of you who geek out about what frequency that is it is 430-770THz. For those of you unfamiliar with tetrahertz, know that kilohertz represents 103  Hz, megahertz equates to 106 Hz, while gigahertz amounts to 109 Hz, and lastly, terahertz stands at 1012 Hz.

With both single and three-chip cameras, a color filter sends a certain bandwidth of visible light for each pixel. There is a pattern of what colors are in what order, which is called the Bayer Filter.

The Colors of Red, Green, Blue

Most all of the single-chip cameras use a Bayer pattern on the pixels to get the Red/Green/Blue color variation. For a 1 million pixel image sensor, 500,000 pixels are green and 250,000 each of red and blue. Why double the green pixels compared to the other two colors, you ask? That's because the human eye is most sensitive to green light. From there, the imaging processor must fill in the void areas with the information to make a complete image.

So why does this matter?

Well, for one, the Bayer pattern leaves "holes" in of each of the colors that the image processor has to interpolate and fill in the details of the blank areas. In the three-chip setup, there is pixel for pixel accuracy for each color. What this equates to in the real world is much higher color sensitivity and also accuracy. The color accuracy has two notable factors pixel fill factor and the tightness of the color filters used. The prism of the three-chip camera has high quality dichroic filters with a steeper cutoff where the single chip camera uses a less precise filter. That makes for more overlap where the camera's image processor must guess as to which color it is.

What to Look for in a 4K Camera Versus a 1080 Camera?

One such question I regularly see in my Facebook feed, is "Which one is more future proof for my church?"

People see 4K and immediately think it will offer a better image than a 1080 camera. But let's go back and compare if we have a single-chip 4K camera versus a three-chip 1080 camera. The 1080 camera will offer better color depth and image sharpness with a 1080 image. If you were to sensor crop the 4K to get a 1080 image, the three-chip camera would eat it for lunch in low light performance, and also color accuracy.

Several questions also come up regarding the short depth of field with a large image sensor. If you have a large 4K sensor that you are using for a 1080 image, though, that crop would be a much smaller area on the sensor. This also leads to the discussion of the Bayer pattern and that is in an optical sense the traditional Bayer pattern, which is more along the lines of a 4:2:0, where the rows of pixels alternate red-green and then the next row alternates blue-green. This is true of your smartphone camera, up to the most expensive 4K or larger single-chip cameras.

The algorithm of deciphering this is a process called debayering. This process is different from camera to camera and as such, the image processing available on today's smartphones is frankly quite good, compared to point and shoot cameras of yesteryear.

Now the three-chip camera can provide a 4:4:4 signal that is not interpolated like the single chip which means it has more data about the color and pixel accuracy. So a lower resolution three-chip camera, can actually render more accurate images when you compare on a pixel-by-pixel resolution.

So enough with all the technical mumbo-jumbo. What does this all mean if I want to upgrade cameras for my church? Particularly, if we are looking to stream or broadcast to another campus?

I will personally advocate that if a church is just starting out with streaming, or are looking to get online, that you first work on the product of what you are trying to present.

The first thing is get the lighting right, as no camera will make up for poor quality lighting. Then get your program worked out, so that what you are presenting is well organized.

Second, start small and see where it gets traction. In all seriousness if you are on a shoestring budget, start with Facebook Live, streaming from a mobile device.

When the time comes to getting up to speed with a multicamera system that has a switcher and is either streamed, broadcast or sent to another campus via Sneakernet, i.e., not streaming but transporting by way of flash drive, CD or DVD, make sure you carefully evaluate the camera selection. The three-chip camera will offer better color rendering and low light performance than a similar size single-chip camera.

So an example of where what is spelled out in the third image matters, is where AVE two years ago installed a set of 4K "studio" cameras, that were single chip into a church. While the resolution was great, their performance during a normal worship service left a bit to be desired. Most notably with the low-light level.

The church opted to turn the gain all the way up on the camera, which added noise to the picture. The results were still acceptable, but it left no margin as the gain was at maximum. The only other option was to increase the light level.

On the flip side, another project that the integrator did, involved a set of three-chip 2/3-inch 1080 cameras. Even with the very low-light levels during worship, the image was still crisp and rendered accurate color. So much so, that if compared side by side with the aforementioned 4K cameras, the choice was easy for anyone to see that the "lower resolution" camera was the better choice.


Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.