BobDiaz
The Professor
We want all the advantages WITHOUT any disadvantages, but as any good engineer will tell you, "Design is sometimes a series of tradeoff choices." The Panasonic HVX-200 is no exception to this rule and all the other $10,000 HD Cameras must have several design tradeoffs. As potential buyers we have to be aware that none of these cameras are 100% perfect and there will be some tradeoff in selecting one over another.
Let's start with something simple; for a given CCD size, the higher the resolution, the smaller the size of a CCD's pixels, and the worse the low light performance will be. In addition, the CCD noise level is increased and dynamic range is worse.
It would be nice to have a 1/3" CCD with a resolution of 1920h x 1080v, BUT with so many pixels, it would be hard to even do 24p. You can forget about doing 30p, 60i, or even 60p with that many pixels. We solved the resolution issue, BUT the noise level, low light performance, and dynamic range will be VERY poor.
At the other end of the design selection, we could have a CCD with 720h x 480v. The gives us the best noise level, low light performance, and dynamic range, BUT resolution is VERY poor.
The best we can hope for is to select a resolution somewhere in-between those 2 extremes. Each manufacturer needs to decide which is the best choice for their design tradeoff.
JVC chose 1280h x 720v CCDs (about 0.88MP). When reading the CCDs at 60Hz, the data rate is too much for a single A/D (Analog to Digital) Converter. JVC's solution was to use dual A/Ds per CCD. Because A/Ds may not track exactly, there is a risk of the SSE (Split Screen Effect) when the camera is cold and the gain is increased. Had JVC read the CCDs at 30p or 24p, a single A/D could keep up and the SSE would never occur. I've read, BUT HAVE NOT VERIFIED, that the JVC does have problems with CCD noise and low light performance.
Canon uses a 1440h x 1080v image sensor (about 1.48MP). Because it is read in interlaced mode, it is reading 0.74MP 60 times a second. I would love to see the noise level of the CCD when the Canon's gain is set to +12dB. Maybe the images are out there, but I don't have the time to look for them.
Here's the deal, we know that 0.88MP requires dual A/Ds, but 0.74MP can be done with a single A/D. If you are given 777,600 pixels, give or take a few thousand pixels, to lay out in the CCD, what resolution would you chose? Say the pixels don't need to be square pixels. (777,600 pixels = 0.74MP) Just remember this is ONLY a WILD guess on my part as to pixel count.
As you try different numbers for horizontal and vertical pixels, you'll find that an increase in horizontal resolution results in a decrease in vertical resolution. Likewise, an increase in vertical resolution results in a decrease in horizontal resolution. While there is no one best choice, the safe ground is the middle range.
In the same way, Resolution, CCD Noise, Dynamic Range, and a whole list of other choices may be selected with one parameter's advantage working against another's advantage. Thus we have engineering tradeoffs.
Barry is going to do a 4 camera test where the footage of one camera is compared to another. Let's assume the test has several categories, like: Horizontal Resolution, Vertical Resolution, Image Contrast, Color Saturation, Low Level Light Performance, Noise at +0dB Gain, Noise at +6dB Gain, Noise at +12dB Gain, ... It seems likely that no one camera will be number 1 in all the categories. In that case, how does one decide a winner?
We all live under that same laws of physics and it's not that Canon, JVC, or Sony can do some magical thing that Panasonic can't do. What we do have are a series of design choices that make one camera more attractive than another is some areas. To look at only one parameter, like horizontal resolution, and ignore all other parameters is foolish. The design decisions come as a package deal and you'll need to accept the good points with the bad points. Just remember, EVERY camera has its limits.
Bob Diaz
Let's start with something simple; for a given CCD size, the higher the resolution, the smaller the size of a CCD's pixels, and the worse the low light performance will be. In addition, the CCD noise level is increased and dynamic range is worse.
It would be nice to have a 1/3" CCD with a resolution of 1920h x 1080v, BUT with so many pixels, it would be hard to even do 24p. You can forget about doing 30p, 60i, or even 60p with that many pixels. We solved the resolution issue, BUT the noise level, low light performance, and dynamic range will be VERY poor.
At the other end of the design selection, we could have a CCD with 720h x 480v. The gives us the best noise level, low light performance, and dynamic range, BUT resolution is VERY poor.
The best we can hope for is to select a resolution somewhere in-between those 2 extremes. Each manufacturer needs to decide which is the best choice for their design tradeoff.
JVC chose 1280h x 720v CCDs (about 0.88MP). When reading the CCDs at 60Hz, the data rate is too much for a single A/D (Analog to Digital) Converter. JVC's solution was to use dual A/Ds per CCD. Because A/Ds may not track exactly, there is a risk of the SSE (Split Screen Effect) when the camera is cold and the gain is increased. Had JVC read the CCDs at 30p or 24p, a single A/D could keep up and the SSE would never occur. I've read, BUT HAVE NOT VERIFIED, that the JVC does have problems with CCD noise and low light performance.
Canon uses a 1440h x 1080v image sensor (about 1.48MP). Because it is read in interlaced mode, it is reading 0.74MP 60 times a second. I would love to see the noise level of the CCD when the Canon's gain is set to +12dB. Maybe the images are out there, but I don't have the time to look for them.
Here's the deal, we know that 0.88MP requires dual A/Ds, but 0.74MP can be done with a single A/D. If you are given 777,600 pixels, give or take a few thousand pixels, to lay out in the CCD, what resolution would you chose? Say the pixels don't need to be square pixels. (777,600 pixels = 0.74MP) Just remember this is ONLY a WILD guess on my part as to pixel count.
As you try different numbers for horizontal and vertical pixels, you'll find that an increase in horizontal resolution results in a decrease in vertical resolution. Likewise, an increase in vertical resolution results in a decrease in horizontal resolution. While there is no one best choice, the safe ground is the middle range.
In the same way, Resolution, CCD Noise, Dynamic Range, and a whole list of other choices may be selected with one parameter's advantage working against another's advantage. Thus we have engineering tradeoffs.
Barry is going to do a 4 camera test where the footage of one camera is compared to another. Let's assume the test has several categories, like: Horizontal Resolution, Vertical Resolution, Image Contrast, Color Saturation, Low Level Light Performance, Noise at +0dB Gain, Noise at +6dB Gain, Noise at +12dB Gain, ... It seems likely that no one camera will be number 1 in all the categories. In that case, how does one decide a winner?
We all live under that same laws of physics and it's not that Canon, JVC, or Sony can do some magical thing that Panasonic can't do. What we do have are a series of design choices that make one camera more attractive than another is some areas. To look at only one parameter, like horizontal resolution, and ignore all other parameters is foolish. The design decisions come as a package deal and you'll need to accept the good points with the bad points. Just remember, EVERY camera has its limits.
Bob Diaz
