Maybe I don't "get" f-stops

[Paul Anderegg]

Camera 1: JVC HM790 with f1.4 Fujinon lens, f10@2000 1/60 0db

Camera 2: Panasonic SPX800 with f1.8 Fujinon lens, f13@2000 1/60 0db

f1.4 to f1.8 is 2/3 of a stop faster. F13@2000 is 2/3 of a stop faster than f10@2000.

So I would assume that the JVC would be equal in low light to the Panasonic, but the Panasonic is 6-9db faster than the JVC. Can anyone explain in simpler terms why this is, and why my assumptions are incorrect?

For giggles, would a Sony PDW700 (f11@2000) with an f1.8 2/3" lens be the same in low light as a JVC HM890 (f11@2000) with an f1.4 1/3" lens? Lets for a moment say they have the same Fujinon HA/HAs18x glass in different mounts.

For even more giggles, if the Canon KH21ex 1/2" lens has a wide open f1.4 iris value, and you put it on a Sony PMW320 that states it has an f1.6 optical block, what would the camera have to weigh on Mars to be rated at f22@2000 using a dichroic filter.........sorry, my brain broke. :-\

Paul

[Brian Drysdale]

The sensitivity of the camera and its sensor, just like film speed ASA setting, will affect the f stop that you need to use. If a camera is more sensitive, the more you'll need to close the lens aperture in order to reduce the amount of light falling onto the sensor(s).

That's what you can see happening in the manufacturer's sensitivity figures the Panasonic has to to be stopped down at the same light level compared to the JVC, so it's a more sensitive camera.

[Noah Yuan-Vogel]

First of all, I assume you are giving the sensitivity specs quoted for 2000lux on each of those cameras. Why would it matter how fast the lens is wide open, if the spec specifies a value for them stopped down.

There are a lot of variables in those actual numbers...First off I assume those numbers are for the camera with that given lens? But what is the T-stop of the lens? that could easily vary by 1/2 stop with differently designed and coated zooms with so many elements. Also, those sensitivity ratings depend a lot on what is considered to be the middle grey of the gamma curve of each manufacturer's different gamma curves. That part is quite arbitrary. And dependent on the particular gamma curve and how the manufacturer specifies the gamma curve to be used.

[Paul Anderegg]

That I understand, if the JVC were stopped down to f13, it would not see a full 2000 lux.....which is why it seems odd to me that you can stop the Panny down to f1.8 vs f1.4 for the JVC, but that doesn't mean it looses light. :-|

On a side note, the Panasonic HPX370 is a 1/3" f10@2000 camera just like the HM790, but it as well does better in low light at the same gain settings. I think JVC lies about their camera ratings. They rate their HM650 at f11/12@2000 (NTSC/PAL), but they neglect to tell you that you have to enter EXTENDED mode to achieve that, which ads a base 6db gain, added underneath the 0db you see on the switch in the LCD display. Scoundrels!

Paul

[Daniel Epstein]

Paul,
I think what you are seeing is the variance between different manufacturers ratings of sensitivity or ASA. I have been noticing this over the years. This isn't like the days of old (film) where people relied on their meters to set exposure so the ASA accuracy was critical. Video gain and signal to noise are related so the manufacturers can play with what ratio they want to publish. The better the signal to noise on the camera the higher one can go on gain and still have a useable picture. Usually numbers within a manufacturers line are a better direct comparison than numbers between manufacturers. Also as Noah says different gammas and settings can give you a different looking exposure of the overall picture so you should try and do you own test comparisons

[Paul Anderegg]

The Panasonic had such a high (65db?) s/n ratio, perhaps it affects the base "news gamma", allowing much brighter dark areas at 2000 lux. I can see the JVC has such horrible s/n ratio that it must be electronically tuned to not crank up the signal in the dark patches, for fear they will noise up!

Paul