IRE and Stop correlation

M

morisato

Active member
Can someone explain the correlation between the IRE value and stops?

I don't believe IRE points have an equivalent value (in the sense that each IRE point adds an equivalent amount of light) with eachother but I could be wrong. The reason I think so is because...

Everyone knows 1 stop is a difference of 1/2 or 2 times the light depending on whether you are stopping up or down. If you light a face properly to 70-80 IRE and it is advised that the background be 2 stops below your foreground (1/4 the light), the IRE of the background would be, generously, around 20 IRE. With 7.5 being absolute black... 20 seems awfully low. Can anyone enlighten me?
 
As I understand it IRE represents the comprehensive total luminance range represented by a video signal. Stops are obviously how the camera handles exposure. Now where the two meet is where it gets iffy becaus the range from where one camera clips highlights at 100IRE and clips shadows at 7.5 may only be 6 stops on some cameras, and as high as 9 or 10 on another camera. Generally IRE is useful as tool for judging where certain exposure values fall in the image, you can judge against each part of the image on the scale where it falls. To convert that to stops you would have to know the Dynamic Range of your camera in stops from black to white and then mathmatically derive on an IRE scale what some approximate equivalent would be. Once again your talking about two different tools that measure light different ways. Anyways thats my take feel free to crrect me anyone.
 
Yeah, that sounds right to me. Where the IRE/stop correlation seems to come up is, if you were shooting something to 70 IRE for example, and you needed the background to be 2 stops lower, you wouldn't be able to come down two stops on the camera and expect the correlation between the IRE and f stop to stay the same because the IRE works on luminance, black, to complete white, and f stops work upon the size of the aperture. For example, you may stop down to f/8, but your IRE could still be much higher (or lower) than what seems like it should be, because different cameras see luminance differently, as well as exposure. Latitude would play a huge part in this.

Basically I just said the exact same thing as Kendall...Ah well.
 
certainly, on the panasonics, what scene files your using will add or subtract from your cameras lattitude....correct? when adding up panasonic's entire lattitude you'll need two or more scene files to accomplish this? Does any 'one' scene file cover the entire dynamic range [ that your camera can reproduce] as opposed to dividing it between scene files?
 
Last edited:
Pretty sure that shooting cinelike d is gonna give you the widest dynamic range at the expense of more noise because the camera is boosting and stretching the signal to cram the most stops of exposure into the chips.
 
As Tatsuo said Cinelike "D" is the most dynamic range with "V" losing maybe 1 stop of light but being significantly cleaner in the signal. Adding up the total of the camera is the total given amount it can capture at the highest settings, its not spread across scene files. The dynamic range is measured from the top of the gamma curve where the highlights are clipped down until the bottom is no longer linear and begins to curve into the noise floor. The "D" setting captures the widest part of that curve that camera is capable of. Other gamma settings are simply manipulating that same curve to get different responses. For example apllying a subtle curve to the bottom in "B Press" curves the bottom of the gamma, effectively crushing the blacks at the lower end of the dynamic range and effectively you would loose a stop or two of dynamic range. However as you said to know the correlation between IRE and stops you would effectively need to know how every gamma curve affected the dynamic range of the camera. I would recommend working with a light meter to set the lights and then confirming with the waveform monitor that everything falls where you want it.
 
Gamma changes things, but roughly it's like 1 stop = 15-20 ire.

is this true for the dvx or another camera?
is there a thread, somewhere here on dvxuser, on this F-stop / IRE correlation...Id like to read that
 
There is no correlation. As LighthouseMedia said, it's all dependent on what your camera can render.

Take an example of, say, an old VX1000. It might have a total dynamic range of 5 stops. And let's compare that to a new HPX500, which has a total dynamic range of 10 stops.

If you point both of those at a grayscale chart, they're both going to render an image that goes from 0 to 110 IRE, but those IRE will be completely and thoroughly unrelated from one cam to the other. Where the VX1000 wouldn't be able to hold onto a medium dark gray it'd render it at 0 IRE (because it's off the bottom of the VX1000's latitude) but the HPX500 might still see it as a strong gray and render it at 20 IRE. And the same thing will happen on the top -- the VX1000 might render a medium light gray as 110 IRE (blown out) but the HPX500 could easily hold it at, say, 80 IRE. So in this highly theoretical case, the difference between four stops on the VX1000 would be 110 IRE (or, say, somewhere around 27 IRE per stop) whereas on the HPX500 the difference between those four stops would be 60 IRE, for something on the order of 15 IRE per stop.

So it's not something you can judge empirically -- there is no set correlation. It'll vary for every camera, and even vary within the same camera depending on how you have the image processing settings set.
 
You must log in or register to reply here.
Back
Top