F-stop calculator

[Vidgamer]

Hey guys, check out my latest f-stop calculator! I made it into an offline app for the phone -- you can set it onto your home-page and it looks just like a built-in app! It'll work on your desktop computer's browser just as well.

http://www.computingbits.com/photography/fstopOffline/

Anyway, let me know if you like it. If you don't like it, don't let me know.

[Greg Beetham]

What I want is a Bv (brightness value in the EXIF) to EV calculator so I can try to get a sense of how a particular camera is metering and how it’s calculating exposure as related to the particular ISO setting. Most of the calculators on the web are still back in the film era where it was all based on 100ISO. Now sensor gain (as related to ISO) is a variable just like f-stops and shutter speeds, and I suspect has been a variable with some degree of flexibility for some time despite having set a particular ISO (which is shown in the EXIF as a fixed value, but I have doubts about how ‘fixed’ it actually is).
Greg

[Vidgamer]

There are two problems with digital ISO as I see it:

• The manufacturers don't stick to the real ISO standard (so ISO 100 isn't really ISO 100).
• Things like DRO can pull up the shadows, leading to more noise, as if you were using a higher ISO.

So, even if you calculate a correct number, you'd find that the ISO didn't match.

Then there's the additional problem where if you have a very fast lens, the manufacturers cheat on that too -- I don't know that you can even tell that they are by looking at the embedded EXIF values.

[Greg Beetham]

Yep quite so, but I guess you have to start somewhere, and the baseline is the metering and the derived value the camera started with to calculate exposure, whether that actually matches the value if it was measured independently is another question too.
And then there is the different types of measuring within the camera to take into account, spot, center weighted and segmented and how the Bv is arrived at with each different type. I think sensor gain + metering + camera processing is a can of worms that not even DXO has really got on top of with their style of analysis because now there is apparently different sensor gain applied to different areas of an image automatically (or is in the works)…how do you measure that even if you wanted to check and what standard and baseline would you use?
And yes T values don’t necessarily agree with f-stops either, but the metering is supposed to easily take that in its stride, we never get to see the actual T-stop of a given lens, we only see the f-stop recorded in the EXIF, and seeing as how the camera controls the f-stop now instead of the lens we can take that with a pinch of salt too, by that I mean as to what f-stop it actually did use as compared to what was set by the photographer and noted in the EXIF.
Minolta brought us step-less shutter speeds and variable apertures 25 years ago and I guess if the photographer set one or the other it would have been roughly adhered to by the system if it felt obliged.
Greg

[Vidgamer]

DxO (and some other site I don't recall) has enough of a grasp on things to detect that manufacturers were faking the ISO when using wider apertures like f1.4. Apparently the steep light rays are not as easily detected, so it doesn't gather as much light as it would have on film. (By the way, this makes sensor size even more important, as, presumably, you can't rely on a large aperture from giving you an equivalent look.)

Anyway, what is a camera? Film did complicate things in that the film was fixed at a certain ISO. With a digital camera, the ISO can be whatever. So, what makes one camera different from another? When I'm using a camera, the only things that matter are:

• The shutter speed
• The amount of noise (and detail) in the resulting photo

Ok, you can also compare things like dynamic range, but for the most part, people are concerned about noise -- perhaps looking at the resulting detail is really the key. Using RAW and a good conversion program and careful use of NR, you can extract a lot of detail even from high ISO photos.

I may not have control over the aperture, or it may only go but so wide. I may be limited by sensor size or quality. But all the camera does is gather light, but in some situations, you need to keep the shutter speeds up. It doesn't matter if the manufacturer fakes the ISO, if I use the same shutter speed with two cameras, I can compare the results.

On DPR recently, someone posted an argument that the LX5 produced better low-light photos than the A6000, but the shutter speeds were vastly different. Thus, the comparison made no sense to me.

So, the way I look at it, it's just a light-gathering box.
______________________________________________________________

There's a free iPhone app (ad-supported) called Light Meter that will calculate the (film-style) camera settings for you -- ISO, shutter speed, and aperture, and allow you to make modifications, using information it gets from the camera. It can (optionally) also display an EV value, but as you say, it's based on ISO. You can change the ISO and get a different EV value. Maybe you can explain to me what that means -- I agree, it would be handy to know the actual brightness of light, but I guess it has to be tied to some scale.

Anyway, if there's a formula for what you want, it can be programmed. If you had enough examples, you could probably derive the formula.

[Greg Beetham]

Yes of course the camera is a light box in simple terms, my interest is in predictability of the light box and what it’s likely to do in a given situation barring mechanical snafus of some kind, and another interest of mine is trends.
I saw somewhere that someone claimed that a conversion of camera Bv to EV was as simple as adding 5 to the Bv value, but I have yet to do some tests to find out, I have my doubts about that one.
The EV is just a number that represents the chosen f-stop and shutter speed combination, it’s only useful to gauge the level of light that was present, the higher the number the brighter the ambient was, the only problem with it is a particular EV can represent a whole string of different possible f-stop and shutter speed combinations.
According to the Wiki chart an EV of 0 is equivalent to a shutter speed of 1s and an f-stop of 1.0 regardless of ISO, but you can also have a shutter speed of 64minutes at f64 and still have an EV of 0.
You can also have a 1s shutter speed at f64 and it’s EV is 12 (EV table here http://en.wikipedia.org/wiki/Exposure_value ) but looking at the EXIF and checking the EV matches the listed shutter speed and f-stop (which it should of course) doesn’t help understand how a particular camera is metering and setting exposure in conjunction with what’s going on with ISO and internal processing at a given light level.
You can get cameras now where you can set the shutter speed and f-stop so they are fixed for a given exposure and the ISO is the remaining floating variable chosen by the camera to make the exposure correct, I was interested in that process too although my cameras might only regulate ISO up to a certain point when it’s set to [Auto].
My thinking initially was to do some snooping with the relationship of the metered light value (Ev) (not the same as EV) and see if there was any kind of discrepancy or not with the resultant EV, taking the ISO setting into account as well.
Assuming that the more correct term (Ev) is equivalent to the EXIF camera term Brightness value (Bv) that they use, and I don’t know if that’s the case either.
http://en.wikipedia.org/wiki/Luminous_e ... c_exposure
Problem is I don’t know enough about how one would apply those light measuring values to a camera light metering system, as far as I know Luminance (Lv) is the measured ambient light at the subject area, and Illuminace (Ev) is the light value in lux at the film or sensor plane, the difference being the light loss due to the lens.
But then those are only useful if you can measure and verify the Ev and Lv independently anyway.
So at the moment I’m not really getting anywhere with my quest there are too many unknowns, I might shelve it for a while.
Greg

[Vidgamer]

How is the light measurement any different than in the film days with a light meter? The meter read some number, and that number wasn't fixed to ISO 100, was it? I mean, you had to then figure out your shutter, aperture, and ISO, but you do that anyway even if you "sunny 16" or use other guesstimates to calculate it.

[Greg Beetham]

No the ISO wasn't 'fixed' per se then or now, but the calculators are all based on 100ISO, the ones that I could find that is, and that might be ok when dealing with film but now ISO is just another variable so I want a calculator that you can enter full stops of ISO (or 1/3rd stops of ISO) into along with either of the other variables and find the missing one or convert the result into an Ev value. That's not asking much...is it?
Greg

[Vidgamer]

That iPhone app I mentioned ("Light Meter") does that, but using the camera's measurement to obtain the EV value(s). So, it would be useful while at the scene you wanted to photograph, but not as useful if you wanted to do lots of "what ifs". So, it depends on what you're looking for. For what I want (using a completely manual camera), the app works for me, and does display the EV values as an option.