Focal length 35mm equivalent, but not F-stop?

Started Jul 24, 2014 | Discussions thread
bobn2
bobn2 Forum Pro • Posts: 62,716
Re: Let's correct it, then.

Tim Tucker wrote:

bobn2 wrote:

Tim Tucker wrote:

bobn2 wrote:

Lee Jay wrote:

Tim Tucker wrote:

Erik Magnuson wrote:

Tim Tucker wrote:

Erik Magnuson wrote:

Tim Tucker wrote: -

Is this the same as saying all f2 lenses transmit the same amount of light, but different sensor sizes collect different total amounts for the same exposure?

Amount is not a precise term. A 55gal drum and a 20oz red solo cup left out in the rain might both collect 1" of water. Is the same "amount" of water collected in each?

Eh? Total amount IS a precise term. Measure your volume in litres and you have a precise answer. But that is not what I asked.

It was the answer to your first question and use of "amount." All f/2 lenses do not transmit the same amount of light or my f/2 cell phone lens would transmit as much light as a 200mm f/2 lens. Even if I managed to put a 24x36mm sensor behind the call phone cam lens, it would not help much. What the lens transmits and what the sensor collects may be related, but they are not the same thing.

Sematics, as corrected by Lee Jay (my original statement above). Lenses for different formats do not have the same coverage therefore the total amount of light will not be the same but the light/square metre of sensor size (transmitted?) is the same at f2. A Super Angulon 90mm at f8 will produce the correct exposure regardless of whether I put 4"x5" film or 35mm film behind it.

More semantics...

"A Super Angulon 90mm at f8 will produce the same exposure regardless of whether I put 4"x5" film or 35mm film behind it."

It may or may not be "correct", whatever that means.

Here's a thought on this 'correct' exposure nonsense. Let's go back to film, since it is film based methodology which has confused people. Suppose we added to the equivalence requirements the condition 'equal granularity', and were looking at equivalence between 110 (roughly Four Thirds) and 135 (what we'd now call 'full frame'). If we wanted equal granularity in the image enlarged to the same output size, we'd want the same number of grains in each shot, which means that the 135 grains would need to be four times the area. Since each grain requires two photons to reduce it, both shots would need the same number of photons to expose. Thus the exposure of the 135 would need to be one quarter of that for the 110, or in other words, the ISO of the 135 film would need to be four times higher (which it would be in any case, since the grains were four times larger in area, so it would produce the same density for one quarter the number of photons).

Ah, but grain is a function of the film emulsion and development

Film emulsion, yes. A film works because an incoming photon liberates a photoelectron which reduces a molecule of silver chloride to an atom of silver. At least two silver atoms are required to catalyse the chemical reduction process which will turn the whole halide crystal to silver. These need to be at the same locality in the grain, so the grain needs an 'electron trap' so that the silver atoms are likely to occur at the same place. The process of increasing the quantum efficiency of emulsions involved (chemically) engineering better electron traps and also manipulating the shape of the crystal so that it presented a bigger surface area (compared with its overall size). So, that's much the same as digital, films improved in QE as emulsion formulations improved, but for a given formulation the basic change in speed depended on grain size, basically how much silver got produced per photon. So in the end 'film speed' was a function of QE and grain size, it could be increased either by increasing QE or grain size. So, my comment was assuming film of similar QE, where the speed would be controlled by grain size. So far as development goes, the silver atoms catalyse reduction in the sense that they speed it up. Leave the halide in a reducing agent, its all going to become silver in the end. In the end, development has to be controlled so that the proportion of exposed halide to the proportion of halide that gets reduced anyway favours the development of a good quality image. So more development (by means of a more aggressive developer, longer development time or higher temperature) increases the cahnces of all exposed grains being developed out (thus increasing 'speed') but also increases the chance of non-exposed grains being developed, this providing 'fog' in unexposed parts of the image, and reducing the overall DR.

(development produces the grain not exposure),

No, manufacture produces the grain. It is in the film. Exposure and development only determines which grains are turned into silver. Development can also effect the shape of the final developed grain.

but is also affected by focus

No

and over/underexposure.

Also no. What can happen is that printing to compensate for a thin or thick negative makes the grain in the final image more apparent.

If you scan you have artefacts of noise, nyquist considerations, sharpening artefacts, if you print then you have development and over developed/under exposed/temperature etc. considerations. Noise in film is not translated into the final image in the same way that digital noise is.

The starting process is exactly the same, a photon releases a photoelectron. The difference is because film is more digital than digital is. While a pixel will continue to accumulate charge so long as its FWC is not exceeded, and can thus count tens of thousands of photons, a silver halide grain can only ever count a few photons (that's why film is non linear, as exposure increases the chance of any photon finding an unreduced grain reduces). So, in the end, what I said is true. Keep the same film technology and development, and to produce the same granularity with 135 as 110 you need a film with four times the grain area, which will have four times the ISO. And I can be allowed 'same film technology and development' if the f2=f2=f2 brigade can be allowed to insist on the same film.

The point being that it as a pointless argument because it doesn't apply to grain size.

It certainly does, you need to brush up on your film chemistry.

I know this is supposed to be an analogy but this is getting out of hand. Stick to the facts.

I am, not my fault if you've got the wrong impression of the facts.

Good grief! I'm sorry, and no disrespect because I'm sure your theory is right... But in the real world? Equivalence in film grain? Just what is the point? With 30 years of experience developing I stand by what I say. It is a completely pointless exercise. Development affects grain size, try push processing, out of focus areas look a lot grainier, - no this is pointless...

There is indeed 'equivalence in film grain' as most film users who used multiple formats knew, and as the film manufacturers knew, reserving the slow, fine grain emulsions for smaller formats. On what development does and doesn't do, of course you observe what you observe, but then its worthwhile understanding what you are observing. A film has a spread of grain sizes, they aren't all the same size or the same shape. Both exposure and development affect which grains get developed and the size distribution of the developed grains. After printing, that can be observed as an apparent increase in grain size, but in reality the size of grain was fixed when the film was manufactured. So, my assertion about grain size holds true,

-- hide signature --

Bob

Post (hide subjects) Posted by
(unknown member)
tko
tko
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow