Understanding 16mm and 35mm type lenses/cameras

[Chris Davis]

Hello. Planning on moving up to Prosumer Land, so, of course I’m looking at the Scarlet 2/3”. To start I’m giving myself a crash course on DSLR still photo lenses to try to understand how this kind of thing works.

Edit: Please see my follow up post in this thread titled "Crash Course in Optics.":help:

1. Can anyone please provide me with some links describing how 35mm type lenses compare/contrast with 16mm type lenses?

2. On a 16mm film camera, is the film closer to the base of the lens than on a 35mm film camera? If I just knew this, much of this stuff would start making sense.

3. If the film is indeed closer to the base of the lens in a 16mm film camera than in a 35mm, would you suppose that the RED Scarlet 2/3” would have about the same distance from the base of the lens to the sensor as a 16mm film camera has between the base of the lens to the film, and that the Red One would have about the same distance from the base of the lens to to the sensor as a 35mm film camera has between the base of the lens and the film?

4. Is a 16mm type lens projecting the same amount of light onto/around the sensor as a 35mm type (with the 16mm just projecting the light into a smaller space), meaning that the 16mm area is getting much more light than the 35mm area (with lenses that provide the same field of view and same amount of zoom)? Could one then use less sensitive sensors to achieve the same low light performance?

A clearer way to ask may be, on 16mm film cameras, do people tend to use slower film (lower ISO) to get the same low light performance as on 35mm film cameras? (I know it depends on the lens, but I'm not yet sure how to compare 16mm and 35mm type lenses).

Many thanks,
Chris

 

[Juhan Malmstein]

Hey, I can't really answer all of your questions, but can try to give you some overview on what I know.

1) The main difference between 16mm lenses and 35mm lenses is the size - to cover the 35mm frame those lenses have to be physically considerably bigger. Another thing is that since the frame itself is about ~2x smaller than a 35mm film frame the focal lengths change. When in 35mm film a "normal" lens is 50mm, then the same field of view on 16mm film can be achieved with a ~25mm lens (50/2). Another thing to consider is that smaller sensors gain depth of field over larger sensors(ever think about why all the point and shoot cameras with tiny sensors have extremely deep depth of field ). So if you have a 25mm lens on a 16mm camera set at f4.0 and next to it you have a 35mm camera with a 50mm lens set at 4.0, both cameras get the same field of view(roughly), and the same amount of light, but the depth of field will look different. the 25mm lens will look like the 50mm lens set at f5.6, except it isn't darker if you can follow what I'm saying. What I wanted to say is that on a 16mm sized sensor you lose about 1 to 1.5 f-stops of depth of field compared to 35mm cameras, but only in depth of field, not loss of light.

2) probably yes but I'm not really knowledgeable in this area

3) again probably yes.

4) No you can't cheat by using lower ISO's on 16mm cameras. Again, if you have a 35mm camera set at for example ISO 400, shutter speed 1/50 and the lens at f2.0 and a 16mm camera set at ISO400, shutter speed 1/50 and the appropriate lens at f2.0 your images will be identical in the terms of light, but as I already mentioned there will be a small difference in the picture due to the smaller frame having a slightly deeper depth of field.

I hope you find some of this stuff useful, the world of glass is a big and confusing one.

 

[Jerrod Cordell]

Juhan pretty much summed up the major difference between 35mm and 16mm. Also, 16mm has a digital equivalent of about 2-3K and 35mm has a digital equivalent to about 4-5K, even though most film projectors only show about 1-2K in theaters.

 

[Stephen Williams]

Hi,

There is no reason that the lens is closer on 16mm cameras, often 16mm cameras have a PL mount so use 35mm lenses for longer focal lengths.

Stephen

 

[Chris Davis]

1) The main difference between 16mm lenses and 35mm lenses is the size - to cover the 35mm frame those lenses have to be physically considerably bigger... So if you have a 25mm lens on a 16mm camera set at f4.0 and next to it you have a 35mm camera with a 50mm lens set at 4.0, both cameras get the same field of view(roughly), and the same amount of light, but the depth of field will look different. the 25mm lens will look like the 50mm lens set at f5.6, except it isn't darker if you can follow what I'm saying.
4) No you can't cheat by using lower ISO's on 16mm cameras...
I hope you find some of this stuff useful, the world of glass is a big and confusing one.

Many thanks. I think I follow you on #1, though I would not have known.

There is no reason that the lens is closer on 16mm cameras, often 16mm cameras have a PL mount so use 35mm lenses for longer focal lengths.

Thanks, Stephen. Are you saying that the distance between the base of the lens and the film in indeed the same on 16mm and 35mm type film cameras, or that it is different and therefore an adaptor is required to put a 35mm type lense on a 16mm film camera?

5. Added question: I had read that the small sensors on consumer cameras are placed very close to the base of the lens to allow maximum light, with the proper corresponding lens. Is there some way that placing a sensor close to the base of the lens allows more light to hit it, or am I just misunderstanding here?

-Chris

 

[Chris Davis]

Crash Course in Optics

Crash Course in Optics

4) No you can't cheat by using lower ISO's on 16mm cameras. Again, if you have a 35mm camera set at for example ISO 400, shutter speed 1/50 and the lens at f2.0 and a 16mm camera set at ISO400, shutter speed 1/50 and the appropriate lens at f2.0 your images will be identical in the terms of light...

I am sure you are correct. But, I’m still wondering, “why?”:blink: Perhaps someone here could explain or give me some links.

Intuitively, it would seem that if I shot a 10’ x 6’ space on a wall from 15’ away (with a lens that would capture only this area from this distance) on a 35mm type film camera, the image would be projected onto a 35mm area onto the film. If I shot a 10’ x 6’ space on a wall from 15’ away (with a lens that would capture only this area from this distance) on a 16mm type film camera, the image would be projected onto a 16mm area on onto the film (these numbers are just made up for the sake of comparison – I realize the light is actually projected in a circle on and around the film frame). With a 16mm camera, the same image and thus the same amount of light would be projected onto a smaller area than the 35mm. As the 16mm area is about ¼ the area of 35mm, it would seem that the film would get 4x the light. It would follow then that one could use film that is 4x’s as sensitive (two stop lower ISO) under identical lighting conditions when shooting 16mm than when shooting 35mm.

The people on this site shoot with both 16mm and 35mm cameras and know that one must use the same speed film with either type under similar lighting conditions. So I know that my logic is incorrect. I’m just hoping that someone can explain to me “the why” or provide some links. Does it have something to do with the opening at the front of a 35mm type lens being larger and thus gathering more light than the opening of a 16mm type lens?

Thanks again,
Chris

 

[David Mullen ASC]

If bigger lenses at the same f-stop put out less or more light than smaller lenses with the same f-stop, then a light meter would be useless.

I believe that the distance to the lens plane is about the same for 16mm and 35mm cameras. A lens designed for 16mm just projects a smaller image because it doesn't need to fill a larger 35mm film area; conversely the lens made for 35mm projects a larger image and the 16mm gate crops this more than a 35mm gate would. This is why the same focal length has a narrower field of view on the smaller format.

So the amount of light at f/2.8 (or T/2.8 to be more accurate) is the same for 16mm or 35mm lenses, and the film stocks used in both formats are usually the same, just cut to different sizes.

Generally one tries to use a slower-speed stock in 16mm to compensate for the graininess of the smaller format due to the degree of image enlargement. So you need to use more light to compensate for the slower-speed stock.

 

[Pietro Impagliazzo]

Generally one tries to use a slower-speed stock in 16mm to compensate for the graininess of the smaller format due to the degree of image enlargement. So you need to use more light to compensate for the slower-speed stock.

Not a problem with Scarlet 2/3"? Right?

Or is there a caveat I'm missing?

 

[Jeff Kilgroe]

Distance from lens to film plane...

Distance from lens to film plane...

The distance from the lens base to the film or sensor plane is a property of the lens mounting system and not anything to do with 16mm, 35mm, 65mm, etc... Lens designs differ from one to another, depending on manufacturer, capabilities and specs, etc.. But what remains constant, in relation to a given type of lens mount, is the FFD (flange focal depth). This is the distance from the face of the mounting flange surface to the image plane. For example, Nikon F mount is 46.5mm, Canon EF is 44mm and the FFD of a PL mount is 52mm.

There is no set standard for the distance of the rear lens element to the image plane, but most don't travel too far back, otherwise they may interfere with mechanical shutter mechanisms or optical low pass filters (OLPF) on a digital sensor. Some newer "digital-only" lenses, like the Angenieux Optimo DP Rouge set, extend further back in this way. This helps them use different internal elements and bring costs down. But it makes them incompatible with film cameras and, unfortunately, even some digital cameras that use a mechanical shutter, also incompatible with some 35mm lens adapter systems, etc..

 

[David Mullen ASC]

Grain isn't a problem with a 2/3" digital camera. The only issue affecting how much light you need, assuming normal frame rates and shutter speeds, is the sensitivity of the camera (and the speed of the lens, i.e. how wide-open it can go). I suspect a RED 2/3" sensor will be similar in sensitivity to the current Mysterium sensor, and I'd guess that the Monstro sensor will be more sensitive (or at least, less noisy and thus allowing higher ASA ratings.)

Now technically a smaller sensor would need a lens with a higher MTF, higher resolving power, than one on a larger sensor.

 

[Jeff Kilgroe]

In a way to confirm what David just said about a higher MTF, the photosites or pixels on the 2/3" Scarlet sensors are said to be 3 microns in size or 0.003mm. The photosites on the current RED One Mysterium are 5.4 mictrons or 0.0054mm. So we're talking about an 80% increase in resolution density or lines/mm over the RED One.

The upcoming Mysterium-X sensor is said to have more sensitivity and better range than the current Mysterium, but the S35 sized X revision will have Photosites that are 5 microns. So just slightly smaller than the current Mysterium. But it will be interesting to see what sort of sensitivity is possible on Scarlet 2/3" vs. the S35 Mysterium-X version. The 2/3" cameras should have much less sensitivity due to their smaller photosite size. Assuming they are truly based on the same sensor design.

 

[Daniel Browning]

both cameras get the same field of view(roughly), and the same amount of light, but the depth of field will look different.

I prefer to think of it this way: it's the same light intensity per area, but the total amount of light is greater because 35mm has more area. I think it's a useful distinction. It's like how a small studio and football stadium can both be lit to f/2.8, but the total amount of light required for the stadium is greater.

The 2/3" cameras should have much less sensitivity due to their smaller photosite size.

I don't subscribe to the idea that small pixels have less sensitivity. The common mental model of that position goes like this: a single pixel, in isolation, when reduced in size, has less sensitivity, more noise, and lower full well capacity. (So far, so good.) Therefore, a given sensor full of small pixels is worse than the same sensor full of large pixels. (That is where I disagree.)

I prefer a mental model like this: forget about pixels and think about light. The amount of light falling on a given sensor area does not change, no matter the size of the pixel. Large and small pixels alike record that light falling in certain positions. Both reproduce the same total amount of light when displayed; one just does it with greater accuracy.

My research and experiments bear that out: when small pixels and large pixels are compared in the same final output, smaller pixels have slightly higher sensitivity and slightly lower noise at base ISO for a given output size. (Some of them have higher noise at high gain compared to large pixels, but I think that's technology dependant.)

However, I do find that sensor size is highly correlated with noise, sensitivity, and dynamic range, as well as technology. I think the reason that "small pixels are worse" has become so widespread is that it is so common to use Unequal Comparisons and the fact that many are used to non-linear formats such as video and film, which are not like raw cameras.

I described the Unequal Comparisons while discussing another topic in the ER thread:

http://reduser.net/forum/showthread.php?p=380476#post380476

Based on that I think Scarlet 2/3" will have similar (or better) sensitivty as the same-size crop from the S35 Mysterium-X, even though it will have smaller pixels.

Kind regards,

 

[Jeff Kilgroe]

Daniel, not sure what research and experiments you speak of, but there are several technical factors at play here. Actual sensitivity of the 2/3" Mysterium-X vs. the S35 Mysterium-X remains to be seen.

We have to consider that not 100% of the surface area of a photosite is actually receptive. There is a pitch gap between photosites in most cases and we are asking a lot for that pitch gap to shrink along with the photosites, to maintain a consistently proportional receptive surface. Not to mention that individual photon emissions do not always register consistantly if they cannot impact the full width of their wavelength onto the receptive area. With smaller photosites we get into smaller targets. Depending on the pitch gap or size of the non-receptive portion, at 3 microns (3000nm) the chances also increase that larger wavelength emissions could partially register across multiple photosites.

While light seems like it's infinitely small, it's not... Visible spectrum frequencies range from ~400nm to ~750nm. So the higher "red" end of the spectrum having a wavelength that is 25% of the width of a photosite.

I don't recall seeing any comments from RED about the sensitivity between the 2/3" Mysterium-X and the S35-sized one. The original announcement had a sensor comparison chart that claimed both versions of Mysterium-X had the same dynamic range, which is entirely believable. But no ISO ratings were ever mentioned. It has however been stated that the larger 6 micron photosite size is implemented for a few reasons, increased sensitivity being one of them.

 

[Daniel Browning]

There is a pitch gap between photosites in most cases and we are asking a lot for that pitch gap to shrink along with the photosites, to maintain a consistently proportional receptive surface.

Agreed, when relative fill factor stays the same despite a shrink in pixel size, the sensor designer deserves some credit. As it happens, that is exactly what sensor designers have been able to do in most cases. For example, you can take a look at Sony's published specs for these sensors:

Perhaps it's related to the fact that the size of the non-photo-receptive portion of the sensor is one of the design factors that benefits (slightly) from advancements in semiconductor "process" technology, allowing it to shrink. Whatever the reason, we can examine the relatively long history of raw still cameras to see that in almost every case where the pixel size was reduced, total sensitivity remained the same or went up, even in cases where the sensitivity per-pixel went down.

Not to mention that individual photon emissions do not always register consistantly if they cannot impact the full width of their wavelength onto the receptive area. With smaller photosites we get into smaller targets. Depending on the pitch gap or size of the non-receptive portion, at 3 microns (3000nm) the chances also increase that larger wavelength emissions could partially register across multiple photosites.

I agree that optical crosstalk is an issue for relative distribution of spectral sensitivity and therefore color accuracy, but I don't think I understand how it would impact total sensitivity.

I don't recall seeing any comments from RED about the sensitivity between the 2/3" Mysterium-X and the S35-sized one.

I don't either. All I can recall is Jim saying that Scarlet 2/3" Mysterium-X would have more dynamic range than a RED ONE, but that doesn't necessarily reflect on sensitivity per se, nor the difference between Scarlet S35 and 2/3".

It has however been stated that the larger 6 micron photosite size is implemented for a few reasons, increased sensitivity being one of them.

I think RED, in those instances, is talking about sensitivity per pixel, not something that will represent the total capabilities of the camera. For example, if EPIC 28K was resized down to EPIC FF35 size (6K), don't you agree that the 617 could be rated higher than the EPIC FF35? Even though they have the same 6 micron pixel size?

Thanks for the discussion, Jeff.

 

[Chris Davis]

If bigger lenses at the same f-stop put out less or more light than smaller lenses with the same f-stop, then a light meter would be useless... So the amount of light at f/2.8 (or T/2.8 to be more accurate) is the same for 16mm or 35mm lenses, and the film stocks used in both formats are usually the same, just cut to different sizes...

Thanks for the response. I was able to make sense of why a 16mm type lens does not put more light the smaller sensor than a 35mm type by something perhaps too basic to mention – the 35mm type is a lot bigger in the front, so it’s pulling in more light to begin with.

PS: When talking about 16mm type lenses on 16mm cameras, what focal length is a "normal" lens -- that is, one that gives the same field of view as a 50mm 35mm type lens mounted on a 35mm type camera?
Thanks,
Chris