Does FOV affect image quality?

[Jacek Zakowicz]

Mgnification

Mgnification

So a larger sensor requires more magnification, hence a longer lens, to achieve the same field of view at the same distance as a smaller sensor?

Let me explain here the term - its the size of the object to the size of its image ratio.
If you photograph a 1" ruler and its image generated by the lens will be .1" long that's 10:1 magnification. Whether you generate this .1" image of the 1" ruler with the 10mm lens or with the 1000mm lens as long as the stops match the dof will match as well..

 

[David Rasberry]

Let me explain here the term - its the size of the object to the size of its image ratio.
If you photograph a 1" ruler and its image generated by the lens will be .1" long that's 10:1 magnification. Whether you generate this .1" image of the 1" ruler with the 10mm lens or with the 1000mm lens as long as the stops match the dof will match as well..

But to maintain the 1" ruler image at the same relative size to the frame on a 30mm wide image requires 3 times the magnification needed for a 10mm wide image at the same object distance correct?

 

[Thomas Wright]

That's some interesting stuff, cos intuiting from shooting from one position as is normal, I mean who's gonna hire a helicopter to fly a few hundred miles away to use a ridiculously long lens to shoot a bleedin' cactus?, one comes to the conclusion, Justin and most of us would think that DOF is tied to FL. But to get back to the initial question on FOV: a smaller chip extracts less of the same image from the same lens (the chip/film being approximately 5cm from the mount in cine-style cameras) Just that lenses made for larger area extraction formats have more glass so a 25mm Distagon manufactured for 16mm film will give you the exact same image as a 25mm Superspeed (also a distagon but naming conventions are what they are) manufactured for 35mm film if your red is set to 2K

There is a theory though that the resolution of the Superspeed would be better since the light is travelling through the "sweet spot" of the lens. I've not tested this but it kind of makes sense when people speak of "sweet spots" (you can see it when you project a lens) but I have apprehensions since I imagine the curvature of the glass must be the same and I assume there is just less of it; making the lens cheaper.

Does anyone know what is actually the case?

 

[Thomas Wright]

If that is the case then yes, FOV would affect image quality cos you'd be using the dodgy edges of the lens...

?

 

[Shane Kelly]

time for Matt's diagram:

 

[David Rasberry]

Most high quality lenses cover an image circle a bit larger than the format image circle so as to maintain the image in the lens sweet spot. Lens quality is more critical for smaller formats too, so a premium quality 16mm film lens for instance will typically be sharper than most 35mm still camera lenses. You can see some good examples of this at www.oldlens.com. Scroll down to the bottom of the page for the section on c-mount 16mm lenses on 4/3 format cameras. The 4/3 image circle is almost twice the size of 16mm, yet the best of these lenses cover it very nicely. That Kodak cine ektar f1.4 is pretty impressive.

 

[Justin O'Neill]

Ok after a good nights sleep I understand what I was missing. Given the same focal length and f-stop, lenses will exhibit shallower depth of field the closer the subject is to the lens. Like Thomas says, in narrative filmmaking we tend to forget about this fact because of the close quarters and usually perspective and field of view are considered first.

Daniel and Jacek, let me try this description and tell me if it is simple enough while still adhering to the facts:

Depth of field is controlled by 3 factors: focal length, f-stop, and the distance of the subject from the lens. The following hold true on all sensor formats, be it 2/3" or FF35:

1. Given the same f-stop and subject distance, longer focal lengths have shallower depth of field
2. Given the same focal length and subject distance, smaller f-stops have shallower depth of field
3. Given the same focal length and f-stop, the closer a subject is to the lens, the shallower the depth of field

If these three points are true on all formats why do we think of small format cameras (HVX200, 2/3" Scarlet) as not being capable of shallow depth of field?

As we can see on the chart below, smaller sensors "see" a narrower field of view then their S35 and FF35 counterparts. You could could achieve a wide shot by moving the camera further back form the subject but there are often objects in the way and this distorts perspective. What we usually do is go with a wider focal length. An 8mm lens on a 2/3" sensor has the equivalent field of view of a 20mm lens on an S35 sensor. Shorter focal length equals deeper depth of field (we can't move back and f-stops rarely go below f1.2) and this is where the stigma comes from.

 

[Terry VerHaar]

[*]Given the same focal length and subject distance, smaller f-stops have shallower depth of field

Hmmm... this seems 180 degrees wrong to me. Isn't it the opposite? Everything else being equal, the smaller f-stop will yield DEEPER depth of field, won't it?

Terry

 

[David Mullen ASC]

Yes, stopping down leads to more depth of field.

The other controlling DOF factor is the Circle of Confusion figure. If you are assuming that all formats will be displayed at the same size, then you have to use a more critical figure as the formats get smaller and smaller.

 

[Justin O'Neill]

Hmmm... this seems 180 degrees wrong to me. Isn't it the opposite? Everything else being equal, the smaller f-stop will yield DEEPER depth of field, won't it?

Yeah it is confusing because "f-stops" and "stops" go in opposite directions. As the f-stop number gets smaller, the aperture is actually getting wider. When you are "stopping down" the f-stop number is actually going up.

 

[D Fuller]

There is another factor that is seldom talked abut in DOF discussions, but that matters a great deal. Display size.

DOF is a perceptual phenomenon. "Acceptably sharp" is very different in SD, HD, and on a 30-foot screen. So the viewer's sense of the impact of DOF will be very different if they watch a shot as web video or in a theater.

The difference is easy to see in a program like Lightroom. Look at an image full size, then look at the same image as a thumbnail. Everything will look sharp.

 

[Justin O'Neill]

The other controlling DOF factor is the Circle of Confusion figure. If you are assuming that all formats will be displayed at the same size, then you have to use a more critical figure as the formats get smaller and smaller.

There is another factor that is seldom talked abut in DOF discussions, but that matters a great deal. Display size.

DOF is a perceptual phenomenon. "Acceptably sharp" is very different in SD, HD, and on a 30-foot screen. So the viewer's sense of the impact of DOF will be very different if they watch a shot as web video or in a theater.

The difference is easy to see in a program like Lightroom. Look at an image full size, then look at the same image as a thumbnail. Everything will look sharp.

My problem with circle of confusion: is it really useful when considering depth of field and which lenses to use on what formats?

The cinematographer has zero control over what size the viewer watches the video at and circle of confusion is also effected by how far from the screen the viewer sits. People in the front row of the theater will see slightly shallower depth of field then the people sitting in the back.

It is valid for the viewer to consider CoC when watching a video but has a shooter ever thought "most people will be viewing this as a tiny little web video, I'd better really make the DoF shallow"?

 

[David Rasberry]

My problem with circle of confusion: is it really useful when considering depth of field and which lenses to use on what formats?

The cinematographer has zero control over what size the viewer watches the video at and circle of confusion is also effected by how far from the screen the viewer sits. People in the front row of the theater will see slightly shallower depth of field then the people sitting in the back.

It is valid for the viewer to consider CoC when watching a video but has a shooter ever thought "most people will be viewing this as a tiny little web video, I'd better really make the DoF shallow"?

A cinematographer does have some notion of the intended viewer and format in planning a productiion though, and should at least take that into account in planning how to compose his shots.

 

[Sanjin Jukic]

Useful links:

Understanding Depth of Field Tutorial >>>

FOV calculator >>>

 

[Stephen Williams]

Yeah it is confusing because "f-stops" and "stops" go in opposite directions. As the f-stop number gets smaller, the aperture is actually getting wider. When you are "stopping down" the f-stop number is actually going up.

That is because F stops are fractions 1/1.4, 1/2, 1/2.8, 1/4, 1/5.6, 1/8 etc

 

[D Fuller]

My problem with circle of confusion: is it really useful when considering depth of field and which lenses to use on what formats?

Not really. It's useful in coming to an understanding of why DoF is perceived as it is -- what the optical properties in play are, but even at the same focal length and aperture, different lenses render out-of focus areas and the transition to sharpness differently. That's why numbers are not such a good reason to choose a lens. Lenses are better chosen by experience. Try them out and get to know how the look of a Nikkor 85mm differs from a Zeiss 85mm, or a Cooke 85mm.

The cinematographer has zero control over what size the viewer watches the video at and circle of confusion is also effected by how far from the screen the viewer sits. People in the front row of the theater will see slightly shallower depth of field then the people sitting in the back.

No, you don't have any control over that once it leaves your hands. But this is nothing new. A still image made for a 16x20 inch print gets reproduced in a book at 4x5 inches. Some people stand closer to the wall to look at prints than others. And it completely changes the impact of the image -- including characteristics like DoF, but not limited to them. Have you ever seen a painting by Jackson Pollack in a museum? The experience is totally different from seeing it in a book. And standing 5 feet from the painting is a totally different experience from standing across the room.

You can't control how the end viewer sees the video, but you do have to think about what you intend that experience to be. And you usually do know how it's going to be seen at least the first time. I shoot most of my work for TV, and I think about the reproduction in those terms, with all their plusses & minuses. If you are shooting for the cinema, the viewing experience is different, and that should be part of your thinking in choosing lenses, f-stops, etc. If you're shooting for the web, there's a different set of considerations.

It is valid for the viewer to consider CoC when watching a video but has a shooter ever thought "most people will be viewing this as a tiny little web video, I'd better really make the DoF shallow"?

Of course shooters think about that. The more mastery you gain of the craft, the more of those things you are able to consider as you make images.

As for viewers, it's not really their responsibility (nor is it that much fun) to think about CoC when watching a video. That wouuld be like having to build a watch in order to tell time. I suspect that the only viewers who think about CoC as they view a video are lens designers and student cinematographers.

 

[David Mullen ASC]

My problem with circle of confusion: is it really useful when considering depth of field and which lenses to use on what formats?

The actual number may not matter so much but if you are going to talk about why one format has more or less depth of field than another, you have to assume that you'd be comparing them at the same final size, and that assumes you are adjusting the CoC figure to account for the different degrees of enlargement.

I realized this when I tried comparing DoF merely looking at a DoF chart for a lens. Let's say you are comparing two formats, one twice as large as the other, so that on the larger format, you'd use a 50mm lens and on the smaller format, you'd use a 25mm lens to get the same field of view at the same distance.

Well, if you look up those depth of field figures for a 50mm lens and a 25mm lens at the same distance, you find a 4-stop difference in depth of field, not a 2-stop difference as you'd think and as experience tells you. Why is that?

Well, because the 25mm lens on the smaller format needs to have a twice as critical a Circle of Confusion figure to compensate for the image being seen at the same size as the 50mm image on the larger format. Twice as critical in this case causes a 2-stop DoF loss so your 4-stop figure becomes a 2-stop figure in terms of the extra depth of field with the 25mm lens on the smaller format compared to a 50mm lens on the larger format (which matches practical experience).

So while I agree that none of us can know the degree of enlargement our images will be seen at, if we are going to talk about comparative differences in DoF between formats, we have to factor in that smaller formats use a more critical CoC figure when being compared to larger formats for the same size final presentation. In some ways, this is a good thing in that the more critical CoC compensates partly for the big increase in DoF as you go to shorter focal lengths when switching to the smaller formats.

 

[Rick Burnett]

So let me ask a question that I do not have the resources to answer yet. On my 7D I bought a 30mm lens that I use frequently at F1.4. I based this off of a 50mm lens giving the same FoV roughly as a 5D (or full frame camera).

Since I do not have a 5D, I haven't been able to test this, what should I expect the difference between two images that are shot from the same distance and effectively, let's assume the lens size equally cancel out the crop factor.

My thinking was that since the 5D was a larger sensor and the aperture the same on each, that the 5D would have a shallower DoF.

I ask all this as I am trying to understand what to expect when I start shooting on a 2/3 scarlet.

 

[David Mullen ASC]

Yes, a 30mm on a 35mm cine / APS-C sensor would be similar in view to a 50mm on a FF35 camera. But the 50mm version would have less depth of field.

As a rough ballpark figure, I used the magnification factor as the difference in effective depth of field, in this case, about a 1.5-stop difference in practical DoF between a 35mm cine and a FF35 camera once you match FoV, distance, and actual f-stop.

So if your 50mm / FF35 shot is at f/4, you'd have to shoot your 30mm / 35mm cine shot at f/2.0-2.8 split to get an approximate match in DoF.

The difference between 2/3" and 35mm cine is 2.5X, so roughly a 2.5-stop difference in DoF.

 

[Rick Burnett]

Thank you, I am glad my thinking was in the right direction. Heh, there are so many variables here to think about, sometimes I just have to sit down and think about it.

This is why I will be happy when I eventually get my hands on a S35. I have used the APS-C sensor so much now for video, I know how to get what I want intuitively.