It's a good theory but I can get a T1.2 lens for a 2/3" sensor and there are T1.0 lenses for 35mm, interiors for Barry Linden were shot with one but they are extremely rare. I think it's easier to make a fast lens for a smaller imager as much as it's easier to make a zoom for a smaller sensor, all else being equal but there's not too much of the 'all else being equal' when it comes to optics. Those T1.2 2/3" lenses are seriously expensive as they're still sharp at T1.2 wheres the lenses I'm trialling do get softer at T1.3.
The whole shallow DOF is a bit of crock in my opinion, most movies are shot at around T4 or 5.6, some even stopped down further to get a deeper DOF despite the blow out in the lighting costs, more so on fine grained stocks.
But there's one thing that's going to blow the whole thing out of the water, 3D. With 3D you need the biggest DOF you can muster or you have some really wierd things happening. 3D looks real and it invites the eye to look around the frame and it expects to be able to focus on anything, just like we do in the real world. If it cannot focus on something our brain seems to figure something is wrong. Having the audiences brains going into panic from not being able to resolve part of the frame might not be a good thing in several ways.

Bob.

Bob....

you're confusing t number with f/stop with simple aperture, which is the effective diameter of the lens. T/stop is a number used to define the light transmission thru a lens. DOF, optically, is a function of aperture, distance to the subject and focal length, but not t/number. f/stop, which is the ratio of aperture diameter/focal length is a way of non-dimensionalizing those two parameters, so people understand that DOF increases as you stop down, mainly because the effective diameter is decreasing at a constant focal length(ignoring zoom factor).

Beleive what you will. people spend a great deal of money to simulate the shallow depth of field of 35mm lenses. I MUCH prefer the pull focus I can get from a 35mm lens over a 1/3" lens. 1/3" lenses have such a deep depth of field that they make very sad artistic devices for film.

> Is the DOF with the EX1 and its .5" sensors good
> enough to make cinema lens adapters unnecessary?

No.
1/2" is a little "better" than 1/3", but the 35mm adapters' screens are usually > 1" (diag) and that's hard to beat if you're looking for a shallow DOF.

Thanks Tim for answering my question, but I ask if you can supply evidence, numbers to back up your claim that 1/2" is "little better than 1/3"" because in the Sony blurb or a review I read somewhere, similarly without numbers, it is claimed that shalow DOF is much improved. A shalow DOF is important to me in a buying descision and I'm trying to qantify these "much" and "little better" opinions to see just what I am to believe.

Mike

At 12.7mm, the EX1 doesn't even come close to a 16mm lens DOF. I consider 16mm marginally acceptable when trying to pull focus.

Phil Bloom, in his blog with the EX1, is using a Letus 35 adapter with a Nikon 35mm lens. His images are truly stunning!!
http://www.philipbloom.co.uk/Philip_Bloom/Home.html

Sorry, but .5 inch won't cut it for pull focus and critical object attention for filming.

I think the lens on the EX1 might be a bit faster as well which might account for what Sony is saying.
If you really need to use a 35mm adaptor my gut feeling is you might do better to wait for the Z7 which will have an interchangeable lens. Then a 35mm adaptor could be added that used a simple relay lens and that would avoid a large part of my objection to how many are using 35mm adaptors.
If you're using a camera in the wilds then the EX1 with an adaptor and decent lens in front of that would seem to add upto a front heavy and probably hard to use package.

Bob.

There's a thread on this forum from the summer that goes over m,uch of this, but...

The three things on a camera that affect depth of field are:
- Lens T-stop or F-stop setting (f/1.2, f/2.8, etc.)
- Size of recording surface (sensor / groundglass / film)
- Focal Length of lens (25mm, 50mm, etc.)

The smaller the T-stop or F-stop, the shallower the depth of field.
The larger the recording surface, the shallower the depth of field.
The larger the focal length, the shallower the depth of field.

Any combination of those factors will result in more or less depth of field.

F-stop is what matters, T-stop doesn't.

Take two f/1.4 lenses, one with T1.6 max and one with T1.8 max.

Their DOF will be identical.

T-stops are for the exposure meter, they reflect light loss in the elements which changes exposure.

The diagonal of a 1/2" sensor is about 50% bigger than a 1/2" sensor, so there is a substantial difference in DOF.

In addition to the three primary factors mentioned, there is also the CoC (Circle of Confusion) to affect the DOF.

Much simplified, a "blurry" lens can have a greater perceived DOF, all else being the same.

But 1/2 and 1/3 sensors both are tiny compared to a 35mm (24mmx36mm) frame. The DOF difference betweeb them is negligible compared to shooting with a 35mm adapter.

I ordered the Letus Extreme for the EX1 cause I want the really shallow DOF.

Sami

The key to really understanding DOF is to understand how you get a DOF in the first place. There's only one point that is in focus. Any points outside that may appear in focus due us not being able to see that they're out of focus. The parameter you need to factor in to understand this is Circle Of Confusion. That's determined by pretty well everything in the chain, right up to how far you sit from the screen. You can see this pretty easily, watch a movie on a SD TV or an iPod and it's DOF wise very, very different to watching the same movie in a cinema.

I keep making this point because there's a big trap to this shallow DOF thing unless you're very careful. You might well believe you have all of the talents face in focus, it looks that way on set. And yet when you finally see it on a large HD monitor you see that only their ears were in focus and as you racked focus to follow the talent it went from ears to nose.
Now I don't have decades of experience with this, I'm partially relying on what those that do have had to say and my own limited experiences. It does seem to me that working in HD with a shallow DOF can be a two edged sword. Without lots of experience I believe you really need monitoring on set that gives you the same experience as your target audience, preferably better.

Bob.

> F-stop is what matters, T-stop doesn't.

In my experience lenses are either marked with T-stops or F-stops, not both. T-stops are considered a more accurate method of measuring the light that makes it through the lens, but their values are most often just a hair different than F-stops.For the most part they're interchangeable. T-stops start to matter when you're using multiple lens on the same scene and you want exposure to match perfectly from shot to shot. f/2 might be a bit different on different lenses, but t/2 will expose the film/sensor exactly the same regardless of lens or manufacturer.

For example, professional cine lenses are where you'll fine T-stops, however professional light meters are calibrated in F-stops (and footcandles)

>>>>At 12.7mm, the EX1 doesn't even come close to a 16mm lens DOF<<<

Actually the diagonal of a standard projected 16mm frame is 11.9mm, so they're comparable. Of course agree that the DOF will still be fairly deep but I could pull focus on 16mm using a 50mm lens. Phil Bloom's short should be seen.

DOF must be one of the most discussed topics on camera forums, and I'm always surprised by the debate and misinformation because the principles and arithmetic are laid out in any serious manual. No magic or mystery involved, although it does involve a subjective judgement about when a disk becomes seen as a point.

It has nothing to do with the size of the sensor, except in that it affects the magnification of the viewed image. Aperture, focal length, magnification, viewing conditions. The same image on your Nokia will show very deep DOF compared to your cinema screen.
Suggest reading http://en.wikipedia.org/wiki/Depth_of_fieldWiki[/link]
and/or http://www.cambridgeincolour.com/tutorials/depth-of-field.htmtutorial[/link]
and/or http://www.cs.mtu.edu/~shene/DigiCam/User-Guide/950/depth-of-field.htmlDOF[/link]

or even your American Cinematographer Manual. I've seen people seriously ask manufacturers for the diameter of the circle of confusion for a lens, which would seem to me to be a nonsense question. But I'm always up for education!

> It has nothing to do with the size of the sensor,

Then, using the Depth of Field Calculator on the page at the "tutorial" link in your post, please explain the reason for the "Camera Type" pull-down menu.

I can't directly answer that question, I'll leave it up to Serena.
However much of what you'll read about DOF relates to film.
Film has a fairly fixed resolution and is viewed under pretty well defined viewing conditions. HD Video is not the same as film, in fact video period is not the same as film. You can keep the sensor the same size and change the resolution. With film you can go from a high speed stock to a fine grained stock and yes, that alone can alters the perceived DOF out of the exact same camera and lens.

To give you further food for thought. SMPTE did some exhaustive tests and found that for the typical photochemical process a 35mm print delivers under 800 lines of res on the screen. Yet you can scan the 35mm neg and get 4K lines of res. Project the 4K DI digitally at 4K onto a monster screen and assuming the camera lens was top shelf you'll see a shallower DOF than you would have from a 35mm print done traditionally.

Bob.

The size of the sensor has EVERYTHING to do with the size of the lens...duh!!!
The bigger the sensor, the bigger the lens NEEDS to be to cover the diagonal dimension of the sensor. That's a NO BRAINER folks. The bigger the lens, the more it costs to polish and coat...another NO BRAINER folks. You guys are really something, if not stubborn to the point of being infuriating. Oh well, nothing I can do about it. Denial is not a river in Egypt. Manufacturers are all about profit. You think they're gonna make a lens higher quality than they absolutely need to...HAH! You're more naive than I thought. Sometimes, this forum is a huge waste of my time. Maybe if bellybuttons were made out of optical glass, some of you could see where you're goin'. But then, some of you are also too stubborn , cheap and wound around your own egos to use anything of quality.
Ya know Bob is right about one fundamental fact. Experience experience experience. All the book learning in the world is worthless without a little experience. (sigh)

>>>The size of the sensor has EVERYTHING to do with the size of the lens...duh!!!<<<

That's correct, of course, but wasn't an issue I was including.

The reason for "camera type" in pull down menu? Magnification of the image for the "standard" viewing condition. The set down standards of diameters of circle of confusion for different formats reflect the magnification of the image required. It is 0.025mm for 35mm film and 0.015mm for 16mm and these are the figures employed in calculating the DOF tables.

Really the resolution of the sensors/film and lenses have little to do with the DOF under the agreed conditions of presentation. If they are inadequate then the CoC won't be achieved and the image will look out of focus. If better than that you won't see it under the standard conditions because it is a human judgement relying, in the end, on the resolution of the eye. Should you choose to view at greater magnification then your acceptable CoC will be smaller and your DOF less.

The good thing about book learning is that it gives understanding on which you can base experience. Then you can know how a particular task requires a variation on the rules. But maybe these days in film school they emphasize the New Age touchy feely over hard knowledge, which might be why DOF is such as contentious issue.

Serena,

It sounds to me like you're saying an image projected on a very small screen will have more depth of field than the same image projected on a very large screen.

That is absolutely correct.

Easy to test at home if you have a video projector (or film ditto of course).

For example, professional cine lenses are where you'll fine T-stops, however professional light meters are calibrated in F-stops

Yikes.

Yes, pro cine lenses show T-stops, because that's the only way you can get consistent exposure between lenses.

Light meters usually show "f-stops," but when the meter shows "2.8" at 24 fps, that means T2.8, NOT f/2.8.

Why? Because exposure meters are for measuring exposure, and the exposure is dependent on the exposure time for each frame and the T-stop (sometimes referred to as "effective f/stop"). T stands for transmission, i.e. how much light actually gets through to the other end of the lens.

The f/stop is of interest for DOF and related calculations.

Since T/stops and F/stops are close in most modern lenses, they are often used interchangeably by amateurs especially.

I think the full answer is a little more complex.
For the same viewing angle it doesn't matter.
Conversely get far enough away from a large screen and the resolving power of your eyesight can become what determines the CoC. Get really close to an LCD and the resolution of the panel could become the factor determining the size of the CoC etc.

Of course we're talking about relatively small differences. I think for many of us talking about shallow DOF we're talking about where the subject is in focus and the background is pretty much a blur. Under those conditions you need to get the in focus subject as soft as the background to see an increase in DOF i.e. the whole thing looks out of focus.

Bob.

Yes, the angle subtended by the image is the thing that determines the necessary CC, and that is all about resolution of the eye. Worth repeating that: resolution of the eye.
It is important that the structure of the screen cannot be resolved at that viewing distance, or the image will all fall apart. I've seen it argued that the size of pixels or line interlacing determines the CC. Obviously the size of a pixel determines the minimum CC that the screen can indicate, and if the desired CC is smaller than a pixel then your eye will resolve the screen structure and not see a sharp image. I don't think it a complex issue. The science is simple, and application follows so long as we understand that the viewing environment affects DOF, and that DOF is about human perception rather a physical fact. Achieving a desired CC involves all the elements in the process from recording to display, so it's pretty easy to become sidetracked from the "big picture" by concentrating on particular elements (e.g. confusing the resolving power of a lens with CC; different matter). The actual CC achieved is a product of the chain and hopefully (if the images are to look sharp) that is at least as small as needed for the viewing conditions.

Hopefully we're not going to start on airy disks and diffraction effecting optical resolution. For the purpose of video or film DOF a point is a point.