DOF ~ CCD Size

[Colione]

In asking about the XL2, a friend sent me this little talk relating problems with the depth of field and the CCD. The smaller the CCD the greater the depth of field. Also, you loose quite a bit of width in your shots with such a small CCD. I think this would be a fairly big concern, especially the DOF. Instant film-look, get a blurry background. The smaller CCD may pose problems. Anyways, here is what a friend said, let me know what you think ...

Allright, here I go.

I kind of think the xl2 is okay but it has some problems. The CCD is still 1/3 inch. Now this is okay, the dvx is also 1/3 inch, but it crops the top and bottom of the chip to get the 16x9 effect. Yes, it does have true 16x9, but on a chip area of about 1/4 inch (after the cropping). Whereas if you used an anamorphic adapter on the dvx, you would use all of the 1/3 inch chip.

The smaller your chip, the more depth of field, which isn't what you want when you're trying to replicate film. Shallow depth of feild is the nice out of focus effect apparant to 35mm film.

This also means that the smaller the chip, the more telephoto your lenses become. So the widest you can get with the new lens in 16x9, equivalent to 35mm film, is 42mm, which isn't very wide at all.

The dvx widest setting is 32.5mm. Much better I think. More depth and perspective in your shots. Just my opinion. I was just reading american cinematographer and almost all of the new harry potter was shot with 24mm lenses!

Add to this the fact that is is US$1500 more than the dvx, and you know how I feel. And it's still 4:1:1 dv. Nothing new there.

So unless you have invested huge cake in xl lenses already, it just isn't worth it.

But then again, I haven't seen it in action! This is all theoretical techno-babble.

-Mitch

P.S. If you can afford the rental, SDX900 is the way to go. True 16:9 shaped 2/3 inch chips, pro lenses, 4:2:2 color. And big price tag. But a third of the price of HD.

But on a really really small budget, the dvx100 is the champ, for now. Your friend made a wise choice.

[Barry_Green]

Sounds about right, except the actual CCD size thing. The XL2 in 16:9 mode has effectively a 1/3.3" CCD. In 4:3 mode it has a 1/4" CCD.

As far as shallow DOF, at equivalent focal length in 16:9 mode the two cameras will be the same, but the XL2 can go MUCH longer in the telephoto range, and telephoto is a key component in establishing shallow DOF. There will be limited circumstances where you could use that full telephoto, but when you can, the XL2 will deliver a significantly shallower DOF than the DVX at anything above 45mm.

In 4:3 mode, the DVX will exhibit a much wider field of view and correspondingly shallower DOF at equivalent focal lengths... I don't know if the longer telephoto will overcome that.

I don't think the XL2 would be my first choice for 4:3 shooting. For 16:9 shooting it looks very interesting: potentially shallower DOF and none of the restrictions imposed by using the anamorphic adapter. But then again, it's $1500 more.

[Zoomforce]

but the XL2 can go MUCH longer in the telephoto range, and telephoto is a key component in establishing shallow DOF. *There will be limited circumstances where you could use that full telephoto,

Ha ha.. I can definately see it now, guys are gonna be shooting movies while the camera is 50' back in the woods.

[Antoine_Fabi]

absolutely......and i am curious about the image quality in real world...color, dynamic range etc...

[Policar]

Yes, I want so much to see a dynamic range test!

And I bought a boom pole just so I could zoom to full telephoto to get more DOF and still get clean audio! Don't make fun! (This isn't entirely true, btw...)

Shallow DOF can mask bad lighting in the background. Blown skies and chromatic abberation on the edges aren't a problem if the background is blurred out! So, sadly, it does mean a lot to me.

[Guest]

Remember though. the Telephoto means nothing, if you're trying to achieve the same framing that you would without the telephoto.

EG. Standing 2 feet from someone and framing so their entire head files the frame, will give the same DoF as standing back and framing identically.

Aaron

[hal9000]

Dear Aaron, you are wrong here.

If you use a wide angle in the first shot and tele in the second framing the same space (face of the talent) you will get different DoF (shallower in the second take). This is clear for me. I assume same aperture and same CCD size in both cases.

[Barry_Green]

Yeah, I tested that, and even though theoretically Aaron's explanation is correct, it doesn't work in the real world. Telephoto definitely gives the "look" of shallower DOF.

[XCheck]

I am not sure even the 'theoretical' explanation is correct (not according to the formulas that I am using).

But suppose it is, could DOF appear to be shallower because of the perceived magnification of what's outside of it?

[Barry_Green]

The theoreticals are correct as I understand them, which is that the focal length effect on DOF is ostensibly canceled out by the proximity to subject effect. *However, what you said is exactly right: even though the background on the telephoto shot may technically be just as out-of-focus as it would be on a similar-subject-size wide-angle shot, the fact that the telephoto lens magnifies a smaller area of background to fill the screen makes it LOOK a lot more out of focus.

Here are two shots, taken at identical aperture and (as near as I could get) identical subject size. *The difference is the zoom setting and the focus distance, which theoretically should have canceled each other out and yielded identical DOF, but obviously the tele shot looks a lot shallower:

wide-angle:
http://www.icexpo.com/dvx100/DOF-Test2Wide.jpg

telephoto:
http://www.icexpo.com/dvx100/DOF-Test2Tele.jpg

[hal9000]

For me is not only a matter of the magnification effect, because if you look again to Barry's pictures, the "out of focus" factor (circle of confussion) is significantly higher in the second picture (telephoto) than in the upper one, which has a sharper background

I still think that telephoto gives still more "fading" to the background, independently of the magnifycation degree.

[Guest]

Thanks for those images Barry. I was talking about the DoF in the form of calculations. i.e. You plug the numbers in, and you get near and far plane distances.

Sure, the zoomed stuff does look like a shallower DoF and if you can handle the extreme distance compression etc, then go for it! ;)

Aaron

[scharky]

Actually I think that barry's concept is correct. *I took barrys wide angle shot and blew up the background so that we can see what it would look like if the wide angle background was seen at the same size as the telephoto background, and guess what, the leaves are just as soft on the wide angle shot as they are in the telephoto, they are just so much smaller that you don't notice it as much. *Things that are smaller will look sharper until seen close up. *Sure this is a digital manipulation, but getting the exact same fuzziness does say something about the Barry's concept
http://www.stefweb.net/dof.jpg

[XCheck]

You plug the numbers in, and you get near and far plane distances.

What calculations do you use? I use the following:

Hyperfocal distance:

H = f^2/(N*c) + f

where:
f... focal length
N...stop number (i.e., 2.8 )
c...circle of confusion diameter

Near and far plane:

Dn = s*(H-f)/(H+s-2f)
Df = s*(H-f)/(H-s)

where:
s...subject distance.

Here are my calculations for c=0.012 mm (roughly a 1/3" CCD?), but doesn't really matter what c value you use:


Focal Length (mm):15 45
F/stop:2.8 2.8
Hyperfocal distance (m):6.822 61.306
Subject distance (m):3 9
Inverse of magnification:199 199
Dn (m):2.085 7.852
Df (m):5.343 10.541
D-Dn (m):0.915 1.148
Df-D (m):2.343 1.541
DoF=Df-Dn (m):3.252 2.689


In my mind, that shows that even theoretically, depth of field gets shallower as you move away from the subject and increase focal length while maintaining the same magnification.

Jerry