Something new from Canon on Nov. 3rd...

[David Chilson]

LAKE SUSPENSE, N.Y., October 27, 2011 – Canon U.S.A., Inc., a leader in digital imaging, today announces the long awaited 5D MKIII. In addition to superb low-light performance and increased overall speed, this camera has the new DIGIC 5 image processor, HS SYSTEM, Full HD video capture, High-speed Burst HQ and Intelligent IS providing clearer, smoother looking images. Limited numbers will be available for the Holiday shopping season.......

If Chris is right that history repeats itself the last time a new processor was introduced (G10 I think) it was followed up by the 5D MK II. The soft S100 release, current discounts and dealer batch pricing would support this.

As far as the Nov 3rd date, two new cameras one taking the new Canon PL mount and the second that utilizes EOS lenses, to be released around NAB. After that my crystal ball gets a little cloudy.....

[Chris Hurd]

Originally Posted by Jon Fairhurst

Canon. New York. October 26th. Pixma. EOS....

That's the eve of PhotoPlus Expo.

It's always a nice little photo trade show. Not as big as PMA used to be, but still a very good show. I'm planning to be there, as usual.

[Jon Fairhurst]

I think David might have nailed it. 5D3 at PhotoPlus. Shipping almost immediately. EF/PL cinema cam(s) announced in Hollywood. Shipping in April, 2012.

We will see if the Hollywood glamor rubs off on the stills line.

My guess is that a 1D5 will follow in Q1 2012. It will be full frame and include a (possibly windowed) sports mode with higher burst rate, merging the 1Ds and 1D into a single camera.

[PURE CONJECTURE MODE - OFF]

[Edward Mendoza]

If the Nov. 3rd Canon announcement (or Oct. 26th announcement for that matter) is in regards to a single-sensor cinema cam to compete with the likes of the F3, FS100, or Scarlet, it would be really nice if they offer both a "just-the-cam" option and an option that includes a set of fast primes, like Sony did with the F3. The unfortunate thing is it'll likely be up in the price range of a Scarlet or F3, which'll likely gear us poor-folk video guys to the 5D Mark III, that is if it's one of the releases and is similar to the Mark II.

[Jim Martin]

In the past, as far as I can remember, Canon offered the XL-1s or XL-2 as a body only and it didn't sell that well. Now, times have changed, and if a new cinema camera is on it's way, I would think they would offer it with and without a lens ala the Sony F3. A bigger question would be, would they offer a motorized zoom for the camera? Can you make a motorized zoom that works and not interfere with the communication between the EF lenses and the camera....because I would think the most important thing would be for all the 5D & 7D owner's lenses the seamlessly work with a new video/cinema camera....if that's whats coming.

Jim Martin
Filmtools.com

[Edward Mendoza]

That would be awesome to have a motorized zoom lens as an option. It stinks having to invest in two cameras--one for run-n-gun and another for staged videography. How sweet would that be to be able to take your camera from your commercial shoot, switch lenses, and now be able to shoot sports?

[Brian Drysdale]

A sports option for the S35 sensor already exists with the adapters that allow you to use 2/3" video lenses on the large sensor cameras. Of course, they're not that cheap, but neither are the lenses and the max aperture becomes f 4.

From a cost point of view on a S35 sensor a modified stills zoom lens could be your best bet, but it won't be a piece of glassware with a wide aperture and a wide zoom range.

[David Heath]

Originally Posted by Brian Drysdale

A sports option for the S35 sensor already exists with the adapters that allow you to use 2/3" video lenses on the large sensor cameras. Of course, they're not that cheap, but neither are the lenses and the max aperture becomes f 4.

I believe the diagonal on a 2/3" camera is about 11mm, and on S35 more than double that?

So it's worth underlining that any such adaptor is not a simple mechanical coupling, but must have optics to spread the image to cover the bigger sensor. Which is why your f1.8 2/3" lens translates to f4 (maybe less) when used with such an adaptor.

The other issue may be powering of the zoom. 2/3" cameras will be designed to electrically drive the lenses - will your S35 camera have the same ability?

And as far as a dedicated motorised zoom lens goes, then yes - anythings possible. But it will need a focal length range more than double that of an equivalent 2/3" lens for the same angle of view. It may be possible - but expect it to be big, heavy and expensive. Or compromised in terms of zoom range and/or maximum aperture.

This is why in the broadcast field 2/3" has ended up as the norm - a good compromise between versatility, cost and such as dof control. Larger formats may bring benefits if you want more dof control, but expect it to be at the cost of other factors.

[Daniel Browning]

Originally Posted by David Heath

I believe the diagonal on a 2/3" camera is about 11mm, and on S35 more than double that?

Yes -- around 25-28mm. If Canon does an S35 sensor for this November 3rd announcement, I wouldn't be surprised if they used the same sensor size as the 7D (22.3 x 14.9mm active area).

Originally Posted by David Heath

So it's worth underlining that any such adaptor is not a simple mechanical coupling, but must have optics to spread the image to cover the bigger sensor.

Agreed. Optically, it's the exact same thing as a rear teleconverter, and has the exact same downsides.

Originally Posted by David Heath

It may be possible - but expect it to be big, heavy and expensive. Or compromised in terms of zoom range and/or maximum aperture.

I do not consider a larger format with a narrower f-number to be "compromised" in any sense of the term. For example, if the November 3rd announcement is a S35 camera with a 20-400mm f/7-15 lens to match the 4.5-90mm f/1.6-3.5 lens on 1/3" formats, I would consider it equal. If anything, the lens would have fewer aberrations and lower cost. But they could make it much better (and not just equal) for only modest increases in weight/cost (e.g. 20-400mm f/4-8). I explained the whole reasoning here:

http://www.dvinfo.net/forum/canon-eo...eeper-dof.html

For example, a 70-200 f/4 lens on the 5D2 is sharper, lighter, and cheaper than the 35-100 f/2 on Four Thirds. And it has the same (or less) noise at ISO settings that are two stops higher (e.g. 1600 vs 400).

Basically, the reason comes down to the fact that 200mm f/4 and 100mm f/2 have the same aperture diameter (200 / 4 = 50, 100 / 2 = 50).

Originally Posted by David Heath

This is why in the broadcast field 2/3" has ended up as the norm - a good compromise between versatility, cost and such as dof control. Larger formats may bring benefits if you want more dof control, but expect it to be at the cost of other factors.

The reason 2/3" ended up as the norm is because of the issues/limitations imposed by the dichroic beam splitter, at least according to what I was told by a guy who designed three industrial beam-splitter cameras.

Speaking of 2/3", if the Nov 3 announcement turns out to be an XL-series replacement, I would prefer it to be 2/3".

[David Heath]

Originally Posted by Daniel Browning

I do not consider a larger format with a narrower f-number to be "compromised" in any sense of the term. For example, ...............

I'm familiar with your reasoning, Daniel, and (in general) tend to agree with it. By "compromised" I meant a direct comparison in so far as f nos and zoom range etc go, and I'll stick by that. (And I don't think you'll disagree with that, in so far as it goes?)

The wider issue is that you're saying that if sensor A has (say) 8x the area of sensor B, then it will give identical depth of field at a 3 f stops less? So if B is at f2, A will be at f5.6? I agree with that.

I believe the follow on is that the bigger sensor gathers more light, therefore better sensitivity, hence A at f5.6 will need the same lighting level as B at f2 - hence equivalence. With all else equal, I agree with that assumption.

But in reality, the thing that will NOT be equal is that the 2/3" system is likely to be 3 chip, the s35 likely single chip. And a single chip of a given size will inevitably not be as sensitive as a 3 chip system of the same size, due to the filtration loss. In the example above, for B at f2, A is more likely to be around f4 at best. And once you start to consider other factors, it's likely to be even less, that includes using sensors designed for stills, such as DSLRs and the AF100. (Which is why the FS100 has a better than expected sensitivity advantage over the AF100.)

[Daniel Browning]

Originally Posted by David Heath

I'm familiar with your reasoning, Daniel, and (in general) tend to agree with it. By "compromised" I meant a direct comparison in so far as f nos and zoom range etc go, and I'll stick by that. (And I don't think you'll disagree with that, in so far as it goes?)

Yep, I agree here and with the rest of your post too.

Originally Posted by David Heath

But in reality, the thing that will NOT be equal is that the 2/3" system is likely to be 3 chip, the s35 likely single chip. And a single chip of a given size will inevitably not be as sensitive as a 3 chip system of the same size, due to the filtration loss.

You're quite right; that's a good point that I forgot about. Increased sensitivity is one of the major advantages of 3-chip systems. The lens on an equivalent Bayer camera would have to be at least a stop faster just to equal the sensitivity, and that means heavier, more expensive, and thinner DOF (which is often not desired).

[Emmanuel Plakiotis]

Do you account for the loss of light due to the prism splitter. Maybe what the 3 CCD gains from the larger area is lost due to the prism in front of them.

[Daniel Browning]

Originally Posted by Emmanuel Plakiotis

Do you account for the loss of light due to the prism splitter?

No, but I think it's less than 1/3 stop. Best case scenario it has no additional air-glass surfaces compared to Bayer, and only 2% loss of desired spectrum from the dichroic mirror coatings.

Worst case, it has four extra air-glass surfaces (one air gap inside the prism and another between the prism and each sensor), bigger loss of desired spectrum from the dichroic mirror coatings, and additional loss from an extra color filter over the sensor (to remove undesired spectrum). If you guess 5% loss at each extra surface (i.e. no anti-reflective coatings at all), and, say, 15% loss of desired spectrum from the dichroic mirror and color filters, that's still less than 1/2 stop of light loss.

[David Heath]

Originally Posted by Emmanuel Plakiotis

Do you account for the loss of light due to the prism splitter. Maybe what the 3 CCD gains from the larger area is lost due to the prism in front of them.

Good point, but those losses are small compared to the difference between 3 chip and single chip due to such as Bayer filtration. "A stop" is a rough approximation for the difference, and allows for the prism loss. Practically, I'd expect somewhat more than a stop rather than less. Because of the dichroic principle, the great majority of the incident red light should go down that route, and ditto for green, blue.

In a Bayer sensor, half the green light is lost to filtration, (1/4 in the red, 1/4 in the blue photosites) and 3/4 of the red and blue. It's difficult to easily equate that to simple sensitivity terms because of the matrixing, but "about a stop difference" is a decent ball park figure.

What this means is that (with all else equal) something like a 3 chip 1/2" system will have roughly comparable sensitivity to a single chip 2/3", and the same for any two systems where the single chip is twice the area of the 3 chip.

[Steve Kalle]

EDIT: I don't want to get anyone in trouble; so, all I can say is that the camera 'should' be a large sensor video camera. It should be a major competitor to the F3. It should have a sensor size similar to the 7D with electrical EF/EF-S support and 50Mb 422 codec.

These 3 features alone should should make the camera extremely popular. Many people have been waiting for a real video camera to use their Canon glass with.