Pellicle Mirror
 

Another feature that would be KILLER would be the addition of a pellicle mirror. That would allow vision through the optical viewfinder, as well as high-quality auto-focus.

Hopefully, we will see that on future Canon DvSLR technology cameras, whether in a still or video format body.

Pellicle mirror - Wikipedia, the free encyclopedia

It's not my idea. Credit where credit is due:

There was a Canon SLR that had a Pellicle mirror, think it was the EOS1RS, had super fast motor drive because the mirror could stay put. You lose some light though.
Steve

According to the video and Wikipedia, Canon has made at least three such models: the Pellix, the RT, and the F-1 high speed (or 1-RS, according to the video). That means that Canon would have minimal royalties to pay if it brought Pellicle mirror technology back to its lineup.

I am very much with Chris on this one - the issue is very much regarding the lens.

For a DSLR it's not necessary for the lens to track at all, and to keep costs reasonable they generally don't - they don't need to for stills work. Hence it's a question of zoom to frame, focus, then take the picture. For cinema type work, that may not be too much of an issue, you accept a fixed focal length during a shot.

General video work is a different story. It's firmly expected that it's possible to zoom from one end to the other and for the lens to hold focus throughout - and this the average still camera lens just will not do for very good design reasons. (They are commonly not real zoom lenses, rather variable focal length lenses. A true zoom holds focus throughout the range, with the latter you have to refocus every time the focal length is changed.)

It's possible to build a lens to meet various criteria relatively easily - to be wide angle, or to be fast at max aperture, to hold focus accurately throughout the range, to have a good zoom range. The trouble is when it's desired that a lens should have all of these desirable qualities *AT THE SAME TIME*.

And the bigger the chip imaging size, the more difficult (ie expensive) it is. That's before we even consider other factors like zoom servos etc.

Hence the popularity of 1/3" chips. It's not because they are good in themselves, it's because they mean that most of the desired design criteria can be met reasonably well at not too high a cost or size/weight.

It's also why the EX series have taken off so well. It's quite a triumph of design to be able to use chips with twice the area of 1/3", yet still keep overall size/weight/cost to the levels that have traditionally meant 1/3".

Bolex 16mm cameras had a semi-silvered mirror (same thing more or less as a pellicle I think). Can't remember whether there was a downside, but Arri and Aaton didn't use one, so they obviously didn't think it was the way to go.
Steve

Quote: who would use an AF lens on a video camera?

'Push to focus' is a very very handy thing to have!

True, but only if your viewfinder is awful (as it is on most small cams, notably the Canon XL1 types).
Steve

More than a decade ago, Panasonic made a three 1/3" chipper that was really small - the EZ1 (and it even had a lens with a 10x optical zoom). Looking at the EZ1, I'm not so sure the EX cams are that much of a triumph in size.

If you take something like a DSR500 as representative of a typical 2/3" camera, and something like a PD150, PD170 or Z1 as typical of a prosumer grade 1/3" camera, then the EX cameras are far closer in size/weight/cost to the latter than the former, whilst having chip areas halfway between 1/3" and 2/3" in area terms.

Yes, it's possible to get 1/3" cameras smaller etc than the examples above, but I don't think you're then comparing like with like.

The 5D/7D show that it's possible to get a 35mm sensor in something smaller than any of these cameras. That's not the issue. The issue is where to put all the connections (2-XLR, HDMI, SDI, large capacity battery, headphones, etc.) and all the switches (ND, gain, shutter speeds, zebra, etc.) when you have a smaller footprint. And you're going to need room for a the media to go. And a place to tuck away an LCD because no one likes the fixed one on the DSLRs.

There is only so much real estate to be had. A 35mm sensor in something the size of a DVX would be most welcome. From Canon or anyone else.

No, the killer feature is a mirror. So you don't lose much if any light. And it's already been done. Just not at this price point. If you want to see the newest iteration, Google "Arri Alexa".

Sorry, apples and oranges... David refers to a three-chip block built around a prism, while the D-SLR is a single-chip camera.

I agree that single-CMOS is the Next Big Thing, but the topic of this particular discussion centers around a three-chip design.

Ah... yep.

:)

Maybe the key is to have the pellicle mirror retractable. Use it when you want to use the optical viewfinder and AF. Retract it when you want more light, don't need/want AF, and will use the LCD or other monitor.

Also, you could feature it as a built-in ND. :)

The down side is that you don't have a bright viewfinder - Arri & Aaton V/Fs are in a different league with the rotating mirror shutter design. Plus you lose some light going through to the film.

I always quite liked Bolex viewfinders - the benefit over Arri/Aaton is that the image doesn't flicker due to the spinning mirror. You get used to this flickering after a while though.
Sorry, we've gone a bit off topic here - guess I'm just missing my old Arri!
Steve

As I explained in my article, manual focus -- if it's to be done properly -- requires a second person (the 1st AC) whose sole job is to pull focus manually while the camera operator does everything else. Many times it's not possible or practical to have a two-person camera crew; quite often the camera operator is working alone, and these situations can call for AF, or at least some way to drive focus remotely from the tripod pan handle or elsewhere. On this new Canon, remote manual control of focus will most likely happen through LANC, which calls for a focus motor on the lens, which means it will have AF. And as Graham Hickling has already pointed out, Push AF is tremendously helpful for quickly finding focus.

I think that's going to be very much a thing of the past, what with selective face tracking, zone tracking, phase detection technology (instead of contrast), etc. I think we're going to start seeing the high-end AF processes that are currently used on the broadcast side work their way down into the $5K class of cameras.

What broadcast cameras are you talking about that are using AF? None of the broadcast cams I use have it.
If you're talking about servo focus to allow the use of pan bar mounted focus demands then OK, but for serious video work AF is just useless. You talk about focussinf "if it's to be done properly", crikey AF is the least "proper" way to do it. In broadcast wildlife we've been focussing manually with massive telephotos for years and there haven't been many complaints. There is just no way AF on an HJ40 would work with fast action subjects like wildlife. With stills it's different as you only need focus to be spot on for an instant.
Steve

Lenses like the Canon DigiSuper 100AF are auto focus.

These are all AF:

Canon DIGI SUPER 100 xs AF

Canon DIGI SUPER 86 xs AF

Canon Canon DIGI SUPER 27 AF

The HJ40 you've mentioned is Servo focus, not "just manual" (that is, it's motorized for remote focus control):

Canon HJ40 x 14B IASD-V (and anything else with an "A" in the model number is Servo focus, not just Manual).

I'm not doubting that AF is improper for wildlife videography. I'm only saying this:

The forthcoming Canon replacement of the XH series (that is, the topic of this thread) will *most likely* allow for remote focus control.

At the $5K price point, that remote control interface will *most likely* be LANC.

If it has LANC, then it will *most likely* have AF.

And also since it appears to be a "tapeless XH," then it will *most likely* have AF regardless.

The usefulness of AF for any pro videographer will be, again as Graham has mentioned, at the very least the ability to quickly find focus with Push AF. I'm also willing to bet that selective face tracking AF and selective zone AF will be included in the new camcorder, primarily for the benefit of wedding & event videographers. That's all I'm saying here.

I know the big box lenses are AF, but that's not to say they're used in AF.
I know all about the HJ40, it's one of my standard lenses. I find the servo focus doesn't work too well either, the lack of contact with the lens itself seems to lessen your control over fine focussing.
I have no doubt at all that the new Canon will have AF, all the semipro models do. Because the lenses use servos rather than helicals AF is incredibly easy to implement so why not.
Steve

AF not useless for wildlife
 

I'll donate my 2 cents for filming wildlife with an ultralight tripod - too often more robust equipment would be just too much to carry. Focusing by hand while shooting results in shaky footage, then. While at it, I wish AF could be taught to follow a particular animal. These systems already detect shapes so why not. If AF would be better, I would use it in most cases.

The latest Sony face detection can learn a particular face for priority even if it goes out of frame and then returns!!!!!!
I use Spot focus all the time on the SR11 and XR500 and was a little disappointed it was not implemented on the NX5.

Ron Evans

Quote "The latest Sony face detection can learn a particular face for priority even if it goes out of frame and then returns!!!!!!" - yes, theoretically! It's the same principle as all the Nikon 3D focus tracking etc., but it's far from foolproof.
Steve

For a thread based on speculation this has taken on a life of its own.

I think the predictions Chris suggested in his excellent article will come to pass, and we'll see a new contender for the prosumer AVCHD market that could see off the competition at this price point. But it doesn't look like Sony need to feel too threatened re its EX range of cameras, and IMO, the most likely competition will come from Panasonic bringing out an AVC Intra version of the 171, a fixed lens hpx300, which i for one would be very tempted by, even with the 1/3 chips.

Having sold my XHA1 last year (for a 5d mkii) i'm interested in buying another video camera this year. I've a project coming up which ideally would require a small camera and i think the HMC40 looks like a bargain, assuming it would complement the 5d. But if i were looking at a larger camera, i can't see any real competition to the EX1. Also, there are dozens of EX1's on ebay going for less than Canon's new AVCHD is likely to cost, and you can always sell it for pretty much what you paid. So for the time being at least, the EX1 looks to me like the best of the bunch, and i'm mystified why none of the other big companies are attempting to match it, let alone better it.

Now P2 card prices are becoming a realistic alternative, any thoughts on the likelihood of a fixed lens hpx300 - along the lines of an hpx171 with AVC Intra- turning up in 2010?

A year ago I said I wouldn't trade my EX1 for any camera under $10k. I feel the same way today. It's a HECK of a camera for what it costs. As to your comment about P2 becoming more realistic... I think Sony really caught them off guard with SxS and it's pricing. Now that P2 prices have fallen more in line with SxS, Sony dropped the bomb this week that they are supporting the SDHC option in the EX line of cams. So now you have manufacturer supported recoding in the EX line at about $45 per hour. Not sure how Panasonic plans on responding to that.

While I think Canon has had remarkable success in their DSLR product (seemingly by accident for the most part), they seem to be lagging badly in the prosumer video market and I really do fail to see how this new offering is going to find itself in the hands of anyone other than Canon loyalists. Those who prefer other brands already have viable options and honestly, looking at what is speculated for the Canon, doesn't seem to offer enough to compel someone to switch. Maybe I am wrong about that, but that's how it looks from where I'm standing.

Even though I was brought up through the decades on a diet of MF stills cameras and prefer to use MF on my video cameras when I can, there are times, even in nature & wildlife photography & filming when AF can come in handy.
There is no advantage of having an MF-only lens when you can use the same price & quality AF lens which has a simple ON-OFF switch.

"A year ago I said I wouldn't trade my EX1 for any camera under $10k. I feel the same way today."

I don't blame you. I would too.

When Canon responded to the Z1 with the XHA1 i was a happy camper, but the EX1 has been out ages now, and though there have been some good cameras from Panasonic and JVC, none have really tempted me, particularly since there are so many used EX1's kicking about at half price. I think we were all hoping Canon would come to our rescue, and Chris's article - though based largely on assumptions - has left us feeling disappointed.

Having said that, Canon has always been great for glass, so it's possible that they'll come out with a camera with superlative 1/3 chips, some nice features and a fantastic lens and affordable media at an attractive price-point. If that camera produces a great image, then i'd buy one over the Sony.

Agree with that Tony, no harm in having it if you can, why not.
But for video of wildlife is a cmplete non-starter. Do you know anyone working on high quality broadcast wildlife progs that has used it? I certainly don't, and I just can't ever imagine it happening. If you look at all current wildlife output it's invariably Varicams with HJ40 or HJ18, and before that it was Arris with 300s and 150-600s, all MF.
Steve

If you had to design a camera as an upgrade to the excellent XL-H1 what would you do to it? Problems I see are these (not used one much but enough to see problems). Viewfinder is awful (related to the AF discussion going on here as it's needed becasue MF is nigh impossible - only sabing grace is massive depth of field because of 1/3" chips). No slomo, needs some way of getting at least 60fps. Tape rather than solid state - actually a mixed blessing! No true progressive mode, but Frame mode sort of is the same. No NDs on the body, so using 35mm stills lenses is problematic. HDV codec is hopeless, needs at the EX 35mb/sec codec. Lens controls - yuck, continually rotating focus rings etc. The EX lens controls are (for the price) excellent.
That's the list that springs to mind so far, and if all they do was rectify those things they'd get my vote.
Steve

I don't know how smart the AF is, but a lot wild life has some form of camouflage which could cause problems for a system that's set up for human beings rather than animals with counter shading etc.

The EX1 left a lot of room for improvement. It got most of the basics right. Great sensors, SDI port, awesome LCD. But the VF was weak, the codec needed a boost in bit rate, it really needed HDMI, some of the buttons were awful, the power switch was a joke, etc. The EX1r has addressed nearly every issue outside of the codec.

If Canon could offer something similar to the EX1r with a better codec, they'd give Sony a run. The NXCam is really going to sew up the $5k space I think. Once you get north of that, the EX1 is the heavyweight in the room. South of that, the HMC150 is the obvious target.

What baffles me is that these manufacturers seem to have all the stuff.. just not on one camera. It would take NOTHING for Sony to turn up the bitrate of the EX1r. It's offering HDV and 35Mbps XDCam. Just allow 50/100Mbps XDCam and the #1 knock against the camera just fades into oblivion.

It's not AVC-I that would make so much a difference, as putting in the full raster imaging chips. The HPX170 just doesn't use imaging chips that even come close to offering the resolution of the imaging chips in EX camcorders, and Panasonic's AVCCAM AVCHD H264 encoding is apparently as good or better (image fidelity wise), compared to XDCAM EX MPEG-2 encoding from the EX cams. I'd venture a guess that, with typical footage, AVCCAM AVCHD image quality might exceed image quality of the 50Mbps flavor of AVC-I even (certainly for very low motion footage).

Could be that Sony don't want to step on the toes of their higher end cameras by putting the higher bit rate into the EX1r.

No doubt about that whatsoever Brian.
They are shamed by people like Convergent Designs who prove it to be true.
Steve

Um,

Based on what Sony put in the EX1(r) it doesn't appear they care about stepping on their more expensive cams at all. Moving 1/2" sensors down the range. Moving SDI down the range, etc. What's the next model in the Sony line that does over and under cranking? Replaceable lenses on the EX3. The image shaping on the EX line is right off the $50k level cams.

If anything, Sony seems to be more willing than anyone else to take features from the big boy cams and move them into the prosumer line.

It sure looks like the NX5U is going to be doing some stepping on the EX cams, with some of it's features (like very elegant and affordable redundant recording). Sony is competing with themselves more than any other camcorder manufacturer is.

Perrone, they are stopping short of the higher end cams by not putting 50 or 100 mb/s codecs in the EX - no reason at all other than marketing.
Steve

If Canon's new cam offers H264 encoding at 32Mbps (or higher), along with an imaging block that can resolve 1000 lines of detail cleanly, Sony may have some incentive to offer higher bitrates with the EX cams. Right now they don't.

Not sure I agree. The primary reasoning is that they have had data rate limitations. The Expresscards can do the job, but their external recorder and the SDHC cards that have been available wouldn't hold up to the bit rates. The new class-10 SDHC cards should solve that problem, at least on the EX1r, but this may still not be possible on the EX1.

It's the same reason that overcranking failed on the SDHC cards in the original EX1/EX3 once you got to a certain rate.