Aspect ratios and CCDs on XL2 and DVX100A

[Haakon]

I am getting more and more into shooting motion picture, and sometimes discussion over aspect ratio really confuses me. After seeing Canon's recent announcement of "true 16x9" CCDs on the new XL2, there have been a lot of comparisons to other cameras (most obviously and notably the AG-DVX100) about "effective pixel count" and "true 16x9" that have left me a bit in the dark.

First off, I found http://www.dvinfo.net/canonxl2/articles/article06.php with information on the CCD used in Canon's new XL2 camcorder. I am a bit confused as to why a camera with "true 16x9" chips would look 4x3 in shape, and why they'd choose to make 1/4 of the area completely "inactive." I guess this all comes down to the age old debate of what aspect ratio is ultimately "better," but it seems to me that if you've got a 4x3 sized chip, then it's native, or "max" resolution would be in 4x3 mode. Yes, I realize that a rectangle (let's say 16x9) is wider than a square (say 4x3 - I know this isn't really a "square," but just play along), so it would appear that a "widescreen" 16x9 image has "more data" to the left and the right. People are conditioned to this response because they're used to seeing 4x3 pan and scan versions of widescreen movies that were originally shot with an intended 16x9 (or wider) aspect ratio, and when it's "formatted to fit their TV," they lose a bunch of the picture. But really, if the chip itself is 4x3 in shape, then by making 1/4 of it "inactive," aren't you in reality just "letterboxing" the chip to get your "true" 16x9 area? And if so, why on earth would you make the 4x3 area even SMALLER by "double letterboxing" it?? The 4x3 aspect ratio that is actually used by the camera looks like it's only using about 2/3 of the total CCD (maybe even less?), which seems incredibly wasteful to me. And considering everyone is so big on resolution and having the "maximum effective pixel count" possible, why would a manufacturer choose to make the system work like this? Is it so that the 16x9 mode actually looks "wider" when you use it and the manufacturer can claim that's it's "true" mode? Sorry, but if I'm going to spend $4,000 on a camera, I want to get my money's (CCD's) worth. To my understanding, the Panasonic AG-DVX100 "16x9" mode is more of a "letterbox" mode, which means that the CCD is 4x3 in nature, and it renders the top and bottom "inactive" to shoot in "16x9." Which, if the chip actually is 4x3 in physical size, makes sense to me. This Canon chip looks like it's already letterboxed to begin with, but then letterboxes again on the left and right to make up a 4x3 image. This means they've thrown out all kinds of resolution and have effectively wasted half of that CCD when making up a 4x3 image. I'm not making any claims or pointing fingers, I'm just trying to understand why it is being done this way. Are my observations anywhere close to what's really happening?

Also, "16x9" is not a common motion picture (movie/film/whatever) ratio as far as I'm to understand. (16x9 works out to roughly 1.78x1, whereas most motion pictures are shot at either 1.85:1 or 2.35:1... right?) Now I realize that HDTV has an agreed upon aspect ratio of 16x9, but if the 24p cameras are more aimed at independent movie makers and not HDTV producers (they aren't HD-resolution, anyway), then why on earth are the "widescreen" modes in these cameras 1.78:1 ("16x9") and not 1.85:1?? Frankly, with the options that have been presented to me, I feel like always shooting in 4x3 mode and then letterboxing it to 1.85:1 in post, because then at least I can crop/recenter the image vertically to meet my aesthetical preferences if I feel like something wasn't shot perfectly - with no loss in resolution. (If I shoot in "16x9 letterbox" with the Panasonic, then I'm then stuck with the image I have, and if I shoot 4x3 with the XL2 then the resolution is seemingly much worse than the AG-DVX100, so there's really no point in using that camera at all!) Essentially, unless the actual chip in the camera is physically 16x9 in size, then it's never really "truly" 16x9... (and the chip should really be 1.85:1 in size anyway.)

These may be basic questions, but the answers seem really unclear to me. It's very exciting that 24p digital cameras are now accessible in the "prosumer" domain, but the choices are slim (and with my confusion over the aforementioned topics, I'm not 100% ready to jump on this.) Since I don't have $60,000 to spend on a camera that's "professional-level," I'm trying to find the best solution for my needs that will actually allow me to make the best quality films I can on my own until I get to that point. I've been very impressed with the capability of the AG-DVX100 (I've seen both good and bad footage shot with it, but the best of the good looks quite acceptable to me - especially Ralph Oshiro and Ted Lederman's work), and I'm definitely leaning in that direction. With the advent of the XL2, obviously I want to do my research before shelling out four thousand dollars, but the design of their CCD just makes me feel like it's a huge step backward - and it's all so they can claim that it's "true 16x9" since they've hardcoded it in a way that a 16x9 ratio yields the highest resolution that the camera is "allowed" to shoot (regardless of the fact that the CCD isn't "truly" 16x9 at all). Ug!

If anyone can touch on any of my questions and provide some clarity for me, I'd really appreciate it. I feel like the confusion I'm having is legit in at least some remote way. It certainly would be much easier if there weren't so many choices of aspect ratio to begin with!

[jessecoane]

i'm not sure about the ccd thing either, but 16x9 is the aspect ratio of dvds. 2.35 x 1 dvds are 16x9 with the top and bottom matted for 2.35. Thats why plasma screens still have some black on the tops and bottoms for 2.35 x 1 programs. Hope that helps.

jesse

[Mike_Donis]

16x9 chips are very expensive - and they'd be 2/3 chips, not 1/3 chips. I don't think Canon has ever said it was TRUE 16:9 chips, only that you can shoot in true 16:9 resolution - and they're not lying. The fact that they crammed more pixels into a 4:3 CCD means that we're getting a smaller use of the CCD, but we're still getting the full RESOLUTION.

As far as the 16:9 vs 1.85 thing goes: all NLEs support 16:9 editing, because they're made for video releases. So we're cutting in video, we get to use 16:9. The difference between 1.78:1 and 1.85:1 is so small though, anyways, that it doesn't really matter. There's more of a difference between different theatre's safe zone markings, and the overscan on widescreen television sets - if you're shooting in 16:9, projecting it at 1.85:1 won't screw you over at all. And because the NLE's support 16:9, it makes everything all the more convenient.


The XL2 may not be immensely better than the DVX, but it seems (on paper) to be a good rival. Canon hasn't really screwed anybody over here, and I can't wait to see what it looks like up against my DVX!

[Haakon]

The fact that they crammed more pixels into a 4:3 CCD means that we're getting a smaller use of the CCD, but we're still getting the full RESOLUTION.

I can see that Canon is using a higher-pixel (more dense) CCD than the DVX, so even though parts of the CCD are "inactive," the resolution of the 16x9 image is higher than that of the DVX's. *But if ANY of the CCD is inactive, how can you possibly say you're getting the "FULL resolution" out of it? *You're wasting a bunch of the CCD... I don't see how that's "full" anything. You might have more resolution, but you're not using all of it. Also, more pixels in the same-sized CCD translates to more picture noise and higher susception to visible abberation. *On top of that, let's not forget while the 16x9 image contains more pixels, their doublecropped 4x3 image uses only 345,600 of them (whereas the DVX uses 380,000). *So in 4x3 mode, the XL2 has worse resolution than the DVX. *I just don't like how it's being touted as this "true 16x9" camera when it's CCD isn't 16x9 and their 4x3 mode has less resolution than their strongest competitor (and we haven't even gotten into 8-bit vs. 12-bit DSP yet). *If Canon had said "more resolution in 16x9 mode than the AG-DVX100" that would be one thing, but I don't see how they can make claims of "full" or "true" anything with this camera.

[Mike_Donis]

Well, it's not full resolution of the CCDs, but it's full resolution according to the DV specifications. Actually, the CCD image area extracting the 16:9 shape is higher resolution than what's actually recorded to tape (720x480 in NTSC).

So while it's not as good as 2/3", 16:9 native chips, it's still better (on paper) than letterboxed footage (resolution-wise).

[Barry_Green]

Haakon, I don't really see where you're confused -- you seem to have a pretty good grasp of the whole situation. You're basically right on all counts.

(except the aspect ratio thing: Mike explained that, there's really practically no difference between 1.85:1 and 1.78:1, and 1.78 is what TV is going to, so that's what the standard is).

[Haakon]

Okay, that makes more sense Mike. *I still think their terminology is misleading, but I doubt they care what I think. *:P

I guess my question is, Barry, if they've got this huge 680,000 pixel CCD that's 4x3 in shape, why aren't they using all of that resolution for their 4x3 mode and just letterboxing it to create a 16x9 image? *I know everyone seems to cringe at thinking of anything as "letterboxed," but let's face it - that's already what they're doing. *The 16x9 image uses 460,800 out of the total 680,000 pixels regardless of if they use the full CCD for the 4x3 image or if they crop it even smaller than their 16x9 field. *In doing it the way they chose to, they make the 4x3 image even lesser quality... and that is what is making no sense to me.

Here's another way to look at it: if then entire CCD was used for 4x3 mode and I recorded in that mode, I can then matte (letterbox, whatever) the image in post to whatever aspect ratio I desire (since 16x9 isn't a "film" aspect ratio as we discussed earlier). *Then I'd have full use of all of the resolution in the chip at my disposal, to crop or shape in whatever way I wish. *THAT would be "FULL resolution" in my opinion. *The way they chose to do it, they've hard limited it to 960x480 when they could make another good 25% more resolution available to users using the exact same CCD chip. *Does it make any sense to do it the way they're doing it at all?

[Mike_Donis]

It would be cool if they gave you the full surface of the chip in 4:3 mode...but I'd bet that their 16:9 mode would still look better...because remember, they're still scaling down to 720x480. So a gazillion megapixel chip would only become 720x480 in the end....and their 16:9 shape from the centre could be scaled to 720x480 ALSO....

the inital image would be better in 4:3, but when recorded to tape, you probably wouldn't notice a difference.

[Haakon]

Okay, I'll buy that. But then if the image on the tape is only 720x480 (which isn't a 16x9 aspect ratio), how does the 16x9 footage look on the tape? Is it letterboxed or will it look vertically stretched (anamorphically squeezed)?

[Mike_Donis]

The viewfinder displays it as letterboxed, so you can see it in its proper dimensions...but it's recorded stretched to tape.

[Haakon]

That sounds like it works almost the same way as the "squeeze mode" on the DVX100A. *I know you'll say that the difference is that the XL2 ultimately has higher resolution than the DVX, but since the DVX has a full 720x480 lines of resolution, then following your "gazillion megapixel" example, isn't it all a wash in the end?

[Barry_Green]

I guess my question is, Barry, if they've got this huge 680,000 pixel CCD that's 4x3 in shape, why aren't they using all of that resolution for their 4x3 mode and just letterboxing it to create a 16x9 image?
Your understanding is correct. *As to why they have done so, I cannot fathom a guess. *All I can do is assure you that (as far as I know) your understanding of WHAT they have done is accurate. *The WHY remains a mystery.

It could be as simple as pandering to the marketing department... by cropping the sides in 4:3, they can then demonstrate that the picture gets noticeably wider when switching to 16:9, therefore "see what a great benefit we have given you?" *I would be most disappointed if that was the truth, because it would mean (in my opinion) that they sacrificed 4:3 performance in order to create a lie to sell cameras to unsuspecting consumers. *Let us hope that there is a better reason.

In doing it the way they chose to, they make the 4x3 image even lesser quality... and that is what is making no sense to me.
Agreed. *We do not understand why they have chosen to do this, and on the surface it appears to make no sense. *I do hope that in the four years it took them to design this camera, that much engineer time was spent on the problems and that they made the right choice (for whatever reasons, hopefully someday to be made clear). *But at first glance, it looks like they just sacrificed 4:3 mode in order to make 16:9 sound more appealing.

Furthermore, sacrificing 4:3 performance makes no sense to me. We live, as I have said frequently, in a 4:3 world. While Europe is closing in on 50/50 as far as 4:3 vs. 16:9 TV's, America remains about 97% 4:3. Why design a camera that's purposefully hampered to deliver video in the overwhelmingly dominant television delivery system? I remain puzzled.
since 16x9 isn't a "film" aspect ratio as we discussed earlier).
Yeah, but don't get hung up on that. *The difference is fairly miniscule.

Then I'd have full use of all of the resolution in the chip at my disposal, to crop or shape in whatever way I wish. *THAT would be "FULL resolution" in my opinion. *The way they chose to do it, they've hard limited it to 960x480 when they could make another good 25% more resolution available to users using the exact same CCD chip. *Does it make any sense to do it the way they're doing it at all?
In our limited understanding of the process, no it makes no sense. *The aforementioned "pandering to marketing" is one possible explanation, although the most unsatisfying one. *I hope there are better reasons forthcoming.

It appears they have adopted an identical strategy to JVC's implementation of the HD1, a 4:3 chip that extracts a 16:9 patch for its HD and SD modes, and then extracts an even smaller 4:3 patch from the 16:9 area for its "DV" mode. *I have heard no justification from JVC as to why they implemented it that way, although at least JVC does have one mode (still photo capture) that uses the full chip.

The ball is in Canon's court, as far as an explanation goes. *Until we get one, it appears that they're implementing 16:9 in the same way as the PDX10 and other Sony and Panasonic and Canon cameras: using a megapixel CCD and sampling a 16:9-shaped patch from it. *Which can give you "native 16:9" resolution, but the rest of the video quality suffers (particularly low-light performance and image noise, and in the case of the PDX10, artifacts such as smearing). *I do so hope this isn't the case. *But I fear it may be.

[theos]

SO what about the dvx plus the anamorphic adapter filling the 4:3 fully through optics? How does that compare with the stretch? Are we going back to those old endless posts of anamorphic vs squeezed vs letterbox? ;D

[Mike_Donis]

Well, the DVX has 720x480 resolution to start with, but when squeeze mode is used, its effectively 720x360 (approximately). So that being the case, the XL2s "squeeze mode" equivalent actually does have 720x480 effectively (and started out, before being recorded to tape, with even more resolution on the sides).

The XL2 scales *down* to get 16:9 stretched. The DVX scales *up*...so in theory the XL2 should have about 33% higher resolution than the DVX in 16:9.

It uses the same size chip, though...that should be taken into consideration as well, when thinking about quality. The truth is, nobody has seen a side-by-side comparison yet, so nobody knows for sure. But on paper, it does indeed seem that one benefit to the XL2 will be its resolution in 16:9 mode.

[Barry_Green]

Nobody knows, that's for sure.

The DVX with optical anamorphic adapter would *probably* deliver better overall video quality, all other things being equal, because you'd be using a larger chip, with larger pixels, better light-gathering capability, and no scaling/resizing artifacts (such as might be expected when an XL2's 960x480 CCD pixels are resampled to create a 720x480 DV frame).

However, that's paper talk. Just theory. *In reality, the XL2 will probably be much easier to work with (no worries about minimum focus distance, or keeping the aperture within a certain range to deliver adequate sharpness, also autofocus would work properly). *If the chips deliver the goods (which remains to be seen) then the XL2 could be the superior acquisition hardware for 16:9 (again, all things being equal, and we do not know if all things are equal yet).

The cameras need to be stress-tested, put through smearing and noise and color rendition tests and low-light and shadow and highlight and blowout and latitude tests, before we'll know if they're comparable. *If all else shakes out comparably, then the XL2 would be the easier way to shoot 16:9 at full chip resolution.

[theos]

Any news re the AA guide yet Barry? I am looking out!

[Barry_Green]

I posted an announcement -- it's available now. Go here to order:

http://www.dvxuser.com/articles/anaguide/

[theos]

THANKS!!!

[theos]

Wtat does an autographed copy cost ;D ;D ?

[Barry_Green]

For you, Matthew... no additional charge! ;D

[jrv3034]

Hello, all. This is my first post here, though I've been reading this forum for months.

Not to beat a dead horse, but would it not have been better to leave it as the full resolution 4x3 chip and use an anamorphic 16x9 adapter to use ALL the resolution for widescreen purposes? No one seems to have mentioned that Panasonic's 16x9 adapter also works with the XL1s, and by default I would assume with the XL2, right? Using the full 4x3 680,000 pixel resolution with said adapter should "theoretically" produce the optimum results. Or am I missing something?

Anyways, we still have to wait for hands-on reviews to make a final judgement, but I'm as baffled as anyone else here about the whole "true" 16x9 methods of the XL2.

[MovieSwede]

Well i dont know if the adapter will fit or not, but the XL2 uses only a small part of its CCD to give 4:3 picture. You will not gain any advantage of using an adepter on the XL2. Unless you shoot in 16:9 and wants 2,35:1.

[Barry_Green]

The horse is definitely dead, but the benefit to beating a dead horse is that it can't fight back. Try that with a live horse and you could get into trouble.

Anyway, the question about why Canon chose to saddle the XL2 with an effectively 1/4" CCD in 4:3 mode is likely to always remain a mystery. I'm afraid it may be due to the marketing issues, as demonstrated in the simplydv review, where they show a shot in 4:3 and in 16:9, where 16:9 adds width to the sides... it makes people think "wow, the picture's wider", when in fact, no, it's just that the 4:3 is crippled.

Megapixels hurt image quality, that's all there is to that. So using an anamorphic adapter on a megapixel CCD would return worse results than using the adapter on a non-megapixel CCD. Canon may have made a decent choice here by offering full-res in 16:9, but it's the same technique used on the PDX10 and other cameras. There's nothing special about the implementation, and there's certainly nothing new about it. The freedom afforded by not having to use an additional adapter may make up for the limitations imposed by using a megapixel CCD... only time will tell.

But if you want the best 4:3 24P camera, I doubt the XL2 is going to offer any image quality improvement over the DVX, as the XL2 is a 1/4" CCD (in 4:3 mode) vs. the 1/3 of the DVX.

[Kidster]

Barry,
If megapixel CCD's hurt image quality, then why do manufacturer's use them?

[Phil]

Consumers are caught in the "the higher number megapixel the better" web...

[MovieSwede]

Well its nothing wrong with using megapixels CCD IF you have the right lens for it.

When George Lucas did AOTC. He used a 2 megapixelcamcorder (2/3) with lenses for over 10 000$

Guess what happens when a consumer uses a 5 megapixel camera with a 100$ lens ;)

As TheAmPro said, consumer wants as much pixels in so small camera as possible.

[Barry_Green]

Um, but that analogy isn't quite accurate because he was recording a 2-megapixel frame. *And the chips were much bigger, so the relative size of each pixel was still plenty big. For video you want the biggest pixels you can get, as long as you have enough to cover the full frame. For a standard-def camera that means about 350,000 pixels. For a 16:9 camera, that means about 500,000 pixels. And for a high-def camera, you want about 2m pixels.

Cramming megapixels onto a CCD makes each pixel tinier, which reduces its ability to gather light and also reduces its potential latitude. *If you want bad low-light performance and no latitude, get a megapixel video camera. *Look at the JVC HD1: terrible latitude, lousy low-light, and it's got a megapixel 1/3" CCD. *Look at the TRV950: weak latitude, and a full stop worse low-light performance than the model it replaced, the TRV900. *The difference? *TRV950 adds a megapixel CCD.

Megapixel CCD's are good for digital still cameras, where you need the additional resolution for the still frame. *But they're lousy for video, where everything gets sampled down to a 720x480 frame anyway.

You'll never see a "megapixel" CCD in a broadcast video camera. *You'll only ever see CCD's with just enough pixels to cover the frame. *

[Phil]

When someone comes in looking for a prosumer camera, and they're looking at a consumer camera and asks "Whats the difference? They're both 3ccd"

I say:
3 million tiny dots on 3 1/6 inch ccd's means each pixel has to be very small to fit all of them onto those small chips. As a prosumer camera has less pixels, so each ones bigger Plus the chips are bigger 1/2 1/3 more than and twice the size, hence more quality.

This also effects LUX or low light level ratings...right?

[Phil]

I just saw you said almost the same thing... :D :-/

[Kidster]

OK, so I wonder why Canon uses a megapixel CCD in the XL2, if can can achieve a better quality image with a CCD that produces less but larger pixels. I just don't get why they would do that and charge the consumer a premium price. Did they have to use the megapixel technology in order to achieve their "TRUE 16:9". Or could they have used a smaller pixel count with larger size and still achieve their 16:9. I think if they would have done this, the 16:9 would look good , but the 4:3 would still suffer. Talk to me Barry.

[Mike_Donis]

Well if they used a 1/3" chip, with smaller pixels, they couldn't get full resolution for 16:9...there's going to be a tradeoff somewhere along the way when trying to fit both aspect ratios in SD regardless, and they just chose the method they did.

They could have used 2/3" 16:9 chips, but that would have been WAYY to expensive...

[Barry_Green]

Exactly.

They used a 4:3 chip. It's not 16:9. But they really really really want to market it as 16:9. So they compromised and crammed more pixels into the chip so that they could sample a full-res 16:9-shaped patch off that 4:3 chip. It's the only way they could go and still claim to be full-res 16:9.

The compromises mean that you get a 1/4" CCD for 4:3, and that's a 1/4" MEGAPIXEL ccd... I question how well it will perform in 4:3 mode. The 4:3 mode got double-compromised, and it didn't have to (at least according to the Sony PDX10 implementation of the technique) so I think it's a marketing decision.

Still, the megapixel CCD was the only way they could make full-res 16:9 mode, because 16:9 CCD's do not exist in the small 1/3" size. So they chose the path they chose. Now we'll just have to test the camera and see how well it performs in the real world, because everything else is just theory and talk.

[gcaus]

"they crammed more pixels on the chip".

Hmmm, being in the computer industry, and having seen chips being produced as well, I'm not sure how you do this. :D

Do you think they might have more pixels on one side - where they did the cramming? ;D

-Jerry

[scharky]

"they crammed more pixels on the chip".

Hmmm, being in the computer industry, and having seen chips being produced as well, I'm not sure how you do this. * *:D

Do you think they might have more pixels on one side - where they did the cramming? *;D

-Jerry

They cram more pixels on the chip by making the pixels smaller. This, BTW, has proved to be a not so great thing in past "consumer" camcorders.

[TangMotorPictures]

I am trying to get something straight. So the more pixel you have the worst the quality? I thought the more pixel you have the more resolution. And the less pixel you have the more you can see the pixel, hence it looks not as smooth and it looks like an old "square" arcade game in the 80s.

[Kidster]

I have both the DVX and the XL2. Even though the XL2 doesn't use the entire CCD, in 4:3 I cannot tell any diffenence in quality, between the DVX 4:3. Colors tweaked in both camer's, seem the same to me in 4:3. So whatever Canon did, the did it good.

[Barry_Green]

I am trying to get something straight. So the more pixel you have the worst the quality? I thought the more pixel you have the more resolution. And the less pixel you have the more you can see the pixel, hence it looks not as smooth and it looks like an old "square" arcade game in the 80s.
You are correct. However, resolution is *not* the only thing to factor in. A properly-designed camera would have enough pixels to create the frame, and no more. That way the pixels would be as large as possible (to gather more light, better s/n ratio, etc) while being as small as necessary to deliver maximum resolution.

Look at broadcast cameras: you'll just about never find a megapixel ccd in a broadcast camera. They have just enough pixels to fill the frame, and no more. More pixels than necessary just leads to aliasing and resizing issues, which can result in *softer* pictures (i.e., less actual resolution!)