SMPTE Timecode on XL2

[Steve McDonald]

It's good to have the professional SMPTE timecode on the XL2. However, how does this mesh with DV camcorders and VCRs that use the consumer timecode? If tapes shot in the XL2 are played in these units that have only consumer timecode capability, will that leave them without any usable timecode? Is it possible that consumer timecode is also encoded on the XL2 recordings, which would eliminate this problem?

Another question is, if the XL2 has a miniplug mike input, along with the XLR inputs?

From what the press release said, the 1.6X extender
piece from the XL1, that fits under the lens, will work with the XL2. I wonder if a new 2X extender will be available, such as they had for the L-1/L-2? This would give the XL2 a nice, vignette-free, 40X lens power, instead of the paltry 32X with the 1.6X extender. It's also good news that all the lenses that fit the the XL1 will work with the XL2. Canon should be commended for allowing this carry-over compatibility. Another large manufacturer with which I'm familiar, for sure would have set it up so you'd have to buy all new alternate lenses and lens accessories.

If I sit on a sidewalk downtown with a bucket and a sign that says, "Desperately Need Donations for an
XL2", will I be able to raise five Gs?

Steve McDonald

[Rob Lohman]

Steve: ofcourse no-one has here has actually hold the camera or
shot anything with it yet so I'm trying to go and answer the first
official question (whoohoo).

From the press release it sounds like the XL2 uses plain DV. If
they want to stick to the standard they will have to include
consumer timecode. I'm definitely not a timecode wizard, but I
never understood what the difference between DV timeocde and
SMPTE timecode is (other then the means to get it in and out of
a camera).

DV timecode stores the most essential things exactly like SMPTE
timecode does. I'm not sure yet how these "userbits" are going
to be incorporated. Keep in mind that the DV format has a little
area that camera manufacturers can actually use to store extra
information like settings and possible such a thing as SMPTE
timecode with extra information if they like.

How this would work is anyone's guess at this point in time I'd say.

That the best I can do, I'm afraid. There should be more info
coming later in the day.

[Chris Hurd]

I think Adam Wilt said it best when he said that for all practical purpose, there is no difference between SMPTE timecode and DV timecode. There should be no compatibility issue whatsoever.

To quote Adam's site, "The bottom line is this: don't worry about whether or not the timecode recorded on tape is 'SMPTE' or not. What matters is whether or not you have timecode, period (and DV does have timecode)."

For more info, see his TimeCode page at http://www.adamwilt.com/DV-FAQ-editing.html#DVTimecode -- hope this helps,

[Steve McDonald]

Nothing on the XL2 press release mentions the size of its CCDs. It says that in the true 16:9 mode, the full width of the CCDs is utilized. Since the older 1/3-inch lenses are compatible, this likely means that the new CCDs are 1/3-inch, as well. If so, then the narrower segment of the new CCDs used for standard 4:3 ratio images, would be smaller than that of the CCDs in the XL1.
A smaller CCD causes there to be an effect of more magnification of the optical image that strikes it. This should cause the older, 16X lens to produce more than 16X on the XL2, when in 4:3 mode. The smaller effective area for 4:3 images in the XL2 might account for much of the increased magnification of the new 20X lens.

However, it isn't clear yet if the physical proportions of the new CCDs are 16:9 or 4:3. It could be either way, with blanked zones along the sides and the top and bottom, as needed, to accomodate the two aspect modes. It's even possible that the part of the new CCDs that provides the extra side panels for 16:9, could go beyond the 1/3-inch size, if the optical image from the lens out in that area is clear and undistorted. I'd bet that there's enough of a safe margin in the image the lens throws to allow this. If this is the case, the 4:3 sensing area wouldn't be smaller and there wouldn't be an extra magnifying effect. One problem you'd find in using a wider part of the optical image from the lens, is that vignetting would be more likely, under some circumstances.

There's also the question of low-light performance, if the 4:3 sensing area is smaller.

The Century Precision Optics 1.6X telextender lens that has 75mm mounting threads, was designed, in part, with the various Canon "L" series lenses in mind. They all have 72mm outer filter threads and with a 72mm-75mm step-up adaptor ring, the Century extender lens can be mounted.

If the XL2 had the Canon 1.6X extender under the 20X lens and the Century 1.6X extender on the outer end, the total magnification would be 51.2X. Of course, this would likely result in some vignetting in the lower end of the zoom range.

The Century 1.6X extender also works with the Panasonic DVX100. The Century lens costs U.S. $900., MSRP.

Steve McDonald

[Rob Lohman]

Steve: as you can see in my summary thread and in the
original release info it states the following:

" 51.8mm to 1036mm (in 4:3), 42.2mm to 844mm (in 16:9) "

So your assumption about magnification seems to be correct.

It is indeed unknown at this point in time what the physical size
of the CCD's. However if you read the release carefully it is
easy to see that it is a true 16:9 image. So either the chips are
of a bigger 4:3 size to allow for 16:9 (some wasted size) and
traditional 4:3 (more wasted size) or it is 16:9 (no wasted pixels)
with 4:3 using the middle portion (some wasted pixels).

The latter is the most likely since otherwise you would be paying
for pixels you will never end up using.

[Steve McDonald]

Rob, as I noted previously, the press release said that in the 16:9 mode, the full width of the CCDs is used. In thinking about that more deeply, I infer it to mean that in 4:3 mode, the full width is not used. That would indicate that the XL2 CCDs are proportioned in the 16:9 aspect ratio and their 4:3 sensing areas would be smaller than those of the XL1, if their size at their widest diagonal measurement was just 1/3-inch. But, if the new CCDs are wider than 1/3-inch diagonal, because of sections that provide the 16:9 mode side extensions, then their 4:3 sensing areas wouldn't necessarily be smaller. If this latter supposition is true, then all the old and new XL lenses must be able to throw a clear and usable image that is wider than 1/3-inch. Did I just say anything that I didn't mention in my last message?

Steve McDonald

[Rob Lohman]

Well I'm not sure <g>

I know what you are saying. Either they increased the chip's
width or they decreased the 4:3 area. Anyway what I was
trying to say is that the magnification indeed chances between
the modes. The rest is still a bit of speculating.

My "feeling" says the 4:3 area is actually smaller than it used
to be. That might explain the 680K pixels which is more than the
amount use for 16:9. Perhaps they use it still retain a high
resolution for 4:3 even if the sensing area is smaller. In this
case the pixels themselves and the size between them would
be smaller (better lowlight as well?)

[Jeff Donald]

SMPTE timecode comes in two types, VITC and LTC. LTC is recorded in a special TC track (usually an unused audio track) and is an 8 digit number specifying hours, minutes, seconds, and frames. It can be either drop frame or non-drop frame. VITC records the information in unused lines in the vertical interval portion of the video signal. Vertical interval is the portion between the end of one frame and the beginning of the next. During that period the electron gun is turned off to allow it to return to the top of the screen for the next field. DV timecode is neither but is frame accurate and serves the same purpose as LTC and VITC. DV timecode is multiplexed with the video signal.

Until more information is released I would assume that SMPTE is being written in the sub-code portion of the tape that Rob refers to and DV timecode is being provided for convenience and compatibility.

[Rob Lohman]

Jeff: DV timecode is actually also written in the sub portion better
known as VAUX in technical DV terms. It just sounds like you can
set a part of the timecode (hour?) and that's about it. Which
would save the exact same signal to tape but just allow you to
set it. We'll have to wait and see what it all really means.

The only difference I can find between SMPTE timecode and DV
Timecode is the way it is stored. Both are frame accurate and
contain hours:minutes:seconds:frame. The new cam seems to
allow you to set part of that and allow you to specify drop versus
non-drop etc.

[Nick Hiltgen]

Does anyone know if there is the equivalent of a "lens return" button on the camera?

(It's that button you press that allows you to go back to your last recorded image and the start right at the end of it, so that you don't have broken time code when you go to edit in post.)

[Rob Lohman]

I think that is called "index search"

[Zack Birlew]

I hope there isn't any problems with 24p carried over from the DVX100A, like this "strobe" effect I've been hearing about. But just to get one thing clear, the 30p mode is Frame Mode, right?

But yeah, no HD, I guess people will have to stick with JVC for awhile if they want HD. I don't think Sony's attempt at HD is going to hit one fish in the barrel. I've already seen some PD170 footage that just, to me, looked too much like extremely sharp home video. And all I bet I'll see with their HD version is just higher resolution extremely sharp home video. Shoot, I'd rather stick with my GL1! >=D

But I guess we'll have to wait for the next generation of DV cams for truly good HD. But now that I think about it, what's next for the GL line? Maybe they'll have the GL3 be Canon's HD cam? That'd be neat. I'd get both in that case, but now is not the time for rumor starting, let's get back to the XL2!

One thing I do like a lot about the XL2 though, is that you can have four channels of audio when you add the MA-300 XLR adapter to it, nice one Canon! =D

Can't wait to see what people with the P+S Technik 35mm setups can do with this camera.

[Rob Lohman]

Well, no, 30p is "not" frame mode. Since frame mode was the old
way of doing things. 30p on the XL2 is just full blown 30 frames
progressive per second (since the XL2 has true progressive scan
CCD's which the XL1 did not).

[Holly Miller]

Does this mean the XL2 will work with the SMPTE TC Slates? (Deneke)

[Rob Lohman]

I've not seen any timecode in or out ports, so this would have
to be a NO. Unless some other manufacturer makes a firewire
to timecode converter. Then again, those slates are pretty
expensive anyway.