Chris, I notice that recently your signature changed...Was it due to this forum? ;)

Any NTSC people who want to put up a still frame with a short
DV movie (couple of megabytes / seconds)?

Hi,

I'm a new user to this forum and I just joined in after reading this thread. I thought I'd add my experiences with these black lines (I have a PAL XL1s). I learned about these black lines the hard way. Here's the story:

I shooted some footage in 4:3 frame mode to later do some 3D camera tracking. After days of work the tracking came out pretty good and it was time to add some special effects in the scene. My intent was (among other things) to add some raytraced water in the scene which reflects the original footage. After hours of work it was time to do the first test rendering. When I finally saw the outcome I was shocked. Where the he** do these stripes and other strange blemishes on the water surface come from?

Because I'm still learning the process I thought I'd made a mistake somewhere and I started to tweak the 3D scene. Again after days of work I finally realized that the stripes were coming from the footage that had the black bars (just like in Rob's example earlier in this thread). The bars got raytraced to the water. Because the cam track was already done and the 3D scene already modelled, I just can't go and crop/resize the footage (the perspective gets screwed up).

So days of work went down the toilet. Well not exactly but yet again I had to do some more work to get rid of the stripes. The only positive thing here is that there isn't any money involved in this. Imagine if this happened on a real production. Someone would be pulling hair off his head :/

The bottom line. The black bars are a real problem to me. If standards says that a frame should be of certain size, then why doesn't Canon obey this? Please Canon, Fix this and the XL1s can prosper again...

karibrown: "If standards says that a frame should be of certain size, then why doesn't Canon obey this? "

That's -precisely- the point! Arguing, as some here have, that this is not a problem because the lines fall outside of the normal NTSC and PAL viewing area, is rubbish in an age where televised viewing is only one potential destination for footage. Also, contrary to some assertions, this is not a wink-and-nod problem shared by all manufacturers.

Definitely not a "wink and not" problem. This is strictly and purely wrong. No excuses from a $4000 camera when a $900 dollar camera does not have the same problem. Not even the XL-1's main competitor - the Sony PD150 - has the problem.

I sit here and watch this "little" problem mushroom and cost me thousands of dollars in editor time, equipment time, and just plain time. We were able to fix the problem in the rough cut by blowing up the picture 102% but somehow that effect was not uniformy rendered in the final cut, so it's back to a re-do of the final cut after the sound has already been mixed. Not a small problem at all and I await some sensible solution from Canon. Hell, if this were a car there would have been a recall ages ago.

If Canon is billing the XL1 line as a inexpensive professional piece of gear, which it seems to be doing, this must be a top priority before it can be called "professional."

<<<-- Originally posted by Ken Tanaka : karibrown: "If standards says that a frame should be of certain size, then why doesn't Canon obey this? "

That's -precisely- the point! Arguing, as some here have, that this is not a problem because the lines fall outside of the normal NTSC and PAL viewing area, is rubbish in an age where televised viewing is only one potential destination for footage. Also, contrary to some assertions, this is not a wink-and-nod problem shared by all manufacturers. -->>>

Exactly Ken. I too just purchased the XL1s and it sickens me to hear that this problem exists. I expected to be able to shoot stuff for the web as well as TV, but man what to do now. It really ticks me off how some people are too concerned with defending Canon.

I wonder if a class action lawsuit would get Canon's attention?

I'd have so much more faith in the company if they step up and fix this mistake at no charge to their loyal customer base. I've been purchasing Canon products for over 15 years.

For Mr. Cisco.

<< It really ticks me off how some people are too concerned with defending Canon. >>

Please try to understand, I am most certainly NOT "concerned with defending Canon." Instead, my responsibility to you is to tell you how it is. How it really is, not how it should be, or what I think, or how I feel. This is how it *is* and whether or not I agree with it doesn't really factor into this equation. If all the sudden I became emotional and started carrying a banner for the people complaining about this, it would accomplish exactly nothing and would do a serious disservice to you as well as myself.

And here's how it is.

Canon USA doesn't view this as a problem. That is, not yet. I'm working right now on getting them to issue some kind of FAQ page, or statement, or something at least acknowledging all of the attention this is getting, but don't hold your breath. They are not known for moving very fast.

The reason why Canon USA doesn't perceive this as a problem is because the issue is at the edge of the frame, outside the TV-safe area, meaning that in very general terms, in "most" situations you'll never see it (doing a picture-in-picture is one exception I can readily think of).

With regard to web video delivery or video CD's, Canon USA doesn't perceive this as a serious problem here, either. Their point of view is, why wouldn't anyone *not* crop the image down to the TV-safe area anyway, using Media Cleaner Pro or any other toolset for this purpose. Their line of thinking is, surely an image crop down to TV-safe is already an established part of the media output process. In other words, from their point of view, cropping down to TV-safe is a given. Something you'd want to do anyway. So where's the problem.

Now. Please note that I'm reporting to you their take on it. I'm not agreeing with it. Nor am I defending it. All I'm doing is reporting it. At the same time I'm telling them, "look, this is being perceived by the customer base as a serious flaw." I'm reporting to them your take on it.

And now another thing. Unlike a lot of other message boards out there, this one is not a platform for venting anger or frustration in the way it usually comes out on other places around the net. I can understand your anger and frustration, but please, consider reacting in another way. Vote with your dollars.

If this is an issue you simply can't live with, you can certainly say so here. I was really impressed with Ozzie Alfonso's explanation of how his production suffered a serious bite as a result of this. That's something that will get their attention. The other thing that will get their attention is if a bunch of slightly used XL1's show up for sale on our Community Marketplace boards, or Ebay, or other trading posts around the web. That always sends a signal. The Sony camcorders (past the VX1000 and a few others) don't have this issue.

This whole thing is a function of the Panasonic CCD's used in the XL1 and other camcorder models by Canon as well as other manufacturers. I'm trying to build a list of those too. If you think you can organize a class-action lawsuit, go for it. I salute you. If you can secure legal representation and move forward with it, then I'll even report about it.

But please, let's try to hold back a bit on the passion and not accuse each other of being on one side of the fence or the other. All that accomplishes is, drawing lines in the sand and hurt feelings and no progress whatsoever. Instead let's talk about this from a technical standpoint, and discuss the impact on workflow and business. That's what we really need here in order to be productive. Hope this helps,

Chris has noted earlier that Canon routinely cruises through here on reconnaissance missions. If true, they've very likely already read this thread and their marketing group (the true driver for all product-driven manufacturers) -may- be considering what their response might be. This, of course, is my speculation. We have no evidence of Canon's cognizance whatsoever.

To be sure, however, this flaw was the result of conscious decisions tempered somewhere in the interests of costs and profitability of the product. Like it's peers, Canon's engineering group is far too sophisticated to have simply overlooked such a fundamental aspect of the XL1/1s' imaging system. That said, only market pressure is likely to force the issue out of the internal memo -pushing cycle and into an external response. Legal maneuvers would probably be fruitless and only enrich the lawyers since such action would basically have to challenge the product's warranty claims, a generally dicey proposition that would get bogged-down (in the U.S.) at individual state attorneys offices and, internationally, even worse. In the end, nobody's camera would get fixed in our lifetimes.

No, the more effective course would be to widely publicize this flaw, both inside and outside the trade press, and then let the competitive forces of the free market persuade Canon to earnestly address this problem both retroactively and prospectively.

But, skeptical as I am that Canon has any intentions of offering remedy, we should give them a reasonable chance to do so.

p.s. This post hit the boards at the same time as Chris' post, above. I earnestly agree that we absolutely must refrain from any internecine conflicts on this subject. Further, I agree that continued reporting of the discomforts and inconveniences that this problem has inflicted on you may build a powerful body of evidence for Canon to face.

>>No, the more effective course would be to widely publicize this flaw, both inside and outside the trade press, and then let the competitive forces of the free market persuade Canon ....<<

I think this is a very effective idea. For all the great reviews the XL1 received when it was first introduced, isn't there a mag out there with the guts to update their review with this story? Probably not, because they all take Canon advertising money.

Not only is the lawsuit idea bad for reasons stated above, but also because I don't think there is a strong case.

The bruden of proof for the plaintiff would be to show that Canon has committed fraud and that plaintiffs has been harmed.

Trouble is, I am not sure that Canon does say that the camera does support a standard of 720 x 480 pixels. I took a quick look at the Specifications section of the Xl1s Instruction Maual (p. 120) and it says only:

"Television System: EIS standard (525 lines, 60 fields) NTSC color signal.

"Video Recording System: 2 rotary heads, helical scanning system DV system (Consumer digital VCR SD system) Digital component recording."

Now unless anyone can say that the 720 x 480 pixel standard is implied by these statements (and they may, I am new to this area) these statements don't say anything about such a standard.

Canon warranty says that the camera is "warranted against defective material and workmanship" It's unclear that a design flaw would be covered under such language.

No doubt Canon has a good business practice requirement to address this issue.

I haven't checked the thread to pinpoint where the lawsuit idea originated but, like most legal issues dealing with language, it's splitting hairs. I don't know off-hand how NTSC defines the color signal but I do know there are very clear NTSC standards defining the size of the picture raster (frame size and dimentions.) I'm not of a litigious nature but it would seem to me there is at least a case of deceptive advertizing if meeting NTSC standards is implied but not really adhered to when parsing the language.

I think someone made the point earlier that the best argument is to point out the XL1's competitor, the Sony PD150, fortunately doesn't have this "feature". Neither do most other DV cameras, even cheaper ones. My contribution to the argument would be the XL-1 series cannot be called "professional" as long as they don't meet all the established "professional" standards, which in this case implies adherence to NTSC video standards. As long as that's not met, the XL-1 and 1S are nothing more than very high end and very good consumer cameras, but "buyer beware" - not for serious professional use.

Since I'm rambling on let me pose a valid analogy. I'm a private pilot. In the flying world every little piece of gear has to, for the manufacturer's protection, specify all the limitations of the gear. I own a Garmin 295 GPS. This piece of avionics costs $1444 but, even though it is perfectly suitable for instrument approaches, it is not designed as such and that limitation is clearly stated everytime you turn on the instrument. For me to use the 295 for an approach in poor visibility is not only reckless but illegal. I would have to go to the instrument certified Garmin 430, costing around $4500, in order to use the GPS for instrument approaches. Okay, it can be pointed out the problem with this analogy is that, in the case of the GPS, the difference is literally a matter of life and death. In the case of the XL-1, it's not. I've just lost several thousand dollars because of a deceptively stated claim. My company could go out of business. So where's the difference?

I emailed Canon's customer service and said that the frame was short by two pixels. Here's the reply:

Dear Peter Wiley,

Thank you for your inquiry. The XL1S has a EIS standard, 525 lines at 60
fields for the NTSC color signal. The monitor may need to be adjusted to
the standard to get a full image.

Thank you for your interest in Canon. We look forward to assisting you
in the future.

Sincerely,

Todd
Product Support Representative

Okay, I've been able to run some capture tests.

Test 1
(Analogue DPS Perception board)
- JVC GY-DV500
- XL1
- Sony DCR PC9
(I also captured the footage by swapping tapes to cameras)

Each capture exhibited the EXACT SAME black bar on the left side of the screen.

Test 2
(Captured using FCP3, Firewire)
- XL1
- Sony DSR500 (16x9)

The XL1 had the same black bar at the left of the screen.

The Sony DSR500 had black bars at the left and right edges of the screen. The were about the same total number of pixels as the XL1 black bars.

As I've stated before my VX1000 had the same "problem".

So as you can see, even the 15 grand DSR500 has these "issues". It's not uncommon.

I believe that since these cameras base functions are for NTSC (no PAL units were tested) use, then you'll NEVER see this bars (overscan). It's only the advent of online, or computer disk viewing where the standard for NTSC becomes noticeable.

Justin...

Thanx for your input, but, there's really nothing new in what you've "discovered". This "feature" is not something I'd be proud of, if I was a Canon employee. It certainly is anything but professional....and whether pro cameras have it or not is irrelevant. Maybe if this were advertised and sold as a NTSC broadcast camera only, but it is not. The web and computer displays are getting to be bigger and bigger everyday....replacing television in many cases. This "feature" is, plain and simple, unacceptable....there is no rationale to excuse it.

Justin,
Thanks so much for taking the time to investigate further and report back.

But I want to be sure I understand what you're reporting. As I read your post I gather that footage shot with the XL1, JVC GY-DV500, Sony DSR500 and Sony DCR PC9 -ALL- exhibited the black-bar(s) somewhere in their frames. Correct?

Ken:
Correct

Bill:
Like I said, I believe this is an NTSC issue, not a Canon issue. But I'm not the expert in this area. And I believe the PD150 has the same thing as well, despite what was said in this thread. I'll try and run a test when I can. But go ahead and flame away (or better yet, post an example).

The only reason I posted this was to add some "facts" to the thread. I was really only seeing anecdotal evidence (except of course your image post bill) and some assertions that $900 dollar cameras don't have this problem. I happen to have a $900 Sony so I tested it, exact same issue.

Also, there were assertions that "professional" equipment shouldn't have this issue. Hence, my check with the DSR500. Again, the same problem. But in this camera they dealt with the blanking issue differently. I believe my Sony VX1000 did the same thing (except I think it favored the left edge).

When I have more time I'll pull out some VX1000 tapes, capture them, rent a PD150, grab a friends VX2000 and post those as well. Sorry, I don't have much time.

Got to go paint my office.

I'll confirm that my own VX-1000 has this problem on both vertical sides of the image frame.

Also, WRT a cheaper camera that doesn't have this problem....my handheld for "snapshots" is a progressive scan JVC JY-VS200U that DOES NOT have this problem. Admittedly, it has other problems, but at least the image frame is clean.

Please understand that my BIG objection to the blank pixels is in regardsto the havoc it causes with modern compression algorithmns like MPEG2. A great deal of effort and time was spent to develop these algorithms, but, they can be negated by black bars in the image. Cropping is not an option because of the resolution loss, potential moire, and requirement to re-render all cropped image data.

All this is fine for a consumer level camera. NOT for a prosumer. I'm shocked to learn that the hi end cameras display this "feature".

For the record, my GL1 seems to have vertical lines on both sides of the frame, but not on the top or bottom. (Sounds identical to Bill's VX1000.)

I wonder if anyone's looked at JVC's "Streamcorder" camera yet, since it's marketing angle is laser-focused towards full-frame platforms such as the Web.

Don't know about the Streamcorder, *however* I thought you guys might be interested to know that the GL2 does *not* have this issue and indeed records a *full* 720x480 image. This is one of the first things I checked, and CUSA confirms it. CUSA also confirms there are no plans to change the XL1S. Figured y'all would be glad to know that the GL2 is a full 720x480.

Now -that's- interesting! To a degree I think it will make some people even more irate with the XL1s problem since it's apparent that Canon knows how to avoid the problem.

I noticed that there are a few (more) XL1s' offered for sale here. ;-)

Dunno if this is just coincedence but in my many month long search for a camera, I have never seen any XL1's or XL1s' for sale in our local trade and exchange paper. Yeterday I looked and found 1 XL1 and 2 XL1s' Haha. The xl1s' were being sold for more than what I can get them new unfortunately..

Does anyone know if this "black bar" will show up if the video is transfered to film or does one have to specifically tell the transfer company to crop it out?


Revolver1010.

When you deal with a film transfer house, due to the amount of money involved, you'll be discussing many things several times over. I believe that most if not all places have to crop the upper and lower portions of the image anyway. So I don't see how this is an issue for going to 35mm film. Hope this helps,

hello,
i know this might be a bit dated, but i came up with this thread for searching for something (can't even remember what exactly now). anyways looking at some things i decided to go to an xL1s. while this problem probably wouldn't effect me i do wonder about something.

chris mentioned earlier in the thread that canon's take about web/cd/multimedia delievery of video would be to "crop to tv safe areas". ok i could understand that view. but if this is the case, why offer/boast about the "edge to edge" , "fullframe" clarity on the gL2? it just seems that canon would be a bit hypocritical with that take on it.


ps. i don't know if there is another thread regarding any final word from canon, but is there? thanks.

matt

Matt,
Chris summarized Canon's position on this matter in the following thread:

http://www.dvinfo.net/conf/showthread.php?s=&threadid=5629

Basically, it seems to be a completely dead issue on the current XL1s. Since the GL2 does not seem to have this problem we can only hope that the XL1s successor will not either.

ah, thank you for the link. i think i was searching for cases or warranty and some how came up with that. it still looks like a great rig for my purposes though. thanks again :)

matt

Back from the grave, and posting on very old threads. ;)

This topic always interests me, don't know why. I guess I'm a geek.

Anyway, like I've stated earlier, I've found that your capture format might effect this issue.

I captured the first image from an old broadcast quality analog card, with its capture software. The second is a firewire in Premiere.

http://www.monsterrocket.com/mini35/test/AA1analog.BMP
http://www.monsterrocket.com/mini35/test/AA1digital.BMP

As you can see they look different.

So not be beat a dead horse, but can someone capture GL2 and an XL1 using the same capture device and playback deck?

Welcome back from the "dead", Justin and Happy New Year!

Very interesting. I'd love to help but (a) I've no analog capture device and (b) the GL2 is free of this problem. I get wall-to-wall image with mine.

The only thing that's going on here, Justin, is the analog card image is sampled (-16,-1) with respect to the digital image, so you've got more black columns on the left of the analog capture, a row of black pixels at the top, and you lose a row of useful pixels at the bottom and a 10 columns of useful pixels on the right side of the image.

Additionally, you've actually provided two slightly different frames, probably no more than 1 frame (or 0.5 frames) apart. Note the slight motion in the man's hand and the woman's left tendrils when you align the images up in an image editor and shift back and forth between the two.

I wouldn't capture with the analog card any more. Not only do you lose image integrity, you lose useful pixels.