XL2 image problem

Barry Goyette · November 16th, 2004, 01:05 PM

Max,

Have you tried any of the methods discussed in this thread and others to rectify the problem, (unfortunately this thread turned into a mudslinging affair, but I think there is at least one other that might offer some advice)

The moire artifacts you are seeing are a by-product of higher resolution and the progressive scan system employed by the xl2, and are not, in fact a defect of the camera. In truth, they are more typically caused by the output device (monitor, TV) than the camera itself, but the higher resolution of the xl2 will cause more moire with more output devices than the xl1s (simply because it is more capable of "seeing" fine patterns than the xl1s is.)

The way to make the image look like the xl1s is to, well... lower the resolution of the camera by a variety of methods including the sharpness and detail settings, adding some light diffusion (or low pass filter) over the lens, or moving/zooming to alter the frequency of the offending pattern.

Hope this helps.

Barry

Marty Hudzik · November 16th, 2004, 01:57 PM

I agree with all that Barry suggested but have to say that if the problems you are describing are indeed as severe as you say, there could be a problem with your partiular unit. Obviously there is no way to know for sure by just your description. Do you have access to a store that has an XL2 that you can compare it to so that you can assure yourself that the camera is indeed working properly?

Kenneth Irvine · November 18th, 2004, 02:09 PM

It's not a particular xl2 unit issue. Moire is definitely more noticeable in 24p. It does jump out at you, expecially in a handheld situation.

I for one am not worried about it. This thread has been a hoot to read! Good to see lot's of heated discussion. Hopefully Canon will weigh in at some point.

Ken

Marty Hudzik · November 18th, 2004, 02:39 PM

I know it is not an XL2 issue but the way it was described sounds much worse than what I have experienced with the DVX for 2 years and now the XL2.

Max Morris · November 18th, 2004, 05:21 PM

The problem is not just the moire. The camera is also having serious aliasing issues. I was shooting an event and some kids glasses were jagged. The straight lines were dancing around. I tried adjusting everything on the camera but it did not change.

I have looked at the footage I have shot on an HD monitor and I am still seeing the same moire and aliasing.

I have returned the camera and went back to my old XL1s for now until I make the switch to HD here soon.

I have worked with higher end camera's that had a higher resolution than the XL2 and did not have a problem.

I am really disapointed with the XL2.

Marty Hudzik · November 18th, 2004, 08:21 PM

I am sorry to hear that. I really feel that there must have been something wrong with that particular camera. Good Luck!

Max Morris · November 18th, 2004, 09:23 PM

Thanks.

I think I am going to switch to ocuar implants that records in real life ;)

I think I will see if they will swap the camera out and see if it is the camrea.

Kenneth Irvine · November 19th, 2004, 07:52 AM

If anyone has some space where I can post video, I can provide few xl2 (uncompressed) avi examples of this moire problem at 24 fos. It's not just a minor, "sometimes it happens" issue. It is something a dp has to be aware of and must work around.

It is also not just one camera. Many people have experienced it, as I said before, if you shoot 24 fps with an xl2, the moire jumps out at you. Anyone who says it doesn't has solved the issue and should share their solution here.

On the plus side, the camera is still incredible. Not perfect, but incredible.

Thanks, Ken

Barry Goyette · November 19th, 2004, 09:52 AM

Kenneth,

If the issue is moire and/or aliasing, we should be able to see it in a still frame. If you want to email me a few grabs, I'll be happy to post them for you.

I'm not sure I agree with you though. I've been using my xl2 for almost as long as anyone, and I'm familiar with moire in its many forms. Any issue I've seen can be tied to two things...the higher resolution of the xl2 (as compared to the xl1s) and SD monitors when dealing with progressive footage. But really I have almost never seen a problem, except when I know I should...like shooting bleachers at a baseball stadium, or a vent grating...things with definable patterns of specific frequencies.

As for the "aliasing" issue, I would go as far as saying that it has the lowest aliasing of any camera I own (3 different canons 3chips and a DVX).

And I've never made a single adjustment to affect the camera in terms of controlling moire or aliasing.

In reality, most of the issues I've seen have ended up being almost completely related to the monitor's processing of progressive footage. I shot and posted footage of san diego a few months back. When viewing the raw footage in my hotel room, I experienced exactly what everyone is commenting on. My shots of an empty Petco Stadium were a moire makers dream. But when I viewed them on my HD set at home, the problem was almost a non issue, even less on my computer (remember that this was a situation that should cause moire, and would have on any camera capable of focusing). That tells us something about the nature of moire's: always caused by at least two overlapping sets of patterns ---stadium seats>>CCD Block>>monitor array.

Yes you can argue "so what, I'm still getting moire, and I never saw it before". There are two answers to that. The first is that all of us who saw the crazy moire coming off the very expensive HD broadcast cameras at the Olympics track and field events this year are beginning to understand that moire is a fact of life with higher resolution cameras. (I doubt that the camera operaters were all begging NBC to give them back their old cameras...they'll learn to deal with the moire in exchange for the incredible detail that the new HD cameras can provide)

The second answer is this...now that you've seen it (it was always there before, you just weren't looking for it)...and you're probably very unlikely to be able to look past it now...so you can take steps to resolve it in the situations that warrant it...and because of that you are a better videographer with a better tool than you ever had before.

Anyway, please send along your frames...I want to see what you are seeing.

Barry

Kevin Gilvear · November 19th, 2004, 12:10 PM

Hi everyone, what an interesting thread. A lot sounds complicated to me as I'm still learning the ropes but informative nonetheless. I have yet to shoot on tape but I have been switching on the camera and looking through the viewfinder monitor. I have a PAL cam by the way. To me when you view in wide mode it looks a little "jaggy". I'm not sure if this is just down to the little monitor or not though. is there an option on the cam to shoot in progressive or non progressive, or does it shoot in the best method as standard?

I'd definetly like to see more pics and this has made me want to go out and just film footage of brick walls as a test.

On a final note, I just wanted to say I respect everyone's opnions here. They know far more than I and I don't think there was any need for that fella (who's name I've forgotten) to be so rude to everyone trying their best to help. It is a worrying thing, but one that I hope is easily fixed through learning about the cameras abilities.

David Lach · November 19th, 2004, 12:13 PM

There's always the obvious steps to avoid moire and similar type of distortions. No fine line patterns, of any kind, at any time, on the set.

This was actually my first request to the girl that is in charge of costumes and accessories on my next project (that and no white).

No fine stripped ties, pants, shirts, etc. No blinds (they're fine in close up but when shot wide they become a nightmare) and as a general rule of thumb, nothing with a strict and regular pattern (lines, grids, etc.).

Organic and irregular patterns are fine with video (to some extent) but repetitive and identical patterns tend to cause problems so far from what I've seen.

If you're shooting fictional stuff, this is something you should test BEFORE the shoot, since you have the ability to change it if it causes problems.

Kenneth Irvine · November 19th, 2004, 12:29 PM

Thanks for the offer Barry, but...

When I play the captured aviin Premiere Pro 1.5, the moire is there. If I still the picture from the captured avi in Premiere Pro 1.5, I can see the moire patterns. When I choose Export Frame and open the bitmap, the pattern is gone. This is very interesting. Can someone explain?!

Marty Hudzik · November 19th, 2004, 12:47 PM

<<<-- Originally posted by Kenneth Irvine : Thanks for the offer Barry, but...

When I play the captured aviin Premiere Pro 1.5, the moire is there. If I still the picture from the captured avi in Premiere Pro 1.5, I can see the moire patterns. When I choose Export Frame and open the bitmap, the pattern is gone. This is very interesting. Can someone explain?! -->>>

Are you viewing on the PC or on the NTSC monitor?

Barry Goyette · November 19th, 2004, 01:03 PM

Kenneth,

this is exactly the issue that colin first encountered when he started this thread...he was seeing it on his monitor, but when he looked at a still frame it wasn't there...at which point he produced deinterlaced frame to show what the problem looked like....as if this proved something.

De-interlacing is like halving the resolution(vertically)...essentially you are placing another lower frequency pattern over the the existing pattern...this will bring out moire's where none actually exist in the original pattern (you may like to try this with your image...I tried it with some that I did myself trying to figure out what created the effect colin was seeing.)

can someone explain?...this issue is the one I keep harping on...in clinton-ese...its the monitor, stupid...(no offense meant here)...why you are seeing it in premiere and not in the still frame is that thay are most likely being viewed at different resolutions...let me offer you an illustration

go to my public folder

http://homepage.mac.com/barrygoyette/FileSharing23.html

in it is a file called moirebit.tif

open it in photoshop or some viewer that will allow you to zoom. Start by viewing it at 100%. you'll note that it is just lines...no moire...just lines, none in the image...this is equivalent to a camera image that doesn't have any moire in it. Now zoom out step by step...and you'll notice that gradually small moires will start to appear, and the more you zoom out that the moire will gradually get more noticeable and change its form. This is a very coarse example of what is going on in your case...the moire isn't in the image...its in the monitor's interpretation of the image based on certain pattern frequencies (image and monitor).

I hope this helps.

Barry

edit: don't try this in apple's preview program, it won't let you zoom out far enough, and it has a neat little anti-aliaser that will smooth things out...the effect is much more obvious in photoshop.

Kenneth Irvine · November 19th, 2004, 02:47 PM

thanks Barry none taken :)

So...the recorded image is fine and the moire is simply a monitor interpretation? If so, then the viewfinder on the xl2 has the same problem. I just played back the tape and the patterns are visible in the viewfinder. What type of monitor would not interpret the footage as having moire?

Ken

November 18th, 2004, 05:21 PM	#125
Max Morris New Boot Join Date: Nov 2004 Location: Salt Lake City, Utah Posts: 12	The problem is not just the moire. The camera is also having serious aliasing issues. I was shooting an event and some kids glasses were jagged. The straight lines were dancing around. I tried adjusting everything on the camera but it did not change. I have looked at the footage I have shot on an HD monitor and I am still seeing the same moire and aliasing. I have returned the camera and went back to my old XL1s for now until I make the switch to HD here soon. I have worked with higher end camera's that had a higher resolution than the XL2 and did not have a problem. I am really disapointed with the XL2. __________________ me thinks me likes

November 18th, 2004, 09:23 PM	#127
Max Morris New Boot Join Date: Nov 2004 Location: Salt Lake City, Utah Posts: 12	Thanks. I think I am going to switch to ocuar implants that records in real life ;) I think I will see if they will swap the camera out and see if it is the camrea. __________________ me thinks me likes

November 16th, 2004, 01:05 PM	#121
Barry Goyette Trustee Join Date: Oct 2001 Location: San Luis Obispo CA Posts: 1,195	Max, Have you tried any of the methods discussed in this thread and others to rectify the problem, (unfortunately this thread turned into a mudslinging affair, but I think there is at least one other that might offer some advice) The moire artifacts you are seeing are a by-product of higher resolution and the progressive scan system employed by the xl2, and are not, in fact a defect of the camera. In truth, they are more typically caused by the output device (monitor, TV) than the camera itself, but the higher resolution of the xl2 will cause more moire with more output devices than the xl1s (simply because it is more capable of "seeing" fine patterns than the xl1s is.) The way to make the image look like the xl1s is to, well... lower the resolution of the camera by a variety of methods including the sharpness and detail settings, adding some light diffusion (or low pass filter) over the lens, or moving/zooming to alter the frequency of the offending pattern. Hope this helps. Barry

November 16th, 2004, 01:57 PM	#122
Marty Hudzik Trustee Join Date: Jul 2004 Location: Canton, Ohio Posts: 1,771	I agree with all that Barry suggested but have to say that if the problems you are describing are indeed as severe as you say, there could be a problem with your partiular unit. Obviously there is no way to know for sure by just your description. Do you have access to a store that has an XL2 that you can compare it to so that you can assure yourself that the camera is indeed working properly?

November 18th, 2004, 02:09 PM	#123
Kenneth Irvine New Boot Join Date: Oct 2004 Location: Ottawa, Canada Posts: 23	It's not a particular xl2 unit issue. Moire is definitely more noticeable in 24p. It does jump out at you, expecially in a handheld situation. I for one am not worried about it. This thread has been a hoot to read! Good to see lot's of heated discussion. Hopefully Canon will weigh in at some point. Ken

November 18th, 2004, 02:39 PM	#124
Marty Hudzik Trustee Join Date: Jul 2004 Location: Canton, Ohio Posts: 1,771	I know it is not an XL2 issue but the way it was described sounds much worse than what I have experienced with the DVX for 2 years and now the XL2.

November 18th, 2004, 08:21 PM	#126
Marty Hudzik Trustee Join Date: Jul 2004 Location: Canton, Ohio Posts: 1,771	I am sorry to hear that. I really feel that there must have been something wrong with that particular camera. Good Luck!

November 19th, 2004, 07:52 AM	#128
Kenneth Irvine New Boot Join Date: Oct 2004 Location: Ottawa, Canada Posts: 23	If anyone has some space where I can post video, I can provide few xl2 (uncompressed) avi examples of this moire problem at 24 fos. It's not just a minor, "sometimes it happens" issue. It is something a dp has to be aware of and must work around. It is also not just one camera. Many people have experienced it, as I said before, if you shoot 24 fps with an xl2, the moire jumps out at you. Anyone who says it doesn't has solved the issue and should share their solution here. On the plus side, the camera is still incredible. Not perfect, but incredible. Thanks, Ken

November 19th, 2004, 09:52 AM	#129
Barry Goyette Trustee Join Date: Oct 2001 Location: San Luis Obispo CA Posts: 1,195	Kenneth, If the issue is moire and/or aliasing, we should be able to see it in a still frame. If you want to email me a few grabs, I'll be happy to post them for you. I'm not sure I agree with you though. I've been using my xl2 for almost as long as anyone, and I'm familiar with moire in its many forms. Any issue I've seen can be tied to two things...the higher resolution of the xl2 (as compared to the xl1s) and SD monitors when dealing with progressive footage. But really I have almost never seen a problem, except when I know I should...like shooting bleachers at a baseball stadium, or a vent grating...things with definable patterns of specific frequencies. As for the "aliasing" issue, I would go as far as saying that it has the lowest aliasing of any camera I own (3 different canons 3chips and a DVX). And I've never made a single adjustment to affect the camera in terms of controlling moire or aliasing. In reality, most of the issues I've seen have ended up being almost completely related to the monitor's processing of progressive footage. I shot and posted footage of san diego a few months back. When viewing the raw footage in my hotel room, I experienced exactly what everyone is commenting on. My shots of an empty Petco Stadium were a moire makers dream. But when I viewed them on my HD set at home, the problem was almost a non issue, even less on my computer (remember that this was a situation that should cause moire, and would have on any camera capable of focusing). That tells us something about the nature of moire's: always caused by at least two overlapping sets of patterns ---stadium seats>>CCD Block>>monitor array. Yes you can argue "so what, I'm still getting moire, and I never saw it before". There are two answers to that. The first is that all of us who saw the crazy moire coming off the very expensive HD broadcast cameras at the Olympics track and field events this year are beginning to understand that moire is a fact of life with higher resolution cameras. (I doubt that the camera operaters were all begging NBC to give them back their old cameras...they'll learn to deal with the moire in exchange for the incredible detail that the new HD cameras can provide) The second answer is this...now that you've seen it (it was always there before, you just weren't looking for it)...and you're probably very unlikely to be able to look past it now...so you can take steps to resolve it in the situations that warrant it...and because of that you are a better videographer with a better tool than you ever had before. Anyway, please send along your frames...I want to see what you are seeing. Barry

November 19th, 2004, 12:10 PM	#130
Kevin Gilvear Regular Crew Join Date: Oct 2004 Location: Milton Keynes, UK Posts: 126	Hi everyone, what an interesting thread. A lot sounds complicated to me as I'm still learning the ropes but informative nonetheless. I have yet to shoot on tape but I have been switching on the camera and looking through the viewfinder monitor. I have a PAL cam by the way. To me when you view in wide mode it looks a little "jaggy". I'm not sure if this is just down to the little monitor or not though. is there an option on the cam to shoot in progressive or non progressive, or does it shoot in the best method as standard? I'd definetly like to see more pics and this has made me want to go out and just film footage of brick walls as a test. On a final note, I just wanted to say I respect everyone's opnions here. They know far more than I and I don't think there was any need for that fella (who's name I've forgotten) to be so rude to everyone trying their best to help. It is a worrying thing, but one that I hope is easily fixed through learning about the cameras abilities.

November 19th, 2004, 12:13 PM	#131
David Lach Major Player Join Date: Oct 2002 Location: Montreal, QC, Canada Posts: 570	There's always the obvious steps to avoid moire and similar type of distortions. No fine line patterns, of any kind, at any time, on the set. This was actually my first request to the girl that is in charge of costumes and accessories on my next project (that and no white). No fine stripped ties, pants, shirts, etc. No blinds (they're fine in close up but when shot wide they become a nightmare) and as a general rule of thumb, nothing with a strict and regular pattern (lines, grids, etc.). Organic and irregular patterns are fine with video (to some extent) but repetitive and identical patterns tend to cause problems so far from what I've seen. If you're shooting fictional stuff, this is something you should test BEFORE the shoot, since you have the ability to change it if it causes problems.

November 19th, 2004, 12:29 PM	#132
Kenneth Irvine New Boot Join Date: Oct 2004 Location: Ottawa, Canada Posts: 23	Thanks for the offer Barry, but... When I play the captured aviin Premiere Pro 1.5, the moire is there. If I still the picture from the captured avi in Premiere Pro 1.5, I can see the moire patterns. When I choose Export Frame and open the bitmap, the pattern is gone. This is very interesting. Can someone explain?!

November 19th, 2004, 12:47 PM	#133
Marty Hudzik Trustee Join Date: Jul 2004 Location: Canton, Ohio Posts: 1,771	<<<-- Originally posted by Kenneth Irvine : Thanks for the offer Barry, but... When I play the captured aviin Premiere Pro 1.5, the moire is there. If I still the picture from the captured avi in Premiere Pro 1.5, I can see the moire patterns. When I choose Export Frame and open the bitmap, the pattern is gone. This is very interesting. Can someone explain?! -->>> Are you viewing on the PC or on the NTSC monitor?

November 19th, 2004, 01:03 PM	#134
Barry Goyette Trustee Join Date: Oct 2001 Location: San Luis Obispo CA Posts: 1,195	Kenneth, this is exactly the issue that colin first encountered when he started this thread...he was seeing it on his monitor, but when he looked at a still frame it wasn't there...at which point he produced deinterlaced frame to show what the problem looked like....as if this proved something. De-interlacing is like halving the resolution(vertically)...essentially you are placing another lower frequency pattern over the the existing pattern...this will bring out moire's where none actually exist in the original pattern (you may like to try this with your image...I tried it with some that I did myself trying to figure out what created the effect colin was seeing.) can someone explain?...this issue is the one I keep harping on...in clinton-ese...its the monitor, stupid...(no offense meant here)...why you are seeing it in premiere and not in the still frame is that thay are most likely being viewed at different resolutions...let me offer you an illustration go to my public folder http://homepage.mac.com/barrygoyette/FileSharing23.html in it is a file called moirebit.tif open it in photoshop or some viewer that will allow you to zoom. Start by viewing it at 100%. you'll note that it is just lines...no moire...just lines, none in the image...this is equivalent to a camera image that doesn't have any moire in it. Now zoom out step by step...and you'll notice that gradually small moires will start to appear, and the more you zoom out that the moire will gradually get more noticeable and change its form. This is a very coarse example of what is going on in your case...the moire isn't in the image...its in the monitor's interpretation of the image based on certain pattern frequencies (image and monitor). I hope this helps. Barry edit: don't try this in apple's preview program, it won't let you zoom out far enough, and it has a neat little anti-aliaser that will smooth things out...the effect is much more obvious in photoshop.

November 19th, 2004, 02:47 PM	#135
Kenneth Irvine New Boot Join Date: Oct 2004 Location: Ottawa, Canada Posts: 23	thanks Barry none taken :) So...the recorded image is fine and the moire is simply a monitor interpretation? If so, then the viewfinder on the xl2 has the same problem. I just played back the tape and the patterns are visible in the viewfinder. What type of monitor would not interpret the footage as having moire? Ken