XH-A1 chromatic aberration..blue meanies? at DVinfo.net

Dennis Wood · December 8th, 2006, 10:40 AM

I had posted some footy from the Brevis35 and the A1 yesterday, and there were some nasty chromatic issues in part of the clip. I did a quick test with the camera and found that CA shooting with a wide open aperture was not good. However, it disappeared as I stopped down to f4.8.

I was using the CINE. V preset, with NR on low, and gain at -3db, and I have not tested to see if these settings play into things. It is something we'll have to be aware of when shooting with the A1.

Here's the clip.

Tony Tremble · December 8th, 2006, 10:45 AM

Doesn't look remotely like CA.

There are plenty of other causes of colour fringing caused, for example, by overdriving CCDs.

TT

Dennis Wood · December 8th, 2006, 10:51 AM

Tony, there is a blue halo to the left of the entire plant stalk at f 2.4. I may be incorrect in calling this chromatic aberration, usually it's magenta, (I'm not Adam Wilt :-) but if it's not CA, what is it? I was thinking blue fringing from HDV, but it disappears as the lens is stopped..perhaps a contrast induced artifact?

Tony Tremble · December 8th, 2006, 11:04 AM

It is probably CCD blooming. Excess charge is spilling over into adjacent pixels around an area of extreme contrast.

It shouldn't be a problem as you were seriously over exposing the scene to get the effect.

TT

Dennis Wood · December 8th, 2006, 11:25 AM

It would be very cool if that is all it is...either way, I'll be a lot more carefull with exposure shooting adapter footage. I'll have to look at what IRE levels this shows up at and adjust the zebras to match. Either that, or invest in a decent HD on-cam monitor. It definitely is worse when the plant stock is at the edge of the frame, and better when the stalk is at frame center. From that perspective, it behaves just like CA.

With respect to the adapter footage, the left side of the man's head is not at 100 IRE, but definitely at the edge of the frame. The blue halo was not there as I panned (not in the clip) but does increase as his head approached the edge of the frame. In other words..with head at center frame, no blue halo...with head at edge of frame..huge one. This is exactly what I'm seeing with the bare cam footage at f2.4.

Chris Hurd · December 8th, 2006, 11:57 AM

Not all fringing is chromatic aberration. As has been extensively discussed previously in our XL H1 forum, CA is only one of many potential causes.

Barry Green · December 8th, 2006, 01:52 PM

Quote:

Originally Posted by Dennis Wood

In other words..with head at center frame, no blue halo...with head at edge of frame..huge one. This is exactly what I'm seeing with the bare cam footage at f2.4.

If it's the exact same exposure and everything, but in the center of the lens it's clean and at the edges it's fringing, then yes that would be some optical issue such as chromatic aberration. I've also noticed a distinct red/green separation on extreme wide shots that looks almost like a prism error, but just may be an extreme case of chromatic separation. Always on the edges of the frame, never in the center.

But yes, as Chris and Tony said, there are many causes of fringing, not all are due to lens error, many times it can be simply overdriving the green CCD but the blue and red CCDs are still delivering signal, which shifts that section of the image to become purple. The way to tell if it's the lens or if it's exposure or whatnot is to see if it happens elsewhere in the frame, as you tried. If it is consistent throughout the frame it's not likely a lens issue; if it only happens on the edges then it's definitely a lens issue.

Dennis Wood · December 8th, 2006, 02:14 PM

Yes, check out this clip as I pan. It's definitely worse at the frame edge...and that's why I was thinking along the lines of a chromatic issue.

I can reproduce this exactly with the bare cam ... and it disappears completely as I close down a few stops. I should stress that I have not tested this without NR (set to LOW) or anything but -3db gain and the camera is set to CINE. V for a preset (tweaked as above).

Agreed on the fringing issue Chris. Red fringing, for example, in the world of DV is all about the codec. From what I've read, chromatic abberration is not typically blue either. To be honest, I'm not overly concerned about this. If it is something common to the cam, it's just a matter of paying attention and shooting accordingly. Ever cam seems to have its idiosyncrasies, if indeed that's what we're see here. If it canned be tuned out by turning down sharpening etc, great.

Bill Pryor · December 8th, 2006, 02:49 PM

Quote:

Originally Posted by Dennis Wood

Every cam seems to have its idiosyncrasies, if indeed that's what we're see here.

A well-said point. I have yet to see a camera that did NOT have some idiosyncrasy, including film cameras.

Steven Fokkinga · December 8th, 2006, 04:25 PM

Are you sure you don't have some setting wrong or it is an error in your specific model? Not saying it's impossible that all a1's have this but i haven't seen it in any other footage, take for example rafaels footage where he is sitting in front of a glass door, no blue halos...
Also, did it matter what opening the lens in front of the brevis was on, or was it just the a1 lens?

Dennis Wood · December 8th, 2006, 04:32 PM

If I could reproduce the adapter shot exactly, I would. All I'm testing now is the bare cam as I need to understand what conditions, f/stops, presets etc either contribute, or ameliorate the fringing. It is possible it's just this camera, hopefully not.

Because the prototype diffuser is not set up properly, I had assumed it was this that was causing the problem. I was a bit surprised to try that test today and see the fringe on the bare cam. That test is typical of what I would try testing the adapter to evaluate CA.

Noel Evans · December 8th, 2006, 06:48 PM

If you are using NR one on low you WILL see some form of ghosting. Try NR2 on LOW, its effect on Noise is much less but 0 ghosting.

I have shot a lot of footage on NR1 and NR2. NR1 probably isnt something you want to use.

Dennis Wood · December 8th, 2006, 07:05 PM

Thanks for that tip Noel. I was thinking NR would be the first thing to try, given my tweaks to the CINE. V. preset. I'll try the same test tomorrow with/without NR1 and 2. I've got issues with noise, but I'd rather see that over the blue halo :-)

Dennis Wood · February 6th, 2007, 09:49 AM

I thought it fair to the A1 to summarize a few things I've discovered since shooting this footage.

1. I don't believe NR had anything to do with the blue fringing. In fact, with a fairly static subject, NR1 or 2 even on high, would have been fine in this clip.

2. The A1, does show some CA, with wide open aperture, at full wide, in limited conditions. However, reducing exposure, zooming in, or stopping down to f2.8 when at full wide eliminates it. This was shown in my own subsequent tests with the bare cam. No show stopper there.

3. One of the test grabs from the m2/sgpro/brevis test by Phil Bloom showed what appears to be CA from the SGpro that was not there with the other adapters. However, a Canon FD 50mm f1.4 was used on the Sgpro ... the same lens I used in this A1 clip. Recognize the blue meanie? http://www.dvinfo.net/conf/attachmen...1&d=1170214446

So although the A1 does show chromatic issues under limited conditions, there is evidence to suggest the blue fringing problem was coming from the Canon 50mm f1.4 lens being used on the adapter.

December 8th, 2006, 10:40 AM	#1
Dennis Wood Major Player Join Date: Mar 2005 Location: Ontario, Canada Posts: 916	XH-A1 chromatic aberration..blue meanies? I had posted some footy from the Brevis35 and the A1 yesterday, and there were some nasty chromatic issues in part of the clip. I did a quick test with the camera and found that CA shooting with a wide open aperture was not good. However, it disappeared as I stopped down to f4.8. I was using the CINE. V preset, with NR on low, and gain at -3db, and I have not tested to see if these settings play into things. It is something we'll have to be aware of when shooting with the A1. Here's the clip.

December 8th, 2006, 11:57 AM	#6
Chris Hurd Obstreperous Rex Join Date: Jan 2001 Location: San Marcos, TX Posts: 27,368 Images: 513	Not all fringing is chromatic aberration. As has been extensively discussed previously in our XL H1 forum, CA is only one of many potential causes. __________________ CH Search DV Info Net \| 20 years of DVi \| ...Tuesday is Soylent Green Day!

December 8th, 2006, 10:45 AM	#2
Tony Tremble Regular Crew Join Date: Oct 2006 Location: United Kingdom Posts: 451	Doesn't look remotely like CA. There are plenty of other causes of colour fringing caused, for example, by overdriving CCDs. TT

December 8th, 2006, 10:51 AM	#3
Dennis Wood Major Player Join Date: Mar 2005 Location: Ontario, Canada Posts: 916	Tony, there is a blue halo to the left of the entire plant stalk at f 2.4. I may be incorrect in calling this chromatic aberration, usually it's magenta, (I'm not Adam Wilt :-) but if it's not CA, what is it? I was thinking blue fringing from HDV, but it disappears as the lens is stopped..perhaps a contrast induced artifact?

December 8th, 2006, 11:04 AM	#4
Tony Tremble Regular Crew Join Date: Oct 2006 Location: United Kingdom Posts: 451	It is probably CCD blooming. Excess charge is spilling over into adjacent pixels around an area of extreme contrast. It shouldn't be a problem as you were seriously over exposing the scene to get the effect. TT

December 8th, 2006, 11:25 AM	#5
Dennis Wood Major Player Join Date: Mar 2005 Location: Ontario, Canada Posts: 916	It would be very cool if that is all it is...either way, I'll be a lot more carefull with exposure shooting adapter footage. I'll have to look at what IRE levels this shows up at and adjust the zebras to match. Either that, or invest in a decent HD on-cam monitor. It definitely is worse when the plant stock is at the edge of the frame, and better when the stalk is at frame center. From that perspective, it behaves just like CA. With respect to the adapter footage, the left side of the man's head is not at 100 IRE, but definitely at the edge of the frame. The blue halo was not there as I panned (not in the clip) but does increase as his head approached the edge of the frame. In other words..with head at center frame, no blue halo...with head at edge of frame..huge one. This is exactly what I'm seeing with the bare cam footage at f2.4.

December 8th, 2006, 02:14 PM	#8
Dennis Wood Major Player Join Date: Mar 2005 Location: Ontario, Canada Posts: 916	Yes, check out this clip as I pan. It's definitely worse at the frame edge...and that's why I was thinking along the lines of a chromatic issue. I can reproduce this exactly with the bare cam ... and it disappears completely as I close down a few stops. I should stress that I have not tested this without NR (set to LOW) or anything but -3db gain and the camera is set to CINE. V for a preset (tweaked as above). Agreed on the fringing issue Chris. Red fringing, for example, in the world of DV is all about the codec. From what I've read, chromatic abberration is not typically blue either. To be honest, I'm not overly concerned about this. If it is something common to the cam, it's just a matter of paying attention and shooting accordingly. Ever cam seems to have its idiosyncrasies, if indeed that's what we're see here. If it canned be tuned out by turning down sharpening etc, great.

December 8th, 2006, 04:25 PM	#10
Steven Fokkinga Major Player Join Date: May 2004 Location: The Netherlands Posts: 204	Are you sure you don't have some setting wrong or it is an error in your specific model? Not saying it's impossible that all a1's have this but i haven't seen it in any other footage, take for example rafaels footage where he is sitting in front of a glass door, no blue halos... Also, did it matter what opening the lens in front of the brevis was on, or was it just the a1 lens?

December 8th, 2006, 04:32 PM	#11
Dennis Wood Major Player Join Date: Mar 2005 Location: Ontario, Canada Posts: 916	If I could reproduce the adapter shot exactly, I would. All I'm testing now is the bare cam as I need to understand what conditions, f/stops, presets etc either contribute, or ameliorate the fringing. It is possible it's just this camera, hopefully not. Because the prototype diffuser is not set up properly, I had assumed it was this that was causing the problem. I was a bit surprised to try that test today and see the fringe on the bare cam. That test is typical of what I would try testing the adapter to evaluate CA.

December 8th, 2006, 06:48 PM	#12
Noel Evans Regular Crew Join Date: Feb 2006 Location: Tokyo, Japan Posts: 157	If you are using NR one on low you WILL see some form of ghosting. Try NR2 on LOW, its effect on Noise is much less but 0 ghosting. I have shot a lot of footage on NR1 and NR2. NR1 probably isnt something you want to use.

December 8th, 2006, 07:05 PM	#13
Dennis Wood Major Player Join Date: Mar 2005 Location: Ontario, Canada Posts: 916	Thanks for that tip Noel. I was thinking NR would be the first thing to try, given my tweaks to the CINE. V. preset. I'll try the same test tomorrow with/without NR1 and 2. I've got issues with noise, but I'd rather see that over the blue halo :-)

February 6th, 2007, 09:49 AM	#14
Dennis Wood Major Player Join Date: Mar 2005 Location: Ontario, Canada Posts: 916	I thought it fair to the A1 to summarize a few things I've discovered since shooting this footage. 1. I don't believe NR had anything to do with the blue fringing. In fact, with a fairly static subject, NR1 or 2 even on high, would have been fine in this clip. 2. The A1, does show some CA, with wide open aperture, at full wide, in limited conditions. However, reducing exposure, zooming in, or stopping down to f2.8 when at full wide eliminates it. This was shown in my own subsequent tests with the bare cam. No show stopper there. 3. One of the test grabs from the m2/sgpro/brevis test by Phil Bloom showed what appears to be CA from the SGpro that was not there with the other adapters. However, a Canon FD 50mm f1.4 was used on the Sgpro ... the same lens I used in this A1 clip. Recognize the blue meanie? http://www.dvinfo.net/conf/attachmen...1&d=1170214446 So although the A1 does show chromatic issues under limited conditions, there is evidence to suggest the blue fringing problem was coming from the Canon 50mm f1.4 lens being used on the adapter.