Wow... Rolling shutter

Tom Hardwick · December 20th, 2008, 08:11 AM

Quote:

Originally Posted by Perrone Ford

AND that flash is shorter than the time required to scan the sensors, then you get banding.

So we need flash photographers to go back to the old days before thyrister controlled very short duration flashes. Then we'd het the whole CMOS chip read out at the higher light intensity.

Ken Ross · December 22nd, 2008, 01:52 PM

Quote:

Originally Posted by Perrone Ford

Intensity of light and duration.

When the flash constitutes the bulk of the light hitting the sensors AND that flash is shorter than the time required to scan the sensors, then you get banding.

So are you saying the slower the shutter speed the less likely a CMOS will see this issue?

Jeff Harper · December 22nd, 2008, 03:00 PM

NewBluFX Video essentials has a flash remover...could it work to correct this?

Tom Hardwick · December 23rd, 2008, 02:13 AM

It might do Jeff but what would it do? Remove selected frames? That would leave horrible jump cuts. Darken them appropriately? Ug. The whole point of a wedding is that's it's a couple in the spotlight of all the attention, their union being witnessed by family and friends. The plethora of flashes is what makes them look special, and I like the effect.

I don't think we can vary the camera's shutter speed Ken - all I'm saying is the longer the flash duration the more likely the entire CMOS frame will be over-exposed rather than just odd parts of it. Thing is modern flash guns give shorter and shorter flash durations in an effort to save battery power and speed up re-cycling times.

tom.

Jeff Harper · December 23rd, 2008, 06:32 AM

I don't know what it does, Tom. In it's demo it appears to work perfectly. I have intended for ages to download the demo..how that I have a CMOS cam I suppose it's time. (Actually I don't know if there is a demo)

Tim Akin · December 23rd, 2008, 07:13 AM

Quote:

Originally Posted by Jeff Harper

NewBluFX Video essentials has a flash remover...could it work to correct this?

I use NewBlue and I've never had much luck getting the flash remover to work as good as the demo shows. Seems like it softens the flash some but doesn't completey remove it. I really like some of the other tools though.

Jeff Harper · December 23rd, 2008, 07:18 AM

Tim, I'm thinking if it softens it that it would be better than nothing.

Ken Ross · December 23rd, 2008, 07:44 AM

Quote:

Originally Posted by Tom Hardwick

I don't think we can vary the camera's shutter speed Ken - all I'm saying is the longer the flash duration the more likely the entire CMOS frame will be over-exposed rather than just odd parts of it. Thing is modern flash guns give shorter and shorter flash durations in an effort to save battery power and speed up re-cycling times.

tom.

Tom, I actually meant varying the Sony's shutter speed. Perhaps in areas where you know this will be a major issue, a somewhat slower shutter speed might make things a bit better? I guess it would take some experimenting.

Tim Akin · December 23rd, 2008, 10:04 AM

Jeff, I think it does help and I use it, sometimes it does remove the flash. I haven't figured out why sometimes it works and sometimes it doesn't and you really don't know until after the render.

Jeff Harper · December 26th, 2008, 04:20 AM

Thanks Tim, I might give it a try. I've not had a big issue with flashes, myself. Annoying at times, but not much more than that.

My attitude may change after seeing the effects of rolling shutter. Despite what someone in another thread says about the rolling shutter issue being rubbish, in some videos I've seen it looks pretty bad, particularly in dark environments.

Ken Ross · December 26th, 2008, 10:44 AM

Jeff, in looking at those videos from the one guy who posted on Vimeo, it seems to me the worst case was the slo mo clip. It was probably a poor choice in editing to slow that piece down precisely at the point of the manifestation of the rolling shutter.

But as I mentioned in another thread, I sat my wife down and had her watch those videos and she really didn't notice anything that would have caused her concern until I specifically pointed out the slo mo piece. So my point is that we may notice it, but I suspected most brides/customers would not. It's not particularly beautiful to see the flashes going off with a CCD sensor equipped camera either. Exposure goes to pot there too. So my thinking is that although this isn't 'rubbish', it may be less of an issue than we're making it out to be.

Jeff Harper · December 26th, 2008, 12:11 PM

That's encouraging Ken.

Robert Petersen · December 26th, 2008, 01:06 PM

There are always pros and cons. I don't like the banding during some of the flashes, but I also don't like the CA (color fringing) on the XH-A1 (which is only noticeable in some cases). I'm curious to see how the HDR-FX1000 does on CA, and also the new XH-A1s. I did some close observations on sample shots provided on a German site; I looked at a small portion of the provided full resoultion screenshots of the same scene, and found that the HDR-FX1000 vs the XH-A1s provided slightly more detail and also had a much cleaner image (especially on a red object). The XH-A1s had more video compression noise in the image. Overall the images were very close, and probably wouldn't be an issue for most people. The FX1000 image was not properly white balanced, but once corrected the colors were equal to the XH-A1s. I plan on purchasing one of these cameras in 2009.

Ken Ross · December 26th, 2008, 01:42 PM

Bob, do you have a link to that German site? That's the kind of comparison I'd love to see.

Ken Ross · December 26th, 2008, 06:10 PM

Well I think I found the site and I see what you mean about the pix from the 1000 and A1S. I do prefer the 1000 and it does look cleaner. The thing that I find confusing about the site, is that when you bring up pix from the A1 and the A1S, they look radically different.

This leads me to believe that the pix from the A1 have a profile loaded and they are not from default settings. If that's not the case, one might feel the A1S has taken a step backwards...at least from the default settings.

The other interesting thing about that site is that although they make no mention of it, one could come to the conclusion that the A1S has higher resolution numbers than the A1 based on their rez charts.

Here's the link for those that are interested, but it is in German:
www.camcorder-test.com

December 23rd, 2008, 06:32 AM	#20
Jeff Harper Inner Circle Join Date: Jun 2005 Location: Cincinnati, OH Posts: 8,425	I don't know what it does, Tom. In it's demo it appears to work perfectly. I have intended for ages to download the demo..how that I have a CMOS cam I suppose it's time. (Actually I don't know if there is a demo) Last edited by Jeff Harper; December 23rd, 2008 at 07:25 AM.

December 26th, 2008, 01:06 PM	#28
Robert Petersen Regular Crew Join Date: May 2006 Location: Girard, Ohio Posts: 103	What's important There are always pros and cons. I don't like the banding during some of the flashes, but I also don't like the CA (color fringing) on the XH-A1 (which is only noticeable in some cases). I'm curious to see how the HDR-FX1000 does on CA, and also the new XH-A1s. I did some close observations on sample shots provided on a German site; I looked at a small portion of the provided full resoultion screenshots of the same scene, and found that the HDR-FX1000 vs the XH-A1s provided slightly more detail and also had a much cleaner image (especially on a red object). The XH-A1s had more video compression noise in the image. Overall the images were very close, and probably wouldn't be an issue for most people. The FX1000 image was not properly white balanced, but once corrected the colors were equal to the XH-A1s. I plan on purchasing one of these cameras in 2009.

December 26th, 2008, 06:10 PM	#30
Ken Ross Trustee Join Date: Jan 2005 Location: New York Posts: 1,945	Well I think I found the site and I see what you mean about the pix from the 1000 and A1S. I do prefer the 1000 and it does look cleaner. The thing that I find confusing about the site, is that when you bring up pix from the A1 and the A1S, they look radically different. This leads me to believe that the pix from the A1 have a profile loaded and they are not from default settings. If that's not the case, one might feel the A1S has taken a step backwards...at least from the default settings. The other interesting thing about that site is that although they make no mention of it, one could come to the conclusion that the A1S has higher resolution numbers than the A1 based on their rez charts. Here's the link for those that are interested, but it is in German: www.camcorder-test.com Last edited by Ken Ross; December 27th, 2008 at 10:36 AM.

December 22nd, 2008, 03:00 PM	#18
Jeff Harper Inner Circle Join Date: Jun 2005 Location: Cincinnati, OH Posts: 8,425	NewBluFX Video essentials has a flash remover...could it work to correct this?

December 23rd, 2008, 02:13 AM	#19
Tom Hardwick Inner Circle Join Date: Mar 2003 Location: Billericay, England UK Posts: 4,711	It might do Jeff but what would it do? Remove selected frames? That would leave horrible jump cuts. Darken them appropriately? Ug. The whole point of a wedding is that's it's a couple in the spotlight of all the attention, their union being witnessed by family and friends. The plethora of flashes is what makes them look special, and I like the effect. I don't think we can vary the camera's shutter speed Ken - all I'm saying is the longer the flash duration the more likely the entire CMOS frame will be over-exposed rather than just odd parts of it. Thing is modern flash guns give shorter and shorter flash durations in an effort to save battery power and speed up re-cycling times. tom.

December 23rd, 2008, 07:18 AM	#22
Jeff Harper Inner Circle Join Date: Jun 2005 Location: Cincinnati, OH Posts: 8,425	Tim, I'm thinking if it softens it that it would be better than nothing.

December 23rd, 2008, 10:04 AM	#24
Tim Akin Major Player Join Date: Nov 2008 Location: Newbern, TN Posts: 414	Jeff, I think it does help and I use it, sometimes it does remove the flash. I haven't figured out why sometimes it works and sometimes it doesn't and you really don't know until after the render.

December 26th, 2008, 04:20 AM	#25
Jeff Harper Inner Circle Join Date: Jun 2005 Location: Cincinnati, OH Posts: 8,425	Thanks Tim, I might give it a try. I've not had a big issue with flashes, myself. Annoying at times, but not much more than that. My attitude may change after seeing the effects of rolling shutter. Despite what someone in another thread says about the rolling shutter issue being rubbish, in some videos I've seen it looks pretty bad, particularly in dark environments.

December 26th, 2008, 10:44 AM	#26
Ken Ross Trustee Join Date: Jan 2005 Location: New York Posts: 1,945	Jeff, in looking at those videos from the one guy who posted on Vimeo, it seems to me the worst case was the slo mo clip. It was probably a poor choice in editing to slow that piece down precisely at the point of the manifestation of the rolling shutter. But as I mentioned in another thread, I sat my wife down and had her watch those videos and she really didn't notice anything that would have caused her concern until I specifically pointed out the slo mo piece. So my point is that we may notice it, but I suspected most brides/customers would not. It's not particularly beautiful to see the flashes going off with a CCD sensor equipped camera either. Exposure goes to pot there too. So my thinking is that although this isn't 'rubbish', it may be less of an issue than we're making it out to be.

December 26th, 2008, 12:11 PM	#27
Jeff Harper Inner Circle Join Date: Jun 2005 Location: Cincinnati, OH Posts: 8,425	That's encouraging Ken.

December 26th, 2008, 01:42 PM	#29
Ken Ross Trustee Join Date: Jan 2005 Location: New York Posts: 1,945	Bob, do you have a link to that German site? That's the kind of comparison I'd love to see.