Lin Evans: It's nice that NHK is producing a sensor to go with their new high resolution display, but it would be really great if they and the camera manufacturers would be realistic about "resolution" and stop the foolish and misleading use of equating resolution to pixel count. Doing so tends, as the discussion here shows, to confuse people who have only a "marketing" understanding of the term "resolution." Resolution is not defined by pixel count, it is only roughly correlated to pixel count and then only somewhat useful as a metric when comparing like sensors (CFA or Foveon). Increasing pixel count by 16x absolutely does not increase "resolution" by 16x; it increases resolution by 4x and then only in theory because in practice there are intervening variables which serve to reduce the actual optical resolution to below the theoretical limit. "Resolution" is measured by photographing resolution charts then counting the line pairs per millimeter visible viz a viz the printed numeric value.
Apparently we "don't" all know what it means to have a "4 MP sensor" - otherwise there would not be the discussion. It's time for the term "resolution" to be used properly. "Resolution" is not defined as the "number of pixels" produced by a sensor's output, nor by the number of photosites on the sensor. It's clearly defined by the amount of detail possible to "resolve." Read the reviews of cameras here on dPReview. Go to the Resolution tests. Why do we bother "testing" resolution at all if we simply use the number of pixels on the sensor as a metric?
Direct link |
Posted on Feb 25, 2012 at 16:38:02 UTC
It's nice that NHK is producing a sensor to go with their new high resolution display, but it would be really great if they and the camera manufacturers would be realistic about "resolution" and stop the foolish and misleading use of equating resolution to pixel count. Doing so tends, as the discussion here shows, to confuse people who have only a "marketing" understanding of the term "resolution." Resolution is not defined by pixel count, it is only roughly correlated to pixel count and then only somewhat useful as a metric when comparing like sensors (CFA or Foveon). Increasing pixel count by 16x absolutely does not increase "resolution" by 16x; it increases resolution by 4x and then only in theory because in practice there are intervening variables which serve to reduce the actual optical resolution to below the theoretical limit. "Resolution" is measured by photographing resolution charts then counting the line pairs per millimeter visible viz a viz the printed numeric value.
Direct link |
Posted on Feb 24, 2012 at 17:54:14 UTC
as 6th comment
| 4 replies
Henry M. Hertz: http://en.wikipedia.org/wiki/Image_resolution http://en.wikipedia.org/wiki/Spatial_resolution
the resolution is 4x times that of HDTV... who ever made that 16x should get some basic math education or stop lying.
Absolutely - this is very misleading. The "actual" optical resolution is less than 4x that of HDTV. 4X would represent a "perfect" linear relationship between pixel count and "optical resolution" increase which, of course, never exists in the real world because of CFA and AA filtering losses. I suppose they are simply trying to maximize the shock effect on the public - but like big screen display manufacturers, tend to push the envelope with misleading advertising.
Direct link |
Posted on Feb 24, 2012 at 16:41:26 UTC
Gene Hack: the problem is not only the steep price of the cam. If it were a superior camera to all existing DSLRs, people would at last accept it.The fundamental issue is, that they have issues with the sensor. 7F stops dynamical range is something NO PROFESSIONAL photographer, let alone ambitioned amateurs accept.Weak greens, smeared color transitions and lifeless color balance worsen this. No listening at all to the customer, and having a pitbull defending shilled user base over at the Sigma forum with passive support for this behaviour by the company kill this product.PR at worst.
Unfortunately, Gene, you are wrong on multiple levels - your response indicates you have no experience with Foveon sensors.
Direct link |
Posted on Jan 19, 2012 at 18:29:41 UTC
Lin Evans: It's nice that NHK is producing a sensor to go with their new high resolution display, but it would be really great if they and the camera manufacturers would be realistic about "resolution" and stop the foolish and misleading use of equating resolution to pixel count. Doing so tends, as the discussion here shows, to confuse people who have only a "marketing" understanding of the term "resolution." Resolution is not defined by pixel count, it is only roughly correlated to pixel count and then only somewhat useful as a metric when comparing like sensors (CFA or Foveon). Increasing pixel count by 16x absolutely does not increase "resolution" by 16x; it increases resolution by 4x and then only in theory because in practice there are intervening variables which serve to reduce the actual optical resolution to below the theoretical limit. "Resolution" is measured by photographing resolution charts then counting the line pairs per millimeter visible viz a viz the printed numeric value.
Apparently we "don't" all know what it means to have a "4 MP sensor" - otherwise there would not be the discussion. It's time for the term "resolution" to be used properly. "Resolution" is not defined as the "number of pixels" produced by a sensor's output, nor by the number of photosites on the sensor. It's clearly defined by the amount of detail possible to "resolve." Read the reviews of cameras here on dPReview. Go to the Resolution tests. Why do we bother "testing" resolution at all if we simply use the number of pixels on the sensor as a metric?
It's nice that NHK is producing a sensor to go with their new high resolution display, but it would be really great if they and the camera manufacturers would be realistic about "resolution" and stop the foolish and misleading use of equating resolution to pixel count. Doing so tends, as the discussion here shows, to confuse people who have only a "marketing" understanding of the term "resolution." Resolution is not defined by pixel count, it is only roughly correlated to pixel count and then only somewhat useful as a metric when comparing like sensors (CFA or Foveon). Increasing pixel count by 16x absolutely does not increase "resolution" by 16x; it increases resolution by 4x and then only in theory because in practice there are intervening variables which serve to reduce the actual optical resolution to below the theoretical limit. "Resolution" is measured by photographing resolution charts then counting the line pairs per millimeter visible viz a viz the printed numeric value.
Henry M. Hertz: http://en.wikipedia.org/wiki/Image_resolution
http://en.wikipedia.org/wiki/Spatial_resolution
the resolution is 4x times that of HDTV... who ever made that 16x should get some basic math education or stop lying.
Absolutely - this is very misleading. The "actual" optical resolution is less than 4x that of HDTV. 4X would represent a "perfect" linear relationship between pixel count and "optical resolution" increase which, of course, never exists in the real world because of CFA and AA filtering losses. I suppose they are simply trying to maximize the shock effect on the public - but like big screen display manufacturers, tend to push the envelope with misleading advertising.
Gene Hack: the problem is not only the steep price of the cam.
If it were a superior camera to all existing DSLRs, people would at last accept it.The fundamental issue is, that they have issues with the sensor. 7F stops dynamical range is something NO PROFESSIONAL photographer, let alone ambitioned amateurs accept.Weak greens, smeared color transitions and lifeless color balance worsen this.
No listening at all to the customer, and having a pitbull defending shilled user base over at the Sigma forum with passive support for this behaviour by the company kill this product.PR at worst.
Unfortunately, Gene, you are wrong on multiple levels - your response indicates you have no experience with Foveon sensors.