timmciglobal
08-03-2009, 02:33 AM
So I've been reading more and more people posting 50D, D300, D90, T1I posts about "bad noise" and how "high iso sucks" and I've been consistently printing stuff at ISO 1600 for a while now and it got me thinking...
Are we too used to pixel peeping to determine "noise" versus printing/resizing?
As the pixel density goes up there is a good argument for "printing" to see what you truly think because when you had dSLR's are 8 MP a 8x10 printed at 300 DPI would show 7.2 million out of 8 million pixels and you could make an argument for "seeing the noise" because you were actually printing it.
Fastforward to a 15 megapixel camera... an 8x10 is showing only showing about 45% of the pixels printed at 300 DPI, 11x14 is still short of 15 megapixels printed at 300 dpi, so your noise is being down sampled which almost always makes it appear far less servere.
Should the standard for "noise" be relative now? Should we be downsampling to 8x10 print size and then asking ourselves "how does this look" versus saying "here is a 100% crop from an image almost no one will print at pixel level ever"
To demonstrate my point:
Which appears to have less noise:
http://pictures.divergentservices.com/dcresource/16001.jpghttp://pictures.divergentservices.com/dcresource/16002.jpg
Tim
P.S. If you guessed "neither" your right because one is just the same 50D downsampled to 8x10 printing resolution and one is 100% crop at a 15.7" print resolution.
Are we too used to pixel peeping to determine "noise" versus printing/resizing?
As the pixel density goes up there is a good argument for "printing" to see what you truly think because when you had dSLR's are 8 MP a 8x10 printed at 300 DPI would show 7.2 million out of 8 million pixels and you could make an argument for "seeing the noise" because you were actually printing it.
Fastforward to a 15 megapixel camera... an 8x10 is showing only showing about 45% of the pixels printed at 300 DPI, 11x14 is still short of 15 megapixels printed at 300 dpi, so your noise is being down sampled which almost always makes it appear far less servere.
Should the standard for "noise" be relative now? Should we be downsampling to 8x10 print size and then asking ourselves "how does this look" versus saying "here is a 100% crop from an image almost no one will print at pixel level ever"
To demonstrate my point:
Which appears to have less noise:
http://pictures.divergentservices.com/dcresource/16001.jpghttp://pictures.divergentservices.com/dcresource/16002.jpg
Tim
P.S. If you guessed "neither" your right because one is just the same 50D downsampled to 8x10 printing resolution and one is 100% crop at a 15.7" print resolution.