Canon Mark IV officially announced

[ElFenix]

Originally posted by: Deadtrees
No wonder. As seen in this thread alone, some people are willing to draw conclusions based upon a few samples pictures that doesn't even have EXIF data. Even worse, it was a pre-production unit. :disgust: It gets even worse when that person compares such images to sample images carefully selected by other brand's.

i wouldn't expect it to be up to the D3s as the sensor is about 60% the size and has a native sensitivity about a stop lower. so at each particular sensitivity rating the output is already amplified by a stop and the signal was likely weaker to begin with (though the crop of the 1D allows gapless micro lenses which are impossible on FF which needs offset for vignette control, so the 1D is capturing a larger % of the light that hits the sensor). then once it's been hardware amplified it gets software pushed another 3 stops (so shot noise gets amplified then both shot and read noise get pushed)

 

[Kaido]

Originally posted by: Eug
If I had a 1D I'd probably never use 102400.

51200 maybe, but it seems to take a big turn for the worse at 102400.

I have an iPhone for those types of shots :thumbsup: :laugh:

 

[jpeyton]

Originally posted by: Eug
If I had a 1D I'd probably never use 102400.

51200 maybe, but it seems to take a big turn for the worse at 102400.

I guess CNET Asia was correct with their samples; at ISO 102400, the 1D4 cooks the output to a point of no return. Even for a post-card sized print, it's too noisy.

 

[PurdueRy]

Originally posted by: jpeyton

Originally posted by: Eug
If I had a 1D I'd probably never use 102400.

51200 maybe, but it seems to take a big turn for the worse at 102400.

I guess CNET Asia was correct with their samples; at ISO 102400, the 1D4 cooks the output to a point of no return. Even for a post-card sized print, it's too noisy.

Lets put it into perspective:

http://i417.photobucket.com/al...pp252/VVFF/pic_003.jpg

http://i417.photobucket.com/al...252/VVFF/A00A07801.jpg

Does the Nikon look better? Yes. Is the different any more than I expected? No. The Nikon is probably a stop better or so but again we're comparing a 16MP camera to a 12MP and a FF sensor vs. a 1.3 crop so an increase in noise is to be expected.

To draw any better comparison we'll have to wait for similar test shots from both cameras that will be easier to compare than two shots taken in different settings. One by the camera maker one by a random site and entirely different lighting.

 

[angry hampster]

Originally posted by: Kaido

Originally posted by: Eug
If I had a 1D I'd probably never use 102400.

51200 maybe, but it seems to take a big turn for the worse at 102400.

I have an iPhone for those types of shots :thumbsup: :laugh:

I highly doubt that.

 

[Deadtrees]

Originally posted by: ElFenix

Originally posted by: Deadtrees
No wonder. As seen in this thread alone, some people are willing to draw conclusions based upon a few samples pictures that doesn't even have EXIF data. Even worse, it was a pre-production unit. :disgust: It gets even worse when that person compares such images to sample images carefully selected by other brand's.

i wouldn't expect it to be up to the D3s as the sensor is about 60% the size and has a native sensitivity about a stop lower. so at each particular sensitivity rating the output is already amplified by a stop and the signal was likely weaker to begin with (though the crop of the 1D allows gapless micro lenses which are impossible on FF which needs offset for vignette control, so the 1D is capturing a larger % of the light that hits the sensor). then once it's been hardware amplified it gets software pushed another 3 stops (so shot noise gets amplified then both shot and read noise get pushed)

Nowhere I implied 1d mkIV would match noise performances of D3s. I simply pointed out how direct comparions are meaningless and that's what my post was about.

 

[Kaido]

Originally posted by: angry hampster

Originally posted by: Kaido

Originally posted by: Eug
If I had a 1D I'd probably never use 102400.

51200 maybe, but it seems to take a big turn for the worse at 102400.

I have an iPhone for those types of shots :thumbsup: :laugh:

I highly doubt that.

For fuzzy, pixely shots that look like crap? ps - y so srs?