ISO speed inconsistencies: different shutter speeds vs different headroom

Started Mar 17, 2012 | Discussions thread
ForumParentFirstPreviousNextNext unread
Flat view
BJL
BJL
Veteran MemberPosts: 8,740
Like?
ISO speed inconsistencies: different shutter speeds vs different headroom
Mar 17, 2012

I think I should try again; please ignore rmy other thread!

In the debate about variations between ISO settings on different cameras, there seem to be two different ways that cameras vary, and DxO confuses the two when adjusting its graphs that compare DR and SNR at various speeds.

1. Different shutter speed at the same f-stop, ISO setting and scene lighting.

2. Same shutter speed in the above scenario, but different placement of the tonal levels, such as more or less headroom between mid-tones and maximum level.

The first is arguably cheating when a lower than normal shutter speed is used, as seems to be the case in some Canon vs Nikon comparisons. This needs adjustment in high ISO performance comparisons: I would like graphs where the scale at the bottom is actual shutter speed rather than ISO speed (these two scales should roughly agree in a situation like bright sunlight at f/16, or four stops dimmer than bright sunlight with f/4.)

The second situation is just a matter of there being legitimate room for different decisions about tonal placement and highlight headroom. It would help in image comparisons to equalize the tonal placement, but no adjustment is needed in graphs of signal to noise ratio or so-called "dynamic range at a higher than base ISO speed".

In particular, DxO measures ISO speeds on the latter basis of highlight headroom, not shutter speed differences, but then shifts its DR or SNR curves in a way that only makes sense as a correction for a shutter speed discrepancy. This incorrectly penalizies a camera that has equally good low light SNR but less risk of highlights being blown by hitting the maximum digital output level: the left shift applied to a camera with more highlight headroom artificially pushes the DR or SNR curve down.

So how do the E-M5 and G3 compare in these two different aspects of ISO speed discrepancies?
--
Smaller lenses, better in low light, more telephoto reach:
you can have any two at one time.

ForumParentFirstPreviousNextNext unread
Flat view
Post (hide subjects)Posted by
ForumParentFirstPreviousNextNext unread
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow