Started Jan 16, 2013 | Discussions thread
Mark H
Mark H Veteran Member • Posts: 3,540
ISO etc...

Steen Bay wrote:

Mark H wrote:

marike6 wrote:

Mark H wrote:

marike6 wrote:

Luego wrote:

My concern, I have yet to see any images shot with ISO 100.

Here are some more interesting X20 images:




I'm not sure why that's a concern. Base ISO of 200 is not inherently noisier than a base ISO of 100.

No, that's quite wrong - even if it's not always readily visible in JPEGs, because of variable noise reduction levels.

The overall exposure at ISO-200 will be only 50% of the light level of ISO-100, so the photon shot noise (which is the dominant noise source at low ISO) will be 40% higher - everything else being equal.

That is simply not true.

Nonsense - it is absolutely true.

An exposure at ISO-200 provides half the light to the sensor of an exposure at ISO-100 - that is almost indisputable (only excepting inaccurate/misleading ISO values).

Since the exposure is halved, the noise (relative to signal), resulting from photon shot noise, increases by 40% (the square root of 2 = 1.4 = 140%) - (assuming same fill-factor/quantum efficiency, etc, as previously stated).

Base ISO is quite simply the "native" or "base" ISO is the sensitivity you get without amplifying the analogue signal you get from the sensor.

If a give base ISO is 200 ISO, then that is simply the sensors sensitivity that receives no amplification to the voltage from the sensor which then goes to the A/D converter.

Firstly - there really is no such thing as an absolute "native" or "base" sensor ISO.

Secondly - the analogue signal is always amplified, even at the lowest ISO (so called base) - the base ISO is just the lowest amplification (obviously).

The base ISO of the camera system as a whole is 'decided' by the camera designers' choice of the amount of highlight headroom that is allocated between the sensor's saturation point and the mean/mid exposure level.

E.g. For a given sensor - one designer might allocated 2.5 stops of highlight headroom whereas another designer might allocate 3.5 stops of highlight headroom. If in the first case the camera system was a base ISO-100, the later case would have a base ISO-200 - for exactly the same sensor. [Note: Although the later base ISO-200 camera would have greater highlight range, it would also have higher noise in lower tones, than the base ISO-100 camera].

But both sensors would have the same measured base ISO on DxO, and if using ETTR, then the results would be the same,...

That is true - but, quite evidently, camera designers/manufacturers simply don't abide by the same rules and measurements as DxO do (and for very good reason too).

...so it doesn't really tell us much what the lowest ISO on a camera is.

The "...lowest ISO on a camera..."  - is the lowest value setting you can select from that camera's ISO setting control/menu.

Of course, anyone can choose to override a camera's default metering by various means, in order to 'ETTR' etc, as long as they are also prepared to adjust the image(s) as necessary in post-processing.

For example, the lowest ISO on the X10 is ISO 100, and the measured base ISO is ISO 124, and the lowest ISO on the E-M5 is ISO 200, while the measured base ISO is ISO 107, so the E-M5 can actually 'handle' a higher exposure than the X10, despite the E-M5's higher 'camera' base ISO.

Based on those DxO values, X10 ISO-124 versus E-M5 ISO-107, that certainly appears to be so (just about).

But that just confirms and reinforces my original point - the E-M5 having a higher 'camera' base ISO of 200 enables it camera to have a greater highlight capture range than the X10, apparently by over 1EV as measured by DPReview...

...and that is actually something that I have speculated about previously (ref the end of this message ) re the reason for the X20's 'possible' base ISO-200 (?).

Maybe the X20 has two base ISOs, so to speak. ISO 200 as the 'normal' base ISO, and ISO 100 as an extended base ISO, and the two ISO will have the same measured ISO on DxO. Only the cameras metering and tone curve will be different.

The term 'extended ISO'  usually just refers to ISO changes achieved without any change in analogue amplification, but achieved by digital/numeric scaling instead - it doesn't necessarily relate to any definition of 'base ISO'.

As I said above, camera manufacturers just don't really follow DxO /ISO Standard 12232 at all.

Even DxO state, quote: "As tests show, the ISO settings reported by camera manufacturers can differ significantly from measured ISO in RAW. This difference stems from design choices, in particular the choice to keep some “headroom” to avoid saturation in the higher exposures to make it possible to recover from blown highlights."

It's certainly possible that any camera's lowest ISO could (whether 'extended' or not) have more limited/compromised highlight range (compared to higher ISO) purely because it is the nearest to the sensor saturation point 'by definition' - again, it's just a 'design choice', a compromise between highlight range versus optimum noise.

It doesn't matter if it's ISO 100 like the X10 or ISO 200 like the X-Pro1 or ISO 160 like the Panasonic GH2. That's why they are called "base".

It does matter - because, everything else being equal, the ISO value determines metering/exposure level - and exposure level affects signal to noise ratio, resulting from photon shot noise.

Everything else being equal, an ISO-200 exposure will have 40% higher noise (w.r.t signal level) than an ISO-100 exposure - so very clearly, it matters.

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