E-M5 Firmware update... V 2.0 today..

Started Jan 29, 2014 | Discussions thread
Shop cameras & lenses ▾
texinwien Senior Member • Posts: 3,320
Re: Incorrect Explanation

Thomas Niemann wrote:

texinwien wrote:

Thomas Niemann wrote:

texinwien wrote:

duartix wrote:

texinwien wrote:

Thomas Niemann wrote:

Two new features were included in the firmware update: ISO 100 and small focus points.

ISO 100. In the ISO menu you'll notice ISO 100 is marked as LOW and ISO 200 is Recommended. This indicates that ISO 100 is actually ISO 200 over-exposed by one stop. For example, if the ISO 200 exposure was 1/100 @ f/4, and you shot it at 1/50 instead, you would be over-exposing by one stop. And 1/50 @ f/4 would be the proper exposure for ISO 100. The result would be highlight clipping and less dynamic range because ISO 200 is the optimal ISO for the sensor. As a benefit, shadows would receive more exposure and, consequently, less noise.

As I have noted elsewhere in this thread, this explanation is incorrect. It is incorrect for the E-M1 and the E-P5, and will with practical certainty also be incorrect for the E-M5.

ISO 200 on the E-M5 (and all of the other Olympus m43 cameras with this Sony sensor) is actually ISO ~100 underexposed. The new LOW settings will be like ISO 100 exposed correctly.

Are you talking about this? http://www.dxomark.com/Cameras/Olympus/OM-D-E-M1#tabs-2

That is correct. As you can see from that graph, the sensor saturation ISO is close to 100 when the camera is set to ISO 100. It's even closer than that on the E-M5.

Then you also mean that ISO400 is also ISO ~200 underexposed and that ISO800 is ISO ~1600 underexposed, and so on...

That is correct, and exactly how the cameras are calibrated to work. They 'underexpose' by one stop, by design, at every ISO (except for LOW / 100) in order to preserve highlights. It's been known and talked about in detail since shortly after the E-M5 was released.

We were talking nominal ISO here, that's why it gets confusing. But if this is your reason, then it's nothing but a JPEG trick.

Whether you want to call it a JPEG 'trick' or not is your call. I'm interested in the effects on RAW, which are, basically, none at all. No (or very minor) loss of DR for RAW shooters, so anyone saying LOW will result in reduced DR is incorrect.

Interesting articles on this topic are here and here.

Both of those articles contain a number of incorrect and misleading statements. They've been discussed in detail here, almost exactly a year ago. I originally defended them as flawed primers on their respective subjects when they were first shared on this board, but subsequently had to admit that they are indefensible and likely do more harm than good.

Nonetheless, I am talking exactly about the difference between sensor ISO and camera ISO. The fact that the E-M5 has a sensor ISO 107 at camera ISO 200 means Olympus can (and did) add a Low / ~100 Camera ISO that has a sensor ISO of 107 without really sacrificing DR.

Note that ISO 12232:2006 includes the Recommended Exposure Index that allows manufacturers to specify an ISO value that produces well-exposed images.

Indeed, I've read that article (and many others on the subject) backwards and forwards many times. It, in no way whatsoever, disproves anything I've said. As a matter of fact, it's at the foundation of what I'm saying.

I wonder why Olympus would recommend ISO 200 when, per your comment, ISO 100 would be better as an increase in exposure would yield less noise.

Well, the error-laden article you linked to talks about it a bit. A manufacturer can set camera ISO arbitrarily either with more care taken to preserve highlights, or more care taken to minimize noise. Olympus made a conscious choice to calibrate camera ISOs on the E-M5 (and its other recent m43 models) with an emphasis on saving highlights at the expense of a little added noise. The new LOW setting is calibrated to have lower noise at the expense of more clipped highlights IF you rely on the camera's meter to set your other exposure parameters (which I hardly ever do - I am almost always in M mode, and use the highlight/shadow clipping "blinkies" to nail optimal exposure).

Can you link to tests that have been done to verify that no highlight clipping takes place when switching from ISO 200 to ISO 100 and increasing exposure?

You're asking the wrong question here. You're more likely to clip highlights using ISO LOW if you rely on the camera's meter to set your other exposure parameters, BUT you will gain an approximately equal amount of DR on the shadow end (i.e. less noise). You're simply shifting the DR window to the right (toward the highlights), but since the E-M5 has an extra stop of highlight headroom, and the LOW ISO only shifts exposure one stop toward hilghlights, there's not a loss of actual DR.

As Anders pointed out elsewhere in this thread, setting the exposure manually (aperture and shutter speed), then taking the same image at ISO LOW and ISO 200 results in the exact same RAW file.

You can test all of this out yourself with a program like RawDigger, or you can do some more reading and get up to speed on what many of us have known about this particular camera since soon after its release (I purchased it the first day it was available in the country where I reside, and have been eagerly learning its ins and outs ever since).

 texinwien's gear list:texinwien's gear list
Panasonic Lumix DMC-GM5 Olympus E-M5 II Olympus 12-40mm F2.8 OnePlus One Canon EOS 300D +20 more
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