Avoiding blown highlights in digital night photogprahy

[Ranchu]

I think you're quite special, Bob.

 

[Rob-F]

I have the X10, which is similar to the X20. I understand they work in the dynamic EXR mode by interleaving two images one atop the other, using half the pixels for each frame, such that one frame is optimized for shadows and the other for highlights. So it's like an HDR setup, except the two frames are recorded simultaneously instead of one after the other. This makes it practical for handheld photography but the tradeoff is a 6 mp file instead of 12.

~Joe

When I set my ISO high enough to get the 400% DR option, is that the "dynamic EXR" mode? When I do this, Aperture still shows 10.7MP, not half. So I must be missing something.

 

[thereabouts]

Personally, I see too many chasing technical perfection as indicated by histograms or other digital data and not paying enough attention to the visual impact of the actual photos themselves.

But maybe I just march to the beat of a different drum.

I agree. But I'm a bit different. I tend to chase technical imperfection.

I like to know the limits of what the technology can do and then I'm happy to misuse, or ignore it. But, knowing the limits of the technology, allows me to choose how to do so.

 

[JoeV]

When I set my ISO high enough to get the 400% DR option, is that the "dynamic EXR" mode? When I do this, Aperture still shows 10.7MP, not half. So I must be missing something.

You have to set the command dial to the "EXR" position and then select dynamic 400 in the menu.

~Joe

 

[Rob-F]

You have to set the command dial to the "EXR" position and then select dynamic 400 in the menu.

~Joe

Ah, I see. Can't wait to try that!

 

[willie_901]

This is new information for me. I didn't know about "ISO invariant." I wonder if any of my cameras have that feature? Nikon D300 and D700; Fuji x10, x20, X100; LeicaSonic D-Lux 6?

My Fuji X20 has a dynamic range selector that allows DR100%, 200%, or 400%. It allows DR400% at ISO 400 or higher; DR200% at ISO200 or higher; and DR100% at ISO 100 or higher. I wonder if that is related to ISO invariance?

Many current cameras can be operated as if they are ISO invariant. I am not familiar with the LeicaSonic D-Lux 6, but the others you list are not.

In my post,I mentioned a site that publishes DR vs ISO data from unrendered raw files for a large number of cameras. The screen capture below compares three

Nikon bodies. This data is from that link.

The D7000 (blue plot) is strictly ISO-invariant. The only way post-acquiition brightness is achieved in-camera is by digital multiplication of the raw-file values. There is no amplification of the DC voltages in between the the sensor and the analog-to-digital converter.

With the D700 (green plot) the DR range is a function of ISO. It is ISO variant.

The D800 (red plot) is essentially ISO invariant even though it uses electronic amplification. This is because its CMOS sensor unit's ISO amplification technology has extremely low read noise at all ISO settings. This means the increase in read noise with amplification is below the minimum detection threshold of the ADC. Note below the D800 sensor's native ISO DR does not increase. This is because the data at native ISO is digitally scaled so the exposure matches the light meter's estimate.

For completeness the DR in these plots is directly proportional to the signal to noise ratio. So I am using the DR vs ISO characteristics. This assumption is confirmed by published (link 1, link 2) measurements of the Fujifilm X-E1 raw SNR vs ISO. In the second the data shows the read noise is essentially ISO invariant for at least 4 stops of under exposure. Post #23 in this thread indicates this is not the same for sensor thermal noise that comes into play during very long shutter times (2 mins.)

------------------------

The Fujifilm in-Camrea JPEG DR Rendering Parameter.

Like many other brands, Fujifilm provides a means for JPEG users to automatically invoke selective pulling of highlight regions and pushing of shadow regions. The marketing name for these methods is usually trademarked and Fujifilm choose the term DR for their process.

The menu parameter name, DR, that invokes JPEG push/pull rendering is an unfortunate, misleading choice.

Fujifilm decided when DR = 200 or 400, ISO has to be at least 400 or 800 - respectively. That is Fujifilm wants minimize the chance the sensor will blow highlights by forcing sensor underexposure (when one let's the meter decide exposure). According to the data, this actually decreases the analog DR when the shutter is open!. At DR = 100 nothing happens.

So what the Fujifilm DR setting actually does is produce a JPEG that is unlikely to have blown highlights while the push/pull in-camera JPEG rendering makes the most of the data's inherent dynamic range. This is convenient for JPEG users at the cost reducing maximum possible analog dynamic range by 2 or 4 stops! Raw file users should never use DR 200 or 400.

 

[Rob-F]

Willie, I did look at the site you linked to. I thought the Df looked superior.

As to DR, now it's starting to sound like DR is just a big crock of cranberries. I'm thinking about what you said:

"So what the Fujifilm DR setting actually does is produce a JPEG that is unlikely to have blown highlights while the push/pull in-camera JPEG rendering makes the most of the data's inherent dynamic range. This is convenient for JPEG users at the cost reducing maximum possible analog dynamic range by 2 or 4 stops!"

That sounds like what DR does is to shrink the original dynamic range. Kind of like volume compression in audio. So it doesn't make the DR fit the picture, it makes the picture fit the DR? Do I have that right?

What about EXR? I tried setting the X10 to EXR. It does reduce the pixel count to just over 5MP. So I guess EXR really does increase the available dynamic range, in exchange for reducing resolution. It takes away all control, though. It wouldn't let me choose the aperture or shutter speed. The camera makes all decisions, apparently.

BTW, the X20 is marked, "SR+" in place of the X10's "EXR". It seems similar in use, but the file comes out 10.7MP rather than the 5.2MP of the X10; and the exposure came out lighter and not as good. Not sure what they have done, there. A step backwards?

I'm still learning . . .