Pushing the Dynamic Range Envelope

I have been experimenting with a pretty cool technique of late, to extend the dynamic range of my camera. I have been dong HDR composites of Architecture with a tripod mounted camera for some time, to combat the often excessive DR needed to avoid burning out windows, and getting detail out of the shadows.

But what about hand held HDR composites? I do not usually take a tripod out with me when I go out for a walk in the mountains. I used Capture One, to make these composites. The HDR does a pretty good job of matching the 3 frame composites. You need to bee pretty careful in keeping the camera pointing in the same place. I usually use the 10 sec timer in my Z7, to automatically trigger the three frame sequence one after the other.

First shot is an architectural subject that was in deep shadow on a sunny day. I managed to get detail in the sky and did not have to do a lot off on the sunlit mountains.

The finished HDR picture

Made up of:

This was a backlit landscape shot, with very bright clouds.

Made up of

Handling scenes with a dynamic range larger than the camera can handle is fairly straight forward and the technique has been around for a long time.

In general, take shots using the maximum exposure* within blur and dof requirements without clipping important highlights for each important element in the scene and then blend them in post.

This technique can range from quick and simple to tedious depending on the number and types of shots taken to combine.

For deep shadow elements you can use ETTR to maximise the SNR and so minimise visible noise in that element before blending it into the final image.

*exposure - amount of light striking the sensor per unit area during a shutter actuation.

I agree that in camera hand held HDR with multiple exposure is effective and easy to do. Good image stabilization in the camera helps as does giving a slight delay to the initial shutter release. This depends on how steady your hands are. Many photographers tend to have a little more hand movement on shot 1 when they do the first press. I prefer to assemble the images in PP rather than have it done by the camera. I think it was Sony, way back on their original Nex APS-C mirrorless cameras who first began taking multiple exposures and then assembling them into a single image. They were looking to reduce noise and give some freedom from tripods. I thought it was odd at the time that they never really promoted these features. They used much the same in camera tech to do what I think were the first panoramas done in camera by stitching edges together.

These were in fact done in post, using the Capture One HDR widget. It makes all those bodges like ETTR redundant, as you do not compromise shadow detail.

ETTR does not compromise shadow detail. It actually enhances shadow detail as described below.

The idea behind ETTR is to maximise the quality of the raw data by maximising the SNR.

That minimises the visible noise after you readjust the shadows back to the desired lightness and blend them into the final image with the other images in the HDR sequence.

There will be less visible noise in the final image if ETTR is used for the "shadow" elements in the scene.

You don't seem to be interested in ETTR when making HDR images and that's your choice to make.

I prefer to maximise the quality of the raw data, especially in the shadows, of shots that will be used to blend into a HDR final image.

What you are describing is the traditional technique for HDR but it does not maximise the raw data quality in the shadows.

This thread is titled "Pushing the dynamic range envelope", so I see no reason why I shouldn't continue to use ETTR when making HDR images.

But isn't the only shot of an HDR range that you can use ETTR on the one for the highlights? Aren't all the others, the ones the shadow values are derived from, even further to the right by definition?

So ETTR is quite redundant in HDR.

ETTR is exposing a single frame that might be underexposed to avoid highlight clipping. A bit like when we exposed slide film, to avoid burnt out areas.

An HDR set is pretty simple, it just extends the dynamic range both ways, usually in 2 stop steps. I then just burn and dodge in post to get all the areas just right. See the example I posted for the makeup of the final HDR file

I am not a great fan of HDR but only because there's not much need for it myself these days.

Having said that, perhaps using "detail" and "hand-held" in the same sentence or paragraph is an oxymoron. By which I mean that hand-holding a camera causes loss of detail, especially when exposing for shadows .(lens aberration at low f/numbers) Also, diffraction could lower detail if shooting at a high shutter-priority speed to keep the effects of camera-shake down (fortunately, it's no big deal if clouds are a bit blurred and shadows are less diffracted due to low f/numbers).

Not to mention the need for image alignment in post so as to not get smeared detail.

If you look at my examples it is needed sometimes and hand held HDR certainly delivers exceptional results, depending on your hand holding technique, and how well IBIS works in your camera.

Modern software is capable of some mind blowing magic these days.

I am envious - after an accident causing nerve cord damage, I am unable to stand straight and my hands shake almost constantly and I shoot a camera with no IBIS and usually M42 lenses.

That it is - I rather like G'MIC these days ...

So choosing purely arbitrary objects. In practical terms no different to just setting exposures at +2 or +3 stops, given that each exposure will also be an ETTR of some other object. Maximising time wasted? 😁

Absolutely, not to say that NCV doesn't do a great job of it.

Absolutely. Not just smeared detail, there are other artefacts often more conspicuous and less easy to live with than the noise we were trying to avoid.

Real nicely kept impression of light, but the group of trees middle left has retained the background brightness that doesn't match the rest of the background. The building is nicely handled. I find an HDR type process nearly always looks slightly unrealistic and so tends to suit the slightly abstract rather than the documentary.

But hey you've done a good job of it.

The process of HDR in Capture One, is pretty quick. The images I posted took about ten minutes to process to my liking.

Ha Ha, LOL, wise words indeed. 😂

Yes, that is the way I handle high dynamic range scenes to maximise the quality of the raw data and when tine to take the individual shots is not an issue.

Bracketing shots in camera is quick and simple but more often than not does not maximise the quality of the raw data for the images that will be blended into a final image.

Yes, noise reduction apps are generally pretty good nowadays but when using ETTR in very low light scenes your noise reduction app will thank you for making its job much easier by minimising the noise it needs to remove.

ETTR is quick and simple to use when using a tripod. Set base iso, set the widest aperture that gives the dof you want and then spin the shutter speed dial until the camera's histogram begins to touch the right side. Press the shutter button. Done.

Blending HDR shots is just as quick in Photoshop and I would expect in any other reputable app.

Not necessarily.

Depending on the scene, someone might set +2 stops bracket for the shadows when they could have actually set +3 or maybe even +4 using ETTR and still not blown anything in the shadows.

You then set lightness in post back to what you want before blending that shot into the final image.

The shadows in the final image that came from the ETTR'd shot will have less visible noise than a shot taken with the +2 bracket.

Which technique someone chooses to take the shots to blend into a hdr image comes down to personal choice, styles and how important maximising the quality of the raw data is.