Exposure bracketing, LR and Photoshop

[Antonio Correia]

At first, my temptation was to post in this thread but I was afraid it could be considered a highjack and decided to start a new one.
-
I use now Olympus/Panasonic cameras with their lenses as they are pretty good and all the equipment is not too heavy as Canon and their L lenses I used before. My Olys are two 5 Mark II and a Pana GX7 which my wife uses all the time. We change lenses occasionally. While she uses the 12-40 I mainly use the Oly 7-14, the Oly 75 and the Pana 42,5 f/1.2, both last lenses absolutely superb. For my level and for what I print it is quite sufficient, thought I became very pixel peeping with time...

I "discovered" the exposure bracketing in my Olys digging into the menus. Other bracketing are possible but let me concentrate on this one for now.
I use most of the time LR for the merge of the bracketing photos, always with good results. However, I take to special attention to the Auto results which, sometimes are a bit overcooked towards the HDR which I want to avoid.
LR produces a DNG file while CC produces a TIF one.

I have tried HDR in CC just after reading the referral thread. The result doesn't please me that much but I must consider that it was the Default parameters which can be adjusted later on.
As CC produces a TIF file with only one layer (!) I a bit disappointed as masks can't be used in the various layers (exposures). Or perhaps I am missing something and TIF with all the layers is possible...
Of course I do know it is possible but with some extra work !

The images I post are self explanatory regarding witch exposures were used and the results on both apps

Perhaps 5 is too many ? Well, it depends indeed...
I have done lots of them already and the most were pretty good. Can show later.

To your consideration, please. Thank you !

[Antonio Correia]

A very successful image with this method !

[Shadowman]

Antonio,

Does the Oly take the 5 exposures simultaneously, the only issue I could see with a capture involving subjects that might move is the merge function not being able to line up correctly? How many exposures are needed is debatable but most of the HDR tutorials I've seen suggest at least 5, I've done a few HDR's in camera which only uses two exposures and the key is being able to get details in both highlights and shadows as well as the midtones or correctly exposed capture and that's only done with a well focused shot.

[Manfred M]

Antonio - Photoshop has two different HDR algorithms. I suspect the one you are using is the older one (Merge to HDR Pro), whereas the one in Lightroom, which can also be accessed in Photoshop through Adobe Camera Raw via Bridge is the newer one

My understanding that most people feel that the newer algorithms give better results than the one in Photoshop.

[DanK]

I tried Photoshop's merge to HDR pro only once. The color distortions were substantial. (I posted this years ago here.)

In the LR Merge to HDR function, the Auto option adjusts tonality. I generally turn it off, but you can undo it after the fact if you want. It's often the case that merging exposures results in a composite that needs a lot of tonality adjustment.

My preference in general is exposure fusion, which is an automated version of manual exposure blending. It does no tone mapping, instead simply selecting areas that are properly exposed in each image. You can get a plug-in for Lightroom that does this, called Lightroom Enfuse. I was surprised to find that the "merge to HDR" function in Lightroom often produces results that are quite similar to Enfuse--that is, colors that are quite natural and similar to exposure blending. So I go back and forth; if I don't like the results from one option, I try the other. But I never go back to merge to HDR function in Photoshop.

[Antonio Correia]

Thank you John !
Oly takes 3, 5 or 7 photographs with different exposures generating the same amount of raw files.
So far, I have been merging in LR which takes care of the areas where movement was recorded, resulting in one simple image. Sometimes an error may occur...
But the tonal range is quite interesting as the one above. However, the first capture is not the best result. But it is also true that the photos have not been carefully taken care of. It was a quick one. Not the best example !
In the one below the result is excellent even if a small movement have been registered.

As we can see in this other example an error occurred because the guy was moving.

But this is a method I do like.
Oh ! The shots are all taken handhold !

[Antonio Correia]

Thank you Manfred.
I rarely - very rarely - use Bridge...
I will check and try that merge through ACR ! Good point ! Thank you !

[Antonio Correia]

Dan thank you !
I knew Timothy Armes already and could not remember Enfuse !
Thank you !
Something else to try !

[Manfred M]

Thank you Manfred.
I rarely - very rarely - use Bridge...
I will check and try that merge through ACR ! Good point ! Thank you !

It should give you exactly the same result as what you are seeing with Lightroom as with the ACR version. Both use the same underlying software engine to generate the HDR image. The version used in Photoshop itself is older technology.

[Antonio Correia]

Thank you Manfred !
I have been trying to go through ACR but I couldn't. Never mind !

In fact I have tried with Exposure and ISO bracketing in both I "asked" 3 photographs but the camera gave me 9 (nine) !
I have to go over this matter again.

[Geoff F]

I do a lot of exposure merging, either with bracketed shots or more commonly with different conversions from a single Raw file. None of the auto merge options which I have tried have given what I want for my photos.

So, working in photoshop I use layers and masks to hand make the personal mix which suits my requirements. Edit the images as required using ACR then transfer them to the photoshop work screen. Auto align the layers and add masks which I edit to give a final merged result.

After transferring my images from ACR I use the Photoshop Window menu to tile the images vertically (or horizontally) then decide which image is to be my 'starting base' and drag the thumbnail images of the other tiled images onto the base image. Then simply shut down the images leaving the stacked thumbnails on the base image. Alternatively, you can store images in a suitable location and open them using Load Scripts or copy and paste each image; but once you get the knack of simply dragging the thumbnails across it becomes much quicker.

The base image then needs to be converted into a layer. Select all layers and use the auto align option then add and edit your masks.

In reality, it is much easier to do it than to describe how it is done.

When creating multiple edited versions from a single Raw file I normally work in a similar manner but do it with Smart Objects instead.

[Manfred M]

I do a lot of exposure merging, either with bracketed shots or more commonly with different conversions from a single Raw file. None of the auto merge options which I have tried have given what I want for my photos.

So, working in photoshop I use layers and masks to hand make the personal mix which suits my requirements. Edit the images as required using ACR then transfer them to the photoshop work screen. Auto align the layers and add masks which I edit to give a final merged result.

After transferring my images from ACR I use the Photoshop Window menu to tile the images vertically (or horizontally) then decide which image is to be my 'starting base' and drag the thumbnail images of the other tiled images onto the base image. Then simply shut down the images leaving the stacked thumbnails on the base image. Alternatively, you can store images in a suitable location and open them using Load Scripts or copy and paste each image; but once you get the knack of simply dragging the thumbnails across it becomes much quicker.

The base image then needs to be converted into a layer. Select all layers and use the auto align option then add and edit your masks.

In reality, it is much easier to do it than to describe how it is done.

When creating multiple edited versions from a single Raw file I normally work in a similar manner but do it with Smart Objects instead.

It's nice to see that someone else does this too.

At "base ISO" many modern cameras have a 14-stop dynamic range, and depending on the source one is looking at the human visual system dynamic range is somewhere in the 20 to 30 stop range, depending on whether this is measured with or without letting our eyes adjust to the lighting conditions. This suggests that we should be able to pull everything we would ever want out of 2 or 3 images. Given our relative insensitivity to highlights, I generally don't worry too much about the brightest values, so long as they are not clipped. We are also less sensitive to shadow details, under normal light levels, so again, I tend to look for identifiable textures rather than a completely linear view of those areas of an image.

The other areas I consider are our output devices. A computer screen, used in a fairly dimly lit room will have a dynamic range of around 10-stops (for a 1000:1 contrast ratio, which is pretty typical). People like me that print deal with a medium that has a 7.5 - 8 -stop dynamic range (matte papers are at the lower end and glossy or lustre papers at the higher end).

Throwing all this into the mix, I find that hand-blending, like Geoff, gives me far more natural results and using the data in my image files, I can get a far better final image for either screen or print than HDRI or exposure blending software can deliver. With a bit of practice, I've found I can do this very quickly and I don't find there is a lot of time difference between tweaking a HDRI or exposure fusion (i.e. using Enfuse) outcome and one that I do from scratch by hand. There is no question at all in my mind, which one gives me the best results.

[Antonio Correia]

Excellent ! Thank you to both of you Geoff and Manfred, for the tex and important tips !
However, I struggle to understand completely Geoff's method though I understand the basic concept.
Thank you guys !

[Manfred M]

Just an example I did some time ago to explore the differences between the three methods of doing a high dynamic range image. Unfortunately, I can't remember which HDRI software that was used.

1. Five image blend using HDRI





2. Image created using Enfuse plugin with Lightroom






3. Image created through manual blending of 5 images






4. Screenshot of the layers and layer masks used to create the hand-blended image

[DanK]

Thanks. Very informative.

The first pair, perhaps because of the particular colors, doesn't make the color distortions in HDR Pro obvious. I'll post my pair below.

The second pair illustrates why I wrote that the composite often needs a lot of work on tonality. For example, you wanted the edge at the top left dark, but the algorithm blindly assumed it should pick a "better exposed" image for that part. I had always done the compositing first and the tonal adjustments second, but I think you have persuaded me to try the method that you and Geoff described.

Here's my pair. The color of the sky and of the rocks at the bottom left are accurate in the second image.

HDR Pro in PS:




LR Enfuse:

[Antonio Correia]

Thank you guys ! Good and challenging informations and methods !
I wanted to do the same but with smart objects but I have not the time now to study...
I have just searched - and found - this photograph and I worked a little on it with masks and adjustments.
At the time I was doing bracketing. Shame on me !
I hope to be able to come back here to post more work done with the suggestions above. I am not going to spend more money in software since what I have is what I am supposed to worth with and take all the profit I can get.

The original raw file

Capture of the work done in CC

[Manfred M]

You have been to Thiksey Monastery in Ladakh, India too?

[Antonio Correia]

Oh yes... gorgeous isn't it ?
Been in India 7 times ! After tomorrow, the 8.th time with wife and daughter !

[Manfred M]

Oh yes... gorgeous isn't it ?
Been in India 7 times ! After tomorrow, the 8.th time with wife and daughter !

Yes. It reminds me very much of the Potala Palace in Lhasa, Tibet. You and your family are very much ahead of me; I've been to India 3 times and my wife has been there 5 times.

[Antonio Correia]

Yes. It reminds me very much of the Potala Palace in Lhasa, Tibet. You and your family are very much ahead of me; I've been to India 3 times and my wife has been there 5 times.

Never been in Tibet