Enfuse parameters to prevent blowout in highlights?

[Alister Ling]

Hi, I've played with the parameters using EnfuseGUI but I am missing something.

I have details in my short exposure that I can't see in the final; they are being blown out to nearly white:

https://sites.google.com/site/alistargazing/home/images_hdr

Advice would be appreciated. I have read the enfuse manual but I don't see anything obvious. It is probably some techno-babble that I don't understand. 

Thanks,

Alister.

P.S. I'm hoping it is something like someone looking for the maximum but they don't realize that there's a parameter that finds the negative of the Lapacian that does exactly what they want. 

[cspiel]

Alister -

On Monday, November 24, 2014 4:15:33 PM UTC+1, Alister Ling wrote:

I have details in my short exposure that I can't see in the final; they are being blown out to nearly white:

Try ` --exposure-cutoff' after grabbing the version of the Enblend
user guide that matches _your_ version of Enblend.

/cls

[cspiel]

On Tuesday, November 25, 2014 8:55:47 AM UTC+1, cspiel wrote:

Try ` --exposure-cutoff' after grabbing the version of the Enblend

user guide that matches _your_ version of Enblend.

Bummer!

Please make that "Enfuse" instead of "Enblend".  Although
the advice of getting the documentation version that refers to
the applications' versions -- be it Enfuse or Enblend -- still holds.

/cls

[Alister Ling]

Have not finished playing with the parameters yet. Last night I ran out of steam and tonight I ran out of steam shoveling 30+cm (12+ inches) of snow. 

I am using 4.1.1 doc and exe (checked with --V). The doc talks about the parameters but does not actually give an example line so I'm guessing.

I've tried variations on 

set enfuse_additional_parameters= --exposure-weight=0.8 --saturation-weight=0.5 --exposure-cutoff=40%:70%

 with a wide range of percent values but cannot see any change of note. I will explore more this weekend. Time to get some sleep :(

Alister.

[cspiel]

On Friday, November 28, 2014 6:22:16 AM UTC+1, Alister Ling wrote:

set enfuse_additional_parameters= --exposure-weight=0.8 --saturation-weight=0.5 --exposure-cutoff=40%:70%

If you are having troubles with the highlights, why are you setting
the lower exposure cutoff?  That makes no sense unless noise is
also a problem.

Moreover, your cutoff values are pretty drastic.  To improve on the
brightest highlights try e.g. 99% or 95% as upper cutoff values.
You may want to set the saturation weight to zero to get a
clean response, while playing with the cutoff.  BTW, option
`--save-masks' gives you access to the blend masks and can be
used to check the effectiveness of the weights, cutoffs, etc.

A 16-bit channel width of the input images probably will help, too.

/Chris

[Alister Ling]

Hi Chris, In answer to your question: "why are you setting the lower exposure cutoff?  That makes no sense.. "  

It's because I am very much a beginner! And the docs, while they explain the syntax of the parameters, they really don't explain the effects of them, so I don't have a sense of what is a reasonable value. So when I did not see much of a change, I tried some drastic numbers.

I really appreciate your suggestions! I tried 0.0 saturation and varied the top end of the cutoff from 80 to 90,95,98,99,99.8,100

--exposure-weight=1.0 --saturation-weight=0.0 --exposure-cutoff=0%:99%

with very little difference from one output to the next: https://sites.google.com/site/alistargazing/moon-rise-sets/enfuse-paramaters-for-the-moon. I put in --save-masks but nothing happens. Maybe something in the droplet batch file...

I am familiar with imagemagick and have it installed on this windows machine, so I can play with masks a bit, as suggested in the docs, but because I am new to this whole masking thing, I don't really know what I'm looking for.

Regards from a freshly 40cm (16-inch) snowed-in Edmonton (and Buffalo, I can only barely begin to imagine what you've gone through).

Alister.

[cspiel]

Alister -

On Saturday, November 29, 2014 3:28:03 AM UTC+1, Alister Ling wrote:

...

with very little difference from one output to the next

https://sites.google.com/site/alistargazing/moon-rise-sets/enfuse-paramaters-for-the-moon.

If these are your input images, then
(a) the output is already pretty good and
(b) exposure cutoff cannot help much,
because you did not feed Enfuse an image where the moon is
tightly exposed, say e.g. Zone 4 to 6, to dig up Ansel Adams.

Enfuse never makes up pixels _outside_ of the luminance range
of all relevant input pixels.  See the fundamental weighting equation
in the Enfuse manual.  So, just feeding Enfuse "overexposed" input
pixels invariably leads to an "overexposed" output pixel.

However, you can try to tilt Enfuse's notion of optimum exposure and
make it prefer darker values instead of exactly the middle of the
normalized luminance interval by passing option `--exposure-optimum'
with an argument less than .5.  Consult the manual for the meaning
of "normalized luminance", if it sounds all greek to you.

I put in --save-masks but nothing happens. Maybe something in the droplet batch file...

 
You should not see any change in the output image, but files named
"softmask-#.tif" will be written for each input image.  These are the
_final_ masks Enfuse uses before it unleashes the multi-resoultion
blend algorithm.  With a bit of exercise they can be used to judge
the settings of all the fusion parameters.


/cls

[Alister Ling]

BINGO! You`re the best Chris!

I don't know where I'd be if it wasn't for the help I get from the internet. BINGO!  Chris said my problem was likely due to "you did not feed Enfuse an image where the moon is tightly exposed, say e.g. Zone 4 to 6, to dig up Ansel Adams."  i.e. not short enough. So I used GIMP`s color levels to fake a shorter exposure, fed it into enfuse, tweaked the values and Got it. Values are Exp=1.0; Con=0.93; Sat=0.0; Mu=0.49; Sigma=.232 Grayscale=l-star. I am betting the city could be brightened up a bit by feeding it an overexposed image (tomorrow). Here is my result: https://sites.google.com/site/alistargazing/moon-rise-sets/enfuse-paramaters-for-the-moon