• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

Amazingly natural looking new exposure blending method

Michael Fontana

pro member
Tim was quite busy:

"Version 2.00, 18th March 2008

* LR/Enfuse can now automatically align your photos by using the 'align_image_stack'
application from the Hugin project.
* The user interface has been reorganised.
* Permission problems experienced by some users should be resolved.
* Error handling has been improved."


Well done, Tim, and a question:

Is the °merging-engine° of LRenfuse of vers. 2 the same as version 1. x?
So you can decide between image aligning or not?

Personally, as I shoot from a stable tripot, it's not to emportant for me. So I might have both versions installed?

Un grand Bravo for the permissions bugfixing, too:

Here we go
 

Tim Armes

New member
Tim was quite busy:
Well done, Tim, and a question:

Is the °merging-engine° of LRenfuse of vers. 2 the same as version 1. x?
So you can decide between image aligning or not?

Yes, it still uses enfuse to merge.

If you turn on autp-aligning then the images are first passed through the 'align_image_stack' application (from the Hugin project) and then they are passed through enfuse.

Tim
 

Michael Fontana

pro member
ok, I see it now, running the first tests.
Looks good!

The implementation - in 2.10, with the autoalign - of a scratch disk makes really sense; the alignement of the images requires quite a bit of temp item, aka disk space; with 5 images, it went to 1.2 GB, so a dedicated partition is very good!

Without autoaligning, it was kinda 50% of it.

Very good work, Tim! Thanks.

A pitty that C1 didn't allows export-plugins!
 

Michael Fontana

pro member
And here is a example of the Auto-align; the panohead had somehow moved between the bracket shots, good job:


autoalign.jpg
 
Profile question

Hi,

I just started to play with the enfuse plugin, version 2.50, under windows xp. Very impressive, both from bracketed sequences as from differently developed versions of the same picture! I've run into one problem, though: the output menu offers a choice of color profile, presumably this is used for Lightroom's export function. But whatever I chose, the resulting tif had an sRGB profile. In particular, I wanted to stay in Prophoto, but opening the tif in CS3 resulted in the dreadful flat and desaturated look familiar from a profile mismatch. And indeed, the picture's profile was only sRGB.
Maybe I'm missing something here?

Nevertheless, thanks to Timothy for a very useful tool!

-- Jürgen

P.S. Why are the posts in this forum not sorted by date?
 
It works for me. Have you turned off copying of metadata? That needs to be on.

Tim,

Thanks for the quick answer! Copying of metadata was checked in the output menu. This seems to be an enfuse problem, since the same thing happens when feeding ProPhoto-profiled tiffs directly to enfuse, actually enfuse-align. I'll try to contact Erik Krause on this matter.
Fortunately, I can later on assign the intended profile. Yesterday I had only looked at the conversion dialog and seen there that the pictures were interpreted as sRGB-profiled.

-- Jürgen
 
Update on profile problem

Erik Krause's suggestion was to add a line to the file exiftool_enfuse_args.txt in the directory Programs/Enblend-Enfuse/bin (this being on Windows XP, of course). Presently the last line should be "-P"; adding another line "-ICC_Profile" makes at least the drag-and-drop operation of enfuse_align_droplet.bat (one of the icons on the desk-top) work correctly: the ProPhoto profile is preserved. However, this still doesn't happen when using Timothy's plug-in. Maybe it does not call enfuse_align_droplet.bat when used with the auto-align option?

Erik assumes that the root cause is a bug in exiftool, which apparently doesn't correctly handle the parameter -ICC_Profile.

-- Jürgen
 
Profile problem solved

Apparently it just took a restart, either of Lightroom or of Windows, to make the patch outlined above take effect. Sorry for the false alarm.

-- Juergen
 

Cem_Usakligil

Well-known member
Apparently it just took a restart, either of Lightroom or of Windows, to make the patch outlined above take effect. Sorry for the false alarm.

-- Juergen
No need to be sorry Juergen. I am very happy you've brought this up. A week ago, I was having a very long session at my PC creating HDR pictures when I got stuck with this problem. For whatever the reason, I did not associate it with the profiles being wrong (duh). So I spent ages looking for anwers here and there and recreating the same HDR using Photomatix in order to compare results with the ones from the Enfuse/LR combo. Then I stopped trying and stayed away from HDR untill you've brought this issue to our attention and everything suddenly fell into place :). So thanks a million!

Cheers,

Cem
 

Cem_Usakligil

Well-known member
Good morning, Cem

So which one do you prefer, the enfuse or the Photomatix-version?
Although Photomatix offers better controls and more extensive fine tuning of options, I'm afraid that Enfuse is my choice for the time being. I have got two pet peeves with Photomatix:
1) Colors can become way too punchy and unnatural looking (compared to Enfuse, that is). But this can be pulled down using the settings.
2) The RAW converter sucks. I get very noisy (especially in the shadows) results. So I always have to convert the RAW files into TIF ones prior to feeding them into Photomatix. Using the Enfuse plug-in of Tim, this is a one step process.

Cheers,

Cem
 

Michael Fontana

pro member
Cem, I agree: enfuse produces more natural looking blendings.

Using Photomatix since a few years, I came to similar or identical results, but its more work, as the HDRt- and tonemapped image needs to be postworked in PS again.

And yes, I did HDRing with tiffs, too.
 
Anyone comaper Bracketeer/ENFUSE and Photoacute? Which is best?

Hi,

I use Bracketeer/ENFUSE and I was wondering if anyone had actually compared the two programs (Bracketeer and PhotoAcute) to see which one did a better job?

Thanks
 

Michael Fontana

pro member
Bennett
yep, I did, its horses for courses.

Meanwhile enfuse is plusminus merging bracket shots only, PhotoAcute has many other options, and works directly and with lens-profiles in the RAW-pipeline.
 

Asher Kelman

OPF Owner/Editor-in-Chief
Bennett
yep, I did, its horses for courses.

Meanwhile enfuse is plusminus merging bracket shots only, PhotoAcute has many other options, and works directly and with lens-profiles in the RAW-pipeline.
Michael,

Why do you say efuse is +/- "for merging bracket shots only"?

Asher
 

Michael Fontana

pro member
enfuse is done here with 16bit-tiffs, and the enfuse app can't do much more than merhing bracket shots, which it does very well, btw.

meanwhile PhotoAcute does CA-correction, lens distortion-correction, etc in the RAW-pipeline and is rather used for supersolution.
 

Nill Toulme

New member
Somewhat but not entirely off-topic... what are the relative merits of merging (with enfuse or otherwise) different exposures vs. different RAW conversions of the same frame?

Nill
~~
www.toulme.net
 

Michael Fontana

pro member
Nill
you will not gain any additional image data when using a single frame:

if there aren't any bits and bytes at the left and right end of the histogramm (clipping), every software has to invent, which isn't the same as having real image data.
 

Cem_Usakligil

Well-known member
Nill
you will not gain any additional image data when using a single frame:

if there aren't any bits and bytes at the left and right end of the histogramm (clipping), every software has to invent, which isn't the same as having real image data.

While I agree, there is still some merit to using the merging on a single frame.
RAW conversion of the same frame at different exposures delivers optimized results which when merged can act as the ideal highlight/shadow filter.
 

Asher Kelman

OPF Owner/Editor-in-Chief
enfuse is done here with 16bit-tiffs, and the enfuse app can't do much more than merhing bracket shots, which it does very well, btw.

meanwhile PhotoAcute does CA-correction, lens distortion-correction, etc in the RAW-pipeline and is rather used for supersolution.
Michael,

Thanks for the answer. So Enfuse is plus-minus (or "so-so") in dealing with bracketed shots. In panoramas, much of the defects you mention might be dealt with by first using autoPano Pro which can batch create, for example, 3 separate but exactly registered layers when images are bracketed at the time of exposure. All the images are used but then only those belonging to each layer appear for that pano at that exposure.

I'm not sure how particular lens info is put into APP. DXO has a look up table. What about APP?

Asher
 

Chris Lilley

New member
I spotted this on the wikipedia page

Use the CIECAM02 color appearance model for blending colors

This implies its bang up to date in terms of colouimetry and takes into account colour apearance modelling. Interested to try this.
 

Chris Lilley

New member
Nill
you will not gain any additional image data when using a single frame:

if there aren't any bits and bytes at the left and right end of the histogramm (clipping), every software has to invent, which isn't the same as having real image data.

Your first statement is partially correct. Your second is correct, but does not follow from the first one. A raw image contains a lot of data, in a wider dynamic range than can be expressed in an 8-bit JPEG. Clipping in the original raw channels is not the same as clipping in the white-balanced RGB channels of the developed RGB image.

So if you take a photo with a wide dynamic range, and develop three shots from it with +- 1 EV exposure bracketing, you do indeed get new information which you would not have got by raising and lowering the lightness from an developed shot. Clearly there are limits to this, and taking multiple exposures allows a wider range of tones to be captured (provided the camera is rock steady and so is the subject and the lighting).
 

Michael Fontana

pro member
I agree that making 3 conversions out of the same RAW is better than just one conversion; but since Nill asked about the difference vs bracket shots ....

Asher: I related these last questions to singleshots, not panos.

While stitching - I have both, but like PTGui better - here's my workflow:

creating 16bit-tiffs out of all the bracket shots, then stitching the frames with the identical exposure (as you want really to keep the same f-stop - don't you) creating a template with that first stitch, and applying it on the other tiffs, ending in having 3 differently exposed panos.

Merging them in enfuse - if necessairy correcting the result in PS with one of the 3 inital stitches - as a layermask in PS.

The advantage of this workflow: you alwith have control of the entire pano. This wouldn't be possible when enfusing the bracket shots first, prior to stitch.

Chris; °Use the CIECAM02 color appearance model for blending colors°

that's what's in use here; consequently, the tiff opens with the correctly tagged profile.
 

Chuck Bragg

New member
Those interested in focus stacking might think about CHDK and the Canon cameras that are compatible (Digic II and III I think). One function in CHDK is changing the focus in every shot of a continuous series. You align the photos (focus shift makes for some differences) and then focus-blend them together with Enfuse. I've seen some very nice macros where an entire insect is in perfect focus - like an electron microphoto but in color.
 
Top