Krawetz repudiates Gradient Map test

Questions, speculations & updates on the techniques and nature of media fakery
burningame
Member
Posts: 109
Joined: Thu Dec 09, 2010 1:21 pm

Krawetz repudiates Gradient Map test

Unread post by burningame »

We know about Neal Krawetz’s Hacker Factor blog and his FotoForensics ELA site… now we can say he definitely knows about this forum.

In his latest post http://www.hackerfactor.com/blog/index. ... -Test.html he basically repudiates any fakery in the ‘fireman in the rubble’ image, and questions the legitimacy of the Gradient Map test iteslf:

Image

He says:
My site admin regularly watches the logs and lets me know when there is a sudden increase in unexpected activity. Recently there has been an increase in people searching for something called a "Gradient Map Test.

A new photo analysis algorithm? I'm intrigued!

The First Clue

Most of the requests were coming from a site called the Clues Forum. As far as I can tell, this site focuses on conspiracies. In some threads they debunk hoaxes, in others threads they promote them.
:blink: what?!

He links to my post http://www.cluesforum.info/viewtopic.ph ... 1#p2359142 where I’d listed all the various tests I’d read about. He’s saying my post attributes the algorithm to him, but I think anyone can see I’m just blundering around in the dark here, and simply added the Gradient Map test to the list. My mistake for not being clearer - sorry, Dr. Krawetz!

he gives us a serve:
As far as I can tell, people in the Clues Forum want there to be a conspiracy so badly, that they will provide a false graphical analysis. Moreover, they have attributed their mysterious algorithm to me in order to give it some kind of credibility.

Finally, I want to make this very clear: I am not interested in conspiracies around 9/11, Kurt Sonnenfeld, empty vaults, missing black boxes, or the death of his wife. I only want to point out that I can find no basis for this "Gradient Map Test" that the Clues Forum attributed to me
his conclusion on the image:

“In contrast, I do not detect splicing. I do not detect blending. I do not detect selective color corrections. I do not detect lighting abnormalities, missing shadows, or inconsistent focal points. Frankly, I'm not seeing anything abnormal or suspicious in this image.”

What do you think, Simon?
lux
Member
Posts: 1913
Joined: Sat Oct 01, 2011 10:46 pm

Re: Krawetz repudiates Gradient Map test

Unread post by lux »

Too bad Krawetz didn't read the "Why are smart people fooled" thread. :P
Mercurial
Member
Posts: 121
Joined: Fri Dec 31, 2010 12:23 am

Re: Krawetz repudiates Gradient Map test

Unread post by Mercurial »

"Want there to be a conspiracy so badly..." I think we'd all rather there was no such thing. What a knob-jockey!
simonshack
Administrator
Posts: 7339
Joined: Sun Oct 18, 2009 8:09 pm
Location: italy
Contact:

Re: Krawetz repudiates Gradient Map test

Unread post by simonshack »

burningame wrote: What do you think, Simon?
Dear Burningame,

Let me just post this preliminary reply for now - there's a lot more I'd like to add as far as photo forensics is concerned - both generally speaking and with regards to my own stance/personal views on this subject. So I'll soon follow-up this post with a more exhaustive reply - to address the more specific/far-ranging aspects of the 'forensic analysis' question.

The gradient map tool - and what it does - is described in the below screenshot of my Sony Vegas video editor :

Image
In the screenshot above (at left) for instance, we see a green section which corresponds to the darker areas of the image. To bring them out, I have simply chosen a given gradient control point. When raising the threshold of that setting, the rest of the image will gradually turn fully green - as the gradient map detects, progressively, the lighter areas. In the above example though, the (sunlit) foreground turns full green - way earlier than any other dark spot in the backdrop.

The gradient map is a tool/plug-in available in many graphic programs / photo-video editors such as Final Cut and Photoshop, etc. Now, the only reason why I put that caption (on that animated gif you posted on top of page) "Gradient Map Test" was because it was, in fact, just a test/experiment I performed which I found interesting to share with the forum readers. At no point have I stated/suggested that this "Gradient Map Test" was some known, widely-accepted (or much less, patented - or 'scientifically solid' ) forensic method.

In fact, all I did was to toggle a certain curve of gradient map control points over a time frame - and record the results in an animated GIF. The point was to highlight one (of the many) curious, recurring problem(s) with the imagery credited to Kurt Sonnenfeld. This one should be obvious even to the naked eye:

( The original "Kurt Sonnenfeld" image I used for my "Gradient Map Test")
Image
Now - in all honestly: Does it not, at very first glance, appear that these firefighters are 'cut-outs' inserted into an unrelated scenery?

Here's another image - of the very same scenery (but with different people walking about the rubble). This time, the backdrop looks even lighter - and more 'detached' from the rest of the picture. Can this be explained by different camera exposures? By 'dust/smoke hanging in the air'? Is it plausible that Kurt just stayed put in (or returned to) the very same spot to snap these two very different dust conditions? To be sure - neither of these two explanations would seem plausible/satisfactory.
Image

Yet, as crazy as it may seem, it seems like Kurt really did fall in love with that spot/view angle.... A bit like a birdwatcher, he must have stayed perched for hours in his chosen photographer's nest - to snap away his pictures as each new creature came by...

... aand as US flags wondrously sprouted out from the tall steel beams...
Image

...uncaring of how much rubble was being shuffled around him...
Image

...Kurt the birdwatcher kept patiently snapping and snapping away...
Image

...the dogged, tireless photographer stayed put in his nest...
Image

...snapping away...
Image

...and away...
Image

...snapping away at the same patch of "Ground Zero scenery" ... well into the night ! :lol: :lol: :lol:
Image


So, what are we to do with this? What would be - in all logic - the very first considerations anyone should make, as we assess this bizarre bunch of images credited to one "Kurt Sonnenfeld" - a purported FEMA-accredited photo/videographer now "mysteriously exiled" in Argentina - in fear of his life"? I'll leave you to ponder this thought - until my next post on this subject.
hollycrap
Member
Posts: 223
Joined: Sun Mar 25, 2012 1:04 am

Re: Krawetz repudiates Gradient Map test

Unread post by hollycrap »

Here's my two cents.

The picture whose authenticity Dr Krawetz doubts can be proved as being released by FEMA. at this site.

=> http://old.911digitalarchive.org/crr/im ... ations.htm

Image

Title: These FEMA rescue workers are dwarfed by the pile of rubble at the site of the World Trade Center.
Description: NY, September 21, 2001 -- These FEMA rescue workers are dwarfed by the pile of rubble at the site of the World Trade Center.Photo by Michael Rieger/ FEMA News Photo
Photographer: Augustino, Jocelyn

Only it does not contain any exif metadata whatsoever. :(

But the same image at http://irapl.altervista.org

Image

Contains the whole EXIF information

Color Space sRGB Components Configuration YCbCr Compression JPEG (old-style)
Date/Time 2002:08:21 16:12:39 Date/Time Digitized 2002:08:21 16:12:39 Date/Time Original 2002:08:21 16:12:39
Exif Image Height 1680 pixels Exif Image Width 2240 pixels Exif Version 2.10
Exposure Bias Value -0.7 EV Exposure Program Aperture priority Exposure Time 1/125 sec
F-Number F4.8 File Source Digital Still Camera (DSC) Flash Flash did not fire
FlashPix Version 1.00 Focal Length 36.0 mm ISO Speed Ratings 80
Image Description OLYMPUS DIGITAL CAMERA Make OLYMPUS OPTICAL CO.,LTD Max Aperture Value F2
Metering Mode Multi-segment Model E-10 Orientation Top, left side (Horizontal / normal)
Resolution Unit Inch Software 42-0120 Thumbnail Data [6614 bytes of thumbnail data]
Thumbnail Length 6614 bytes Thumbnail Offset 6424 bytes User Comment
X Resolution 72 dots per inch Y Resolution 72 dots per inch YCbCr Positioning Datum point

:P :lol:
simonshack
Administrator
Posts: 7339
Joined: Sun Oct 18, 2009 8:09 pm
Location: italy
Contact:

Re: Krawetz repudiates Gradient Map test

Unread post by simonshack »

Image

Holy crap! You're right... :rolleyes: :rolleyes: :rolleyes:

They can't even agree on who snapped that image. What a mess...

On this webpage, the image is credited to ... (??)
Description NY, September 21, 2001 -- These FEMA rescue workers are dwarfed by the pile of rubble at the site of the World Trade Center. Photo by Michael Rieger/ FEMA News Photo
Photographer: Augustino, Jocelyn
http://old.911digitalarchive.org/crr/im ... s/3620.htm
On this other webpage, the very same mage is clearly credited to...
"Fotografía concedida por cortesía de Kurt y Paula Sonnenfeld - copyright Kurt Sonnenfeld."
http://www.flickriver.com/photos/edicio ... 049231458/
*************************************

But - dear Hollycrap, you wrongly mention that Dr. Krawetz "doubts the authenticity" of that image. Not so: He clearly concludes that ["Frankly, I'm not seeing anything abnormal or suspicious in this image."] On the other hand, he does mention that EXIF data discrepancy - and another interesting fact concerning the Olympus E-10 camera identified in the EXIF data :
According to the EXIF meta data, which should represent the camera's data, the picture was taken on 2002-08-21 16:12:39. The problem is, that date is almost a year after the picture was supposed to have been taken. (And about 2 months after he [Kurt Sonnenfeld] was released from jail after being detained relating to his wife's murder investigation.)

The EXIF data also identifies the camera as an Olympus E-10. But this is inconsistent with the photo. Specifically, the Olympus E-10 uses big-endian byte ordering for the EXIF data, but this file uses little-endian. The E-10 specifies three quantization tables, but this file only has two tables. And there are other differences in the meta data. This file fails a camera ballistics test. We can be certain that this is not a camera-original image.
http://www.hackerfactor.com/blog/index. ... -Test.html
In spite of this, Dr. Krawitz's article ends with this sweeping, 'expert assessment' of Cluesforum...
"As far as I can tell, people in the Clues Forum want there to be a conspiracy so badly, that they will provide a false graphical analysis."

What gives, Dr. Krawetz? <_<
nonhocapito
Member
Posts: 2579
Joined: Sat Jul 10, 2010 5:38 am
Location: Italy
Contact:

Re: Krawetz repudiates Gradient Map test

Unread post by nonhocapito »

I think you are all being a little too hard on Krawetz. As I have tried to explain with my comment on his blog, his initial harsh judgment of the cluesforum derived largely from a misunderstanding:

He thought that we were saying that the "gradient map test" used for that GIF was really a forensic tool created or suggested by him, which is what made his whole post quite defensive.


In this sense, the title of this thread is not exact, because no "repudiation" has taken place on part of Krawetz, as far as I can tell. He never claimed the "Gradient map test" as his own, nor did we.

However, I think he has also been intellectually honest in wanting to check the origins and exif of those alleged FEMA pictures. And also, he later concludes in his further response comments that both Sonnenfled and Booher's pictures must have been manipulated.

Granted, this happens because he thinks these images are not "official" but of "conspiratorial" origin: when we know all imagery from 9/11 is equally manipulated regardless, and, very likely, photographers like Booher or Sonnenfeld are non-existing photographers just like all the other official or non-official producers of the fake 9/11 imagery.

But still, this is a good result: I don't know why we should ask for more or mock it, or be defensive ourselves, when it is all too clear that most of people are not enthused to get involved with what they call "conspiracy theories", since they have been told they are for "kooks".

Personally I am still favorably impressed that, even if caused by a misunderstanding, this contact between what we do here and his "professional" forensic approach has taken place. Maybe it will bring nothing, maybe it is a trap, maybe it can be useful -- time would tell. But it can only be up to us and our attitude to keep the dialog open, or close it shut simply because we expect more.
Mercurial
Member
Posts: 121
Joined: Fri Dec 31, 2010 12:23 am

Re: Krawetz repudiates Gradient Map test

Unread post by Mercurial »

Fair enough, my comment was childish. Sorry. Just frustration at always being told we're looking for things that aren't there. I agree about sustaining a dialogue with Dr Krawetz, he doesn't sound to like he'll want to though.
simonshack
Administrator
Posts: 7339
Joined: Sun Oct 18, 2009 8:09 pm
Location: italy
Contact:

Re: Krawetz repudiates Gradient Map test

Unread post by simonshack »

*
Just for the record:

Neal Krawetz (the name rang a bell - but I just couldn't place it) is the "researcher and computer security consultant" (quoting Wired.com - which has asked Krawetz to error-check their article linked below) who made headlines for exposing that the Intelcenter had manipulated images of "Al-Qaeda's N°2, Al Zawahiri" ... only to retract this accusation a while later:
Question of authenticity

Neal Krawetz did an image-compression analysis of the As-Sahab and IntelCenter logos on "a 2006 al Qaeda video of Ayman al-Zawahiri". He originally told Kim Zetter of Wired News that the logos had "the same error levels and that this indicated they were added at the same time" (Zetter's words). IntelCenter boss Venzke subsequently denied that his organization had added the As-Sahab logo. He commented: "just because the error levels are the same for two items in an image, that doesn't prove they were added at the same time, only that the compression was the same for both items when they were added" (Zetter's words). Krawetz then went back on his original statement, saying that "the error levels on the IntelCenter and As-Sahab logos are different and that the IntelCenter logo was added after the As-Sahab logo" (Zetter's words again).

IntelCenter has not revealed how it acquires these videos or explained the discrepancies of the As-Sahab and IntelCenter logos.
http://en.wikipedia.org/wiki/IntelCenter


Image
Article on Wired.com:
http://www.wired.com/threatlevel/2007/0 ... chers-ana/

FWIW...
SmokingGunII
Member
Posts: 557
Joined: Fri Oct 23, 2009 9:34 am
Contact:

Re: Krawetz repudiates Gradient Map test

Unread post by SmokingGunII »

FWIW, Simon, that Sonnenfeld comparison is priceless! I hadn't studied his work before and I don't have time to post what I have just seen, but let me draw your attention to a few pointers.

Column 4 in photo "loose rubble b" is the same as column D in angle 1 & 2. Now, take a look at the leaning column against the lattice work on photo angle 1 - just behind the guy resting his hand on his knee. This same column can be seen in the photo above angle 1 just behind the two members of Village People. The same column is seen in the photo above that. It is also featured in Loose Rubble 1 & 2 and the photos above these.

Now take a look at the 3 columns behind the hands on knee man in angle 1 photo. The top one of the three is the one that gets its own stars and stripes flag at some point!

Now you have these reference points. Please re-examine the proportions of these 4 columns between this photo, the "Village People" photo and the ones above. Pay particular attention to the size of the rescue workers and the relationship between the size of the leaning column to the set of 3 columns.

Could a zoom lens cause such distortion? ;) :P :rolleyes:
Post Reply