FLIR Background

Forward-Looking Infrared, or FLIR, technology originated as a sensor system for fighter aircraft. The first sensors were large, and mounted in a fixed position facing forward, hence the "Forward Looking" part of the name. Today sensors are smaller (some are no larger than a small videocam), but "FLIR" has stuck as a description.

FLIR technology essentially senses heat emission, and creates a video image of heat. This had of course serious military utility, since you could see enemy aircraft against the cooler sky (technically, against the cold background of outer space, since FLIR doesn't see "blue sky.") Vehicles and ships stood out, due to the heat of their engines and equipment. At night, or when temperatures were not especially high, people stood out. The 1993 technology used at Waco formed a greyscale, or black and white, video in which hotter objects were generally white and cooler objects black. (The FBI operator does reverse this occasionally, perhaps because his eyes were growing tired of staring at the same image).

FLIR often looks like regular video. It is important to remember that it is recording heat, not light. In sunlight, a piece of polished metal will often look bright to the eye but jet-black on FLIR. Its polish ensures that the sun does not heat it very well, and it reflects back the coolness of outer space. In technical terms, it is highly reflective and not very emissive. A sheet of black material will often look white on FLIR; it absorbs sunlight and becomes warm.

FLIR sensors are designed to operate at one of two wavelengths -- 3 to 5 or 8 to12 microns. These are "atmospheric windows" where water vapor in the air is much less of a problem. The FBI's sensor was a Marconi, and operated at 8-12 microns. Its recorder also had a soundtrack, which could record sounds inside the aircraft, including radio traffic. The FBI crew switched the sound recording control on and off at various points during the taping.

Exactly why the FBI worried about making a FLIR tape on April 19, 1993 has never been explained. FLIR is useful at night, but not particularly valuable in the daylight--then ordinary video (which FBI insists its aircraft did NOT make) is much more useful. It has been suggested that use of FLIR would make sense if you knew that a fire was going to occur and wanted to be able to document its spread--which is exactly the purpose for which the FBI originally used the FLIR tapes made at Waco.

The Waco FLIR Coverup

For six years after the Waco incident, FBI insisted that had only two Waco FLIR video tapes, the earlier one beginning at 10:42 A.M. Although the gassing assault began at 6:00 A.M., FBI claimed that there were no tapes from that period. In my FOIA lawsuits, FBI said that it had interviewed the agent who operated the FLIR camera and he confirmed that the taping began at 10:42; the chief of FBI's Litigation Unit filed a sworn statement that no earlier tapes existed.

In September, 1999, FBI finally admitted that it did have FLIR tapes going back to 6:00 A.M. The reason they had been withheld was obvious. FBI had been insisting that it had not shot pyrotechnic tear gas projectiles: this was important, since pyrotechnic projectiles burn and are infamous for starting fires when used against buildings. But the sound track of the early morning FLIRs contained radio traffic showing FBI agents asking for, and receiving, permission to shoot military pyrotechnic projectiles. Click here for more data on this coverup.

The Waco FLIR controversy

The real controversy over the Waco FLIRs centers upon certain flashes which repeat at locations outside and behind the building. These flashes begin as the tank at the rear commences its demolition of the back portion of the building (generally known as "the gym") and become very prominent as the fire later begins and spreads. The flashes are not round, but elongated. They appear on an axis pointing toward the building and move in the direction of the building. (There are a few which appear to come out of the building rather than toward it.) They repeat rapidly at the same location. What caused these flashes, or, more technically, bursts of heat?

Dr. Edward Allard, a FLIR inventor and designer, formerly with the Army's Night Vision Laboratory, has given his opinion that these flashes reflect gunfire, from persons outside the building (meaning FBI) directed at persons within the building. He pointed out that item in nature do not heat up and cool down in 1/15 second, let alone do so with repetition. Click here for one of his early affidavits.

Dr. Allard's assessment was confirmed by others, including Ferdinand Zegel, also retired from the Night Vision Laboratory, two different analysts with Infraspection Institute, and Carlos Ghigliotti. Using what are probably the highest-quality copies ever made of the FBI tapes, Ghigliotti found almost 200 bursts of gunfire. More importantly, his careful study of the tapes enabled him to spot human movement and retrace the entire chain of events. At one point there is a massive flash (too big to be a gunshot) near one of the FBI tanks. A person (presumably a Davidian, perhaps the person who threw the object causing the big flash) is seen taking cover and then running back into the building. A hatch opens on an FBI tank and a person descend; flashes come from his position, toward the place where the running person was last seen. At least one more FBI shooter joins him, and the two move about, employing cover, and fire at the building. As the fire breaks out and spreads, the intensity of their gunfire increases. Click here for more on Carlos and his interpretation.

There was also circumstantial evidence supporting the existence of a gun battle. Fired rifle cartridges are visible in photographs of FBI sniper position Sierra One, a house near the Davidian building. Many Davidians were dead of gunshot wounds, but this has been explained as suicide, as the fire drove them to desperation. But one Davidian, Jimmy Riddle, was found dead of a gunshot wound to the head... and analysis of his blood showed no carbon monoxide, indicating that he'd died before inhaling smoke.

Criticism of the gunfire explanation

The possibility that a gunfight raged behind the building became an instant controversy. FBI had for years asserted that its agents did not fire a single gunshot on April 19. Moreover, since the gunfight intensifies as the fire spreads, when visible Davidians might well have been fleeing the fire, the question was posed whether agents might have been pinning Davidians in the flames.

The government initially argued that gunshots do not show up on FLIR. That position quickly collapsed: they show up very nicely. Then it argued that the FBI sensor was too far away to pick up gunshots. That position also collapsed.

The final government position, accepted by Special Counsel Danforth and the House Government Reform Committee, is essentially that the Waco flashes are due to sunlight reflections (more technically, reflections of the sun's heat, but "sunheat" is not a word) off glass or metal debris. They cannot be gunshot flashes because:

1. Gunshot flashes in infrared are very brief events (depending upon which government expert is speaking, either 2 or 8 milliseconds maximum). A standard American videorecorder (NTSC standard, if you want to be precise) such as the FBI was using records in terms of fields (16 ms. apart) and frames (each of two fields, and 33 ms. apart). Thus a gunshot flash can only be one field long in duration; anything two or more fields long cannot be a gunshot.

2. People are not visible near the flashes, and if there are no shooters, there can be no shots. They must be sunlight ("sunheat") reflections.

Based on these princples, Danforth declared "100% certainty" that the flashes were not shots, and the court in Waco dismissed the claims without giving it a trial hearing.

Difficulties with the criticism

As to those arguments we can say, with Danforthian assurance, that there is 100% certainty that they are wrong. Only one of the government experts--VDS corporation, hired by Danforth--ever bothered to test their supposed rules. They relied upon published studies, and in most cases misread them. VDS did attempt a "Waco re-creation" at Ft. Hood. However, we now know that VDS used the wrong ammunition (it tested military ammunition, which has a flash suppressant additive, whereas FBI used civilian Federal Cartridge Co. ammunition) and the wrong weapons (standard 20" barrelled M-16s, whereas FBI used a carbine version with a 14" barrel that generates a much more prominent flash). It also tested on a day about 20 degrees cooler than April 19, 1993, which ensured that shooters stood out as glowing warm objects. At that, at least VDS did try. The remaining experts just pulled figures out of a book.... figures which related to standard military ammo and standard barrel lengths, and in some cases assessed the primary flash (the small flash near the end of the barrel, as opposed to the big secondary flash as the hot gasses expand into the air.)

Experimentation conducted since has shown:

1. Gunshot flashes invariably last more than one field of video, the duration which the government experts claimed was the absolute maximum. Here is a sample of a two-field shot (using, I might add, a rifle with a full 20" barrel whose flash is less impressive than that of the FBI carbines):

When there is dust in the air, which is heated by the powder gasses, flashes can last up to 11 fields of video--11 times as long as the government experts claimed. For details of this experiment, check out Michael McNulty's new documentary, "The FLIR Project. Click here for its webpage. I helped out at these experiments, and can assure you--the flashes went for 11 fields, were very impressive, and match very closely the flashes seen at Waco.

2. On a warm day, human beings do not glow. They are cooler than the earth, and appear as dark rather than light objects. They can easily be invisible if the earth temperature is right.

3. The gunshot flashes have exactly the shape, size, and duration of the Waco flashes. Solar reflection flashes in contrast are generally round, small and weak. This is a product of two factors: (1) glass and metal are mediocre reflectors at IR wavelengths and (2) the sun's output at 8-12 microns is about far less intense than its output at visible wavelengths. Danforth claimed that the reflections probably came from glass rather than metal, so here is a comparison of reflections from a piece of glass about 7 inches on a side:

Below is a closeup, magnified 20x, of the scene. I am holding the glass directly in front of my face, so as to beam the flash directly onto the FLIR sensor. The brightness value of the resulting flash, as assessed in the histogram function of Adobe Photoshop, was within one point (out of 256) of the brightness value of the warm earth around us (and over 100 points shy of the maximum that the tape could record, 255). In order words--yes, glass can generate solar reflections in IR, but they're pretty anemic, and on a warm day may not even be brighter than the background.

And, not to forget....

1. We're not talking one flash at Waco, but one to two hundred, at multiple locations. For each of those to flicker and repeat requires multiple flashing surfaces (that is, as the aircraft is moving it must fly through the cones of reflection of each surface).

2. The FBI also took over 200 aerial photos, in visible light, during the last hours before the fire. In none of these is a sunlight flash visible. But visible sunlight flashes are VERY prominent, whereas sunheat flashes in IR are small and dull. Where is all that reflective material and its flashes in the FBI photographs?

3. The experts on both sides of the question agree on one issue: gunshots do show up in FLIR, and rather prominently. They differ on whether these particular flashes are gunshots, and on what (other than gunshots) causes flashes, but agree that gunshots do register on FLIR. In fact, the government's Fort Hood tests showed that even the M-79 grenade launcher has an infrared signature (which surprised me--it has a unique method of ignition which should have given a very modest IR flash). So every expert agrees--gunshots do cause FLIR flashes.

Now--how do the government experts, who contend there are no gunshot flashes on the FLIR (and all of them indeed contend they can find no, none, nada, by either side) reconcile this with:

a. The FBI concedes that it fired just shy of 400 rounds of M-79 launched gas projectiles;

b. It also fired off several pyrotechnic projectiles;

c. It also claims the Davidians not only shot at its tanks, but hosed them down with full-automatic (machinegun) fire from multiple locations, using tracer bullets (projectiles with a burning flare compound in the base).

I would suggest that a position that "no gunshots can be detected on the FLIR using my standards" is one which calls into question those "standards," since obviously there were hundreds of gunshots, quite independent of the question of FBI firing.


Summary: can the flashes on the Waco FLIR be attributed to solar reflections?

The solar reflection theory has several major, if not insurmountable, problems.

1. Material. There must be a suitable material in the right spot.

No one has documented large pieces of uncrushed sheet metal laying around the building. It is safe to say that any large piece would have been visible, either in the aerial photos or in the FLIR (where it would have stood out as a very dark object at all times when it was not reflecting the sun--see the appearance of tinfoil and galvanized metal ductwork, below).

Wall-insulating tin foil is there, but our tests showed that metal foil is a perfectly awful IR reflector. Most of its surface reflects back the cool of outer space, so it appears black. Crinkles reflect back tiny points of IR, but far too small to have been visible at any distance. The same is true of metal ductwork. Click here for the images.

Glass... the FBI's reports indicate that gas had been inserted into every window early in the morning, with "inserted" here meaning that a 40mm Ferret projectile, travelling at 500 feet per second, had been shot in. Whatever shards remained after that then fell to the ground. In the case of the open area behind the gym, where the first flashes are seen, the area was then criss-crossed by tanks. Any glass shards remaining would have been small indeed. As shown above, even large glass shards are poor IR reflectors.

2. No visible light flashes are noted, although (as shown above) these should be much more prominent than IR flashes.

3. Some of the flashes originate in the shade, although it is hard to see how a source in the shade can generate a reflection of the sun. Here is a scene from 11:34, in which a one-frame flash appears in the shade of the building:

4. Intensity. Our work indicated that reflections off glass are possible, but weak. The sun's flux at 8-12 microns is not that potent to begin with, and unsilvered glass is a mediocre reflector; much of the IR is absorbed and some passed thru. The resulting output may be no "brighter" to the FLIR sensor than is warm earth, and not at all like the bright, contrasty flashes seen on the Waco FLIR. See demonstrations above.

5. Size. Our work indicated that, at 8-12 microns, the solar flash was not only weak, but it appeared no larger than the reflector. So how large an area does a pixel, or "dot" on the Waco tape reflect?

FBI has never released specs on their camera, but FLIRs using the same detector and of the same vintage had Instantaneous Fields of View [IFOVs] of 0.1 to 0.2 mrads. Depending upon the altitude we assume for the FBI aircraft (4,000 to 6,000 feet) and the angle at which it is filming (straight down to 45 degrees off), the smallest image a sensor with those IFOVs can resolve (in colloquial terms, the actual size of each "dot"in the image) varies from 4.8 to 20.4 inches on the ground.

So to span just two pixels (and the Waco flashes are quite a few pixels in size) the reflector would have to be at least nine inches in size, and probably more. That's getting rather large for a piece of window glass that has been shot through, dropped onto the ground, and then run over by a tank. And since the shots come from many different locations, such large pieces of glass must have survived at each.

6. Shape. What solar flashes we did get were round in aspect. The Waco potential gunshot flashes are enlongated--see movie illustration below under item (7). (Note: Mike McNulty's later work found that an elongated IR flash could be generated--by sheet metal, with a sharp bend and otherwise flat edges, and the reflection running the length of the bend. To produce similar flashes at Waco would require a properly-positioned piece of sheet metal two or three feet on a side (which would be, and is not, visible in the aerial photos), with a single sharp crease, and surface not otherwise crumpled. How these pieces of sheet metal came to be at the gunshot locations, without being seen on aerial photos, with surfaces undamaged by the tank demolition except for the single crease, and that crease invariably pointed toward the building, is a bit difficult to explain.

7. Pulsation and repetition. If a flash is produced by the camera-carrying aircraft flying through a reflection of the sun, we would expect the flash to appear, brighten, fade and disappear as the aircraft moved through the narrow cone of light. Instead, the flashes seen at Waco repeat, pulsating bright and dim, without changing location. Here is an example of the flashes, from the events behind the tank [with hot engine comparment glowing] as it is pushing rightwards into the "gym." (If a Quicktime controller does not appear, just click on the movie to make it run).

7. Location and solar angle. This is really the subject of some groundbreaking work by Maurice "Mac" Coxe, formerly of the National Reconaissance Office. He concludes that consideration of the geometry of solar reflections (i.e., their required angles and the angles of the sun at Waco at the right moments) rules out a solar reflection explanation. Click here for his report.

8. Non-FLIR evidence to support the gunshot explanation. For further information, click here: Supporting evidence