Miami Beach, Champlain Towers South apartment building collapse, Part 06 131

[SFCharlie]

A continuation of our discussion of this failure. Best to read the other threads first to avoid rehashing things already discussed.
Miami Beach, Champlain Towers South apartment building collapse
Miami Beach, Champlain Towers South apartment building collapse, Part 02
Miami Beach, Champlain Towers South apartment building collapse, Part 03
Miami Beach, Champlain Towers South apartment building collapse, Part 04
Miami Beach, Champlain Towers South apartment building collapse, Part 05

SF Charlie
Eng-Tips.com Forum Policies

 

[clouditguy]

OK a few more thoughts around the Ring video. I am not at all knowledgeable around structural engineering so cant fully decipher these, but there are a few things that seem very clear to me from intense assessment of this video.

1) I am not at all convinced that the first image in the video is a "bug". This looks more to me like a temporary loss of video feed. The audio continues during this "freeze frame". Ring uses SIP internet protocol to transmit its audio and video, and video is (of course) larger bandwidth in nature. whilst I can't determine it for definite, I would be reasonably convinced that the Ring camera transmits its audio and video feeds separately, and something led to the loss of video during these few seconds - could be coincidental, or could be relating to the failure. There certainly appears to be a "flash" of light at the top left of the image immediately after freeze, which could be something electrical, causing interference in transmission (its a reasonably safe assumption that this camera is wireless not wired, and wireless is highly susceptible to local EM events). My view therefore is that we are not seeing the tail end of something that we dont see the start of - but rather that the first image is chronologically accurate and an event in those first two seconds led to the slanted image we see later.

As an aside on this part - sample size of one admittedly, but I also have a Ring camera and try as I might I cannot replicate this "freeze frame" element in normal usage.

2) At the end of the video, it absolutely appears that the table that the camera is on starts sliding to the right hand side. I am actually not sure this is the case - or at least its not a "free slide" of the table purely due to gravity. Throughout the video, you see a "pillar" of white at the left extreme. I am pretty sure that this is either the inside of the balcony door frame, or it is a jutting out bit of wall (these were in the condos in the corners of this room if you look at some of the sales pictures for these condos). Note that during this "slide" of the table, this part of either wall or door frame never changes its reference point to the camera.

This implies strongly to me that rather than "sliding", this table was being "pushed" by this wall. If this is the case, the other conclusion is that this wall was moving in relative isolation. It would appear this was moving in isolation from other walls (albeit this is clearly hard to work out) but categorically it was moving in isolation to the floor of the condo - there are plenty of reference items that confirm this.

Note that we do not really know that this camera was in fact on a table, its merely on a platform of some kind. Could this be a wall mounted shelf?

3) If the above implications are correct (admittedly could be off) then this has a very interesting further conclusion - that immediately prior to the point of collapse, the outside shell of the condo (i.e. the external door frame and presumably balcony) was moving relative to the rest of the condo. This period is long enough to know that this was not DURING collapse, but just pre collapse. What does this mean? I have no idea, but it doesnt really align to the failure of a column underneath the "body" of the condo I dont think.

4) A quick note on the timings in the video also. I know there have been a couple of comments that at least part of this video could have been during freefall. I categorically dont believe this is the case, for a couple of reasons. First of all any significant downward motion in the condo could not have occurred without inertia moving at least some of the contents/jolting them - all of them. The only item we really see move in this way other than wall forces impacting on them is the empty TV box which judders rather than "lifts". Secondly, we need to consider WHY the camera stopped recording. This could only happen for three reasons (two of which are largely the same). Firstly, the camera lost power. Secondly, the camera lost connectivity. Both of these therefore we can consider to be loss of utilities/services. Thirdly, the camera was destroyed. I don't believe that it can be the third reason - no matter the speed of collapse and destruction, I don't believe that we wouldnt see a single frame of destruction occurring prior to the camera going completely offline. The failure was sudden and immediate. As such, I am pretty sure that this camera stopped recording due to a loss of power or connection - which was almost certainly triggered by the commencement of vertical movement of the condo. There are other reasons as well - right up until the end of the video, the debris from above continues to fall - if the condo was in (or starting to) free fall, you would see a relative slow down of these particle falls - you do not.

5) A note on the "slanting" of the condo, be this in the horizontal plane of the video or the depth plane. On further assessment, I don't believe that theres clear evidence of this (although to be clear not saying this wasn't occurring). A lot has been made of the change in fall of the particles, which initially had me confused as well. However when you look at individual particles very carefully, there is a clear tendency towards the vertical plane, rather than a straight line fall. This is the case on both sides of the camera, and also when particles are dead centre, you dont see this gravity influence effect as if there was a horizontal slant.

I am less sure on a vertical slant. I think you would see more evidence of this in the video - other items moving etc. The particles themselves show no change in luminosity in flight compared to the vertical falls, which you would see if the backwards vertical slant was bringing these particles closer to the camera whilst in flight.

6) I think more likely, given 5, that the SOURCE of the particles was moving in ceiling above the camera. The unusual trajectory I cant fully explain, but this could be due to ejection under stress (for example, concrete failing on a side of a rebar rather than directly below it?).

Equally in line with this, I do think pretty categorically that the particles are concrete not dry wall. We are seeing these being generated for (minimally) 10 seconds, in ROUGHLY the same location. I simply dont think drywall is robust enough to do this. Equally, the location of this camera directly below concrete beams to support the door frame etc would back this up, as well as the lack of any obvious visible cracks in the rest of the ceiling drywall.

7) The deformed pillar beside the door, for what its worth, I have calculated to have "relatively" moved the base of this pillar to the left by around two widths of the pillar.

What does this all mean? I really dont know, but I'm hoping someone else may.

Having read 6 threads on this and been thoroughly convinced of the obvious if "not sensational" pool deck collapse due to water load leading to pulling/other forces causing column collapse, my untrained eye (especially the movement of the camera at the end apparently caused by a wall) does start to lead me to question this is the full story. Did something fall off the roof with force and strike the balcony of 711 or the balcony above, leading to ultimate failure of the outer balcony and wall on this area? It would tie more closely with what is seen in the Ring video. Unsure if this could cause a "pivot" effect - i.e. balcony struck, causing a shift of the slab inside the condo up to balance this downward force and hence deforming the pillar. Of course even if this is the case, this is not necessarily full cause and effect, and this could merely be a symptom in some way of a lower collapse. But I think the duration is significant enough that this may not be the case.

Worth mentioning as well that in the last 1.5-2 seconds (after the box jolt) the camera does not just move right relative to contents of the room, but also DOWN. This is not as obvious, but when you look at it frame by frame theres a definite downward motion compared to contents.

The jolt of the box itself is a little odd in the way it happens, mainly as the box does not move in the obvious way. The rear bottom corner actually moves more into the centre of the room. This is despite the fact that its CoG is outside of its base, it presumably isnt resting on anything behind it (given that theres now more space behind it than there was at the start of the video), and the boxes it is resting on very slightly (but definitely) move towards the centre of the room as well. If the room itself was slanting, then this would further mean that the likely fall would be rear bottom corner towards the wall, therefore towards a flatter profile, but it does the opposite. This would either imply the room is not slanting, or (given the other reasons it should have fallen more rather than become more upright) something has happened to the floor or wall behind it that we cannot see that has pushed on this corner.

Whether all of this has any ring of truth or is just a layman talking rubbish, I really dont know. But it might provoke some thinking.

EDIT - looking back at earlier posts and floor plans etc - does this end movement imply column M10 was the one to cause the actual final collapse? I am still not sure that this was the case until the very end of this video though.

Also - having looked back I can see no confirmation that this camera is actually a Ring camera. This was reported somewhere (cant find where) but not apparently by the resident of 711, so I think its an assumption by a media outlet. However it doesnt change the fundamental analysis.

 

[Demented]

The piling elevations are not in the plans. They’d be attached to the DERM certs. It’s still unknown if it’s all precast driven or the mixture with driven at the boundary and franki for the interior.

I'm assuming based off of dimensional call-outs for 3'6" and 48" of piles, the 2'2" slab elevation for the parking/basement level, 4" pile to cap recess, and 32" pile cap height.

The simple fact that depths of the piles are called out makes no sense to me.
I could completely be reading the plans incorrectly, I don't doubt that one bit. They're a mess and they make the Fontainebleau renovation plans look like a masterpiece.

It was Miami in the 1980s. Augercast piles? Especially with depth callouts, that would be the easiest pile job ever. Drill to depth, insert some corrugated steel tubing to stop the water ingress, pour your concrete, and call it a day. (Old but very common practice here, and wasn't known back then but we know today, water flowing underground around steel pipes leads to open voids). Only deviation being dropping a wired up #7 cage assembly. Few inches of concrete at the bottom so the 72" long rebar cage (Or 78" long rebar cage for the 48" piles, but we're talking about 6" here and you can easily cheat only using 72" long embedded reinforcement, because just like using the incorrect size nelson studs, aint no body going to know once it's in the ground.)

I fully understand this is NOT how it should be. But I'll refer back to my story of the kid using his fall arrest harness to try and swing to a lower level during construction as the level of brilliance building our high rises.



Precision guess work based on information provided by those of questionable knowledge

 

[AutisticBez]

My son plays Portal 2 a lot. Maybe someone was playing with a portal gun, and either managed to make the Penthouse destroy the parking deck, or moved a speeding car on the road outside, end up knocking over a column. I'd photoshop my new theory, but then I'd lose the ability to claim patent rights on the portal gun. Until I lodge that patent, I'm going to have to go silent on the matter. Definitely don't let anyone know about my invention.

Let's just say that if a car did knock the column down, it probably burnt up under the weight of all of that building falling onto it. Maybe you'll only find a drive shaft left?

 

[1503-44]

I have not seen mention of a geotechnical report.

This was done for the
Miami Convention Center
Might be representative of soils in the Surfside area, might not.

https://www.miamibeachfl.gov/wp-con...t-related-to-the-MBCC-site-for-Structural.pdf

 

[clouditguy]

Sorry there is one other observation here as well which I think is important enough to call out itself.

The 711 video has its safe to say NO events contained within it that are more impactful (to this condo at least) than the event which initially triggered the camera, and the event at the end of the camera.

In addition, if we assume (and I think given both the video and audio that this is a certainty) that the end of the 711 video is the final collapse, then clearly the triggering event CANNOT be the same as what occurred to cause the garage area debris, as the timelines do not match up. Absolutely minimally (and in truth longer, but we can say this as fact), the garage debris was created at least 17 seconds prior to the overall collapse due to the length of the TikTok video. In truth this is longer (probably many minutes). The 711 video is only 12 seconds long.

Therefore, what is the event that wakes up the camera 12 seconds before complete collapse? Note that this event appears to be more impactful to 711 than the pool deck collapse, purely by the fact that it didnt wake the camera up (as far as we know). And is it significant (or not) that within this 12 seconds there are other "events" (primarily 7 seconds in) that did NOT cause as much impact? Probably not - but on the assumption mentioned somewhere above that the noises heard are progressive column failures, would these not be minimally as impactful upon the condo in these first 7 seconds as the initial event? Again - probably not, but thinking aloud. The important element rather than this line of speculation is that 12 seconds prior to collapse (and again at 5 seconds prior, albeit not as apparently immediately significant) there was clearly a significant event which affected the building, which was not the pool deck collapse (if we assume that the TikTok video demonstrates the aftermath of the pool deck collapse, which of course I think isnt in doubt).

 

[pellucidar]

1) I am not at all convinced that the first image in the video is a "bug". This looks more to me like a temporary loss of video feed. The audio continues during this "freeze frame". Ring uses SIP internet protocol to transmit its audio and video, and video is (of course) larger bandwidth in nature. whilst I can't determine it for definite, I would be reasonably convinced that the Ring camera transmits its audio and video feeds separately, and something led to the loss of video during these few seconds - could be coincidental, or could be relating to the failure. There certainly appears to be a "flash" of light at the top left of the image immediately after freeze, which could be something electrical, causing interference in transmission (its a reasonably safe assumption that this camera is wireless not wired, and wireless is highly susceptible to local EM events). My view therefore is that we are not seeing the tail end of something that we dont see the start of - but rather that the first image is chronologically accurate and an event in those first two seconds led to the slanted image we see later.

As an aside on this part - sample size of one admittedly, but I also have a Ring camera and try as I might I cannot replicate this "freeze frame" element in normal usage.

Ring cameras do have a snapshot mode, which can only be set up and viewed through the mobile app. The frequency of snapshots is user-configurable. It's not clear that these snapshots get appended to motion-activated video, as some have assumed to explain the initial video frame. BTW, it's recommended to mount this sort of Ring motion sensor to the wall at a height of 7' or 7.5'.

I don't think "something electrical" can explain it either, though, because I'm having trouble imagining an electrical event that would have left the internet connection to the camera intact.

Also - having looked back I can see no confirmation that this camera is actually a Ring camera. This was reported somewhere (cant find where) but not apparently by the resident of 711, so I think its an assumption by a media outlet. However it doesnt change the fundamental analysis.

Good catch! I don't think Ring was mentioned anywhere in the media; possibly it came directly from the eponymous Reddit post.

 

[Penagwin]

I found a longer more complete 911 call log, listening through it now to see if there's anything useful but I sadly doubt it (everyday non-engineers + panic and fear) but probably worth checking anyway.

https://www.youtube.com/watch?v=v2zRuxNjG2I

 

[tmwaits1]

The original plans are, as you noted, terrible. Full of discrepancies and missing and inconsistent information. To clarify, I think the depth you are seeing refers to pile cap thickness, not depth of pile. The depth or length of piles (either 150 ton PIF or the original 50 ton PC) should have been determined by installing a test pile where in the case of the PC would have been driven to practical refusal. How they determined the 150 ton depth of the PIF (aka franki pile) back in 1980 I don't know. Also note that the original pile cap plan which shows closely spaced square concrete piles (PC) was not apparently revised to show a new layout with the PIF. As you can see below, they only provided revised section views for caps with 2,3,4 PIF and the two H shaped shear wall caps. There is no revised cap plan that shows the foundation for the columns that failed (K,L,M). The foundations supporting the main columns that failed cannot be determined from these plans. To my mind, there have always been 2 separate questions regarding the collapse which are; 1. what triggered the collapse? and 2. why was the damage not more localized? The second question comes down to the original design being flawed, and there is certainly evidence in the original design documents and photos of the collapse debris to strongly suggest that the building lacked the type of redundancy that would have ordinarily limited the damage.

 

[clouditguy]

The more I look at it I dont think it is a Ring camera - there would be usually some time/logo stamp from a Ring, which I cant imagine the resident would have taken the time to remove. As I say, probably not that material.

Re: my thinking on loss of video, just to expand. Wireless (which I definitely think its a safe assumption this camera is) is simply an EM wave. An Electrical short of some form would equally cause an EM wave, which (can) cause interference.

Interference in this way would not be an "on/off" thing, it would simply degrade the signal and therefore mean that the packets sent are more likely to be corrupted and so enter what is called a "backoff" due to corruption or collisions. Audio and Video streams are almost always sent seperately, and typically the video stream is roughly 10-20 times the size of the audio, and as such much more likely to be corrupted (as there is more data which can be corrupted). In most such streaming protocols, if the packet is corrupted, the recipient simply rejects it - in many cases it will request a retransmit, but this depends on the protocol - it may simply drop it and "move on".

Admittedly, it is a little unlikely that the device has dropped every video packet, and accepted every audio packet, which is largely what appears to happen from an "end user" perspective. However it all depends on the proprietary implementation by the vendor of the camera - they may be more aggressive with not trying to recompose a heavily artifacted image.

To be clear - I'm not 110% behind this. But I dont fully buy into the initial image being a simple "bug" either, especially given it (equally) is at night and would at least appear to be chronologically "close".

 

[Penagwin]

@clouditguy

I think the camera likely was battery powered (or could be), but it's router, and the router's cabling to whatever it's peered to, and potentially whatever router and cables are peered to that if it's in the same building were likely destroyed rather quickly - especially it's cable/fiber uplink.

I think that first frame was likely one of it's previous frame's it had captured (last one in the video buffer), and that was likely what was being diff'ed to detect motion. There may have been a delay between when it detected motion and started recording, no idea how long that could be.

 

[Auri]

The 711 video has its safe to say NO events contained within it that are more impactful (to this condo at least) than the event which initially triggered the camera, and the event at the end of the camera.

Let's not forget, the camera is a low power device likely to introduce around a second of latency by the time a video frame is encoded and actually sent out within a packet. And while it's eerie and perhaps exhilarating to analyze all the details from... found footage, unless there's "smoking gun" evidence regarding what happened to the building 10 or 30 minutes before recording, it doesn't really matter whether L 10 or M 10 was sinking a few seconds ahead of the other when it's all several minutes past the garage "explosion" which obviously altered the sequence of progressive failure and of which we wish we knew the cause.

 

[Demented]

The original plans are, as you noted, terrible. Full of discrepancies and missing and inconsistent information. To clarify, I think the depth you are seeing refers to pile cap thickness, not depth of pile. The depth or length of piles (either 150 ton PIF or the original 50 ton PC) should have been determined by installing a test pile where in the case of the PC would have been driven to practical refusal. How they determined the 150 ton depth of the PIF (aka franki pile) back in 1980 I don't know. Also note that the original pile cap plan which shows closely spaced square concrete piles (PC) was not apparently revised to show a new layout with the PIF. As you can see below, they only provided revised section views for caps with 2,3,4 PIF and the two H shaped shear wall caps. There is no revised cap plan that shows the foundation for the columns that failed (K,L,M). The foundations supporting the main columns that failed cannot be determined from these plans. To my mind, there have always been 2 separate questions regarding the collapse which are; 1. what triggered the collapse? and 2. why was the damage not more localized? The second question comes down to the original design being flawed, and there is certainly evidence in the original design documents and photos of the collapse debris to strongly suggest that the building lacked the type of redundancy that would have ordinarily limited the damage.

However, we have drawn to scale.

Measure that pile cap to scale off the known 4".

Fuckin oops

Edit: Were the caps supposed to be the 3'6" sections? Because that is not how the plans call it out. As read, and hopefully not as built, that building stood on 6ft legs.

 

[clouditguy]

Thanks both.

To be honest the more I look at this in terms of my first point in my post above, I dont think this really matters all that much in truth. To be clear not just saying this because there are counter views it just probably doesnt change much in truth.

For what its worth I would quite strongly disagree that an internal security camera is likely to be battery powered. This isnt a Ring doorbell (for example).

I do think its of material note to look to understand what happened in the seconds leading up to the actual collapse, as this can give an indication as to the method of collapse. Not to me admittedly without structural knowledge, but others may.

I do think the key points of note above though are:
1) the outer edge of that condo is quite clearly moving DOWN and RIGHT immediately prior to collapse - this is to be clear not only in relation to the building, but in relation to the rest of the condo, which at that time was not showing signs of huge ongoing movement. I do think this is quite interesting information to cross match against the working theories on failure - especially is it would (to me at least) strongly imply a column failure causing final collapse at the edge of the building rather than further within. It could also imply (albeit I'm stretching my already very limited concrete knowledge here) that already prior to collapse the outer wall of the condo had separated from the condo floor slab. In fact from the layman, it doesn't just imply this I must say but really shows it, as otherwise I cannot understand how this camera would move down compared to said slab.

2) Theres an apparent upward, or sideways, movement 12 seconds before collapse, distorting the kitchen pillar. Again I am completely naive to structural issues but I cant think of how this is explainable fully in the current theories.

3) There would appear to be failure of the ceiling above the camera, but nothing obvious (certainly visible) elsewhere in the condo. This seems to be reasonably surprisingly local given the apparent significant building damage already in place by this time, and it is interesting in my view to understand if there is something other than the current leading theory that could cause this.

I get the point though Auri, and I understand there is some lethargy around this extremely detailed video analysis in this thread, looking for colours/objects/shadows/yellow jackets. However, I do honestly think there is a difference between a reasonably static piece of footage which is being assessed to the point of seeing Elvis, and actually looking at a dynamic, internal, piece of footage which shows a progression of events over a period of 12 seconds up to collapse. If this was the 12 seconds prior on the external CCTV, we would be all over it - the clues here are more subtle but I think are still there. In truth I'm a little surprised that there hasn't been more discussion in this thread about this video, as its the single piece of pre collapse objective evidence we have.

 

[MechinNC]

I think really key in this investigation is getting earlier frames from the neighbor property security cam vid. If it exists, as it may not have been recording prior to the collapse.

Or other video evidence from nearby cameras.

And whether the columns central to the initial collapse are at a lower level with respect to grade than pre-collapse. That area is cleared, so should be pretty easy to determine. Remove some of the nearby slab and see if paint line is below top of slab level.

And testing structural elements from the debris for strength, reinforcement, etc.

I suspect all of that is on-going, but being kept very private.

So all we can do is wait and speculate. Frustrating.

I too find it very interesting to analyze the 711 ring (or whatever) vid, but that is probably not too valuable in the forensics. We know what happened to 711. Just don't really know what preceded it.

Bizarre that the hired forensic engr is being kept out of the loop!!! That is some BS!!!

 

[Murph 9000]

Therefore, what is the event that wakes up the camera 12 seconds before complete collapse?

This is mostly hypothetical, the precise details may vary.

Pool & patio deck fails. The 9.1 column line (particularly K,L,M) are fatally damaged by that, but do not instantly collapse, temporarily surviving in heavy overload. The building is under exceptional stress, cracks are forming and spreading, but it's mostly standing up and hasn't moved significantly downwards yet. The damaged columns are in overload, and the building is doing it's best to redistribute loads around them.

After 5 to 15 minutes, M9.1 has a compressive failure in its damaged section, the remaining concrete at the damaged point explodes with a BANG! The bang sends a jolt through the entire structure (particularly everything adjacent to M9.1, like the camera in 711), causing the empty TV box in 711 to tip over and the room to deform, activating the camera. M9.1 slowly descends as its now empty section of rebar cage buckles and the surrounding structure is unable to take all of the load. That stress adds up until L9.1 is similarly overwhelmed, probably failing with the SNAP before the final rumble at the end of the video (note the high velocity rock flying from the L line on the left of the room across to the camera on the right, simultaneous to the snap). K9.1, already damaged by the patio collapse, fails almost instantly.

With 3 adjacent columns gone, I9.1 and N9.1 are just moments behind them, and the I–N8 columns rapidly follow. There's a very brief delay at the 4 line on the north of the corridor, probably from the eastern stairwell shear wall. That shear wall also saves the eastern end for a few terrible seconds, but it's already badly damaged and off-axis to the west.

Hypothetical, as I said, fitting what I believe is a plausible damage and failure sequence around the evidence. The exact details will vary; e.g. maybe N9.1 fails slightly earlier than described, due to its close proximity to M9.1. The SNAP towards the end of the video may be from L8 inside unit 711, moments after L9.1 failed in the basement.

 

[3DSoftwareDev]

I don't want to belabor this because I get the distinct feeling some here think the interior video provides nothing of value (and as non-SEs, we're subject to snipey comments about our background and how we know nothing), but:

you see a "pillar" of white at the left extreme. I am pretty sure that this is either the inside of the balcony door frame, or it is a jutting out bit of wall

I tend to disagree. The camera is sitting on some table, and the table has a column attached to it. The column to the extreme left therefore seems like it's part of the table "furniture" and not built in to the unit. I suspect it's a credenza with a hutch top, featuring a little molding where the tabletop meets the hutch portion, like this:

This implies strongly to me that rather than "sliding", this table was being "pushed" by this wall.

As for the movement, you can see the gap between the box and the left column decreasing over time, suggesting that this furniture is moving separately from everything else. That could still mean the deforming wall is pushing this furniture (I'm a little suspect of this too - the wall isn't deforming as much as the table is moving by the end, and there's a significant gap between the credenza and the wall), but I'm not convinced the column on the far left side of the frame is actually part of the unit wall or door frame. I'm still of the opinion that the credenza is sliding across the floor at the end due to the floor tilting - and tilting mostly on this side of the column line dividing the kitchen from the living room, explaining why the chairs don't tilt and the box moves (maybe).

The 711 video has its safe to say NO events contained within it that are more impactful (to this condo at least) than the event which initially triggered the camera ... Therefore, what is the event that wakes up the camera 12 seconds before complete collapse?

I suspect the cascade of dust was the event that triggered the camera's motion sensors, more than anything else. The box no longer leaning against the wall might have done it, but other changes were probably too subtle. So it's unclear, to me at least, that there were indeed no other events like skewing of the unit, right before the camera turned on. I think the camera turned when the deformation of the unit or whatever event was strong enough to cause enough debris to fall directly in front of the camera to cause it to wake.

Regarding the first frame, it's hard to tell, but there could be a very small amount of debris already on the table surface, indicating the first frame was probably seconds or less before the rest of the video. If so, then I agree that some latency or loss of connection eliminated potentially some movement that happened to cause the box leaning against the wall to be pushed off, and a few moments of perhaps other deformations. This would indicate all the deformations we see happen with one trigger event, right when the camera is woken up, but the camera didn't quite catch everything at the very beginning. Otherwise, the first frame could be minutes before, when a tiny bit of debris was caused by some earlier structural shift, or it could be hours before and there's no debris on the table at all.

That also brings us to:

There certainly appears to be a "flash" of light at the top left of the image immediately after freeze, which could be something electrical,

I think the simplest explanation is:
- The camera was in the pitch black taking occasional samples of the environment (the first frame)
- The dust starts falling, the camera turns on, the scene brightens temporarily to account for the increase of white particles directly in front of it
- Then as the camera acclimates, the brightness is reduced again back to normal

I think the flash of light is just a typical camera thing when a camera is enabled and has to adjust for ambient brightness of the room or objects in front of it.

 

[clouditguy]

@murph - I think that is a really good interpretation and probably very close to what happened.

The compression of the kitchen pillar could be explained by this as well, as (I presume) if the outer wall lost support at the start, and the slab stayed intact, it's feasible I would assume that the slab pivoted, pushing up on the far end of the room.

The part that still troubles me is the relative movement to the body of the condo in the last two seconds, but your explanation is probably the majority of it certainly.

 

[Tropmet]

jbourne8 your mention of column labeled 39/40 in the garage resulted in being able to reinterpret this as an uninterrupted view of the fallen pool deck, but the column is further back in the image. I did lighten and color enhance an image (not remembering now if it’s one I pulled from the forum here or one I took a screen print of from the video). You can faintly see the color of the pool deck tile surrounding that column after this image enhancement, below the arrow. The other image is from the WaPo article, for reference. This is also the correct location for where this column would be seen from this viewpoint.

The column in the foreground labeled for parking spot 27 delineates where the building meets the pool deck. It does appear that the pool deck fell all the way to the edge of the building, before the building collapse. The edge of the pool deck that connected to the building at this location is clearly illuminated on the ground from a light still on in the garage, because you can see the sharp shadow of the broken piece of the deck undeneath it, probably the deck immediately outside unit 111 that was their private deck bordered by the planters. It is located just past column 27, where you would expect it would be. It appears you can then see all the way back to that column labeled for parking spots 39 and 40, behind the fallen planters, partially sticking up through the pool deck tile. This suggests that at this location the pool deck fell all the way from the exterior wall to where it met the building, which would also explain why there are no remaining garage overhead lights in this back part of the garage; we’re only seeing the fallen pool deck. Since it was nighttime there would have been no illumination from the sky, but at this point it is likely there is still some faint illumination of the pool deck from the building lights enabling the view of the distant column.

Admittedly there is a lot of noise in the image so I would not want to overanalyze it, but I was surprised when enhancing the light and color to see that, and felt confident in this conclusion. Also because of the extreme foreshortening in the image it is a challenge to understand the perspective.

 

[jbourne8]

clouditguy, many years ago.. I was actually a VOIP technician and I know a ton about the related protocols. SIP is only used to initiate the stream, and most likely it's actually transmitted with something like RTSP instead. It can transmit video and audio at the same time, and RTSP is UDP traffic so there actually isn't much of a delay to it at all (unlike SIP which is TCP). This is why when you're on a VOIP call or transmitting a live stream there is almost no delay at all.. you can really tell when the delay is even 500ms. When I say the delay was about 100ms that was a very informed opinion based on latency across the US (120ms is around the max delay I've seen when going from east to west coast on DSL or better). And UDP streams will never re-transmit.. they'll just drop packets when they lose connection. Even if it's not an RTSP stream, it has to be UDP traffic. I didn't see any jitter in that in the video. In the last 20-50ms or so of it, the video does seem to cut out right before the audio.. but that's the only real difference in delay because of the size of the video vs audio data. It's really not as significant as you think. I'm pretty confident that the stream cuts out as the building is falling due to loss of power. You can actually see a good amount of stuff in the room start to fall towards the camera at the end too (the dust, that black fridge-like object on the right, and the camera itself in the last few ms).

Someone told me that the "freeze frame" is from these photos that the camera takes every hour and was probably started with that "look back" since they don't transmit video when they don't detect something.. it's also in there for several frames and none seemed to be missing. I'm not sure if that's true too, since that's not how my cameras from a similar company work.. but it fits. While it makes since that something could be pushing the table.. the camera also starts to slide to the right in the same manner and at about the same speed, and I find it hard to believe that something is pushing both items in the same way.

I also think the camera was triggered by the failure of M10.1, or at least I made the video with that theory in mind.. to sort of test it. Seems plausible.

EDIT: There is no M9.1? Pretty sure we're talking about M10.1.

 

[jbourne8]

Tropmet, I would go watch the video again really.. analyzing the photos from tiktok on their own isn't really as clear -

https://www.youtube.com/watch?v=LoNy55PaHlc

The first half is the video I grabbed my drawing from too. That thing you're pointing as is probably a white planter, that's sitting on top of the pool deck.