Background
A couple years ago, Tyler and finally found a helicopter wreck that we had been looking for. It took a lot of detective work and running down a lot of leads before we finally correctly identified it. After we published the story, we were contacted by one of the survivors of the crash, Donald Sanborn. We took a group of people to the site of the crash that included Don and his son along with the widow and some family of the co-pilot, Gus Hermann, who didn’t survive the crash.
When we dove the wreck back in 2021, the first dive was basically to find it and get a preliminary ID and then the second dive with Ben Lair was to gather more evidence to confirm our identification. Both times, I took a lot of photos but wasn’t able to build a “good” photogrammetry model.
Similar to the recent model of the SBD Dauntless debris field, I had a long-standing project to make a “real” photogrammetry model of the Sea King. It has debris scattered about even more so than the Dauntless so I knew I would need to build up my skills before I took on the project.
Also similar to the Dauntless, the wreck is reasonably deep so I wouldn’t have the luxury of time to get the photos.
As a point of comparison, here is a screen capture of the original model from August 2021 and the same viewpoint of the model I just completed. I now have a model of the complete site as well (see further down).
I remember when I built the first, I tried forever to “connect” different parts of the debris field but just didn’t have the coverage. I even tried to build a model from the GoPro footage I got. I also didn’t have the “other side” of the main rotor assembly. The new version of the model is obviously a massive improvement. 🙂
Dive Details
Lauren Martin and I dove the wreck on Fri, Nov 3rd 2023. It was my third time on the wreck and her first time. We had really nice surface conditions with almost no swell, very little current and almost no wind. I was hoping for the deep blue that we sometimes get in the location of that wreck; however, that was not to be.
The first 50-70 feet were murky with a lot of particulate matter. By the time we got to the wreck, the visibility had improved but there was very little natural light. We had planned on about 30 minutes on the bottom and a two hour max run time. Our actual dive was pretty close to that with a runtime of about 110 minutes. I took 1,100 pictures for the model and some natural light photos at the end.
We had a massive number of fish on the wreck. They didn’t affect the alignment of the photos for the model; however, I do believe that they affected the final model in terms of sharpness. Areas of the wreck that didn’t have much fish (e.g., the landing gear) came out really crisp on the model. Areas that had a ton of fish (e.g., the main rotor) weren’t quite as good.
I was also diving with a new camera lens setup that my wife got me for my birthday. I have long used a “wet lens” (Nauticam WWL-1B) which turns a normal camera lens into a wide angle underwater lens. It is a very flexible and relatively small package but it has the downside that it does have some pincushion distortion. It isn’t as bad as a fisheye; however, it is noticeable.
The new camera setup is a Tamron 17-28 full frame rectilinear lens. A wide angle lens like that would usually require a 230mm dome which I just don’t want to carry around underwater. However, for some odd reason, the recommended dome for this specific lens is “only” 180mm which is significantly smaller.
I really enjoyed the lens and the results and I look forward to seeing how it works for non-photogrammetry use.
Photogrammetry Model
Even with the challenges of the lack of natural light and all the fish, I think the model came out superb. I got really good coverage for the model as show below by the camera shots (each blue rectangle is a photo) and the confidence diagram (middle) where blue signifies 100% coverage and red is low coverage.
Below is a link to the high-quality and mobile (low-quality) models on Sketchfab.
Wreck Overview Screen Captures
Below is an overview of the wreck site and some screenshots from different angles. In the next section, I will analyze each individual area of the debris field with closeups.
Wreck Details
Similar to the SBD Dauntless post and details, one of the benefits from a photogrammetry model is that you can really see the whole debris area in one place and identify how the pieces fit together. The diagram below does that.
Below is an example Sea King with some of the parts labeled as well
1. Main Rotor
The main rotor is the “centerpiece” of the wreck site. One of the photos is on the left and a closeup of the mode is on the right. Note all of the fish that kept getting in the way. Also, to the right of the main rotor in the screen capture on the right is one of the jet engines.
2. Main Blades
The Sea King had 5 blades on the main rotor. I can see remnants of three of them in this model. It looks like they are de-laminating.
3. Tail Boom
This is one of those features that only becomes obvious in the overall context of the photogrammetry model. The main parts of the boom are buried but you can see the obvious lines that match the shape of the tail section. At first, when diving the wreck, I had assumed these were rotor blades but they are obviously not.
4. Tail Landing Gear
Tyler noticed this when looking at the model. It is also in the general area of where it would be. You can also see the fish that didn’t move while I was taking the pictures. 🙂
5. Tail Section
This was another key feature of the wreck that we used to identify the specific helicopter. We figured out the tail code and also the paint coloring in order to narrow down the identity.
6. Tail Rotor
Note the blades buried in the sand. Also note how you can see the “red” tips of the blades in the model and photos that match the photo of the Sea King above.
7. AN/AQS Dipping Sonar
The primary mission of these Sea Kings was anti-submarine warfare. They had a “dipping sonar” that would be put into the ocean to listen for sounds from enemy submarines. This was one of the key features that helped us identify the specific helicopter. All of the details are in the original post.
This part of the model came out very well. It is a little hard to tell that the screen capture on the left below is not an actual photo.
8. Landing Gear
There is only one of these at the debris site even though there were two on the helicopter. During the dive, Lauren did find another small debris field and I went to go explore but ended up turning around (you can see this in the screenshot where the sand extends “down” from the landing gear. I have a feeling the other landing gear might be out there.
Natural Light Photo
As we were getting ready to leave, Lauren was headed back to the downline and I was getting my camera put away. I had wanted to take some natural light photos in the beginning, but my eyes hadn’t adjusted to the dark so it was a bit tough. I snapped a few photos just as we were getting ready to leave. I think it is a cool, and a little bit eerie, photo:
Summary
I finally got another one of my long-term projects done and I’m really happy with the results. My photogrammetry results continue to improve and I have a few more projects that I will hopefully complete before the end of the year. I’m not sure when I will next return to this site, but it is one of my favorite dives.
Nicely done Brett.
Thanks, Bill. Much appreciated.