UB88 Photogrammetry Project – Dive 3 (Conning Tower)


A few days ago I posted about my latest dives on the UB88 with the goal of building a complete photogrammetry model in high resolution.

The second objective for the first dive of the day was to finish filling in the “gaps” on the conning tower and to get some better pictures of the gun mount directly in front of the tower.

My original Sketchfab model is posted here. You can see that the link to the gun mount isn’t there (I was able to create one using Markers in the Pro version of Metasoft) but I didn’t have the photographs to do that without markers.

Below is a screenshot of the model before and after this dive. You can clearly see the “missing bits” on the starboard side of the tower on the model on the left vs. the model on the right.

Side Note: In general, I find shooting the starboard side harder. I believe this is because of the currents. Most of the time when I’m on the wreck, they push from starboard to port. So, when you are on the port side, you are either sheltered or you can gently kick to stay in place. When you are on the starboard side, you get pushed into the wreck and you need to swim back out into the open and then take pictures as you get pushed towards your target. Not easy to say the least!

New Conning Tower Model (428 Photos)

As a result of the photos on my 3rd dive for the project(1st dive of the day), I was able to gather the necessary pictures to build a good model of the tower. This is obviously one of the most distinctive features of this wreck (and any submarine). It never ceases to “wow” me.

HERE is a link to the new Sketchfab model with a complete view of the conning tower and gun mount.

Here is a picture that Drew Wilson shot of me taking the photos. It can give you a sense of scale.

Me taking pictures for the model (Photo Credit: Drew Wilson)

I had numerous problems creating the model, the main one being some weird “artifacts” that would show up above the periscope wells. The good folks at Agisoft helped me to get rid of them. I’m still not sure why they are there, but oh well.

Next Steps

My 4th dive on the wreck (2nd of the day) was to photograph the stern torpedo tube and the “ghost nets.” I’m currently working on that model. The software is having trouble getting the orientation and alignment right on the net. That is not surprising at all.

I need to figure out the model for that and then conduct another four dives on the wreck to finish the model. Here is what I think is left:

Dive 5 – Build the port and starboard sides from the conning tower to the torpedo tubes. You can see in the models that this is currently missing:

Dive 6– Build the top deck from the conning tower to the stern. This is where the “hole” is that I went through to go inside the wreck.

Dive 7 – Build the sides of the wreck to the stern of the conning tower

Dive 8 – Finish up the stern & ghost net. I’m pretty sure based on early models that I will need to return.

5 thoughts on “UB88 Photogrammetry Project – Dive 3 (Conning Tower)

  1. Brett. If you have the option to set a date for a dive the UB, check the tide calendar. Strong currents out there can reach two knots—too fast. If you can find a day when the tidal exchange is at minimum, choose that day. When we first dived the wreck, we tried on a day when eddies were circling off the boat transom, that was, until the anchor dragged. We found a day three weeks later where the tidal change was about a foot for the whole day. We dropped our anchor and it went straight down and remained vertical for our first dive. All of the chain piled up on the bottom. The vis was about 80’ on the surface and 60’ on the bottom. Also note that the last couple of weeks and first few weeks of the year we often get some fantastic visibility, given the absence of sunlight and plankton.


    1. Hey Steve. I’m usually pretty flexible in scheduling so it is a good idea and I’ll keep the tides in mind.

      I’m planning a single dive again this Thursday and hope to complete both sides of the wreck forward of the conning tower.

      You would think that the sides of the wreck would be “easy” but I think they are actually going to be hard due to the lack of features for the software to use to align the photos. Also, I’ll need to start the photos somewhere where I know I have a solid reference point so that I can get a good alignment for a few photos and then build from there. We’ll see…. 🙂

      With any luck I’ll have a complete model of the front half of the ship in a week or so.

Leave a Reply to Brett EldridgeCancel reply