I did process all the videos in full side by side just so I could get an idea of what is going on with all of them throughout. Its much easier then viewing them all on their own and I can always see them side by side then go direct to the source for a full screen look if I need to this way. Out of this I am just finding a few shots at each level for an actual comparison video as nobody wants to sit through 2.5hrs of footage even though I found it quite interesting lol.
My PC is likely higher spec then most macs and the price Apple would charge for hardware that could match the performance I would never want to know. If I tried to do what I did on any laptop it would not even allow me to begin as even the best laptops are a long way behind a good desktop setup that will be a fraction of the cost with much higher actual performance. My hardware allows me to do such silly things so its not a real issue and a 2-4hr render is what I used to get for 5-10 minute clips on my old computer setup.
All of those earlier videos were encoded in much faster then real time as they were all pretty simple single layered edits. I made them just as I was scanning through all the footage to give me an idea of what I had. I originally wasnt going to post them all publicly but due to lack of time I probably would have had to hold back a while to get them as good as I would have hoped so ended up just releasing the rougher edits.
Here are a few more screen grabs along with depth data showing how I set it up for my own internal comparison viewing.
These are all in the same order as the earlier ones with top row SRP clip, Gopro dive no filter. Bottom row BS Magic flip, oculus and then the Blurfix with screw in URPro cy on the headcam all 1080p30 wide.
On this next dive I didnt have a clear lens on at all times so top right was Blurfix with clipon filter instead of the clear lens. The auto WB was a bit skewy this dive for some reason and the URPro's didnt do as well as on the other 3 dives with the magic doing quite a bit better here but thats likely due to it being a lighter filter allowing more light even though its the thickest of the bunch. This is much like the shallow water cyan URPro filter I have but didn't use this trip, it allows more light to pass through so works better at shallower depths but reduces actual colour correction effectiveness at around the 6m mark. The magic has a sweet spot of around 15m and the URPro on all other dives did well to 25m.
However the helmet cam was also a URPro CY screw in filter which did quite well throughout but this may be due to the video mode of 960p30 working better then the 1080p30 wide used in the other cams at this light level. I didnt expect this and only changed the mode for that cam as the taller frame will make it less likely to miss align as we found the day before cutting of half the shark that had a go at him.
This dive I had all the cameras going and the order is from top left, URPro clip GoPro dive, URPro clip Blurfix, GoPro dive no filter. Bottom row Magic Flip by BS, Oculus and the URPro cy screw on the head cam looking elsewhere but similar depth levels throughout. There were some air bubble issues early on as I forgot to clear the lenses until well into the dive. The Blurfix with a clip filter had a big air bubble for most of the dive that you will see in the top corner at times and the magic also had a few bubbles in the frame.
If you notice the deeper footage with no filter does most of the time end up in a near green out, and all the filters as they are going beyond their usefull limits have gone quite green too but not as bad as the clear lens. These are all some of the worst examples showing the footage beyond ideal depths in the light levels I had on the trip. Most of the very green videos here would be at total distance of around 30m with depths here 25-28m and a few meters from most subjects. Keeping under the 25m total works well with the URPro's at most times except for the one dive where the 2 GoPro's with the flip were both having more wb issues then the magic, oculus and other urpro on the headcam for an unknown reason.
In terms of correcting colour there is very little that can be done to the unfiltered footage at these depths but with filters I can do quite a bit re balancing off the slate to improve the image and correct the GoPro's wb issues at times. The URPro's do still have the most range for me in post even with some of the shots that were poorly awb by the camera. The oculus isnt far off but does produce some odd colours using my process with magic bullet colorista and premiere pro.
Here are a few more shots with all 6 cameras at various levels, The magic did much better on this dive at these depths in comparison to how it did on the other 3 dives. Most likely the water conditions just favoured it slightly and the lower light was helped by the lighter filter so there are still days where some filters will have better days then others.
Unfortunately having lucked out on the conditions for the trip, they weren't ideal for filter use to deeper depths as Fabrices videos were in his sunny clear waters. I guess these tests do show what you can expect in less then ideal conditions though so they may still prove to be usefull to some.
