Thank you for drawing my attention to this interesting lens, that I overlooked so far (and also to Phil for making the interesting review in UWPMag in 2021)...
The Tamron 17-28mm may not be a direct competitor of the fisheye/WACP lenses, but is certainly very interesting for rectilinear WA. According to Opticallimits, the Tamron performs over the water optically better than the Sony 16-35mm GM f2.8 and (almost) as good as the new Sony PZ 16-35mm f 4.0 (a new 16-35mm mark II is soon to appear in addition)...
Not only you, but also the Nauticam port chards are in favor of the Tamron 17-28: For the Sony 16-35mm GM the 230 dome is recommended, while Nauticam recommends the 250 dome for the Sony PZ 16-35mm. Interestingly, for the Tamron 17-28mm the 180 dome is recommended over the bigger ones (Nauticam once told me that they recommend the smallest dome that gives best performance, e.g. when the 230 does not improve IQ over 180, they recommend the 180 and mark this one with asteriks). This may indicate that the Tamron lens has a strong field curvature that is in favor of the use behind a domeport and works very well UW (as you report). It looks the Tamron is a hidden gem..
You say that the Sea&Sea correction lens is not working well with the Tamron. Did you test this also?
Wolfgang
The tamron 17-28mm doesn’t have particular issues of field of curvature this can be seen in the top side shots at the edges but also from the mtf charts that do not have the typical ripple to correct field of curvature
The sea and sea correction lens is a field flattener it may help other lenses but this one doesn’t need it as the spherical aberration can only seen at the very extreme edges of the frame and is a consequence of the dome not the lens
See the cfwa shots in water with focus on the edge no residual spherical aberration is there
The lens has the classic drop on the meridional at the edges at f2.8 but it is already better than the sony at f/8 by the time you close to f/8 there is nothing to worry
I believe as explained that out of coincidence the nauticam 40mm extension plus the 35.5 adapter works perfectly
The suggested extension for the 230 on this lens is incorrect so I cannot comment on their chart according to my calculations the 230 needs at least 1 cm less but then the lens field of view doesn’t require it and the lens doesn’t need the extra 1cm radius either
Sent from my iPhone using Tapatalk