I have written a number of articles on underwater photography and this one tries to explain why Dome Ports need to be selected carefully and set up accurately in order to work at their best.


Dome Port Theory

Using and optimising wide-angle lenses and dome ports. A dome port, which is actually just a thinnish, curved piece of glass, can be thought of as being a ‘simple’ lens. This means that it is a lens, the curvature of which forms part of a sphere (so it is a spherical optic and as such is often referred to as a concentric dome) and has ‘parallel’ inside and outside surfaces (not absolutely accurate, but I’m sure the meaning is obvious). Concentric dome ports are used on housings designed for underwater photography because they retain the original angle of view of a lens used behind them, and they are used for wide-angle lenses, of both conventional (rectilinear) and fisheye design. Because they are spherical in curvature, concentric domes are ‘relatively’ cheap to manufacture. Non-concentric domes (with glass of varying thickness) have been used (Cousteau had some ground) but are very expensive.
The ‘virtual image’. Using a dome port creates an image known as a ‘virtual image’; which quite simply means one which appears to be in a place where it does not, and cannot, physically exist. The mathematics behind domes shows that for a subject at infinity the virtual image created by the dome will lie in a position which is 3 x the radius of the dome in front of the front surface of the dome. Subjects closer to the dome than infinity will have a virtual image closer to the dome (and all are). However there is an inherent problem with this virtual image, because not only does it lie in this position close to the dome but it also effectively lies on a sphere concentric with the dome and with a radius of 4 x the radius of the dome. So it is a curved, virtual image, and we try to photograph it with cameras designed to photograph planar subjects, so have to rely on depth of field to ensure that all of the virtual image which we need to be in focus actually is. Unfortunately, this all too often means that the corners are not as sharp as we might like because the depth of field can be insufficient.
Alignment. So when we try to photograph anything underwater through a dome port we are actually trying to photograph a curved virtual image quite close to the dome. To do so the camera lens needs to be aligned accurately inside the dome so that its entrance node is at the common centre of the dome. This means using spacers between housing and dome to do this.Correcting for close use As the virtual image is quite close to the dome it is also relatively close to the lens, which in turn adds another need for potential correction. Whilst some lenses will focus relatively closely, doing so rarely helps their performance, unless that is they are designed with enhanced close focus performance as some are. Such lenses are designed with modified optics to increase their close focus performance and some even include ‘floating element’ designs which operates only when the lens is used at closer focus settings.Ideally a lens used behind a dome port should be able to focus between the virtual image of a subject at infinity (4 x the dome port’s radius) and just in front of the dome (the dome’s radius). So clearly there is a trade off between the size of dome and its usability. Too big and it becomes cumbersome and its own radius is a limiting factor on the closest subject. Too small and the it will be a struggle to enable the camera lens to focus close enough at all.Some lenses may focus close enough to use behind a larger dome by themselves and performance will be high if they are also of the ‘close focus corrected’ type. Others will need an aid in order to be able to focus even onto the virtual image. Aids are available in the form of close-up lenses or ‘dioptres’ as they are known in underwater photography.

Again mathematics comes to the rescue and in order to realign the focus of a camera lens so that when it s focussed on infinity it actually focuses on the virtual image of a subject at infinity, we can calculate the strength of the diopter required. The formula is that dioptric power = 1000 divided by 4 x the radius of the port in mm. Diopters are generally available in a variety of sizes for differing lenses and in various powers +1, +2, +3 and +4 being common, although others do exist.

Diopters are also built in a variety of optical configurations the simplest and cheapest having one flat and one convex surface (now rare), through the most common which have one convex and one concave surface and then through to multi element designs (usually doublets) which are highly corrected and are sometimes called achromats.

In theory the simplest design with a flat surface should have another benefit when used behind a dome as its simple design tends to produce image curvature which to some extent opposes the curvature caused by the dome port, and so is considered to help ‘flatten’ the image seen by the camera lens. In practice it is doubtful if using such a simple lens helps because any reduction in image curvature is probably offset by image degradation caused by using another very simple optic (as well as the dome port itself).

Although the mathematics indicates the dioptric power which will fully correct the focus of the camera lens, in practice a lower power diopter is often perfectly adequate as full correction (to allow the camera lens to be set at infinity when focused on the virtual image of a subject at infinity) is not actually often needed. Adding a diopter into the optical system (dome port – diopter – camera lens) may however degrade the overall image quality as the overall optical system is now more complex and is using two relatively simple and poorly corrected components.

Loss of angle of view with diopters. Unfortunately there is another problem which is caused by using dioptres in front of wide-angle lenses. For the technically minded this is because wide-angle lenses designed for dSLRs are of retro-focus design and this in turn means that the aperture (stop) inside them is positioned substantially far back from the front of the lens.

In effect this means that the lens cannot ‘see’ its full field of view when a diopter is attached and will effectively become less wide. This is particularly problematic with large, complex lenses such as fast ultra-wide zooms and in extreme cases might mean that as much as 25% of the field of view is curtailed – so an ultra-wide zoom set at 15mm utilising a diopter may well produce an image which effectively is similar to that of a 20mm lens without a diopter. The problem’s severity depends on the specific lens in use, the position of its aperture within the lens and its optical complexity, and is exacerbated as the power of the diopter increases.

Because it is lens specific, the only way of ascertaining just what the effect of adding a diopter to a very wide lens actually is, is to test it! Fixed focal length lens designs, which are optically simpler than zooms and have an aperture closer to their front, may suffer less significantly from this problem.

The image and the camera lens. Generally dome ports are used for wide-angle lenses. As the angle of view of the lens increases it has to work with more and more of the curved (spherical) virtual image. If the virtual image was simply of an originally planar (flat) subject at infinity, this would be less of a problem, but even so a point would be reached where it would be difficult to retain good corner image quality (this is measurable but in pictorial photography it is also subjective and will always depend on the final use of the image too).

In the real world images are of subjects with depth and often with foreground detail where the image corners are actually physically close to the dome itself. In such cases corner image quality drops off even if the camera lens is used at a small aperture in order to maximise depth of field.

As an empirically derived rule of thumb, good corner image quality is limited with rectilinear lenses to those with a field of view not exceeding 90~100 degrees. Beyond this (and for critical applications, even below this) the image corners will remain of lower quality than the centre of the image. It important to understand that this is due to the optical constraints of using a simple, concentric dome port and has little to do with the camera lens. No matter how good the camera lens is, it cannot compensate for an inherent optical problem caused by using a simple concentric dome port!

The camera lens. Ideally any wide angle lens used behind a dome port should have a fixed entrance node which can be placed accurately at the centre of the dome, should have close focus optical correction and be able to focus close to the front of the dome without using a diopter whilst retaining its optical quality. This would mean using very complex optical designs and few if any such lenses exist. Most have an entrance node which shifts (albeit not much) during focus travel and most lose quality as they focus closer (this has long been a wide-angle design problem although some designs use ‘floating elements’ to try to overcome this). Add to this the fact that many manufacturers now supply ultra-wide-angle zooms which not only have an entrance node which moves during focus but also during zooming and which have varying degrees of close-range optical correction, and it is easy to see why it is difficult, and sometimes impossible, to optimise the lens/diopter/dome port to produce good result all the time.

Conclusions. So to optimise wide-angle lenses and dome ports it is essential to correctly align them. As stated, to do so it may be necessary to utilise a ‘spacer’ or extension tube between port and housing – Seacam produce such ‘spacers’ or extension rings which start at 25mm in length and increase in increments of 5mm (Seacam PVL25, 30, 35, etc.). Custom built ‘spacers’ can be made by Seacam for specific applications too, although these will be more expensive than off the shelf ones.

If a larger dome is used the need for a diopter is reduced because the virtual image is dependant on the diameter of the dome and with a large dome it may not be necessary to use a diopter as the virtual image may fall within the lenses’ normal focusing range. Many lenses may not need a diopter if used behind a Seacam Superdome, however some may benefit from a lower power diopter depending on their focus range, close focus correction and image quality. Smaller domes may necessitate a diopter being used on most lenses.  Generally speaking, larger domes such as the Superdome will work better with wide-angle zooms than smaller domes. Smaller domes may be a better travel alternative but should be used at smaller apertures whenever possible to maximise depth of field.

For best results a dome/extension ring should be set up for each lens being used. Diopters should be used if and when they are beneficial but in an ideal world lenses should be used which focus close enough so that they can focus on the virtual image without needing a diopter. Zooms can usually be optimised at one focal length (often the widest) and may not perform quite as well if used at other focal lengths. Smaller apertures will increase depth of field and retain better corner detail than wider apertures, especially with smaller domes.

Fisheye lenses will give better corrected corners and provided the distortion they produce is acceptable (there are few straight lines underwater so often distortion is not noticeable) then they may be a useful alternative to conventional (rectilinear) ultra-wide lenses for underwater use which remain problematic if corner detail is of great importance and shallow apertures are essential.

 Back to Info