Tag Archives: SONY RX100

Wide Angle Rectilinear Lenses for Underwater Imaging myths vs reality

The subject of rectilinear wide angle lenses and underwater optical performance has been beaten to death.

Most recently some photographers and videographers have done without rectilinear lenses altogether citing the horrible performance at the edge of the frame as the primary reason to shoot lenses with barrel distortion being those fisheye or standard lenses with an added water contact optic.

Is it all justified? Should you stay away from rectilinear lenses altogether?

Of course not. Rectilinear lenses have a place in underwater imaging that is there to stay but…

There are many many buts so let’s dive into some demistification and general considerations.

Topside Performance of Wide angle lenses 17mm use case

The vast majority of underwater shooters do not perform any type of topside imaging being that photo or video and use their set up only underwater with the exception of the odd event or those that like to shoot macro above and below this is where we stand in most cases.

In order to ascertain if the performance of the wide angle lens in water deteriorates it is appropriate to determine the performance of said lens topside. This is something that not many people are in fact able to ascertain. You can read equipment review but it is never the same thing.

I have recently invested in a Sony A1 and underwater housing and I decided to purchase a rectilinear wide angle lens. I always try to buy lenses that are good topside and underwater and my choice has fallen on the Tamron 17-28mm F/2.8 Di III RXD.

Several reasons for buying this lens I will list the most important are:

  1. Close minimum focus distance (19 cm wide – 26 cm tele)
  2. Lens does not extend when zooming
  3. Reasonably compact (99 mm and 420 grams)
  4. Not too wide

Items 1,2 above are important underwater and 3,4 are also important topside when you use the lens on a gimbal or when you shoot interiors or close up and you do not want apparent perspective distortion.

I have been shooting the Tamron topside on trips and I have been very pleased with it. Before taking it underwater I wanted to understand what to expect.

I built my scene using my underwater props on a table top and started taking a series of shots from f/8 to f/22 to see the results. What follows are a set of images taken on land with the focus on the eye of the chick.

Topside f/8 subject

At f/8 the first row of props is blurry and also the leaves behind are not sharp. The edges are soft.

This is due to lack of depth of field not to the bad performance of the lens.

Topside f/11 subject

At f/11 we have more depth of field however the props at the edges and the nearest part is still soft.

Topside f/16 subject

By f/16 all the props are in focus, the edges of the math are a tad soft but overall this is the right place to be.

Topside f/22 subject

By f/22 pretty much everything is in focus but the image quality has dropped considerably.

In conclusion topside we need to close down to f/16 to have all the props in focus on this scene.

Just so we are clear there is not an issue of edges or else there is simply lack of depth of field as we are shooting a close up.

Underwater Performance of Wide angle lenses 17mm use case

After rigging my camera and lens with the Nauticam 180mm dome it was time to hit the pool and try underwater.

What follows are a series of shots from f/8 to f/16 with focus on the chick as the topside shot.

Tamron 17mm f/8 Focus Mid

The image at f/8 looks very much like topside and lacks depth of field, however interestingly the bush behind the chick is relatively sharper to the topside scene. The dome port is increasing the depth of field behind the subject.

Tamron 17mm f/11 Focus Mid

By f/11 there is a considerable improvement all across the frame the items at the edges are still soft pretty much like the topside image.

Tamron 17mm f/16 FocusMid

By f/16 we are where we should be. Note how the entire set of props is in focus exactly like the topside image and only a slight deterioration on the very left of the frame prop.

Move your focus…

Dome ports increase the depth of field of the lens as infinity becomes nearer however they also shift the depth of field of the lens because the virtual image is closer and most of the depth of field shifts behind.

With that in mind I focussed on the first prop to see what happens.

Tamron 17mm f/8 Focus Near

At f/8 the first line is sharper even the props at the edges are substantially better and I would say acceptable however the leaves behind are out of focus.

Tamron 17mm f/11 Focus Near

At f/11 the edges are good and so are the props behind the chick with the exception of those really further back where we are running out of depth of field.

Tamron 17mm f/16 FocusNear

By f/16 focussing on the near prop we have pretty much everything in focus.

As final example this is a shot at f/11 with focus just behind the fake coral red and yellow on the right.

Tamron 17mm f/8 FocusSide

With the exception of the very very edge of the right frame which is bordering the dome edge the image is sharp throughout.

Why are the images not blurry?!?

You are now starting to ask why some images you see on the web look horrible and mine don’t? This is a legitimate question to which there are at least five answers.

Lens working distance

The Tamron 17-18mm has a working distance of 19 cm and is 9.9 cm long. From my calculation the entrance pupil is around 26mm from the front of the lens. This means that a dome radius of 10cm is sufficient to be able to focus right on the port. All other lenses for the Sony E-mount have working distance of 25 to 28 cm and would need extremely large radius to be able to focus on the glass.

Wrong Extension Ring

The Nauticam 18809 wide angle dome port is not a classic dome but has design without a flat base. The port has 11cm radius of curvature and is 83mm deep 180mm wide this means the entrance pupil needs to be 27mm behind the extension.

Nauticam 18809 180mm Wide angle dome port

Nauticam recommended extension is 40mm when combined with the 35.5mm N120 to N120 port adapter. Due to the shape of the lens this cannot be used with the N100 port as the zoom is close to the front of the lens.

Tamron 17-28mm with Nauticam zoom gear

Underwater housing manufactures unfortunately do not apply any science to the selection of domes and extension for a lens but out of pure coincidence the 40mm extension ring is what this lens requires.

Tamron 17-28mm 35.5 adapter and 40mm extension

Nauticam criteria for the 180mm port seems to be the lens needs to be more or less flush with the extension ring. In our case the entrance pupil is 26mm behind and the extension ring is a few mm over so we are more or less spot on. THIS IS ENTIRELY A COINCIDENCE!

With the 40mm extension the glass port will be exactly 11 cm from the entrance pupil and focus right on the surface.

Unfortunately the principles applied by manufacturers which are either to be flush with the ring or the dome mount or to extend until it vignettes generate loss of angle of view and distortion.

One of the worst offender is the Nauticam 230mm glass dome, where the rule of extend until you can generates pincushion distortion (edge pulling) effect.

Software Correction of Distortion

The lens correction has a warping effect on the edges even when the dome is correctly placed.

Exteme edge at f/11 with distortion correction off
Extreme edge at f/11 with distortion correction on

If you have a full frame camera or a camera where you can disable the correction make sure this is set that way.

For systems where lens correction is baked in the lens profile (micro four thirds) ensure to use a program that can disable the correction such as DxO Photolab for best results.

Constant Autofocus with subject tracking

As we have seen in most cases it is better not to focus bang on the subject but to focus closer or even at the side of the frame and work out the depth of field.

Many users use subject tracking that may seem a wonderful idea and works very well with a flat port however underwater results in blurred edges due to the depth of field distribution of the dome port unless you close the aperture until you can care less.

If you want to ensure the edges of your rectilinear wide angle shots are sharper use single autofocus and position the focus strategically in the frame the dome compression will do the rest.

Superwide lenses (<16mm)

When your lens is superwide you may have all effects on top of each other: pincushion distortion because your extension is calculated with the ‘go until vignette’ method, distortion correction in software, subject tracking, lens with long working distance and to make it worse perspective distortion and spherical aberration which occur when the lens is very very wide.

Choosing your rectilinear wide angle lens

I have computed all lenses avaialble for my A1 and calculated the ideal radius of a dome in this table. Unfortunately while the 230 port has a wider angle of view it has a radius of 12 cm which is only 1 cm more than the 180mm port. The choice of port is therefore driven by angle of view and not radius as 1 cm does not change much. Most lenses would need 15 or 16 cm radius to be able to focus close to the glass.

BrandModelWorking DistanceField of ViewRadius RequiredPort
SonySEL1224GM280122152230
SonySEL1224G 280122152230
SonyFE14 1.8 GM250114157230
SonySEL1635GM280107169180
SonySEL1635Z280107169180
SonyPZ1635G280107169180
Sigma1424DGDN280114164230
Tamron1728RXDIII19010499180
Zeiss18mm250100177250
Sony e-mount rectilinear lenses and ports

You can easily see that the Tamron 17-28mm has much better design characteristics to go into a dome port and this is the lens to buy.

From my tests there is no need of any field flatteners and correction lenses it works fine out of the box as the lens has minimal field of curvature and zero spherical aberrations.

Pool Session

If you are not happy with the CFWA studio scene here are some images from a recent pool session. Many f/8 and f/11 images no need to close the aperture more if you don’t have anything close and shot in single autofocus.

Skill training 17/8
Dave Side 17/8
Female Diver Side 17/8
Drysuit Diver Front 17/11
Maddy Portrait 17/11
Rush Diver Side 17/11
Manu Side 17/11
Dad Daughter 17/8

Quite interesting to see the fins at the edges of the f/8 shots.

If there is one challenge of rectilinear lenses is that you need strobe power. The narrower field of view compared to a fisheye means that you are standing further back that in turn means more strobe power and more particles between you and the subject.

Should I buy a rectilinear wide angle lens for underwater use?

My answer is definitely yes but each system will have the special lens, the one that focuses close, does not require a large radius and is not too wide and not too narrow (16-18 is ideal) however examine carefully images of others and check the lens construction as your suggested extension may be wrong.

But when your lens is fit for purpose using the correct size dome and the proper extension you will get high quality images that match or beat fisheye like lenses and water contact optics.

As example here an image shot with a canon 8-15mm at close range f/11. Does it have much better edges?

VideoDiver

And here a WWL-1 image at f/8

WWL-1 f/8

I cannot see any advantages of the fisheye like lenses at the edges at the same aperture in fact in both cases but judge for yourself.

Phil Rudin reviewed the Tamron 17-28mm on uwpmag you can find here some open water images https://www.uwpmag.com/?p=uwp-back-issues&issue=119

Unfortunately Phil used the Sea and Sea correction lens an expensive accessory that this lens does not require but as soon as I am in open water I will post images myself.

Canon 8-15mm with 4.33″ Acrylic Dome for Sony A1

Following from a previous article about not increasing bulk I have considered a few options for the Canon 8-15mm fisheye.

The 8-15mm is not a small lens and due to the different flange distance between Canon EF mount for DSLR (44mm) and Sony E-Mount (18mm) we have a chunky 35.5mm N100 to N120 adapter port that makes the whole set up not that compact.

Dome Options 140mm vs 4.33″

The Nauticam port chart recommends the 140mm glass fisheye dome for the 8.15mm, this port is 69mm radius and is made with anti reflective optical glass and weights 630 grams.

140mm Glass Dome on Scale

There is another dome from Nauticam the 4.33″ acrylic but this does not feature on the port chart for the Canon 8-15mm.

I did some calculations and this dome should require the same extension so I ordered one conscious that this would be lighter but not necessarily increase the underwater lift due to a reduced volume.

4.33″ dome weight

Although there is a difference of 362 grams the smaller volume will result in less buoyancy 348g lift vs 688g lift for the 140mm so overall the additional buoyancy is only 22 grams.

4.33″ vs 140mm

The primary benefit of this smaller dome is that it gets you closer this in turn means that things will look bigger and as consequence depth of field will drop. Depth of field depends on magnification and as you will get closer it will drop compared to other domes. So larger domes have more depth of field not because they are larger when you are at close range but simply because your camera focal plane is standing further back.

To give an idea this is a little miniature shot with the 140mm dome with the target touching the glass port.

140mm dome close up

This is the same target with the 4.33″ dome.

4.33 dome

Side by side shows the difference in magnification.

Left 4.33″ dome right 140mm dome

If we look at the same detail we can see that the 140mm dome image detail is less blurred.

4’33 dome vs 140mm dome

We are on land here there is no water involved and the 140mm image is sharper at the edge simply because it is smaller.

As depth of field must be compared at equal magnification we can also bust another myth of larger domes vs smaller domes there is no increased depth of field you are just standing further back if you compared the front of the port instead of the focal plane.

Building the Rig

The extension required is still 30mm as for the 140mm dome,

Acrylic dome profile

The overall size of this dome means it is flush with the extension ring.

Port details the lens hood must be removed

This is the overall rig with the amount of flotation in this image it is around 600 grams negative in fresh water.

4.33 rig

Now that we know what to expect is time to get in the pool and take some shots. I got some miniature aquarium fixtures to simulate a close focus wide angle situation.

Pool Session

Once in water I set up my artificial reef and got shooting.

I was at the point of touching the props so I had to stand back a little. As expected the issue is depth of field.

Shots at f/11

For starter we try to get as close as possible and focus in line with the chick.

Fisheye f/11 Focus on back

Due to the extreme magnification the front details are quite soft. So from here I start moving backwards a little.

Still focussed on the chick the sharpness improves due to reduced magnification this is a simulation of a larger dome.

Fisheye f/11 Focus on chick

There still is severe blurring of the front detail at f/11. However due to the increased depth of field that the dome brings behind the focus point the rest looks pretty good.

Focussing on the middle of the frame at f/11 results in blurry details for the features in the front of the frame but much less blurry than before and the chick is still relatively sharp.

fisheye f/11 Focus on edge front

Focussing on the pink reef detail results in a better overall result in a counterintuitive way.

Shots at f/16

Stopping down the lens results in increased depth of field so more of the image is in focus however the overall sharpness drops. This is a good place to be if you don’t want to be too sophisticated with the choice of focus point and you are close.

You can get closer but the front detail is still a bit soft but acceptable.

Fisheye f/16 Focus on back

If you move your focus point a bit further in front the situation improves.

fisheye f/16 Focus on middle

At this point I decided to get into the picture with a white balance slate.

Fisheye f/16 Focus on back diver

Although the front is quite blurry due to the extreme close range the result is acceptable for the non pixel peeper.

Shots at f/22

We are here hitting diffraction limit and the image looses sharpness but we are after depth of field so be it.

fisheye f/22 Focus on duck

Now the depth of field is there although the detail in the centre is less sharp.

fisheye f/22 Focus on middle

Moving the focus point makes the image a bit better.

Time to insert the diver in the frame.

Fisheye f/22 Focus on back diver

Overall ok not amazing consider the dome is on the parts.

Conclusion

The small acrylic dome does quite well at close range, the limitations come from the depth of field and not from the water and the dome increases the depth of field behind the focus point. This is something that you can use to your advantage if you remember when you are in open water.

For shots that are further away you can shoot at f/11 and get excellent IQ there is no need to stop down further to improve the edges. Consider however that f/8 may be just too wide on full frame and introduce additional aberrations regardless of depth of field.

VideoDiver at f/11

Some numbers:

Nauticam 140mm Glass dome: £911

Nauticam 4.33″ Acrylic dome: £550

Price difference £361 or 40% however bear in mind that the primary benefit of the glass dome is to resist reflections and ghosting due to the coating and the fact you can keep the 8-15mm hood on.

Fisheye Lenses for Underwater myths vs reality

There are several myths surrounding fisheye lenses when it comes to underwater use.

A quick tour to underwater photography guide usually a good source of information.

Myth Busting

Properties of fisheye lens as per UWPG:

  • They usually focus very close
  • They are small and light
  • When shooting ultra-wide angle, you benefit from a large depth of field
  • You can get very close to large subjects, maximizing color and sharpness
  • They perform well behind dome ports with good corner sharpness, and they don’t need a diopter
  • You usually need at least 2 strobes with good angle of coverage to properly light the entire area.

Some of the above statements are correct in absolute, some are correct but not specific to fisheye lenses and some are just incorrect.

Fisheye lenses usually focus very close -> true for the most recent fisheye lenses, not true for some older models

They are small and light -> Not true. Canon 8-15mm and Nikon 8/15mm are fairly chunky lenses with lots of glass

When shooting ultra-wide angle, you benefit from a large depth of field -> not a property of the fisheye lens but of the focal lens. In fact due to the extreme field of view Fisheye lenses have issues of depth of field.

This is a tea towel shot with a rectilinear lens. Note how sharp the target is at f/5.6

Rectilinear f/5.6

This is the same target at the same distance with the Canon 8-15mm at f/5.6 note how the edges are blurry and the blur starts very near centre.

fisheye f/5.6

You need to stop down the lens to f/16 to start getting coverage for the edges.

fisheye f/16

You can get very close to large subjects, maximizing color and sharpness -> This is a consequence of close working distance and wide field of view however sharpness is another story

As we have seen before fisheye shots at close distance are generally not that sharp especially at the edges.

They perform well behind dome ports with good corner sharpness, and they don’t need a diopter This happens to be true in practice and it is a major benefit for the underwater shooter

We will dive in detail in this topic.

You usually need at least 2 strobes with good angle of coverage to properly light the entire area. Fisheye lenses cover an aspect ratio wider than the format aspect ratio and result in limited vertical angle of coverage. Fisheye lenses are ideal for two strobes except the very far edges.

Canon 8-15mm specifications.

Horizontal field of view: 142 degrees

Vertical field of view: 91 degrees

A barrel gives an idea of the fisheye lens distortion

Let’s ignore the edges and assume we are a one meter.

Horizontal field of view 2*tan(71)=5.8 meters

Vertical field of view 2*tan(45.5)=2.03

Aspect Ratio = 2.85:1

The issue with fisheye lenses is that the frame is really very wide much wider than it is tall. This means some of the edges on the horizontal axis will be normally dark unless you are very very close.

Fisheye lenses and Dome Ports

A dome is simply a lens with a single element that has the property to retain the air field of view of a lens.

A dome is a lens with a lot of field of curvature simply because it is bent.

Using the dome port visualiser we can see that the effect of a dome is to bring the image closer to where it really is.

The net effect of a dome port is to increase the depth of field as infinity focus is reached much sooner.

A dome port has several side effects the main ones are:

  • Spherical aberration
  • Field of curvature

A fisheye lens works opposite to a dome. The centre of the frame is closer to the lens the edges are further away.

Domes, field of curvature and Fisheye lenses

In order to understant how the barrel distortion works in combination with a dome port and a fisheye lens we can build a small simulation in a light box where the edges of the frame are closer than a flat target.

Target in a lightbox focussed head on

We can see that despite the edges are quite blurry this image is actually better than our flat target.

f/11 centre

At f/11 the image is not perfect but we can see that most details off centre are not looking bad at all.

f/11 edge
f/11 detail crop

it is definitely blurry but not as bad as the tea towel as if the way the element are laid out improves the image in the corners.

And this is exactly the point: the items as laid out emulating the curvature of a dome improve the fisheye lens performance.

By f/16 the image is almost all sharp.

f/16 centre
f/16 edge
F/16 Centre 100%

One trick is not to focus in the back of the frame but find a middle point this means we can find additional depth of field in front of the target.

Focus mid way

Let’s see how this goes. at f/11 we already get some better results.

f/11 off centre
f/11 edge off centre

f/11 off centre crop

At f/16 we get some additional improvement but is not as major as the original f/16

f/16 off centre

Looking at the other areas there are some minor improvements but generally less as we close down the aperture.

f/16 off centre
f/16 off centre 100% crop detail

In conclusion the layout of the image elements helps the fisheye lens to achieve better image quality this can be futher improve focussing off centre however closing down the aperture results in the best results regardless.

In short we can improve an image at f/11 by shooting off centre in a strategic point to improve depth of field but ultimately aperture plays a bigger role in improving performance of the fisheye lens.

A similar reasoning can be applied to dome size vs closing down the aperture.

We can plot a scenario in the dome simulator tool.

In the starting example our aperture is 4cm to similate our 15mm lens at f/4.

6″ dome f/4 simulator

We now reduce the aperture to 2cm which is more or less f/8

6″ dome f/8 simulation

And finally to 1cm which is more of less f/14. In reality this is mm not cm but should make you understand that aperture matters more than anything else.

6″ dome f/14 smulation

What we can see is that by reducing the aperture the light rays passing through the dome converge and this means stray light is reduced and as consequence spherical aberrations are decreased.

Let’s now introduce dome size which is the equivalent of depth of field in the mix in our light box shooting off centre.

12″ dome f/8 simulation

We can see that with a double size dome the converging effect on the light rays is not as significant as the aperture is already small, but nonetheless is present. This is consistent with our f/11 off centre use case.

Finally at aperture completely closed.

12″ dome f/14 simulation

Although virtual distance has increased significantly the effect of the large dome on the stray rays is not significant here aperture rules.

What does all of the above mean?

I realise this was a bit geeky.

To summarise a dome has two issues one is spherical aberration for the very shape of the dome. This is mostly cured by closing down the aperture. Dome size has limited effect here unless you shoot wide open and with apertures from f/14 we can see that large dome vs small dome does not really matter.

However when it comes to field of curvature large dome helps the situation but because fisheye lens have barrel distortion and this has a counter effect to dome shape curvature therefore dome size matters much less to a fisheye lens than it would to a rectilinear lens.

Some additional insight in this post. And the summary finding here.

The takeaway message is this: stopping down the aperture improves field curvature and astigmatism somewhat, improves coma a lot, and improves spherical aberration most of all. The sum total of these effects changes our ‘area of best focus’, which is what we photographers really mean when we say ‘field curvature’. 

We could paraphrase this by saying:

A dome port increases depth of field and a fisheye lens, due to barrel distortion, benefits from a dome port. Optical aberrations introduced by the dome are mostly addressed by stopping down the aperture. The size of the dome port does not matter too much when using a fisheye lens and the benefit on aberrations of a much larger size dome is likely to be minimal when we look at that simulator. Focussing appropriately mitigates residual issues of field of curvature of the dome for the fisheye lens.

Underwater proof of concept

I took my Sony A1 with a Canon 8-15mm first and then with a WWL-1 that behaves very much like a fisheye lens.

Let’s have a look at some images shot with Nauticam 140mm dome.

The two buddies at f/8

The image above sees two buddies in the frame almost flat with their fins going back in the frame however the result is much better than the lightbox example as result of distance and dome port increasing field of view and adding curvature to bring the fins in.

This however does not resolve all issues if you focus near like in this example focussed on the eye of the croc

Focus on the eye at f/8

Here the eye is close resulting in the tail being blurred this is an effect of close distance and lack of depth of field despite the dome.

More interesting the nose is even more blurred as the dome brings that even close and blurs away due to field of curvature as the focus point is behind.

In this other example instead of focussing on the eye the focus goes mid frame so the fins are still in decent shape even if deep in the frame at f/8.

Focus midway

In order to prove the concept even more I took some props underwater.

First let’s have a look a shot at f/8 with the WWL-1.

Close up at f/8

As we can see the image is not too bad even in the close area but it is definitely better at f/11

Close up at f/11

What happens if we position the target off centre?

Contrary to our topside example the situation does not improve by focussing on the edge to further prove the issue here is NOT depth of field.

Focus off centre f/8

Here a detail crop the image is still fuzzy despite then focus is right on the spot. Depth of field is not the issue.

Edge focus at f/8

And finally we close down the aperture to f/11.

Edge at f/11

Crop at 100%

Edge at f/11

So here we can see that the underwater interface provides already for the depth of field but moving the focus at the edges does not have such a good effect.

Why? Because this is likely to do with aberrations of the lens itself as shown in my previous post on the Sony 28-60mm.

The combined 28mm with WWL-1 at f/8 means 20/8-2.5 mm aperture when stopped down to f/11 this becomes small enough to cure aberrations (less than 1cm with reduced field of view is sufficient).

For the same reason ASPC and MFT will be able to shoot at wider aperture not because of depth of field but due to smaller lens aperture.

15mm fisheye at f/14 –> 1.07mm physical aperture

8mm MFT fisheye at f/8 –> 1mm physical aperture

Again it is not the depth of field but the aperture size to cure most aberrations.

Conclusion

All Nauticam port chart recommend the 140mm dome and not larger domes. This is aligned with the theory behind this post that dome size ultimately matters but not as much as stopping down the lens and that fisheye are naturally helped by dome port geometry.

This conclusion also extends to water contact optics which are composed by a fisheye like demagnifier and an integrated dome port.

As long as the rear element of the lens is big enough the increased size of the lens does not result in proportional improvement of performance.

To support the empirical evidence of this article you can read this review of the 140mm dome by Alex Mustard.

By coincidence Alex recommends shooting at f/14 or f/16 which means a physical aperture of 1mm which cures all sorts of aberrations.

Considering that the benefit of a much larger dome may be as small as 1/2 to 2/3 aperture stops you may consider going the opposite way and get a very small dome which will result in additional spherical aberration and will need to be stopped down more when shooting very close.

If you use the Nauticam system there are only two ports that are a full emisphere and therefore able to contain a fisheye lens field of view:

  • 140mm optical glass fisheye port
  • 4.33″ acrylic dome port

I happen to own both those ports and in a future article will compare and contrast the two. I will also revisit the topic of dome ports and rectilinear lenses which is obviously different from fisheye lenses.

Getting the best underwater colours for your 4K Sony RX100 Mark IV

It is not a mystery that even the new Mark IV version has issues with custom white balance.

The ergonomics have not changed and you need to go into photo mode to set custom white balance but generally underwater results are poor. Using filters is therefore a necessity also on the new 4K version.

RX100 Mark IV Video Behavior

The RX100 offers now a 4K 100 mbps mode and can use picture profiles.

I have used a modified version of PP6 that use the cine2 gamma curve, I have however changed the colour to the Pro mode and changed a number of other settings in my last video in Puerto Galera.

The water was green and murky but this gives you an idea of what you can get.

Filter Options and Wide Angle

Although the Nauticam WWL-1 is the best lens for the RX100 it does not take filters and therefore is not adequate for video.

In this review clip you can see the options available on the market.

In terms of wide angle you have two options for 4K:

  1. Inon UWL-H100
  2. Inon UWL-100

Both lenses work fine in 4K however the older UWL-100 achromat does vignette in photo mode.

The UWL-H100 offers a very wide field of view also in HD mode with no vignette and accepts the mangrove/deeproof filter.

This filters is loaded with magenta so I suggest adjusting the tint in the auto white balance mode to +2 green.

The UWL-100 works fine in 4K and is wider than the UWL-H100 however has only the M67 mount. If you have one of those lenses you can use the Ikelite 6442 filter. This filter required you to remove the rubber ring on the lens and does work quite well except has a yellow cast to it you can reduce by changing the tint to +2 blue and increasing also magenta to +1.

For flexibility purposes probably the UWL-H100 is better as it takes the bayonet but the UWL-100 is really wide and has a little less fringing. Some people do like the UR/PRO filters better.

I hope you find this post useful and good luck with getting the best colours from your Sony RX100 Mark IV

Nauticam bayonet mount for wet lenses

Nauticam entered the wet lenses market with their SMC close up wet lens that was optimized for DSLR.

Then it released the CMC compact macro converter for compact cameras and micro four thirds and finally the Wet Wide Angle Lens I that is compatible with compacts, micro four thirds and also full frame cameras with 28mm equivalent lens.

Up to now all lenses were using the traditional M67 mount as most of the lenses, even the close up ones, are pretty heavy this means going for the dive with the same lens. Nauticam has developed the flip diopter adapter for flat ports to overcome this issue.

Flip Diopter on Nauticam RX100 IV
Flip Diopter on Nauticam RX100 IV

The flip diopter is a good solution for micro four thirds and DSLR but looks rather cumbersome on compacts as the image shows.

I asked Nauticam for a bayonet adapter and specifically if they could develop something for the Inon LD bayonet system that so far has been the reference for wet lenses for compacts and micro four thirds cameras.

LD mount converter on RX100 IV
LD mount converter on RX100 IV

Edward told me that due to the fact that the WWL-1 lens rear element is so large the Inon LD system was not an option so they went off and developed their own system.

M67 bayonet mount converter
M67 bayonet mount converter

I would like to thank Nauticam again for making those parts available before general availability.

Looking a bit closer to it you can see that due to the specific construction with two concentric rings you need a special tool to apply the adapter on the port.

M67 bayonet mount converter The large item is to attache the mount to the port
M67 bayonet mount converter
The large item is to attache the mount to the port

Obviously as the Nauticam lenses use an M67 thread new adapter needed to be developed.

Mount converter for CMC/SMC
Mount converter for CMC/SMC

Nauticam does not use ABS plastic and uses aluminum for all their parts.

Now that the items have a bayonet adapter there is a need for a lens holder to put on the arms.

Lens holder looks too big for a 5" arm segment
Lens holder looks too big for a 5″ arm segment

The lens holder is too big for a standard 5″ segment but looks in proportion with a longer segment.

Lens holder on 8" arm segment
Lens holder on 8″ arm segment

The adapter is larger than the LD mount and a bit big for compacts to the point that even with a tray the adapter tips the rig back.

Another challenge is that this system is designed for Nauticam lens that have protruding rear element so when used with standard lenses there is a gap between the port and the wet lens that can be counter productive, not the end of the world and frankly the Inon system has the same problem. This however means that if you wanted to use this system with a different wide angle wet lens this would be suboptimal.

I am waiting for Nauticam to ship me back the WWL-1 so I can show how that lens performs on this system.

Another observation of course is that if you use this system for wide angle the super heavy WWL-1 and the fact that the adapter only works on a normal segment means your rig will be very heavy in water. I am going to discuss with Nauticam the possibility to have the adapter on a float arm however their carbon arms do not have any mounting point to be used.

Stay tuned for a full review of this adapter with the new 29 macro port that looks very promising for video.

Demystifying video formats

Youtube now supports double frame rate video 50p and 60p so what?

That is actually a legitimate question look at this example here which is a short clip from a trip to Barbados in 2013, this was originally shot on a Sony RX100 Mark II in AVCHD progressive 1080@50p 28 Mbps

If you don’t see the 50p option is because your browser operating system does not support it. You need the latest version of browser and operating system and a machine fast enough plus enough bandwidth. So for Mac this means OS X Yosemite and Safari and for Windows you need 8.1 and IE9.

I hope you enjoyed the clip now check this other one which is a instead shot at 25p with the same camera at 24 Mbps/

I think you can see by yourself which one looks better and it is the 25p clip despite an overall lower bitrate.

There are a number of reasons:

  1. Underwater clips do not have a lot of action as you may think so extra frames go a bit to waste
  2. The encoding which is how the clip is first recorded by the camera is not really that different.
  3. The human eye does a great job at interpolating missing frames anyway
  4. There is not really much more data in the 50p file compared to its 30p rendition
  5. The image quality if you look at a still frame is better in the 25p clip.

There are of course benefits in shooting at double frame rate if you want to slow down the footage 50% speed but for what concerns your normal shooting you would say for that clip you could not tell.

Let’s think about it in simple terms if you have a clip shot at 25p with 24 Mbps you would expect something not quite double but a bit more for 50p instead you only have 28 Mbps. To be more precise you have 22 Mbps vs 26 Mbps video which is 18% more in Sony’s case. So that is not really much information more.

What is more interesting is the structure of the data what follows now is a bit technical but bear with me.

GOPStructure
GOP Structure Row 1 and 3 Sony AVCHD 25p and AVDHD progressive 50p

The first and third rows are representation of Sony 25p and 50p clips. The green bar are I frames that you can think of like a JPEG image, the red bars are P frames or prediction the only contain a delta from the previous frame not a full image.

You can see that in the first row there are 12 P red bar between each green I bar. This means that the GOP or group of picture is composed of a sequence IPPPPPPPPPPPP that repeats indefinitely.

On the 3 row there is a representation of a Sony 50p clip you can see that now there are 23 P frames between two I frames.

So the increment in full frame is limited however if we look at the sizes we see that the I frames in the 25p clip are 12% bigger and also the P frames are smaller.

So in short if you look at the image quality the 25p clip has more information in the full frames as well as for the predicted frames whilst the 50p clips has more frames but overall with less quality.

Which means that unless you are shooting something that is really action packed or you want to do slow motion there is no actual benefit but instead a deterioration when you shoot AVCHD progressive underwater.

Note: if instead we were shooting at higher bitrates for the 50p the story would be different but at similar bitrate it goes as above.

You will also have noticed stream 2 and 4 in the image above I repeat them again here

GOPStructure
GOP Structure Panasonic AVCHD 25p and AVCHD progressive 50p

The second and fourth stream are generated by a Panasonic camera and they look different. You will notice now the existence of frames with the tag B and also that some of the P frames have a green slice.

This means that Panasonic AVCHD implementation has two features that Sony does not have:

1. It has B frames which not only predict future frames from the past frames but can also reference future frames in the prediction (sounds crazy but it works basically the frames are stored in memory before past ones are saved)

2. It has slices for images so on one frame there can be an element of prediction from a previous frame and another element completely newly generated for example if the prediction was completely in a part of picture where there was a lot of movement.

H264 encoding has motion compensation so things that do not change are referenced and new parts are predicted or in this case partially created from scratch.

So the Panasonic encoding algorithm is much superior to the Sony one for AVCHD this explains why a small camera like the Panasonic LX7 could produce video to compete with a larger sensor RX100 with almost double number of megapixels.

What makes me laugh is when photography magazines jump to conclusion on the video quality of a camera shooting a static frame!

Of course if there is not movement the camera with the best IQ in still pictures will prevail however when you record motion all of that becomes somewhat less relevant as compression impacts the quality.

So the more effective compression algorithm of Panasonic beats Sony to the point that even a larger sensor size seems not to matter.

This explains why when you take a real life clip Panasonic cameras perform better in video despite a worst image quality in still images.

The difference between the 28 Mbps and 24 Mbps follows pretty much the same trend of the Sony clips there is not enough bitrate to justify the double frame rate unless there is a lot of action in your clip.

So to conclude if you are shooting AVCHD the normal 24/25p more will have better image quality and will be more suitable to scenes with a lot of dynamic range, will give more colour and contrast. If there is really a lot of action or you want to slow down the clip shoot in 50p bearing in mind that image quality will actually drop if you look at a still frame in isolation.

Underwater contrary to what you may expect things do not actually go that fast and most of the movement is in a specific part of the frame or in a limited part of it so AVCHD 24/25p gives better results.

Finally when looking at a camera for video check for real clips do not look at resolution charts designed for still images as they give very little indications on the quality of your videos. Also if there are any tests make sure those are on the JPEG images that share similar processing engine not on RAW files are you are not shooting RAW video. And finally consider that at similar bitrate some manufacturer have a clear edge on others when it comes to real time compression in our example Panasonic produces similar quality to a Sony camera with overall a better sensor but poorer compression.

The end of advanced compact cameras for underwater photography

With the arrival of the new Nauticam N50 port system for advanced compacts I believe we say goodbye to using a large sensor compact underwater.

We already saw some signs when Sony released the RX100 Mark III with a 24-70mm equivalent lens, this effectively killed any use of the camera for macro. The new Panasonic LX100 has a similar problem with a 24-75mm lens but on top it has the new N50 port system in the nauticam housing or a severely limited functionality in the Ikelite version. It is obvious that the Panasonic LX100 will have similar sever limitations for macro but also with only 10 Megapixels really not be a strong competitor.

The last kid on the block is the Canon G7X, this camera looks very similar to the Sony RX100 Mark II and has a 24-100 zoom range that on land is very useful with an aperture of f/2.8 at tele end that is really interesting for land use. However this camera in the Nauticam version has the same port system N50.

So why has Nauticam introduced this? Both the Canon G7X and the Panasonic LX100 have fairly long lenses in terms of physical size and the zoom mechanism is such that the lens is retracted at wide end. This means that there is quite a gap between the lens and the port and wet lenses in those conditions have issues of both corner softness and chromatic aberrations. So Nauticam has introduced a short port with zoom blocked so that wet lenses can be used.

Effectively this is like diving with a fixed 18mm lens on your camera behind a dome with no zoom capability, taking the wet lens off results in horrible pincushion distortion so not really appealing.

Pincushion Distortion at 24mm with flat port
Pincushion Distortion at 24mm with flat port

The dome port does not allow the use of the full zoom at least without a diopter, so this is of limited use too for video.

Cost wise the new housing with the port system are around $1,200 in the US with additional $350 for the dome and $180 for the flat wide port. We are approaching the low end of Micro 4/3 Nauticam housing prices and are more expensive than Olympus OEM options but without the same flexibility and quality in terms of lenses.

It is clear that the Sony RX100 Mark I and II will remain one of a kind and this is the reason why they remain popular few years after launch.

Sony RX100 Mark II Rig June 2014 Front
Sony RX100 Mark II Rig June 2014 Front

 

Introducing the Sales Corner

I receive so many queries about equipment that I thought I would put some of my items on sale.

My cupboard is near enough to explode and I need to streamline the equipment I have. Currently I have 3 strobes, a huge selection of ball and clamp arms, 3/4 and 1/2 locline arms, float arms, floats, video lights, tray bases, around 15 clamps and some 6 macro wet lenses and 3 wide angle wet lenses with 3 different mounts.

I guess it is time for a clear up so I will add a section with items on sale where I will list some of the combination items.

The idea is to provide a full set of lenses or a tray + float solution as those seem to be the most asked for items in my set up.

So if you look on the menu you will see the link Sales items with details of the various sets. I can also sell the items individually but I think this makes little sense as I’d rather do that on ebay.

Many people read my posts and then go to gear shops but there there is always a little mistake or problem getting the whole set up from here you avoid that. I priced the items based on 50% of retail UK price and I would ship to UK and Europe, other overseas will attract import duty so it is not worth it in my opinion.

If there is anything you like  use the contact form on the relative page to inquiry

Sony RX100 Mark III – Is it worth upgrading?

UPDATED 23 July 2014

Once again Sony has updated their RX100 camera with a stunning new release that will surely be a market leader at least until the release of the new promised Panasonic large compact sensor with 4K video recording.

So the question is again is it worth throwing our hard-earned money to this new model and housing or should we stay with what we have got? The Mark I and II are still available at reduced price though it is becoming harder to find Mark I housing as new.

The answer to the question is: it depends on what you are planning to do with your RX100, for some users an upgrade may not be required or even not advised let’s see why. no don’t waste any time with the Mark III.

Every time a new camera is released and reviewed I get a bunch of emails with subject: New Camera XYZ what do you think?

The best reviews you find online are made by sites that specialize in land photography and no consideration is given to underwater use. So not necessarily a camera that is a top performer on land will remain such underwater as this depends on specific characteristics that may be different or even opposite to land requirements.

So those reviews cannot be taken as they are they need interpretation. Personally I use two sites for camera reviews imaging resource and dxo mark I use the first to understand ergonomics, performance and to compare images with other cameras side to side, and the second to check sensor quality and lenses. Imaging resource has made a good article to compare the various RX100 on land read it here

http://www.imaging-resource.com/news/2014/06/16/battle-of-the-rx100s-sonys-rx100-vs-rx100-ii-vs-rx100-iii

If you look at a sensor comparison on DXOMark you would conclude that there is no need to upgrade at all if you use your camera primarily for still pictures as the sensor are practically identical in performance as this table demonstrates:

RX100 Sensor Ratings
RX100 Sensor Ratings

As you can see the differences in dynamic range, color depth or ISO are pretty much intangible.

However there are other metrics that are also important let’s see which ones and why.

Flash Recycling

We want our camera on board flash to recycle quickly after a full dump as the RX100 has only TTL flash and the flash can’t set to a minimum manual setting without consequences. Those are the flash recycling times at full output (a full dump will always occur when you don’t aim strobes directly at the subject for example wide angle)

Mark I: 7.2 seconds

Mark II” 4.4 seconds

Mark III: 3.3 seconds

The Mark III is pretty quick and the Mark II is acceptable the first release is definitely too slow.

Minimum Capture Area & Zoom

What is the smallest area that can be captured at wide end at the closest focus distance? This gives you an idea of out of the box close up performance

I do not have precise metrics yet but looking at comparable images on imaging-resource seems like the Mark III is worse than the Mark I and II of at least 20% with a capture width around 90mm versus the already not very good 76mm of the Mark I and II.

To fix this issue we use close up lenses in water that fix to a great extent the focal length and then zoom in to achieve magnification which means longer camera lens more magnification I roughly estimate that the Mark III will be worse of a factor of 0.7x so things will look 1.4x bigger with the Mark III this is terrible news for macro shooters as it means you need to be on top of the subject to fill the frame, this is in some cases not possible.

I have estimated that you need 11 diopters to achieve 1:1 macro so the Subsee 10 and Inon UCL100 that gave real macro on the Mark I and II won’t be sufficient. With two UCL-165 or Dyron 7 you are looking at 2.58″ or 6.5 cm from the back of the first close up lens that means you will be right on top of the subject which is not really an option.

Update July 18th: Nauticam has confirmed that even with their SMC lens the most powerful diopter on the market the capture area is 38mm wide so does NOT achieve 1:1 macro as I suggested…

 

Lens Focal Range

The new mark III has a 24-70mm lens compared to the 28-100mm of the Mark I and II what does this mean?

  • Less magnification with close up lenses and no real 1:1 macro
  • Vignetting or even not possible to use fisheye lenses designed for 28mm lenses

This means that with the Mark III your scene selection will be restricted  compared to the 150 to 24 of the mark I and II.

Update 23 July I have done some tests that confirm my suspicions please look at the following frames. The Mark III once zoom to 28mm actually works fine with the Mark II housing except the power button.

RX100 Mark III with Inon UCL-165 Close up lens
RX100 Mark III with Inon UCL-165 Close up lens

With a single Inon UCL-165 (+6.06 in water) the minimum capture area width is 5.4 cm which is bigger than the Mark II 4.8 cm. Not only that the distance from the top of the lens is only 9cm.

With stacked UCL-165+UCL330 total power 9.09 diopter we get this

Stacked UCL-165+UCL330
Stacked UCL-165+UCL330

4.4 cm width however we start running into problems as the minimum distance from the lens is only 4cm.

We finally achieve 1:1 with two stacked UCL-165 however the distance from the top of the lens is 3cm definitely too little

two UCL-165 stacked
two UCL-165 stacked

I have also tried the Mark III at 28mm and it still vignettes with the Inon UWL-H100 in air until around 30mm. Instead the old UWL-100 28AD is fine at 28mm either way image looks narrower than with the Mark II but this may be an issue with the old housing.

So basically no macro with the Mark III and no decent fisheye as expected.

Underwater Photography Conclusion

Due to the reasons above the best camera for the job is the RX100 Mark II, with the Mark I coming close but being penalized by the strobe recycle time, the Mark III really is not an option for the serious shooter due to limitations of the lens.

Let’s have a look at some other features of the Mark III that are not relevant for stills as much as they are for video.

ND filter

Shooting at the surface with plenty of ambient light or on land can create problems if you want to follow the 180 shutter rule, the ND filter of the Mark III ensures your footage will be smooth on land and in water. Take into account that on land you can apply optical ND filters (I have it for the Mark II) so this is really for underwater use when you are shooting at 1/50 or 1/60 shutter speed and the scene is too bright when you hit f/11 and the lowest possible ISO. Having now tested this feature I confirm it is really valid on land on a bright day to keep the aperture wide.

Clear Image Zoom

This is a special digital zoom with edge enhancement that only worked on JPEG stills on Mark I and II but is now available on video on the mark III. This means an additional 2x zoom is available bringing the focal length to 140mm that is more than adequate for macro. This is not an option on RAW images so irrelevant for still users.

23 July update Having done some tests for macro video (as stacking two diopters is not an option) the image quality suffers but anyway this is all you are left with lacking any other options.

Clear image zoom 1.5x on single UCL-165
Clear image zoom 1.5x on single UCL-165

Image stabilizer

The Mark III adds a further stabilizer mode with additional crop that takes the lens to 96mm equivalent in video mode. This is well worth for super macro hand-held footage and effectively provides an option where you can either use the super steady shot and no clear image zoom when hand-held or leave on only the optical stabilizer and use clear image zoom. If you shoot macro on walls and not on flat sand this is a well worth feature.

Update 23 July 2014

Following a side by side comparison you can find on this link

It follows that there is really no 5 axis stabilisation and only an additional mode with more crop with correction for rolling shutter that anyway does not really work!

Full Sensor Readout Video

The mark III reads the whole sensor not just alternate lines when capturing video this increases should increase the resolution to a great degree and increases the perceived resolution and clarity of your videos. However this does not seem to correspond to real life tests. See this sample with a side by side comparison the Mark III looks visibly worse than its predecessor even on the higher bitrate XAVC!!!

Video Formats

The Mark II introduced a well welcomes 24 Mbps 24/25 fps mode now the Mark III introduces a 50 Mbps mode in XAVC (that can be rewrapped into normal MP4) that produces even better footage although it will be quite demanding on your workstation and most likely some of your home devices will struggle playing the files (my Apple Tv 3rd generation only plays up to 25 Mbps) when uploading files online this will downgrade to 8 Mbps so won’t matter unless you have a good player at home. improves nothing and requires manipulation as the files are actually not MP4 compliant as they have WAV audio (!) they are also inside the AVCHD folder which is painful. As seen above seems like the encoding has got worse on the AVCHD files and only the higher bitrate compares but still does not match the mark II at the lower 24 Mbps!!!

HDMI connection

You can now connect the Mark III to an external HDMI recorder if you are into that kind of semi pro usage. For the normal user this is not relevant: if you are into spending that money probably you have already a DSLR or a Panasonic GH4.

Video Auto Focus

The auto focus in video has got worse you can see an example in the stabiliser test, and actually plenty of frames are blurred on the mark III. Awful!

Underwater Video Conclusion

The Mark I with only a 28 Mbps 50/60p mode is definitely a poor choice, the Mark II is acceptable to most users. and the mark III promises even better performance if you are a serious video user and have another camera to take pictures the Mark III may be well worth it.

The Mark III despite some useful feature is actually a let down and perform worse in practical terms than the mark II that remains my camera of choice.

Summary

If you have a Mark I camera and are frustrated with flash recycling you should be looking at a cheaper upgrade to the Mark II.

If you have a Mark II and your priority is photography upgrading is NOT recommended.

If you don’t have any RX100 get yourself a Mark II or wait for the new Panasonic as the Mark III is pretty much a pile of crap.

Is it the first time I am actually sending back a camera and invoking consumer rights but when I spoke to the camera shop they told me they had other returns and this is not really an upgrade!!!

Look Right inThe above image was shot at 100mm equivalent with one diopter with the RX100 Mark II. You won’t be able to fill the frame with the same close up lens and the Mark III camera

Snell Windows with your compact camera

Snell’s window is a phenomenon by which an underwater viewer sees everything above the surface through a cone of light of width of about 96 degrees [From Martin Edge: Underwater Photography].

During my last Gapapagos Trip there was a bit of an issue in terms of photography subjects, in essence most of them were fairly big like sharks, turtles, eagle rays and of course divers.

After a while shooting or trying to shoot those uncooperative models I had the idea to try and do different things such as silhouettes and Snell windows. You can also combine both as we did in this case.

So what do you need to shoot image of a Snell window:

1. A lens that is wide enough, you need a bit more than 96º field of view to take the whole window

2. Something interesting on the surface (if you have calm water you can see right through)

3. An interesting subject as silhouette to contrast the clear water if there are waves and you can’t easily see through.

A normal flat wet wide angle lens for our compact camera is not sufficient to capture the snell’s windows as in this example

Safety Stop
Divers at Safety Stop – taken with Inon UWL-H100 at wide end

Take into account that even with a fisheye lens on an SLR you can’t capture the Snell’s window on the vertical axis as the field of view will fall short a few degrees.

With a compact camera a semifisheye lens will capture the edges of the window if correctly aimed on the diagonal and horizontal dimensions and will fall short on the vertical.

Snell to boat
Snell window divers exiting the water or ascending – UWL-04 semi fisheye lens Canon S95

I am reporting some of the calculated field of view for the most popular lenses at present for compact at 3:2 image format

Wet Lens Diag FOV Hor Fov Ver FOV Flare/Ghost
UWL-04 151.95 123.22 79.85 Ghost Possible
UWL-100 28AD 149.01 120.97 78.48 No
UWL-28 160.56 129.75 83.81 Both
UWL-H100 157.85 127.71 82.58 No

 

All those lenses will take a good Snell’s window I have not tried the UWL-28 and I have some concerns this lens may flare as it usually does but I do not know for sure.

Both the Inon lenses produce sharp images with no defect.

In general small sensor cameras like the Canon S series will work fine with the fix/idas lenses and should produce good results also of normal shots through the water.

Using the UWL-04 or UWL-28 with large sensor cameras like the RX100 there could be a sharpness issue a larger apertures so make sure you close that at f/8 or smaller instead of using the shutter speed to balance exposure if you want to go through the water.

Snell’s windows are uncommon with compact camera shooters as they are considered an advanced subject but they can be taken. I will be taking more in a next trip to Egypt I hope for calm surface conditions so I can go through the surface with the shots. Failing that this is very easy to practice in a swimming pool.

Have fun shooting snell’s windows!