Following from my previous post I managed to get together a clip out of the 5 dives I did
The first day was somewhat plagued by visibility a bit lower than the norm for the location but the second day was fabulous
For this trip I brought with me the Inon UWL-H100 and the red push on filter from deep roof H20 and the Inon UCL-330 as I was expecting medium size fish and nothing really small
Here is the outcome
I am quite happy how things turned out so let me share the settings with you
First I shot most of the footage in 25p AVCHD mode only some small sections are shot at 50p and actually I did not need to slow down any of the material.
For the wide shots I used steady shot in normal mode and shutter priority at 1/50th. With the filter on I had auto white balance with tint correction G2 A1. I noticed that the camera was giving red tint in some situation and so added a bit of green back. The Amber correction instead is for the RX100 itself the camera does not have vibrant yellow and is a bit blue.
For the first time I use creative mode changing the standard contrast to -3 in order to prevent crushing of blacks.
Exposure was set all along to -1/3 and metering to multi area with AUTO ISO limited 160-800.
I think the results are so good that in fact I have performed no colour correction to any shots in ambient light.
For shots with lights I set up first colour temperature to 6500K and A1 to match the lights but then in some of the far shots this resulted a bit cold so I adjusted very slightly in post the temperature.
Only 40″ are adjusted in the whole video in essence is as shot and the editing took me half hour.
I did a bit of analysis and the camera was operating for most at ISO200 with aperture around f/3.5 – f/4.0 which is really the sweet spot of the lens.
I did have some challenges using the UCL330 for some nudibranches that really required a stronger lens so they look a bit small, there was also surge so I had to fight with focus problems but all in all very happy.
The longer working distance of the UCL330 (20-30 cm or 8″ to 1′) proved challenging on walls as you are too far to hang on to anything or use a stick. I will bear this in mind in the future.
For the close up shots I used steady shot active.
I also wanted to say that a few times the camera did manage to white balance properly however the results were not exciting and frankly not worth the hassle. Using the filter is just so much better with the auto white balance.
I am currently in Malta for few days relaxing and I manage to squeeze in some dives. The Mediterranean sea is nothing sensational (from a pure diving point of view) but does offer clear water, and some brisk thermoclines, and a combination of algae, blue water, caves and silvery fish that is challenging on the dynamic range of our little RX100.
The purpose of this trip is mostly to refine the video settings and go more in depth in few topics. I wanted to try specifically the following:
Medium size fish portraits
Caves and low light
Some of my settings will be the same and I am not intending to changed them those are:
Auto ISO: 160 – 800
Starting off with metering, the first attempt was to try and use the camera on 0 exposure compensation with centred weighted average metering.
Pretty soon I realised this gave issues of banding of the blue water, this was apparent not only in backlit shots but also in normal wide angle of fish in specific cases. So after dive number 1 I changed it to the standard -0.3 from dive 2.
I set a new creative style with contrast at -3 in the hope to recover detail and seem to be working fine with the shots still having plenty of contrast.
I also tried spot metering for close up but it makes no sense the video lights are too wide and ended up with burned highlights at the edges of the frame so back to centred weighted average for close up shots.
So when it comes to metering my settings are:
Wide angle: multi area
Close up: centered weighted average
I did some tests with stabiliser in steadyshot mode, this gives back some field of view and the lens offers 100 degrees diagonal and 90 horizontal, I actually think a bit more anyway with fairly stable conditions this worked fine. At longer focal length for close up I am still using active mode.
I have a +3 diopter for this test as I realised in my last still trip I don’t have a lens for medium size fish, the lens worked very well and I also tried the camera tracking focus but it seems it won’t work with fireworms or similar. So either keep normal focus or manual with peaking.
White balance has been a subject of discussion, I did manage this time to white balance a few times but to be honest it was not worth the effort there is a better correction of the purple hue of the filter but this can be corrected setting Green to 1 or 2 in AWB. I also changed the AWB to include a correction with Amber 1 as per examples. Very happy with the results I think this is the final set up with this filter lacking a proper orange filter.
Shooting in low light was rewarding with ISO maxed out at 800. I am becoming less and less a fan of video lights in cave due to the amount of backscatter am getting. I think I will default at using the lights as dive torches instead of wide beams of even leaving them off for effect.
The time finally came for our trip to Galapagos and I was ready with a set of think rubber suits to overcome the cold water. I just bought a 5mm Oneill Sector (wonderful suit) a 3mm hooded vest, kevlar gloves and a 5mm 4th Element short john. Well the short john was not used and for most I used just the wetsuit and a set of O’Neill Thermo X unders (very recommended) as El Nino is coming and the water was warmer at 25-27C or 77-80F.
This means lack of schooling hammerhead at shallow depth close to the reef, they were in the blue. We had however our good dose of schooling barracuda, tuna (in the 1000s) and even galapagos sharks
This is the video I produced
If you have problems with playing in your country use vimeo instead
The conditions were difficult with low visibility and in some sites very green water as you can see here
Questions will come about what I used etc etc so here is the list
Sony RX100II in Nauticam housing
FIx adapter for Inon LD mount
Deeproof blue water filter
Sola video lights 1200 (just few shots of the barracudas and some close ups)
Combination of arms and sometimes no arms
Ultralight tray TR-DM with extension and handles
The video has been produced with iMovie 9.0.4 no stabilization performed except the scene with the moray eels mating (very mild)
I shot the whole trip in shutter priority 1/50th of a second in the 25 fps 24 Mbps format. After some tests at home I have decided to use this mode as I can’t physically play the 50 fps files the camera produces on any of my devices differently. The 25 fps gives a film look and very smooth footage. This format is only available on the RX100II and not on the original RX100 so the consideration I made at the time for video settings of the RX100 remain valid.
The benefit of 1/50th shutter speed is a full f/stop of light the disbenefit is that at the surface and for backlit shots this is too slow, in those cases I go back to program mode or increase shutter speed to 1/100 or 1/200 or more.
For what concerns the editing I import the raw files in iMovie without using the conversion and then export with x264 using very slow preset and level 4.0 compatibility to use it with my appleTv.
I do minimal color correction in some cases I had to eliminate the red color cast of the filter when shooting upwards, to do that I change the white balance to top yellow until i remove the cast.
In some cases I had to put the green gain to the max for the same problem but in all other cases there is NO color correction in post.
When you work with AVCHD cameras the footage is compressed and the key is to get it right in camera.
As Galapagos conditions were challenging, although less than I expected, I used gloves in some dives to hang on to the barnacles and I also modified the set up to completely eliminate the lights.
I would dive with two 3″ segments and two 6″ Inon Mega float arms when I had the lights on and for other shots in ambient light I would have this set up here that looks odd but it is very effective and almost neutral at only 120 grams in water. The position of the floats means the camera stays upright at all times.
I would put a single Sola light on top of the housing not for video purpose but to signal the dive master when I was a bit far in the murky or dark waters. I used this set up on almost all blue water dives (Darwin and Wolf) and the normal set up with lights for the South and West sites with murky or green water.
There will be a separate post on the photos, I did a total of 18 dives with the camera of which 4 where photo and 14 video. You can see an example in the featured image.
I have now ran some statistics on the final clip that I produced
I used a total of 41 video clips:
ISO average was 273 with the following breakdown
2 clips at 160
30 clips at 200
6 clips at 400
3 clips at 800
f/stop average was 3.1 hyperfocal distance on average 1.16 meters. So if I were able to shoot without filter I would have not gained anything in terms of sharpness as most subjects were further away
I would consider the Galapagos and the dive trip I did not the best in terms of brightness and visibility so I would conclude that the RX100II with the Inon UWL-H100 and a red filter is the best set up for wide angle video in terms of performance, ease of use and flexibility.
Over a year ago I wrote a set of posts for the RX100 and some of the quirks of this wonderful little camera.
Steadyshot – aka Image Stabilizer
The RX100 has a specific Active mode for video not available when taking stills. I described the differences between those two modes in this post
Everyone is so obsessed of having the widest possible field of view that other more important considerations are completely missed out.
If you have ever shot a video with a GoPro underwater you know how bad is the quality of the image in the corners, this is because the flat port combined with the gopro lens create many optical aberrations.
Our RX100, especially the original Mark I, also has an issue in the corners, this is not just an underwater issue is also true on land. The lens on the camera has a lot of distortion and when corrections are applied to the image this effectively creates corner softness.
When we add a wide angle lens the image quality in the corners deteriorates further especially if the lens is flat creating a lot of chromatic aberrations that you can see in pictures with a blue or yellow halo around the edges.
Now the good news the Active steady shot mode crops the image of a factor of 1.15x getting rid of the majority of the corner softness.
There is of course a price to pay which is the loss of some of the angle of view. According to my calculation if you use an Inon UWL-H100 you start from more than the declared 100.8º more around 104º in fact. When the active mode is on this drops to around 95º. Remember all those values represent the largest incident angle that means the diagonal field of view.
A lens with 100º diagonal field of view means 90º horizontal. So after the active mode is engaged our horizontal field of view looks more like 84º which is equivalent to a 20mm lens. This is sufficient for most close wide angle shots and plenty for ambient light videos of large fish or wrecks. I generally suggest to keep the Active mode on, of course if you can be in a fixed position and hold the camera really steady you can also use the standard mode and obtain more field of view. There is a chance though that you will need to crop the extra field of view if you need to stabilize in the editing phase.
For macro shots without a tripod the steadyshot is a must and helps greatly. I do not even see a reason to take it off if you have a tripod as the RX100 does not have a particularly small capture area.
This brings the second subject: digital zoom, if you shoot pictures you avoid it as what you are doing is to crop the image, something you can do yourself in processing. In video though there is very little quality loss as we use just 2 megapixels of the 20 of the RX100 camera. In my test you can use digital zoom until the 2x multiplier is reached, this corresponds to 7.2x magnification and see no noticeable degradation in the image quality. The other benefit is that the depth of field is the same despite the magnification so you can save yourself stacking two diopters with all the difficulties that follow in terms of focus.
Digital zoom is always on in video mode and I recommend to use it with a single close up lens before embarking on dual diopters or a strong single diopter. Also take into account that with a single +6 diopter your working distance is around 4″ which is ideal for most critters except pygmy seahorse and bobtail squid or some shrimps.
In another post I have explained that getting hung up about light angle coverage is not really the only thing to consider, there is also luminous flux and quality of light. With my Sola 1200 I can cover something between 2 and 4 feet away with decent results further away is just back scatter. Generally this is ok for some close portrait work and close wide angle and of course not sufficient to cover part of a wreck or much larger subjects. You may decide not to bother at all with lights for wide angle and just render your deep wreck dives in black and white in this case consider that a pair of Sola Dive 800 at $399 are a high quality macro set up, function as dive light and provide some decent close wide angle portrait illumination. For macro shots you need much less than that, I set my fill light at minimum (300 lumens) and the main light at 2/3 which is 600 lumens, I can shoot at f/11 with this light intensity. You can see me shooting in the feature image.
I thought of writing a brief article about video lights.
In the last years LED lights have become incredibly cost effective and there are many products on sale that look promising and cheap.
A special mention goes to the GoPro a very cost effective option for underwater video plagued by poor performance in low light. The GoPro market seem to have injected new life into the video light marketplace and of course calls for products that are not expensive in line with the typical profile of the GoPro user.
There are many examples of footage with what you would consider powerful lights that actually do not look that bright or even nice, why is that?
Lights are now measured in Lumens which is the SI unit of measure of luminous flux a light is said to emit one lumen if it emits one candela over a solid angle of one steradian. This sounds all very complicated and not very helpful so let’s try to make some sense of this.
Usually when we go an buy a light we are given three key pieces of information:
Lumens – this in a way is considered the power of the light*
Beam angle – this is the maximum horizontal beam that the light can cover – note this does not relate with solid angles but is merely a 2D view
Color – this represent how white is the light, a good light should not exceed 6500K which is what is considered a day light white
*we will see that this is actually incorrect
Let’s have a look at a typical best buy the Archon Video Light W38VR, this light is declared to have the following:
120 degrees beam angle
Looking at the specifications this looks like a very warm (5000K) wide beam light and very powerful so how comes that with a subject at one meter or 3 feet distance this light seems to have no effect on a bright day?
The reason is that lumens are not a real indication of the amount of light that hits the subject but just the total light emitted which includes the light that illuminates the water between us and the subject that we actually are not interest at all in illuminating.
If you had to do a video production in studio you would be looking at intensity of 1000 lux where a lux is measured in lumens/m2. The lux is the density of light that hits a flat surface at a given distance.
With a few calculations we see that the Archon light because of the very wide beam is effective at two feet and produces around 1200 Lux which is great however at 3 feet this drops to 550 Lux which is not that great.
Many times when comparing the new cheaper lights to expensive lights like Sola or Fix despite the less nominal lumens declared the sola or fix seem brighter at the same distance.
So let’s evaluate the lux produced by a Sola 1200 a light much more expensive than the Archon with only a 60 degrees beam.
At two feet the sola produces around 4000 Lux which would nearly tan a fish, at 3 feet it still produces nearly 1800 Lux dropping to the studio value of 1000 Lux at a distance of 4 feet or 1.2 meters. This is because the illumination is in a smaller and narrower beam and is like the light was more dense.
Of course this has the disbenefit of a narrower beam so we need to step away from the subject in order to cover it properly, this diagram gives an idea
The two set of lines contain the video lights beams, we can see that those beams cross and them keep as combined angle of coverage the same beam of the lights themselves, that is the ideal spot for our lights as there are no shadows there.
Another consideration is the angle of coverage that we need for our camera lens to follow the lights, two 60 degrees beam lights actually cover a 100 degrees diagonal lens or 90 degrees horizontal lens typical of a 18mm lens. The GoPro and all compact cameras with a 100 degree lens fall in this group.
Another consideration to be done is that due to the light angle of coverage there is a minimum distance to be kept if we want to avoid our subject to be in the shade, this minimum distance depends on the light beam and the distance between them.
Once you get closer than one foot it is not necessary anymore to point the lights forward as the amount of water is limited and the lights can be pointed directly at the subject like in this example
So assuming we want to be at a distance of one foot or 30cm how far do the lights need to be?
For a 60 degrees beam the lights need to be 36 cm away or around 12 inches. For a light like the Archon with 120 degrees beam the distance between the arms needs to be 1 meters or 40″. Considering a tray is around 30 cm or 12″ this means you need arms at least 13″ in order to cover something at one foot and to cover a subject at 2 feet you need two meters distance that means again 33″ worth of arms.
What happens if you only have the tray with the lights on and your subject is at 3 feet with two very wide video lights of 120 degrees beam?
You illuminate already after 3″ from the camera
The two lights emit 550 Lux that combined are not such a small output
The net result is that a lot of the water in front of the subject is illuminated and this creates backscatter where all the silt in the water gets reflection and ultimately deteriorates the quality of the footage.
If you ever wonder why underwater photographers that shoot super wide have two 16″ arm segments this is why the want to avoid backscatter with their 110 degrees strobes and still be around one foot distance with their fisheye lens.
So what are the conclusions that we can draw:
A wide beam light is only good if you have long arms, the wider the beam the longer the arms otherwise the beams just light up the particles in front of the subject producing backscatter
Lumens are not a real indication of the effectiveness of the light per se but need to be taken into account with the beam angle and a lux calculation needs to be performed you want to aim for 1000 lux at the distance of choice
Light beams need to be evaluated together with the camera lens horizontal field of coverage 60-80 degree beam angles are adequate for most of the video rigs out there and more is only justified for fisheye type of lens not commonly used in video
I hope you find this information useful and get in touch if you want to discuss, the subject is not that simple!