I’ve been studying lighting and video/photography since the last post. At that time I was sure I had found the reason for the lack of red in the recorded video and photographs of paintings. Firstly, I was taking a ‘white balance’ reading from the blank canvas, assuming it was white. But there’s white, and there’s white. Taking a white balance reading is done to measure the colour of the light and allow the camera to deduct this from the image, giving a photo without a colour cast. Its essential that a pure white surface be used and this was the problem. A blank canvas is not pure white. Mine was slightly cream in colour. This meant the red and yellow (cream) were factored out by the processor in the camera. But this was only part of the problem. Continue reading
As any blogger who tries to upload photos or home made videos knows, what you finally see on your blog page never quite looks like what you intended. I constantly have trouble trying to get exposure and colour balance right on the photos and the video. I use the same lights, camera and computer software and still the difference between still photos and video can be enormous.
The exposure is simple, or should be. If the viewfinder is filled with an image of the blank white canvas, automatic exposure will underexpose the image. This is because the camera is programmed to find the average light reading and record it as an average. But the white is not the average. To find correct exposure I take a manual reading from a mid grey card, this is an average. When the white canvas is then placed in front of the camera, with this setting, it will show an overexposed image, that is, a pure white canvas. By the time the painting is finished, if there are equal amounts of darks and lights in the painting, the exposure is indicating correct exposure. If it’s a dark painting it will appear to be underexposed and if a bright painting it appears to be overexposed. This is OK, as dark paintings are supposed to be dark and light coloured paintings light. When viewed on a computer screen, in Photoshop or web browsers, the brightness of the screen or the room lighting where the screen is viewed will affect the apparent brightness of the picture.
Alexander made a comment, on the ‘Dollardstown Wood’ post, about the darkness of the photo. He was right. I had edited the photo in Photoshop in a darkish room. It looked OK to me at the time but when I read Alexander’s comment, I had a look at the post and the photo was, indeed, too dark. He adjusted the exposure and included a link to the photo (here). This brings me to the other problem with photos – colour balance. Although the exposure was correct the colour balance was too red.
I use iMovie to edit the video because its fast and simple to use. I set the colour balance using the white blank canvas. But still the video always looks a little too red. The still photo (at the start and end of the video), has the same colour balance and is close to the actual colour of the painting. But when placed into the video the colour is altered towards green. There are video editing facilities in iMovie but these are too crude to make the fine adjustments needed. To add further to the mess, YouTube compresses and causes further colour changes which are totally beyond my control. I archive the original video until the technology improves to allow a more accurate representation of the painting. Some day in the near future, hopefully, I will be able to make available the full 1 to 2 hour video. For a beginner in painting this might be useful.
Here’s the video. There is more info on this painting in the previous post.
The sweltering heat at hand
and along a stretch of beach
made hard the shadows
and the imperfections on her skin,
footprints on sand,
far from its reach
and its bathwater flows
erasing all but Adam’s sin
lapping at the edge of land.
The above picture is a combination of photography and digital painting. The text is part of the picture and I have included it in the event you cannot read it in the picture.
Its a comment on modern society. Especially the obsession with physical appearance which we see everywhere today. Films, TV, advertising, its everywhere. The human form is presented, flawless as a classical statue. The feeling of being physically inadequate must haunt the susceptible. Especially the young.
Even though we know that photographs can be ‘Photoshopped’ we subconsciously accept the image as real. I think this media was appropriate for this ‘painting’.
The elements in the picture are:
the beach – where the D-Day scene from the movie Saving Private Ryan was shot (Curracloe, Co. Wexford, Ireland)
the statue – Venus de Milo, even she suffered from the ravages of time
the beach litter – Leonardo’s da Vinci’s The Vitruvian Man. Vitruvius (born c. 80–70 BC, died after c. 15 BC) described the human figure as being the principal source of proportion among the Classical orders of architecture (Wikipedia).
The computer application used was Photoshop, appropriately.
When photographing buildings it is sometimes necessary to use a wide angle lens or the wide angle end of a zoom lens. This allows you to include much more of the scene than is possible with a standard focal length. The problem with this is that the verticals are distorted making the building look as if it is falling inwards.
I corrected the verticals in Photoshop using the distort feature in the edit/transform/distort menu. This is a correction which is best done ‘by eye’. Stretching horizontally of the upper parts of the photo alone appears to lower the height of the building so stretching upwards as well will give a more natural look. Leaving a little ‘leaning’ in the building also looks more natural because we expect this in a photo. Before the arrival of Photoshop a special camera with a lens on a bellows was used to correct verticals. This was done by tilting the lens but not the camera body so the correction was ‘in camera’.
The building is what is left of Duckett’ Grove. After extensive and sensitive restoration, by Carlow County Council, the revived walled gardens and wooded pleasure grounds are now open to the public.
In his book A Guide to Irish Country Houses Mark Bence-Jones describes Duckett’s Grove as a “square house of two and three storeys, transformed into a spectacular castellated Gothic fantasy by Thomas A. Cobden, of Carlow, for J.D. Duckett 1830. Numerous towers and turrets, round, square and octagonal. The walls enlivened with oriels and many canopied niches sheltering statues; more statues and busts in niches along the battlemented wall joining the house to a massive feudal yard gateway. The house was burnt in 1933 and is now a ruin”.
As the title of the post suggests, photography is a great resource for the painter. A photograph can be a work of art itself with the limitations of photographic optics often contributing to the artistic effect. One ‘limitation’ of the lens is ‘depth of field’. This onetime ‘limitation’ is very popular nowadays. The photo on the left illustrates ‘depth of field’. The flower is in focus and the background is out of focus. Whatever precise point the lens is focused on, a short distance in front of and behind is also ‘in focus’. This distance is controlled by the aperture setting in the lens (the F numbers, eg 4, 5.6, 8 etc.) and is the ‘depth of field’. A very high F number would have made the background sharp and, in this case, ruined the photograph. It is very much a photographically induced image.
And so to the point of this post. Using a photograph as a reference for a painting is very useful. But including this ‘photographic effect’ in a painting, is painting what a camera sees and not what a person sees. It is not one and the same thing. When I see this effect in a painting it seems to me that the artist copied the photograph and was not using the photograph as an ‘aid’ to producing the painting.
In a still life painting, for example, when the eye focusses on an object in the painting, the brain blurs the other parts of the painting and as one moves from item to item in the painting this is continually happening. Of course you have to ‘isolate’ objects but you do so by more subtle means than ‘photographically’ blurring other areas.
I think this is why a realist still life painting looks more ‘real’ than the photograph of the same objects. I have never heard the comment “I could almost touch these objects” in reference to a photograph but its a common comment when viewing a painting.
Do you remember the experiment, about 40 years ago, in cinematic projection which was called Cinerama. The recent efforts to introduce 3D in cinemas reminded me of this. Cinerama involved projecting the movie on a wide curved screen. 3 projectors were used, side by side, to produce the effect. Depending on where you were sitting in the cinema your entire vision was filled with the screen image. This, of course, was the problem – where you were sitting. Very few of those in the cinema got the full effect. If you were to the side of centre the screen was distorted. Why somebody didn’t think of this before the millions were spent converting cinemas and producing films for same, is a mystery. Of course the whole project was abandoned after a few years. I wonder will the same happen with 3D.
We were in Glenbeigh, a few years ago, looking north across the bay towards the Dingle Peninsula. The vastness of the scene reminded me of a Cinerama scene. Of course I took loads of photos but none captured the scene the way I would have liked. It needed a Cinerama treatment. The above photo was the result of my ‘experiment’ (Click on the photo to enlarge it).
This was the original photo with a Photoshop ‘Watercolour’ filter applied. By sampling the image I painted in the extra width rather than stretching the scene. When printed at about 4 feet wide it looks good as the ‘curved’ effect gives the feeling of been in the scene.
(Printed on an Epson Stylus Pro 9600 on matt paper)
My current camera is a Panasonic GH1. Having used a range of SLR and finally an Olympus E500 the Gh1 had a range of features such as HD video, swivel screen, 10x zoom and more, which the ‘normal’ DSLR did not have. One feature of the GH1 is its iA (intelligent auto). This is not something that the ‘serious’ camera user would use and so did not influence me one way or the other when deciding to acquire a Gh1. This feature, apparently, requires a lot of processing power and although the camera is relatively very small the space for this processor was found by not having the prism/mirror optical system found on other DSLRs. How it works is by analysing the scene and deciding which set of presets it will apply to take a good photo. And mighty good it is. For example, if I have a landscape framed in the viewfinder, the optimal settings for landscape are automatically set. If a person walks into the scene the settings change to ‘portrait in landscape’ and the appropriate changes occur in a fraction of a second. In this instance it finds the face and changes the focus to the face. Exposure, colour balance, sharpness and a host of other items are changed to facilitate the ‘portrait in landscape’. Likewise pointing the camera at a close object changes the camera to ‘macro’ with all the settings you would need to take a good close up shot. The only manual over-ride is the exposure and this leads me to the point of this post.
I learned photography in the era the ‘film camera’. There were no previews. You discovered if your photos were OK when you had the film processed and the prints made. Furthermore, there were a very limited number of shots on a film roll so you had little scope for extra ‘safety’ shots. Although the technology has changed and become very sophisticated the basic problems are the same.
Of all the auto-settings on the camera which needs to be altered the most, its exposure. You zoom to frame the shot, the focus happens automatically telling you what the camera has decided you need sharp, which is usually correct, the exposure is decided by taking the average of lights and darks and you are ready to release the shutter. You can decide beforehand if the calculations are based on the entire area or a particular part of the screen but the final ‘correct’ exposure is an averaging process. This is the problem.
Learn how to over-ride the assigned exposure settings or manually adjust the settings to achieve the effect you are looking for. In the photo above I underexposed the photo by 1 stop which created an underexposed foreground but correctly exposed the sunshine on the grass showing the shadows from the trees. Without making this adjustment the bright green grass would be ‘burnt out’ and the less important foreground correctly exposed.
You decide what is ‘correct exposure’ based on the information supplied by the exposure meter in the camera. Look at the presets, an example is the setting for ‘snow and beach scenes’. Snow, because of its brightness, will trick the camera into deciding the scene is very well lit so it will reduce exposure, darkening the colour of the snow to a mid grey. If you use the preset, the meter measures the exposure and then over-exposes by about 2 stops to produce white snow. This was an issue this winter as we, unusually, had an extended period of snow. When comparing photos I was asked how was my snow so bright compared to those of my fellow photographers. I over exposed the photos! The preset would have achieved the same effect but many of the ‘digital generation’ of photographers were not even aware that this could be done ‘in camera’ and not afterwards in the computer. There is only so much you can do in an application like Photoshop. ‘Burnt out’ or deep shadow areas of the picture have no details and really cannot be recovered.
This advice is aimed at the amateur casual photographers out there recording as JPGs and will probably seem daft to the experts recording in the RAW format.
I hope you derived some benefit from this little rant.