STEVE PEARCE PHOTOGRAPHY
One of the first decisions after taking the plunge and buying a high end digital camera, is which file format to save in. All digital cameras allow a choice of image sizes, but this discussion is concerned with the choice of file type to be used. Most digital SLRs and some high end fixed lens cameras give a RAW file format option in addition to various levels of JPEG and sometimes TIFF.
Generally speaking, each file type has its pros and cons which can be summarised as follows:
|
File Type |
Advantages |
Disadvantages |
| Normal Quality JPEG 1.2MB* |
Small file size, processed and fast-save to memory card. | High JPEG Compression can
delete detail in areas of similar tone/brightness. White balance selected in Camera. |
| Fine Quality JPEG 2.4MB* |
Largest JPEG file size (but still smaller than RAW), still saved reasonably fast. | JPEG Compression can delete
detail in areas of similar tone/brightness. Take up more memory card than low/medium JPEG. White balance selected in Camera. |
| TIFF | Lossless compression format – even in areas of similar tone /brightness, detail will not be lost. | Normally slow to be processed
by camera due to large file size – so take up large amounts of memory
card. White balance selected in Camera. |
| RAW 6MB* |
Lossless format All data captured by sensor saved Image settings can be applied on PC (such as image sharpening, contrast etc). White balance can be chosen later on computer. |
Normally slow to be processed
by camera due to large file size which takes large amounts of memory card
space. Optimising images on PC takes time. |
*Approximate Canon 10D file sizes.
It can be seen from the table, that even though JPEG is a lossy format, there are advantages. It is fast and keeps file sizes to a minimum. The compromise can be the lower quality of the image or getting an image at all.
For a discussion on colour balance and to see the advantages of applying the colour balance later using a computer, see my other article on this site – How Hot is the Light.
The images below show the effects of JPEG compression where an image was purposefully shot underexposed (left). The centre image shows the result when the Low quality JPEG version is taken into an image editor and the levels have been adjusted to try to save the image. You can see that this has not been totally successful. The reason is, that the compression algorithms have examined the image and seen dark areas, then assumed that the detail in these areas is unimportant and cannot be perceived by the eye. The details in the deep shadow areas around Tigger’s feet have therefore been deleted – resulting in a smaller file size. The underexposure has also caused an incorrect colour temperature – leaving a colour cast which is difficult to correct.
The image on the right was saved as a RAW file – so no detail captured by the sensor was lost. Without editing, the image also appeared very dark and under-exposed, but when taking the image into Capture One for development, it is possible to alter the levels so that the detail can be seen (as well as correct the colour temperature) – the data of the detail was always there in the image as it had not been compressed.
It’s still not a great image, but it demonstrates the principle of JPEG file compression.
|
|
|
|
|
Original, 1/20th at F8, 2 stops under-exposed |
High compression JPEG – Levels adjusted in Photoshop. |
RAW File – Levels and colour temperature adjusted in C1DSLR and final levels and contrast in Photoshop. |
Another important aspect in the use of digital sensors is to understand what can and can‘t be saved. In digital imaging at the moment, overexposure spells a lost image. Whilst the shadow detail can, to a certain extent be recovered, highlight detail or overexposed areas cannot. If in doubt, bracket of course, but safe = under-exposure.
A problem resulting from too much image compression, particularly for landscape photographers, is the appearance of what are called artefacts in the image. Artefacts occur when there are large areas in the image of a similar, lightly graduated hue (typically sky areas). High JPEG compression will reduce the graduation in order to reduce the file size, resulting in too great a difference in colour between areas, which can then be seen as distinct patches of colour rather than a gradually changing shade:
|
|
|
|
Jpeg with a very small amount of image compression – File size 46.2 KB |
JPEG with almost maximum compression applied – File size 4.69 KB. |
|
|
|
| The rough effect you can see in this enlarged area is actually the film grain. | Here it's replaced by blocks of tonality due to the compression. |
Despite these compression losses, there are fields where the speed and file size advantages of JPEG’s are more desirable than maximum quality. Sports photographers generally shoot JPEG’s for exactly these reasons – cameras such as Canon’s new 1D Mark II which is aimed at this market, optimise JPEG handling features – 8.5 frames per second for 40 frames – you should be able to capture any action with that! If the image is going to be used in a newspaper, at a maximum size of say 10x8cm, maybe even in black and white, the JPEG image quality is always going to suffice. In practice, for the types of images being shot, a much larger print will also still be very good by consumer standards – the photographer's skill and technique will often be the most influential factor in getting a good image!
For point and shoot digital camera users who only want small prints, or files for web-pages or emailing, JPEG is the way to go. Most consumer level digital cameras only have the option to save files as JPEG’s, as this is perfectly adequate for the requirements of the market they are aimed at.
If however, you are going to want to make large prints and demand the maximum image quality, then TIFF or RAW are the only choices.
If file size and speed are not the most important factors in what you shoot, then RAW is also going to provide the maximum flexibility and be the most forgiving format. It is rather like comparing slide and negative film. With slide film, the exposure must be spot on, as the films dynamic range is low - 5 stops compared to negative films’ 7. Because RAW saves all the sensor data, adjustment on the computer of a couple of stops either side of the shot values will be possible (although a sensors range is only about 5 stops – the same as slide film). This is also similar to negative film in that there is another process step (printing or computer editing) before the final print, where further corrections can be made. The most important thing to bear in mind at the moment, is not to overexpose and blow out highlights. This is why many cameras have an over exposure warning in their preview, as details from overexposed areas will not be recoverable, whereas underexposed areas will.
The images below illustrate this point. The image on the left is the RAW file as it was shot. Had this been shot as a JPEG, the sky would have been mostly compressed to white and the plane and lights to shadow or black. The image on the right shows a combination of working in Capture One to bring the sky detail back, saving as a TIFF and then working in Photoshop to select only the plane and lights and to bring their correct details and colour back. Such a shot would be impossible to produce using the JPEG format.
|
|
|
Once the editing work is done, there is of course no problem in saving the file as a high quality JPEG in order to email it or save as a ‚final version’ to CD for example.
What cannot be done, is further editing and repeated saving in the JPEG format - this will result in a stepped loss of quality with each save (compression) operation. If you think you will need to work again on files, they should be saved in Photoshop file format or as TIFF files.
I hope that by using this article, you are able to identify what you want to do with your images and can choose the appropriate format.
Have fun!
Words and Images Copyright, Steve Pearce, 2004.