RAW images, JPEGs and Processed photos... what does it all mean?

Is a RAW file and Image?

No.

Is a JPEG an image?

Yes.

What's the point??

Digital cameras. When they were invented a number of challenges had to be overcome in order that the result, after all the technical stuff and digital do-dahs, was an image. JPEGs are an image, so digital cameras become things that took light, processed it in magical mysterious ways using maths, engineering and some highly experimental guesswork to output an image.

They were awful.

People with film cameras laughed at them because they were low resolution like an old TV, the images lacked 'beauty' in all but the most careful set-ups and no one had a phone to share them with.

But they got better. They increased in resolution and at about 7 to 10 Megapixels they started to become more pleasant because the pixels were less visible in a small image.

Somthing inside the xamera became important - the Processor. In this case not just a chip but also a method - a way of taking the light sensore information and processing it so that it represented the real-life scene that was captured.

Here is the crux, the division. The light sensor records the light falling on each point after it has been focussed by the lens, moderated by the aperture, limited by the shutter speed and further calibrated by the ISO setting - which in digital cameras is an electrically set sensitivity to light. The point at which this shaped and controlled light hits the camera sensor is the moment when the journey begins.

Guess what? The RAW file is this bit - the beginning of the journey, the 'sensor data'. The point at which the real-world is digitised and becomes discrete values held in the camera's temporary on board memory.

At birth this data weighs in at about 50 megabytes for an uncompressed 24 megapixel sensor. Think of this as some 'scientific measurements' - reams of printed digits grouped and arranged in a manner that does nothing to conjure up the scene that was photographed. And that's really all it is. To get just the basics of the image it all has to be interpreted. Each digit has a meaning that needs to be looked up in tables of other numbers that exist after years of research and refinement. Each sensor has it's own set of tables, and each manufacturer has opinions about how to correctly interpret parts of the data.

Sensors today collect "too much data". That is to say that once the values have been interpreted and calculated some of the data that was originally collected ends up being thrown away - not affecting the final outcome of the image. The thing is, it is in no way obvious which data is not needed. Only processing will reveal this.

Most cameras, including phones, have automated processing built in so that at the point of taking the photograph the software interprets the data, analyses the image that will result and makes adjustments to the processing to meet some pre-programmed criteria about contrast, brightness, depth of shadows etc. Once processed a final image is stored on the memory card as a JPEG. It's probably about 1 to 5 megabytes, which means over 40 megabytes of the original sensor data has been discarded, lost forever.

Most people love this! It's convenient, instant, shareable. If you only need a digital copy and don't want to print it you can reduce the size of the file even more to under half a megabyte.

On the whole 'Professional' photographers don't use the automated processing that's available in the camera. Instead they/we just keep the raw data collected by the sensor in a RAW file. We take that home, load it into a computer and use purpose-built software to process the data into a finished image.

This processing can take on many forms. I like to look at the photos and create a finished look for each event which is derived from how I felt in real life while I was there. It's hard to automate that process because it's based on my own personal experience combined with the experience of the event itself.

After processing the RAW file has gained a few extra megabytes of information that describes how the data should be interpreted in order to create a processed image. The final step is to create a JPEG by taking the RAW data, applying the processing information and building a JPEG of the result.

JPEG is an incredibly ingenious format that is dedicated not to preserving data but to encoding a final image using as little data as possible. It stores how the final image looks, but cares not about the original data that was captured to make it. Once you have the JPEG there is no route back to the original data, that is not included and not derivable since any data from the RAW file that does not contribute to the look is not included in the JPEG.

That's where you arrive finally - a JPEG. This is the image that was the end intent when taking the photo. It's the finished work of art. The RAW data is just a fleeting moment on the journey and 'RAW Image', technically, makes no sense because it's not an image without further processing.