Roughdraft Loc: Maryland

From what I have read on the subject of RAW files my understanding is in RAW format the camera stores everything it "sees", while in jpeg the camera stores what it determines to be the best parts of the image and disregards the trash. Am I close on my understanding thus far?

This then leads to my base question, if RAW is the better format to shoot in why then is it necessary to convert images to jpeg for PP and not just shoot in jpeg and skip the conversion?

Enlightenment please

Longshadow Loc: Audubon, PA, United States

Every manufacturer's RAW is different. Think of it as pure SENSOR DATA. In order for it to be "visible", it has to be converted to a common format, for example, JPEG, that can be viewed by everyone. The conversion doesn't take the "best parts" of the image, it processes the whole image with a predetermined algorithm, to generate a JPEG.

Editing RAW images with the manufacturer's RAW editor usually provides better control over any post processing. I shoot RAW+JPEG so I can see the image on the computer without opening the RAW editor. If I decide to edit an image, I always edit the RAW, with possible tweaking in a JPEG editor for any effects I want to apply. RAW editors usually don't have a vast amount of tweaking and effect options.

rmalarz Loc: Tempe, Arizona

Close, but not quite accurate. If you are shooting in jpg only mode, the camera still captures the RAW data. Then it processes that according to the software installed in the camera. You get the software engineer's best interpretation of the RAW data, based on the camera settings, etc.

The slightly misinterpreted part was that it uses the best parts. It uses the entire RAW file, then deletes the RAW data leaving only the jpg file.

RAW is similar to your film negative. The RAW file is what one would work on in PP. Since RAW is only data, it is not an image per se. It is simply the data necessary for a post processing program to use in making the image.
--Bob

Apaflo Loc: Anchorage, Alaska

You can't "just shoot in jpeg and skip the conversion". It is only a matter of where the conversion is done, because the camera shoots in RAW and generates a JPEG no matter what other options you set.

If you choose to "shoot in RAW", the camera saves the RAW data, including the JPEG embedded in the file, to a RAW file on the memory card.

If you choose to "shoot in JPEG", the RAW data is discarded and only the JPEG is saved to the memory card.

If you shoot in RAW+JPEG, both of those files are save.

imagemeister Loc: mid east Florida

CAUTION: Fasten your seat belts - this is a deep dicey subject !

Suffice it to say, I only shoot JPEG and have never shot raw - I am on Sony APSC A77II.

rmalarz Loc: Tempe, Arizona

I have to disagree with your statement, "RAW editors usually don't have a vast amount of tweaking and effect options."

RAW editors offer a considerable amount of adjustment and refinement options. Adobe Camera RAW is almost a complete editor in itself. I know several photographers who do 95% of their post processing in ACR and only a few tweaks in PS to complete a photograph.
--Bob

Longshadow Loc: Audubon, PA, United States

Yes, the sensor data has to be converted somewhere.

Longshadow Loc: Audubon, PA, United States

Sorry, I was referring to the manufacturer's RAW editor, but I didn't state that. Adobe, or any comparable editor that will accept RAW files, is different. The JPEG editor that I use will not (or didn't used to) accept RAW. It may now though.

John_F Loc: Minneapolis, MN

A sensor is an spacial array of charge centers and each determined the bit value ( 0 or 1 ). A sensor is thus a collection of bit values and, in theory, is the RAW data which is configured into file that is readable by external devices, such as conputers, tablets, smartphones. A JPEG is compressd RAW data and given the lossless efficiency of the compression and the decompression the resulting image will be more or less faithful to the RAW data. But one's camera's RAW might not be as great as one thinks. Example: my Sony a6000 creates 14 bit depth pixels - all of the color and intensity data must be packed into 14 bits. Morever, the Sony is said to compress those 14 bits to 11 bits in the actual RAW file - lossy to lossless value unknown. When Adobe DNG Converter converts such a file or Adobe Camera RAW opens such a file the pixel bit depth becomes 24 - clearly bits are being manufactured. The same applies to the camera JPEG with its compression/decompression efficiency. So does the image seen on the computer screen or printed piece of paper appear faithful to what you remember seeing.

Apaflo Loc: Anchorage, Alaska

Not correct. The spacial array of charge centers (and amusing name) is an analog device and has no sense of bit values. Instead it develops voltages where the amplitude of the voltage is proportional to the number of photons captured. Generally for discussion purposes the maximum output is said to be 1.0000 volts (in fact it might be different). If that is the maximum, then if exposure is reduced by 1 fstop (either a different aperture or different shutter timing or a combination of both) the voltage will then be 0.5000 volts. And another fstop lower would be a voltage of 0.2500 volts. That continues until some very low voltage that does not change with the exposure, because it is at the "noise floor".

That analog voltage is fed to an Analog-Digital-Converter and becomes a digital value made up of bits. Note that it is still not an image though, because it is just raw sensor data with color values encoded using a Bayer Color Filter Array.


Again, that is not at all a correct description.

JPEG is an RGB image format. Each pixel has a value for Red, a value for Green, and a value for Blue. And it is one very specific value!

The original RAW data does not have pixels with values for each color. Any given pixel location also does not even have a specific color, but rather a range of possible colors (all of which are correct, so nearly an infinite number of different images could be produced from one RAW data set). A JPEG produced from the RAW file need not be particularly "faithful" to the RAW data, and in fact it is mathematically impossible to produce the RAW data from a given JPEG image.


None of that is valid as such. Sony may well compress the 14 bit data to effectively 11 bit data, but the 24 bits is from the 8 bit RGB channels for JPEG. Note that 14-bit RAW files only pack tone variations into the 14 bit data. The color is encoded using the Bayer Color Filter Array and is entirely distinct from the digital bit values. That works well, but it is also true that it produces only a very close approximation of the original color.

rmalarz Loc: Tempe, Arizona

John, the sensor is not zeros and ones, after the sensor the image data "goes through" an A to D converter which results in the 0/1 file.

Digital photography is analog and film photography is digital.

Others can deal with the additional errors in your statement.
--Bob

John_F Loc: Minneapolis, MN

Evidently, Apaflo and I are of a greatly different understanding of how semiconductors work and what physically defines a bit value. Is a string of zeros and ones created by electron spin up or down or presence/absence of an electron or something else.

Longshadow Loc: Audubon, PA, United States

rmalarz Loc: Tempe, Arizona

I think the difference is in the understanding of how sensors work. Sensors are not digital devices.
--Bob

Longshadow Loc: Audubon, PA, United States

If they are CCD or CMOS devices, they are digital.