RAW vs. JPEG if exposure isn't an issue?

If you can get the color balance and exposure correct, you're probably good to go. I've said it before. Pros shooting with pro lighting probably don't need raw (but it wouldn't hurt.) People like me, shooting on the run during vacation in a variety of conditions with only minutes if not seconds to set up, need raw.
Yeah, this is why I don’t get the whole, “If you’re a beginner, start with jpegs” argument. If anything, save both. I wish I had raw files from when we started out.
 
Last edited:
If someone is shooting in controlled conditions where getting the exposure correct every time isn't a problem, and if the photos are destined for social media use, does RAW offer any advantages over JPEG?
Depends. Is that someone getting paid for the pics and does the employer care about having very high quality? If not getting paid, who is the audience and will they notice or care about small differences in quality? Does the photographer have the time to waste processing raws if no one cares? if the photographer cares, that's enough, really, so long as they don't mind spending the time they could be using for other things like skydiving or partying like a maniac.
 
Last edited:
If someone is shooting in controlled conditions where getting the exposure correct every time isn't a problem, and if the photos are destined for social media use, does RAW offer any advantages over JPEG?
In simplified terms RAW offers advantages in challenging exposure conditions, in conditions where you need to change white balance, and maybe in situations where you want to pull detail out of shadows. So if none of those apply then it is difficult for me to see any advantage to RAW.
 
If you can get the color balance and exposure correct, you're probably good to go. I've said it before. Pros shooting with pro lighting probably don't need raw (but it wouldn't hurt.) People like me, shooting on the run during vacation in a variety of conditions with only minutes if not seconds to set up, need raw.
Suppose you are making an interiors photograph of a long, wide room. One side of the room is an interior wall. The opposite side has many large windows.

The angle of the sun light coming through the windows is such that hardly any sun light illuminates objects near the interior wall. All the objects on the interior wall are lit by tungsten lamps and the color temperature for that light is ~2000K.

Objects on the opposite side are lit by sunlight and the color temperature for that light is ~5000K.

The in-camera JPEG rendering uses one set of color temperature parameters to render the entire image.

What are the correct color temperature parameters to render an image that resembles what a human sees?

You manually set the in-camera, color-temperature, rendering parameters for tungsten light. The hue for objects near the windows will render blue compared to what a human would see.

You manually set the in-camera, color-temperature, rendering parterres for sun light light. The hue for objects near the interior wall will render yellow compared to what a human would see.

You use the camera's automated color temperature parameters even though you have no idea how those color temperature parameters are computed. Now both sides of the room render incorrect color hues. The wall side is too yellow and the window side is too blue.

How do you render an image with appropriate color hues for the entire room from any of the in-camera JPEG images?

What information is required?

How much of that required information is irreversibly lost due to lossy JPEG data compression using inappropriate, in-camera color-temperature, parameters?
_____________________
“…the mathematical rules of probability theory are not merely rules for calculating frequencies of random variables; they are also the unique consistent rules for conducting inference (i.e., plausible reasoning)”
E.T Jaynes, Probability Theory: The Logic of Science
 
Can you post some examples of this situation with original jpeg and the results of the extensively edited raw for comparison. It does not sound like something that couldn't be tweaked in jpeg to produce an acceptable print. But also depends on the opinion of the person who is judging the finished product. 99% of the folks that don't get this deep into things like white balance in photos will likely be perfectly happy with either version. Most extensive editing of raw files is done to please the photographer not the audience.
 
I don’t know why one wouldn’t record in raw and jpeg unless you are really up against buffer depth problems then you might switch to 12bit / jpeg only.

Generally my cameras ( mostly Nikon ) produce a jpeg output that I like SOOC - with standard WB and picture control settings.

I just dump the jpeg’s straight on servers or a tablet etc without processing - useful and fast. Maybe I will go into the FF raw to tweak it if I have a really nice shot ( happens rarely ) that I want the most out of.

But for m43 and 1 inch sensors in particular- often up against noise eg wildlife in lower light - the raw is essential - denoise like topaz ai, deepprime, prime only work with the raw. Even basic outputs near base iso from a 1 inch sensor usually need a bit of noise reduction.

You can adjust WB in jpeg quite easily in practice for most situations - helps if there is a gray are in the field ir just adjust temperature and tint till it looks right.
 
Last edited:
I don’t know why one wouldn’t record in raw and jpeg unless you are really up against buffer depth problems then you might switch to 12bit / jpeg only.

Generally my cameras ( mostly Nikon ) produce a jpeg output that I like SOOC - with standard WB and picture control settings.

I just dump the jpeg’s straight on servers or a tablet etc without processing - useful and fast. Maybe I will go into the FF raw to tweak it if I have a really nice shot ( happens rarely ) that I want the most out of.

But for m43 and 1 inch sensors in particular- often up against noise eg wildlife in lower light - the raw is essential - denoise like topaz ai, deepprime, prime only work with the raw. Even basic outputs near base iso from a 1 inch sensor usually need a bit of noise reduction.

You can adjust WB in jpeg quite easily in practice for most situations - helps if there is a gray are in the field ir just adjust temperature and tint till it looks right.
No such thing as 12-bit jpeg. JPEG is an 8-bit, 16.7 million color 4:2:0 format.
 
I will link a video at the bottom, but basically pro's use both depending on their needs. Its roughly run and gun versus artsy fartsy.

JPEG allows more images on a card and can produce a finished product good enough for clients, especially when there is a very short time between shoot and product to client. In sports photography, JPEG will shoot faster (fps) than RAW (fps).

RAW is better for landscape and art photography where a higher control than JPEG is desirable.

 
Can you post some examples of this situation with original jpeg and the results of the extensively edited raw for comparison.
I don't have any "original JPEGs" because I never use in-camera JPEGs. If you doubt it is possible for a scene to have objects illuminated with significantly different color temperatures, try it yourself.

Otherwise, here's a some of examples I found with Google. There is a cathedral image about half-way down this page that show how two color temperatures affect indoor hue rendering. You can wager everything you own human vision does not perceive a blue hue for the upper left region in this image. Apparently a large part of this course is how to do selective hue rendering for different regions using different white balance corrections. [1].

This image depicts a difficult situation. Note the different wallpaper gray hues as the light source gradually changes from sunlight to tungsten. Proper rendering requires a color temperature parameter gradient density that matched the shift from sunlight to interior light illumination. The luminance difference requires a gradient as well. This would not a simple post-production fix.

Finally this kitchen photo is resembles a raw file rendered with a single color temperature looked before any selective, regional white balance parameters are applied. Note the blue hue around the sink due to sun light illumination.
It does not sound like something that couldn't be tweaked in jpeg to produce an acceptable print.
A "tweaked JPEG" is more likely to be acceptable as the difference in color temperatures, scene luminance decreases and object detail is absent. At the same time, images are rendered with information. The information content of an in-camera JPEG is significantly less than a raw file. Lossy compression only retains information content (data) required to render an image with the selected parameters. All other information content is removed. In some cases a tweaked JPEG will be significantly inferior when selective white balance parameters are applied after rendering. Where will the information used to compute the different rendering coming from?
But also depends on the opinion of the person who is judging the finished product. 99% of the folks that don't get this deep into things like white balance in photos will likely be perfectly happy with either version. Most extensive editing of raw files is done to please the photographer not the audience.
I hope people who read this agree with you. People with this opinion kept food on my table. This attitude was the reason clients would pay higher rates for my work. I only cared about the opinions of Interior designers, architects and real estate agents selling extremely expensive homes. These clients "get this deep" into image quality because their clients noticed the difference between good enough and professional images.

None of this detracts from the tenants information theory. More information increases fidelity.

JPEGs are convenient. That's it. Each of us decides how to balance convenience and quality. I only used low cost off-camera flash units and post-production rendering to deal with white-balance issues. The best paid architecture and interiors photographers use assistants, multiple, high-power, off-camera lighting sources, sets of lighting fixture bulbs with different color temperatures, gels, scrims, temporary window films and other time consuming methods to light a space with a single color temperature. Ironically they could have used in-camera JPEGs. Guess what - they chose raw.

1/ I have no affiliation with this product. I would not buy it. I don't recommend it. But others might find it useful.
_____________________
“…the mathematical rules of probability theory are not merely rules for calculating frequencies of random variables; they are also the unique consistent rules for conducting inference (i.e., plausible reasoning)”
E.T Jaynes, Probability Theory: The Logic of Science
 
The differences between raw and JPEG are in the accuracy of the recorded data.

This will probably be insignificant unless you edit the image very heavily. The sort of editing that is likely to show up the differences is if you brighten the shadow areas massively. The colours of the JPEG image are recorded much less accurately in deep shadow areas and this will show up when you brighten the shadows a lot.

Here is an example. The processing is deliberately extreme to illustrate the point.

The original image has some overexposed areas (in the brightest clouds) and some very dark shadows as well. Here is the original JPEG image (processed from a raw image in Lightroom):



daf9ff514c2c462b9a11f5c00f0850e4.jpg

Here is the above image processed to show the colour of each pixel at full brightness:



Each pixel raised to maximum brightness without changing its colour
Each pixel raised to maximum brightness without changing its colour

If I export a TIFF (16-bit) image from Lightroom and then process it to show the colour of each pixel, the result is:



The colour of each pixel when the original image was produced as a TIFF (16-bit) file
The colour of each pixel when the original image was produced as a TIFF (16-bit) file

The last image above shows the colours much more accurately than the previous image. A 16-bit TIFF image holds the colours of the deepest shadows more accurately than an 8-bit JPEG image.

If you do not try to peer into the deepest shadows, you will probably never notice this difference!
 
If someone is shooting in controlled conditions where getting the exposure correct every time isn't a problem, and if the photos are destined for social media use, does RAW offer any advantages over JPEG?
jpeg is what you deliver, raw is for editing and/or conversion to formats other than jpeg.

Out of camera jpegs are good when you deliver the images right after shooting, or you can't use a raw processor for whatever reason.
 

Keyboard shortcuts

Back
Top