Jpeg vs raw – Does raw also make sense on a smartphone?

Spread the love

Only a few shoot in RAW with their smartphone, sometimes simply because they do not know about its existence or do not know the setting, but often also because the added value is not clear. In addition, not all smartphones support raw as standard, so in many cases you have to use third-party apps. The RAW format is especially popular among advanced photographers, who are familiar with SLR and system cameras, but does it also make sense on smartphones?

what is raw?

In short, you can compare RAW with a digital negative. It is an alternative to the jpeg image format, which saves photos in a much higher quality – so you can correct the exposure and color deviations afterwards. The name ‘raw’ refers to the fact that the digital data is stored ‘raw’, usually in the dngformat without editing the image. With jpeg this is different and in two ways. First of all, jpeg photos are ‘optimized’ unsolicited, so that they look a bit better. The contrast is increased, the colors are enhanced and the photo is artificially sharpened. Weak points of the lens are also cleverly eliminated by software, such as distortion, vignetting and blur in the corners. However, it does not stop there, because there is a second phase. Compression is used to save the edited image.

Compression

There are two types of compression: lossless and lossy. Lossless works like a zip file; data is stored efficiently in a smart way. Suppose you have the following text: aaaaabbbccccccccccccccc . You can write this literally like this, but also as a5b3c15 . The latter takes up much less space. That is lossless compression; all data remains intact, but the file becomes smaller. This is how raw, using dng as an example, generally works; all image information is retained.

Lossy compression, which is used for the jpeg, but also for the mp3 format, is a form of destructive compression, because ‘excess’ information is thrown away. A blue sky looks smooth to our eyes, but it actually consists of tens of thousands of shades of blue. A few shades of blue can disappear without us noticing a difference. If the sky were reduced to one shade of blue, we would see it, but not if it were a few thousand less or if the compression was cleverly deployed. With jpeg, the compression is applied in blocks of 8×8 pixels, which is also visible in the image on the right. Lossy compression makes a huge difference in the size of an image file. Where a recompressed photo can be 60MB in size and lossless compressed 40MB, with jpeg that can be reduced to just 6MB or less. That saves a factor of ten, so that more photos fit in the memory or on an SD card. This used to be a necessity, because memory storage was limited. Nowadays this is less of a problem, because memory has become cheap. Yet smartphones and cameras still use the 1992 jpeg standard, while there are better alternatives today, with smarter compression and greater color depth.

The degree of compression is adjustable; a lot of compression leads to a visible loss of quality, but with little compression it is often hardly noticeable. With a smartphone, the degree of jpeg compression can usually not be set or only to a limited extent. In all cases there is a loss of details and therefore of image quality. In addition, every time you save the file again, compression is applied again and therefore image information is lost again.

bits

Apart from color loss due to compression, jpeg uses only 8 bit color information, compared to 12 to 16 bit in raw. That difference is huge. As you know, data is kept as ones and zeros. A photo with a bit depth of 1 would consist of 0 and 1, ie each pixel could simply be recorded whether it is black or white. That would make for a very simple picture. Count that up and you realize that with 8bit you get all combinations with a numerical scale from 0 to 255, or 256 brightness gradations per color channel. With three color channels, RGB, that means a display of a total of 16.7 million different color shades: r256xg256xb256.

That seems like a lot and it is also fine for normal use, but it is not optimal for image editing, because not all color gradations are captured, even apart from any harmful compression. More color gradations means that such a photo file contains a larger dynamic range. In other words, a larger area between light and dark can be displayed. This results in photos with more detail, especially in dark and bright areas, i.e. the shadows and highlights. An 8-bit file, such as jpeg, can capture up to approximately 8 stops of dynamic range. With a raw file, that is about 10 to 16 stops, partly depending on the analog-to-digital converter, which is usually 12 to 14 bit. Simply put, the more bits, the wider the exposure range.

Image editing

The earlier examples show a visually large difference, but an important nuance is that this is only visible to a limited extent in practice. In most cases, you can therefore shoot fine in jpeg and you will hardly notice the harmful compression and smaller color depth. That’s different if you use image editing to improve your photos. The loss of color information through harmful compression may not be visible to the naked eye, but for serious image processing it can quickly be catastrophic. In practice, this is especially the case with failed photos, which are, for example, too light or too dark, or situations with very high-contrast light, such as backlighting.

For example, if you want to save details from an overexposed sky, this will only have a limited effect with jpeg. After all, certain color gradations have been lost due to the compression and you can’t get them back. Simply put: if white sky is one hundred percent white, which is the case with overexposure, then you can’t get anything out of this. The sky can only turn gray evenly, but that doesn’t look good. If your photo is saved in raw, the chance of success is a lot higher. That is because of a combination of compression and bit depth. The raw image information stored in the raw file contains more usable image information, which is worth gold in such image processing. This can also be clearly seen in the photos below.

The photos above show a tricky lighting situation in a classic form. There is a dark foreground, with dark clouds in the background through which the sun shines, with a reflection of it on the water. That is an almost impossible situation for a camera to expose properly. Matching the exposure to the foreground by tapping that location on the screen overexposes the sky; if you tune to the sky, the foreground will remain underexposed. HDR is an option, but because three separate photos are then merged, this can lead to ghosting and, moreover, tone mapping sometimes looks a bit artificial.

Shooting in RAW is a better solution, because you can then use the larger dynamic range and the extra color depth afterwards, without harmful compression. As you can see from the second photo in the above sequence, the jpeg version’s overexposed sky won’t budge. There are pixels that are 100 percent white, so they don’t contain enough color information to make overexposed details visible. You can also see this in some image editing apps, Photoshop and Lightroom. The red part in the screenshot below contains one hundred percent white pixels, which cannot be corrected.

The third photo shows the corrected version based on the raw file. The sky can be easily corrected here and the foreground can also be well lit. This photo shows much more detail in the sky. Shooting in RAW therefore offers you much more latitude to correct over- and under-exposed parts in a photo, without visible damage. Similar results can be seen in the photos below.

In principle, the advantage also applies the other way around. An underexposed photo can be repaired more effectively in RAW than in JPEG, but the difference is much less than with overexposure. Unless pixels are one hundred percent black, shadow areas can also be restored quite well in jpeg photos. The big disadvantage of raising shadows is that it leads to more noise. Especially with RAW, color noise quickly becomes disturbing, because the noise has not been reduced. This is the case with the jpeg photo, so it is less noticeable here. Noise can also be removed manually with RAW photos, but you have to put in more effort and this often requires specialized software.

Still, there are certainly situations where you can squeeze more detail out of shadows in raw than with jpeg. Shooting in RAW is most effective when both the highlights and the shadows need to be adjusted, as in the photo below.

Noise

A jpeg photo is always edited by the camera. Think of sharpening, increasing the contrast, but also noise reduction. This is even more the case with smartphones than with SLR and system cameras, because the target group prefers ready-made photos, without mandatory image editing. With cameras with a large sensor, image processing with RAW photos is a major advantage, because in many cases you can optimize photos better than the camera can, for example with regard to noise reduction. With smartphones, the benefit in noise reduction is limited, mainly due to the small sensor. With a 1/3″ or smaller sensor, you will already be seriously bothered by noise at ISO 800, which manifests itself in jpegs in loss of details and large amounts of color noise in RAWs.

Effective noise reduction of RAW photos requires specialized software, such as Lightroom or Noise Ninja. Those who already shoot in RAW with a ‘real’ camera probably already have that software, but less experienced users have to fall back on smartphone apps and they are usually less effective. Incidentally, noise mainly occurs in situations with low light and you only experience it to a limited extent in daylight.

White balance

The extra color gradations and uncompressed image information of the raw format have another big advantage; you can correct the white balance afterwards without loss of quality. Wrong white balance is one of the most common ‘mistakes’ of digital cameras and is mainly caused by artificial lighting. Light, for example sunlight, always has a different color temperature during the day and artificial light also differs per type of light source, from incandescent lamp, via fluorescent tube to candle. A piece of white paper therefore always looks a little different. Human eyesight automatically corrects this color difference, so that we hardly notice a color difference. A camera is less able to detect and correct color deviations, although the automatic white balance is getting better and better.

Color casts mainly occur in an environment with different light sources, such as a city in the dark where different types of artificial light are used interchangeably: white, blue and yellow. Those kinds of images usually become very warm, or orange. This can be prevented by setting the correct white balance in advance, which is standard on most smartphones in the camera app, whether or not in the ‘pro’ mode. However, if you have photographed in raw, you can also do it afterwards, for example in Lightroom (Mobile). You can therefore opt for a pre-programmed white balance afterwards or set it yourself very accurately with a certain color temperature, expressed in Kelvin.

Below you can see a practical example. The first photo shows the original, unedited photo, where the white balance is clearly too warm. In addition, the photo contains overexposed areas due to bright lights. The jpeg version of this photo can be partially improved, but due to the correction of the orange color cast, certain parts become a bit too blue. The highlights can be corrected reasonably well, but overexposed parts can at most turn gray. In the raw version, the white balance and the overexposed parts have been corrected.

The downside of raw

So shooting in RAW has great advantages, but it’s not just hallelujah. In addition to the quality advantage, there are also serious disadvantages to using raw. Most importantly, RAW photos take up five to ten times the space of JPEGs. It is therefore not recommended to shoot in this format by default. For smartphones with limited memory capacity and no expansion options, this is a potential stumbling block.

Furthermore, we are of course talking about cameras with a relatively small sensor, which already have more difficulty capturing a large dynamic range. So if you shoot in RAW, don’t expect results that are comparable to those of a camera with a large sensor, such as a system camera or SLR. And as mentioned earlier, this also applies to noise, because getting rid of color noise in highlighted shadows or low-light situations is quite difficult. Digital zooming is also usually not possible if you shoot in RAW, because the raw image is saved and zoom is a form of image manipulation.

Finally, you need specialized software that can handle dng files, the file format most commonly used for raw on smartphones. Many standard apps ca n’t handle DNGs, which means you sometimes don’t see your raw photos back. It also happens that you only see a low-res jpeg thumbnail, which is usually baked into a dng file. You therefore need special apps to open and edit raw files, such as Lightroom Mobile or Snapseed. But while you’re at it, you can immediately apply some other improvements, such as straightening a skewed horizon.

Importing raw files to a computer is sometimes a bit more work. Third-party apps, such as Camera FV-5, sometimes use separate folders, and iPhones only sync jpegs by default, so you have to resort to Image Capture (Mac) or File Explorer (PC). If you want to quickly share a photo on social media such as Instagram or Facebook, the jpeg version is often more convenient, because you first have to go through a number of steps for the raw version. As a result, it is often practical to save photos in both jpeg and dng format. This method logically costs even more storage capacity.

Support from smartphone manufacturers

Not everyone has the option to take photos in raw format. This depends on the operating system and the manufacturer. Raw support has been available in software since Android 5.0 (Lollipop) and iOS 10, but hardware support varies by smartphone model.

For example, Apple only supports this on devices with a twelve-megapixel sensor, such as the iPhone 6s (Plus), 7 (Plus), SE and iPad Pro 9.7″. The other iPhones and iPads are excluded. With Windows Mobile, the high-end models offer the Lumia 1020 and 950 support raw as standard, but cheaper versions like the 630 do not.

There are big differences even with Android, although raw support has been baked into the OS since 2014. Most recent Android smartphones can shoot in raw, whether or not through a third-party app like Camera FV-5 , but there are still exceptions. A high-end model like the Sony Xperia Z5 and even the recently released XZ do not support raw, not even through other apps. Only a limited number of smartphones, such as the Samsung S7, LG G5/V20 and the Lumia 950, offer raw functionality via the default camera app, although they usually require a ‘pro’ mode to be enabled.

Png and tiff as an alternative

A raw file, such as dng, contains the raw sensor information, but those are ones and zeroes and not yet an image. The image must first be recalculated and interpolated on the basis of the (Bayer) color filter. This is different with a bitmap file, such as tiff or png. This already consists of images that are stored lossless by default in those formats. Some smartphones do not support dng, but can save images as png or tiff. This can be an alternative to raw, although in many cases it is not about the raw, unprocessed image from the sensor, but about image that has been processed by the processor, including noise reduction, for example.

Conclusion

Does it make sense to shoot in RAW? Yes, but not for everyone and by no means in all situations. To start with the latter: in most circumstances jpeg photos offer sufficient quality and the advantage of raw is limited. However, if you photograph a situation with difficult lighting, such as high contrasts, backlighting or a scene with many artificial light sources, the RAW format offers a strategic advantage.

Because a raw photo contains a greater color depth and is not affected by harmful compression, you have much more editing space at your disposal. This allows you to correct overexposure and underexposure and squeeze more detail out of shadows and highlights. You can also adjust the white balance afterwards. This can be done with smartphone apps as well as with desktop software.

In practice, it is best to shoot in jpeg as standard, but it is good to be alert when a difficult lighting situation arises, whether on the road or during a holiday. Then you can switch to raw or possibly shoot in jpeg and raw at the same time. Photographers who are already used to working with RAW, for example through the use of another camera, have an advantage, because they are used to the workflow and probably already have suitable software in house. However, there are also plenty of options for the avid smartphone photographer to get started with RAW, if only because of a good range of camera and editing apps that can handle it. All in all, we think there are enough reasons to give it a try.

You might also like