The new updates of Camera RAW and Lightroom promise to increase the detail of our photographs

0 5

There are new updates within the Adobe photographic universe: Camera RAW 11.2 and Lightroom 8.2. Thanks to the Sensei technology they have managed to increase the resolution of our photographs and therefore improve the detail of the RAW files coming from both Bayer-type sensors and Fuji’s more complicated X-Trans sensors.

From time to time the Creative Cloud surprises us with an update. It usually consists of small fixes and the updating of compatible files (the files of the Sony A6400 and the Olympus E-M1X can now be revealed ). Other times we find small treasures that may go unnoticed by most. This time is good news for all the happy owners of Fuji cameras.

The new function that we can find in both versions of Lightroomand in Adobe Camera RAW is called Details of the improvement , a very bad translation of Enhance Details (Why not call it Improvement of details ?). Before explaining it, let’s see how they tell us in Adobe :

Enhance Details uses an intricate neural network (CNN) with extensive training to optimize the maximum image quality. We formed a neural network for demonstrative raw images using problematic examples, and then we took advantage of the new machine learning frameworks integrated in the latest Mac OS and Win10 operating systems to run this network. The neural network has been trained with more than one billion examples. Each of these one billion examples contained one or more of the main problems listed above that give serious problems to standard demosaicing methods. We make two models: one for Bayer sensors and one for Fujifilm X-Trans sensors. As a result, the improved details will offer amazing results,

It’s a little complicated to understand. It can be explained in another way. Thanks to artificial intelligence (is not this term abused in photography: focus of the Sony A6400 , revealed from various programs …?) The interpretation of the raw information of the RAW file is much more precise in terms of resolution and color, regardless of the color pattern of the sensor filter that our camera has.

The new updates of Camera RAW and Lightroom promise to increase the detail of our photographs
In the process

Personally I think it’s just a new algorithm, like the one incorporated in programs like RAWTherapee , which allows us to read the information in our files in another way. And with it the following problems are avoided:

  • Loss of fine details when forming the image.
  • Little quality in the reproduction of colors at the edges of the objects photographed
  • Ending the worms of X-Trans sensors for misinterpretation.

So is it good news?

Any improvement and anything that allows once and for all to solve the interpretation problems of X-Trans sensors is good news. But not everything could be wonderful. The requirements of computers and operating systems are too specific. And many of us may have to consider a change.

To start it only works if we have an operating system that has Machine Learning . That is, only from the versions MacOs 10.13 and W10 (1809) . For neophytes like me, it seems to be an automatic learning model that incorporates the latest Windows and macOS. Which brings us to the second problem.

The new updates of Camera RAW and Lightroom promise to increase the detail of our photographs
Before / after 100%

Each time we depend more on the GPU of our graphics card. It is no longer worth any moderately powerful model. For example, I have a 2GB Nvidia GeForce GT and I can not enjoy the improvement. I have no choice but to look for a more powerful or modern one. It is no longer worth having enough RAM. Now you have to buy SSDs and graphic cards to match .

 

It seems that the problems in Adobe with Fuji’s .ORF files are finally over. At last we will be able to enjoy all the quality it has, far superior to that of many sensors with a Bayer pattern.

But bad weather good face, as my dear grandmother said. It seems that the problems in Adobe with Fuji’s .ORF files are finally over. At last we will be able to enjoy all the quality it has, far superior to that of many sensors with a Bayer pattern.

How does improvement details work?

As we can read in the Adobe blog , the explanation is very simple. All the photographs that we make with our digital cameras are in black and white. The pixels only see the luminosity, without any color. Only thanks to the RGB matrix that the photoreceptors have on top of and by a process called demosaicing .

Each pixel only records one color. And it lacks the information of the other two. What demosaicing does is generate the final color from that information. This is easy in large areas of color such as the sky or a wall. But it can be very complicated in small areas with details and textures.

If we work with a sensor that has a Bayer matrix, the work is much simpler, because the arrangement is completely logical and perfect, as if it were a three-color chessboard with a repetitive pattern. The subject is much more complex with an X-Trans sensor , with a much more random arrangement of colors.

The new updates of Camera RAW and Lightroom promise to increase the detail of our photographs
After / before 200%

For this reason, if we have photographs with complex motifs, with many details, this new Adobe feature can save the photograph without leaving the secure environment of the suite . However, we can not forget the problems we have discussed. And I can assure you that even though I have tried it, I have not been able to complete the process on my computer. I tell you how to do it as soon as you have the latest version and the corresponding operating system installed:

  • In Camera RAW we select the file and we will go to Slides slide where we will find a horizontal stripes icon that will open the menu where we find the option Details of the improvement .
  • In Lightroom , in the Reveal module , go to Photography> Details of the improvement (Ctrl + ALT + I) .

We just have to wait for the algorithm to do its work and give the Improvebutton to get a copy of the file in DNG format . Any development process we will follow with this new file that has in its name adds Enhanced to distinguish it. Do we have to do it with all the photos? It would be ideal, but the process is slow (depends on the system we have). So the best thing is to use it only in those files that we are going to print or that have a lot of fine detail.

The new updates of Camera RAW and Lightroom promise to increase the detail of our photographs
Final photography

I’ve tried it on three computers. Two PCs and one Mac. In the latter, everything has gone smoothly with a graph that does not reach 2 GB. In a PC I can see in the corresponding window what is the improvement but my computer, my program, my program with the operating system, my graphic card, or all together, is unable to generate the final DNG file due to an unknown error . In a newer everything has been perfect. Yes, with a worse card.

It is still good news. But I do not quite understand why it is such a long process and what is the reason why we do not work with this algorithm (or IA ) directly to differentiate ourselves from the rest of the competitors who are struggling to overthrow the sacrosanct Adobe. And point out that the same corrected file goes from 33 MB to 87 Mb …

If they manage to implement it by default it would be a good blow at the table to make it clear who is in charge. But we’ll see. Interesting months are coming …

 

You might also like

Leave A Reply

Your email address will not be published.

20 − 3 =

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. AcceptRead More