TIPS TO REMOVE PAPER TEXTURE FROM OLD PHOTOS
Minimising or removing the paper texture inherited when scanning an old photo is often one of the challenges photo restoration newbies find time-consuming under conventional methods.
However, some useful methods and algorithms help to minimise or even get rid of patterns and textures prevalent in vintage photographic papers, (especially those used in the late ’70s.)
However, if you are scanning the photo in-house or plan to restore it yourself, find below a few tips to assist you in achieving better results, you can also find useful information about photo restoration and photo restoration pricing in our photography blog,
1. CHECK FOR THE PHOTO NEGATIVE
Before scanning an old photo, Let’s start from the basics, perhaps you or your family member still has he photo negative around. Removing the texture from an old photo is not such an easy task, so if you can take your time to search for the film, you will be saving yourself hours of digital photo restoration.
We are often asked if it is better to scan from negatives rather than from printed photos, and the short answer is: “it depends” anyhow to remove those nasty patterns from a photo negative are the way to go.
However it is vital to assess the condition of a negative, the decision criteria should be to save you time and effort so judge based on the status of both sources, sometimes negatives are also full of dust and scratches so restoring a photo from these can be also a challenge.
We have found that sometimes hybrids are an excellent way to obtain detail when both negatives and prints are damaged.
And yes we can scan negatives for you!
So, checking around for other sources is a good photo restoration practice but also an activity that can connect you with your family history and an opportunity to find other treasures or memories lost in time.
2. SCAN THE PHOTO FOUR TIMES.
Scanning the photo four times but in a different angle is one of the oldest tricks in the book, but still, an excellent way to get the most out of your textured photograph print when home scanning.
Just scan the first time and once the process is complete rotate the photo 90 degrees on the flatbed and scan again using the same settings, repeat this procedure until you have the photo scanned four times in further 90 degree angles.
A best practice is to turn off any auto adjustments such as auto-exposure and auto-levels so you can have more control during the scanning process.
Once we finalise the photo digitalisation, we can combine the four photos in photoshop and use a darken blending mode to get rid of the highlights, which are indeed the problem.
Why? just because despite your scanner quality the light will unavoidably bounce and reflect in a different angle when reaching the texture; therefore each scan will be slightly different and once combined the surface pattern will be minimised. (unfortunately not wholly removed in many cases but worth a try)
3. FFT FILTER (FAST FOURIER TECHNIQUE)
The Fast Fourier Transform (FFT) is used in digital image restoration to examine the frequency components of a picture and remove unwanted noise or artefacts.
The FFT is used to convert a picture from the space domain to the frequency domain. The image is shown as a two-dimensional grid of pixels. It’s simple to locate and remove unwanted high-frequency components in the frequency domain, such as those created by noise or motion blur.
When FFT is utilised, the Wiener filter is usually used to correct an image.
The MSE-optimal stationary linear Wiener filter is used to repair pictures that have been ruined by noise and blur. The Wiener filter is calculated on the assumption that the signal and noise processes are second-order stationary (in the random process sense).
It is used to clean up images by removing noise. The Wiener filter employs the FFT to estimate both the noise and the power spectrum of the picture. A frequency-dependent filter is then used to remove the noise from the picture.
FFT is also used to apply inverse filtering to correct fuzzy pictures and deconvolution to remove blur induced by a specified convolution kernel. It is critical to understand that the FFT is a complex operation that requires a significant amount of time and computing resources to operate on huge pictures.
As a result, FFT-based approaches for recovering digital pictures are frequently combined with additional methods, such as spatial filtering, to reduce the overall complexity of the calculations.
An advanced technique to perform photo pattern suppression is using FFT filters, and some FFT algorithms are compelling when processing and interpreting patterns in photographs, to the point that is a weapon of choice to complex activities such as reverse forensic photography analysis.
This technology is available today to a certain extent for post-photography processing, and it’s advantageous,
Finding and extracting textures, fingermarks, patterns and noise signals from a photograph has never been easier thanks to partially automated processing and professional input when dealing with a photo and interpreting its signal-to-noise ratio.
FFT algorithms and Photoshop techniques such as frequency separation and channel blending are successful strategies to identify, reduce or even remove unwanted textures from old photographs.
However, these strategies have a considerable dependency on human input and analysis.
FFT processing is unique in its capacity to identify, optimise and reverse engineer the image signal to noise ratio by suppressing the obstructive patterns and separating frequencies effectively.
Using an FFT equation filter or algorithm to find regular patterns can successfully assist in finding and then reversing the pattern equations and are the best tool available today to the public.
To know more about FFT or download the plugin/filter freeware follow this link FFT Fast Fourier Transform and pattern suppression
4. FREQUENCY SEPARATION
Another pattern suppression technique that can assist in separating textures and background is Frequency Separation, (basically an inverted high-pass)
This process consists decomposing of the image data into different layers so that we can edit image components in the different frequencies independently by subtracting its elements based on the frequency rate of the pixels.
Therefore we can edit the noise in the foreground and the background fill in the background, this is quite useful to perform patching and healing without compromising tonalities and obtaining shades of colour evenly without interfering with the texture.
For more information on Frequency, separation check on this cool youtube video by PHLEARN.
CHOOSE YOUR PATTERN SUPPRESSION TECHNIQUE
We can remove or minimise repetitive patterns using Adobe Photoshop and a selection of plugins, filters and actions that will let us quickly, and with unparalleled quality, remove periodical repeating patterns from images.
However, photo restoration requires the audacity of knowing the best pathway to take, when in a Photoshop workflow, you can achieve excellent results using a ton of different process.
Our recommendation is to take your time and assess the damage or restoration requirements thoroughly before starting any works, this way you can save yourself some time, avoid getting stuck and achieve the best possible outcome.
IS THERE AN FFT (FAST FOURIER TRANSFORM) FILTER FOR MAC?
There are excellent FFT filter software. plugins or actions to remove or suppress repetitive patterns for PC but unfortunately the ones existing on a Macintosh platform are limited, require an special installation or simply old, however even though some can be quite technical to run, we can list a few applications or extensions that you might consider to give it a try:
Have you tried these plugins before or do you know another plugin that can assist the restoration community speeding up this workflow, contact us at email@example.com and tell us about your experience.