Cellpic Sunday—Testing ON1’s Free Noise Fix

Night view of fire-ON1 Manual

Shortly after returning home from a recent journey, I discovered that ON1 software is releasing a completely free version of its NoNoise AI application. The giveaway runs only through July 31, 2026, and although ON1 offers these end‑of‑cycle freebies fairly often, this one is the latest 24.5 build—even if the “2024” branding suggests otherwise. Claim it before the deadline, and the license is yours to keep, making it a timely opportunity for anyone who wants cleaner cellpics without investing in a full editing suite. You can find the free version here, along with an offer to purchase the developer’s suite of processing tools.

I thought I’d give the tool a try, as the biggest advantage of ON1’s NoNoise AI is that it works on both JPEG and RAW files. Lightroom Classic’s Denoise remains an excellent option, but it’s limited to RAW processing only. When I installed NoNoise AI, I discovered that the free standalone version still includes plugin integration for both Lightroom and Photoshop—handy for anyone already in the Adobe ecosystem. For this evaluation, I chose two images: one especially challenging and one more typical of an average vacation photo, both processed straight from mobile JPEG captures to see how well the tool handles real‑world noise.

On our journey to Kruger National Park in South Africa, our accommodation was near a wildfire that threatened Skukuza Airport and damaged the terminal building’s thatched roof so badly that, when we departed a few days later, the terminal was closed for repairs. As the blue hour waned, the fire’s glow created an eerie scene that I captured with my Samsung Galaxy S23 Ultra. Auto‑ISO pushed the value to 2500, and the 1/8‑second exposure was heavily influenced by Samsung’s computational filtering, producing a painterly look that made noise reduction especially challenging.

Skukuza Fire
Skukuza Fire-On1 – original

The first thing I learned is that even on a 4K monitor, the differences are subtle. Viewed on this webpage, the improvement is essentially indistinguishable. For anyone who really wants to pixel‑peep, I’ve uploaded all versions to my Flickr site (link below), where you can download the full‑size images for closer inspection on your own equipment. I included three files: the untouched original, ON1’s automatic adjustment—which I found a bit heavy‑handed—and a manual edit using the three noise‑reduction sliders, which produced the most natural and effective result. I also shared a screenshot of the before/after comparison on Flickr. For the manual version, I settled on Luminance 25, Enhance Detail 80, and Color 30, a combination that kept the painterly look intact while reducing the worst of the noise without smearing fine detail any more than Samsung’s computational filters already had.

Krimpen aan de Lek -ON1 Manual
Krimpen aan de Lek – ON1 Manual

A more typical vacation photo can also benefit; even a single shot at a clean ISO 64 baseline, like this image of the Krimpen aan de Lek water tower taken near Kinderdijk in the Netherlands, can be improved. The automatic AI settings of Luminance at 100, Enhance Detail at 50, and Color at 100 are overly aggressive. For this image, I set the Luminance to 57, Enhance Detail to 80, and Color to 26, then selected the 2024 model instead of the default 2024 – High Detail model.

Krimpen aan de Lek Detail
Krimpen aan de Lek Detail at 400%

Here’s a screenshot of the image magnified at 400% in the application. On the left of the vertical slider in the photo is the unprocessed sky and tower, and on the right is the version with noise reduction applied. Even in this zoomed-in view, the improvement is easy to spot: the manual settings smooth out the sky cleanly while preserving a natural texture on the tower’s brickwork. Viewed on a 4K monitor, the difference becomes even more apparent, with the noise reduction producing a noticeably cleaner and more natural result.

The biggest advantage of ON1 NoNoise AI works with JPEG files, though you can also run it on RAW images if you prefer it over Lightroom’s Denoise tool. Regardless of which editor you use, noise reduction needs to happen early in the workflow. Running it before cropping ensures the algorithm has the maximum amount of contextual pixel data to calculate its adjustments accurately. Applying it before lens corrections prevents the software from analyzing pixels that have already been digitally stretched, warped, or vignetted by Lightroom’s profile adjustments. For this comparison, I stopped processing the photos immediately after noise reduction, as any further edits would skew the results. Of course, both images could be improved with additional adjustments—but that’s for another day.

As I noted earlier, the images are available at the link below. For those who want to take a closer look, it’s best to download the files and review them on a 4K monitor. Even then, the Skukuza image shows a subtle improvement at 4K, while the sky surrounding the Krimpen aan de Lek water tower benefits much more noticeably from manual smoothing. If you’re interested in evaluating the application on your own computer, be sure to check ON1’s machine requirements first—its AI models lean heavily on GPU performance, and older or underpowered systems may struggle to run the tool efficiently.

While I have relied on Luminar Neo’s Noiseless module for mobile JPEG noise reduction in the past, testing this free version of ON1 shows it to be a superior option. It’s a solid addition to my workflow, especially for cellpics that need a gentle touch. Just remember that this free, unlimited software license is only available until July 31, 2026—after that, the offer disappears.

I invite fellow bloggers to join in by creating their own Cellpic Sunday posts. There’s no set theme. The first rule is that your photo must be captured with a cell phone, iPad, or other mobile device. The second rule is simple: link your challenge response to this post or drop a comment here with your link so others can find it. And remember, despite the name, there’s no penalty for sharing on a day other than Sunday.

John Steiner

2 comments

Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.