Last updated on June 30th, 2020
Topaz Gigapixel AI is an advanced image enlarging software that uses artificial intelligence to upscale photos by up to 600% without losing quality. In many cases, it enhances the final photo by restoring detail with the extra pixels.
In my Topaz Gigapixel AI review, I will provide you important information on how I tested the software and my results.
But first, why would you need to enlarge a photo?
Prints – Large prints require large image resolutions.
Crop Size – Cropping usually reduces the image resolution, especially if you crop to a small sample size of an original image. In this case you may want to upscale the cropped image.
Stock Images – If you sell stock images, you’ll want to offer images in various sizes.
Old Photos – Digital photos from 10+ years ago most likely have small resolutions and you now need a larger version.
Here’s My Topaz Gigapixel AI Review
Now let’s dive into the case studies.
To test Gigapixel AI, I first used a 4032×2268 JPEG image that I photographed with my Samsung Galaxy S8. I enlarged the image by 600%, which is the largest size the software can produce.
I was surprised by the quality; it’s actually good.
I did notice halos along the horizon on the enlarged photo but there were already some artifacts in the original horizon. But it is only noticeable when you really zoom in.
Another thing I noticed is when you zoom in on the original photo, it looks pixelated – as expected. On the other hand, when you zoom in to the same area on the upscaled photo, it looks painterly.
Keep in mind the original image quality is not great in terms of professional photography standards. I did take it with a mobile phone while on a boat (think moving platform). However, this image is good to set a floor quality for Gigapixel AI.
To compare the images yourself you’ll need to see the original and expanded files. The enlarged photo is a whopping 183 MB which I’m not going to load on this blog post.
Lastly, I used a 5496×3670 TIFF image (originally a DNG image) and expanded it by 400%.
The results were amazing. Not only was the final image still sharp, the noise was reduced.
See the before and after. The final image is an insane 1.80 GB.
Topaz Gigapixel AI also has face refinement. When enabled, this feature will enlarge small faces (16×16 to 64×64 pixels) with more detail and sharpness.
A few important notes about Topaz Gigapixel AI:
Gigapixel caps the upscale by either height or width to 22,000 and 32,000 pixels for TIFF and JPEG outputs, respectively. So, you can’t enlarge everything by 6 times.
I tried enlarging a RAW file. The software will not output a RAW file.
Also, Gigapixel’s RAW rendering is not good. The color and contrast look washed out once completed to a TIFF or JPEG file. I tried correcting the color and contrast by playing with the color settings and spaces in Photoshop but it didn’t work.
If you want to enlarge a RAW file, I suggest saving the image as a TIFF file then use Gigapixel to upscale to TIFF.
In addition, it doesnt’ support TIFF layers. If you were to enlarge a TIFF file that has layers, then the layers will be merged.
Gigapixel AI is GPU intensive. It can sometimes take 5 to 10 minutes to process an image. This isn’t a bad thing, since you can do something else in the mean time.
The software has batch processing if you want to process hundres of images. This will take a while.
There is no option to customize resolution ppi during output.
I also tested the facial recognition feature. I used a 640×1030 pixels image of a woman sitting down. The size of the face is small relative to the rest of the image. I enlarged the image by 6 times and compared the details of the before and after face. Gigapixel AI’s results are acceptable.
Topaz Gigapixel AI usually costs $99. Follow this link and use coupon code Aperlust for a 15% discount.
Topaz Labs Utility Bundle
|OS Version||7, 8, 10 64-bit only||macOS 10.12+|
|minimum display size||1024 x 768 pixels||1024 x 768 pixels|
|OpenGL support level||Open GL 3.3 or later||Open GL 3.3 or later|
|minimum||Intel i5 or equivalent (3.0GHz and above)||Ryzen 5 or equivalent (3.0GHz and above)|
|recommended||Intel i7 or greater (4GHz and above)||Ryzen 7 or greater (4GHz and above)|
|minimum||2GB of dedicated VRAM (GT 740 or greater)||2GB of dedicated VRAM (Radeon 5870 or greater)|
|recommended||4GB of dedicated VRAM (GTX 970 or greater)||4GB of dedicated VRAM (Radeon RX 460 or greater)|
|optimal||8GB of dedicated VRAM (GTX 1080 or greater)||8GB of dedicated VRAM (Radeon RX 580 or greater)|
Gigapixel AI does not support Intel HD Graphics 4600 or lower.
A minimum RAM of 8 GB is required. 16 GB is recommended and 32 GB is optimal.
Internet connection is required to log into the software.
On June 25, 2020, Topaz Labs released Gigapixel AI version 5.0.0.
This update includes an improved model for better upscaling, view previews as a single image, new zoom options; and a new man-made mode to better boost the details in typography, architecture, cityscapes, and other images with defined outlines and sharp edges.
I quickly tested the man-made mode on typography. I noticed type was noticeably improved. You’ll notice some smudge on the type, it’s because I used blend modes.
Topaz Gigapixel AI is a great product and it does the job unlike many free online image enlargers which are limited.
I also did a quick test and compared it to Photoshop’s image size tool. I have to give a slight edge to Gigapixel AI because its noise reduction on upscaling is better.
If you’re looking to enlarge a video, then read the Topaz Video Enhance AI review.
Have you tried Gigapixel AI? What were your results?
In digital terms, a gigapixel is one billion (109) pixels or 1,000 megapixels.
Topaz Labs is a software development company that makes a suite of photography applications including plugins for Photoshop and Lightroom.