We put the Xiaomi 13 Ultra through our rigorous SBMARK Selfie test suite to measure its performance in photos and videos from an end-user perspective. This article analyzes the behavior of the device in a series of tests and several common use cases and aims to highlight the most important results of our tests with an excerpt of the acquired data.
Overview
Front camera main specifications:
- 32MP sensor
- Lens with f/2.0 aperture
- 1080p/30fps
Pros
- Good texture/noise compromise in photos and videos
- Good video stabilization when holding the device still and moving during recording
- Good exposure in most situations
Against
- Tone compression quite strong in backlit scenes
- A narrow dynamic range results in clipping of highlights
- The skin enhancement effect may result in unrealistic skin textures
In the SBMARK Selfie tests, the Xiaomi 13 Ultra delivered an overall decent performance and showed some significant improvements over its predecessor 12S Ultra, especially in terms of photo color and exposure. Furthermore, video stabilization was more effective, guaranteeing higher levels of detail. The downside is that the shallow depth of field prevented good sharpness on all subjects in group shots, just like in the 12S Ultra, and in some difficult lighting conditions, we observed strong tonal compression, which resulted in a lack of contrast on the face, as well as unnatural colors.
While the Xiaomi 13 Ultra couldn’t match the overall front camera results of some Ultra-Premium rivals like the Apple iPhone 14 Pro or the Huawei Mate 50 Pro, it did better in some areas. Unwanted image artifacts were better under control than on the Huawei Mate 50 Pro; Noise reduction worked more effectively and flash mode was better than on the iPhone 14 Pro.
Xiaomi 13 Ultra Selfie scores Cons Ultra-Premium
This chart compares overall SBMARK Selfie scores for photos and videos between tested and reference devices. The average and maximum scores of the price range are also indicated. The average and maximum scores for each price segment are calculated based on the SBMARK device database.
Test summary
About SBMARK selfie tests: For scoring and analysis, SBMARK engineers capture and evaluate more than 1,500 test images in both controlled laboratory environments and natural outdoor, indoor, and low-light scenes, using default front camera settings. The photography protocol is designed to take user needs into account and is based on typical shooting scenarios, such as close-ups and group selfies. Evaluation is performed by visually examining images Cons a natural scene reference and performing objective measurements on laboratory-captured graph images under varying lighting conditions from 1 to 1,000+ lux and color temperatures from 2,300 K to 6,500 K. Learn more information about the SBMARK Selfie test protocol, click here. More details on how we rate smartphone cameras can be found here. The following section compiles key elements of SBMARK’s comprehensive testing and analysis. Full performance evaluations are available upon request. Please contact us to find out how to receive a full report.
Photo
123
Huawei Mate 50 Pro
Huawei Mate 50 Pro
Xiaomi 13 Ultra Photo scores Cons Ultra-Premium
Photography tests analyze image quality attributes such as exposure, color, texture and noise under various lighting conditions. The focus range and presence of artifacts are also evaluated on all images captured under controlled laboratory conditions and in real-life images. All of these attributes have a significant impact on the final quality of images captured with the tested device and can help understand the camera’s key strengths and weaknesses.
In photo mode, the Xiaomi 13 Ultra’s front camera produced selfie images with good detail and accurate exposure. However, the skin smoothing effect that is activated automatically could be quite strong. The bokeh mode wasn’t among the best we’ve seen. Subject isolation left some room for improvement, and the lack of a blur gradient resulted in an unrealistic effect that didn’t really provide a sense of depth. When capturing images with flash enabled, noise was well under control on faces, but vignetting around the edge of the frame was quite strong.
Exposure
75
Apple iPhone 15 Pro Max
Apple iPhone 15 Pro Max
Color
87
Google Pixel 8 Pro
Google Pixel 8 Pro
Exposure and color are the key attributes for technically good images. For exposure, the main attribute evaluated is the brightness of faces under various use cases and lighting conditions. Other factors evaluated are contrast and dynamic range, e.g. the ability to make details visible in both bright and dark areas of the image. Repeatability is also important because it demonstrates the camera’s ability to provide the same rendering when shooting consecutive images in a row.
Regarding color, the image quality attributes analyzed are skin tone rendering, white balance, color shading and repeatability.
In our tests, the Xiaomi generally exposed subjects well, except in low-light conditions, where our testers noticed slight underexposure. The dynamic range was quite narrow, which resulted in frequent highlight clipping in high-contrast scenes. Color was handled well, but we often observed severe tonal compression, which resulted in unnatural skin textures and low contrast on the subject’s face, especially when shooting in daylight.
Xiaomi 13 Ultra – Strong tonal compression
Huawei Mate 50 Pro – Natural result
Apple iPhone 14 Pro – Light pink/orange shade
Focus
78
Apple iPhone 15 Pro Max
Apple iPhone 15 Pro Max
Our testers found that the camera provides good focus as long as the subject is a short distance from the lens. However, a shallow depth of field means that the Xiaomi 13 Ultra is not the best option for taking group selfies when subjects are at different distances from the camera. While the closest face will be sharp, subjects further away will be out of focus.
Autofocus tests evaluate the precision of focus on the subject’s face, the repeatability of accurate focus, and the depth of field. While a shallow depth of field can be nice for a single-subject selfie or close-up shot, it can be problematic in specific conditions like group selfies; both situations are tested. Focus accuracy is also evaluated in all real images taken, from 30cm to 150cm, and in low-light or outdoor conditions.
Xiaomi 13 Ultra – Depth of field
Xiaomi 13 Ultra – Blurred background subjects
Huawei Mate 50 Pro – Depth of field
Huawei Mate 50 Pro – Most faces are in focus, the subject behind is slightly blurry
Apple iPhone 14 Pro – Depth of field
Apple iPhone 14 Pro: all faces in focus, even in the background
Structure
71
Asus ZenFone 7 Pro
Asus ZenFone 7 Pro
Texture tests analyze the level of detail and texture of subjects in images taken in the lab and in real-life scenarios. For natural shots, special attention is paid to the level of detail of facial features, such as the eyes. Objective measurements are performed on map images taken under various lighting conditions from 1 to 1000 lux and different types of dynamic range conditions. The papers used are the proprietary SBMARK (DMC) paper and the Dead Leaves paper.
The Xiaomi 13 Ultra managed to limit noise levels on the face, while maintaining a decent skin texture. In low light, it performed better than the Apple iPhone 14 Pro, but couldn’t match the low noise levels of the Huawei Mate 50 Pro.
Evolution of texture sharpness with illuminance level
This graph shows the evolution of texture sharpness with lux level for two holding conditions. Texture sharpness is measured on the Dead Leaves graph in the Close-up Dead Leaves setting.
Xiaomi 13 Ultra – Detail and noise
Xiaomi 13 Ultra – Loss of facial details
Huawei Mate 50 Pro – Detail and noise
Huawei Mate 50 Pro – Good details but unnatural textures in the eye area
Apple iPhone 14 Pro – Detail and noise
Apple iPhone 14 Pro: good details and natural textures
Noise
72
Huawei Mate 50 Pro
Huawei Mate 50 Pro
Noise tests analyze various noise attributes such as intensity, chromaticity, grain and texture on real images and on graph images taken in the laboratory. For natural images, particular attention is paid to noise on faces, but also to dark areas and high dynamic range conditions. Objective measurements are performed on graph images taken under various conditions from 1 to 1000 lux and different types of dynamic range conditions. The graph used is the SBMARK Dead Leaves graph and the standardized measurement such as visual noise derived from ISO 15739.
Evolution of visual noise with illuminance levels under handheld conditions
This graph shows the evolution of the visual noise metric with lux level in handheld conditions. The visual noise metric is the average of the visual noise measurement across all areas of the Dead Leaves graph in the Close-up Dead Leaves setting. SBMARK visual noise measurement is derived from the ISO15739 standard.
Artifacts
82
Apple iPhone 15 Pro Max
Apple iPhone 15 Pro Max
Artifact evaluation looks at lens shading, chromatic aberrations, distortion measurements on the Dot and MTF graph, and sound measurements on the SFR graph in the lab. Particular attention is paid to ghosting, quantization, halos and tone changes on the face, among others. The more severe and frequent the artifact, the greater the point deduction from the score. The main artifacts observed and the corresponding point loss are listed below.
In our tests, the Xiaomi 13 Ultra tended to apply a strong skin-smoothing effect to its images, which could result in an unnatural appearance. In difficult lighting conditions, such as backlit scenes, our testers observed additional artifacts, including halo effects, tone compression, and hue shift near saturated areas of the image. Additionally, anamorphosis may be slightly noticeable in close-up selfies, with the face close to the edge of the frame.
Major penalties for photography artifacts
video
137
Apple iPhone 15 Pro Max
Apple iPhone 15 Pro Max
About SBMARK selfie video tests
SBMARK engineers capture and evaluate more than 2 hours of video in controlled laboratory environments and natural low-light scenes, indoors and outdoors, using default front camera settings. The evaluation consists of visual inspection of natural videos taken under various conditions and performing objective measurements on videos of graphs recorded in the laboratory under different conditions from 1 to 1000+ lux and color temperatures from 2,300 K to 6,500 K.
Compared to its predecessor 12SUltra, the video mode of the Xiaomi 13 Ultra showed improvements in various areas. Video noise levels were lower, stabilization was smoother, and the level of recorded detail was increased at all light levels. On the downside, we noticed occasional underexposure in low-light scenes, and just like photo mode, dynamic range was limited. Rivals like the Apple iPhone 14 Pro or the Huawei Mate 50 Pro fared better in this regard, with a well-managed HDR mode that expands the dynamic range of video footage.
Xiaomi 13 Ultra Video scores compared to Ultra-Premium
Video tests analyze the same image quality attributes as still images, such as exposure, color, texture, or noise, as well as temporal aspects such as speed, exposure uniformity and stability, white balance, and autofocus transitions.
Exposure
79
Apple iPhone 15 Pro Max
Apple iPhone 15 Pro Max
Color
80
Apple iPhone 15 Pro Max
Apple iPhone 15 Pro Max
Exposure tests evaluate facial brightness and dynamic range, e.g. the ability to make details visible in both bright and dark areas of the image. The stability and temporal adaptation of the exposure are also analyzed. Image quality color analysis examines skin tone rendering, white balance, color shading, white balance stability and how it adapts when the light changes.
Facial exposure remained generally accurate in bright light and indoor conditions, but the limited dynamic range resulted in clipping of background highlights in many scenes. In low-light conditions, our testers often found the subject to be underexposed.
Xiaomi 13 Ultra – Stable white balance and facial exposure, highlight clipping in the background
Huawei Mate 50 Pro – Wide dynamic range and accurate face exposure (HDR video), stable white balance
Apple iPhone 14 Pro – Wide dynamic range and accurate face exposure (HDR video), stable white balance
Structure
74
Asus Zenfone 6
Asus Zenfone 6
Texture tests analyze the level of detail and texture of real videos and graphics videos recorded in the lab. Natural video footage is evaluated visually, with particular attention to the level of detail of facial features. Objective measurements are performed on chart images taken under various conditions from 1 to 1000 lux. The chart used is the Dead Leaves chart.
Evolution of texture sharpness with illuminance level
This graph shows the evolution of texture sharpness with lux level for two holding conditions. Texture sharpness is measured on the Dead Leaves graph in the Close-up Dead Leaves setting.
Noise
71
Xiaomi Mi 11 Ultra
Xiaomi Mi 11 Ultra
Noise tests analyze various noise attributes such as intensity, chromaticity, grain, structure, temporal aspects on real video recordings and on videos of graphs taken in the laboratory. Natural videos are evaluated visually, with particular attention to noise on faces. Objective measurements are performed on graph videos recorded under various conditions from 1 to 1000 lux. The graph used is the SBMARK visual noise graph.
Evolution of spatial visual noise with illuminance level
This graph shows the evolution of spatial visual noise with lux level. Spatial visual noise is measured on the visual noise graph in the video noise setup. SBMARK visual noise measurement is derived from the ISO15739 standard.
Temporal evolution of visual noise with illuminance level
This graph shows the evolution of temporal visual noise with lux level. Temporal visual noise is measured on the visual noise graph in the video noise setup.
Video noise was handled quite well by the Xiaomi 13 Ultra, especially on faces. Noise reduction proved to be very effective even in low light conditions, removing noise without sacrificing texture.
Stabilization
78
Apple iPhone 15 Pro Max
Apple iPhone 15 Pro Max
The stabilization rating checks the device’s ability to stabilize footage thanks to software or hardware technologies such as OIS, EIS or any other means. The evaluation examines overall residual motion on the face and background, smoothness, and gel artifacts, during walking and panning use cases under various lighting conditions. The video below is an excerpt of one of the scenes tested.
The Xiaomi 13 Ultra’s video stabilization is solid; no gelling was noticed, and overall performance was on par with class-leading devices, such as the Apple iPhone 14 Pro and Huawei Mate 50 Pro. However, we did notice some camera shake at the start of video recording. Note that this affected the artifact score rather than the stabilization score.
Xiaomi 13 Ultra – good stabilization, slight camera movement noticeable in the background
Huawei Mate 50 Pro – Good stabilization, good background detail
Apple iPhone 14 Pro – Good stabilization, good background detail
Artifacts
76
Apple iPhone 12 mini
Apple iPhone 12 mini
Artifacts are evaluated with MTF and ringing measurements on the SFR graph in the lab, as well as frame rate measurements using the LED universal timer. Natural videos are visually evaluated by paying particular attention to artifacts such as quantization, hue shift, and face rendering artifacts, among others. The more severe and frequent the artifact, the greater the deduction of points from the score. The main artifacts and the corresponding point loss are listed below
Top penalties for video artifacts
Start a new Thread