We subjected the Samsung Galaxy S23 to our rigorous SBMARK Camera test suite to measure its performance in terms of photo, video and zoom quality from an end-user perspective. This article analyzes how the device performed in a variety of tests and several common use cases and aims to highlight the most important results of our tests with an excerpt of the data captured.
Overview
Main camera specs:
- Primary: 50MP 1/1.56″ sensor, 1.0μm pixels, f/1.8 aperture lens, PDAF, OIS
- Ultra wide angle: 12MP 1/2.55″ sensor, 1.4µm pixels, f/2.2 aperture lens, super stable video
- Tele: 10 MP 1/3.94″ sensor, 1.0 µm pixels, f/2.4 aperture lens, PDAF, OIS
Pros
- Wide enough dynamic range
- Accurate color rendering under most conditions
- Good fine detail in all conditions
- Good close-up detail and medium telephoto in daylight and indoors
- Good focus tracking in videos
Against
- Autofocus failures in low light and indoors
- Local loss of consistency in daylight
- Image noise
- Artifacts, including face blending or rendering artifacts
- Underexposed video in difficult low light conditions
With a SBMARK Camera score of 133, the Samsung Galaxy S23 performed well in our tests, thanks to a very well-balanced performance, making it a good option for a wide variety of users. It was capable of capturing high quality images in most lighting conditions and scenes, including night landscapes.
The camera’s 50MP image sensor was capable of capturing high levels of detail, and the images displayed bright, natural colors. However, our testers also noted a loss of local detail in some situations, especially when shooting in daylight, where the Samsung also struggled with skin tone rendering. In video mode the stabilization was effective, and the S23’s autofocus system allowed for smooth focus transitions and good detail, making the device a great option for mobile video creators as well.
Compared to the flagship S23 Ultra, the S23 uses a smaller image sensor and overall different camera hardware. However, in many situations the S23 came very close to the Ultra model in terms of image quality, making it a viable alternative to the top-of-the-range model.
Trial summary
About SBMARK Camera Tests: SBMARK’s camera evaluations are conducted in labs and in real-world situations using a wide variety of subjects. Scores are based on objective tests for which results are calculated directly by the measurement software on our lab setups and perceptual tests where a sophisticated set of metrics allows a panel of image experts to compare aspects of quality of the image that require human judgment. Testing a smartphone involves a team of engineers and technicians for about a week. Photo, zoom, and video quality are rated separately and then combined into an overall score for comparing cameras from different devices. For more information on the SBMARK camera protocol, click here. More details on smartphone camera scores can be found here. The following section compiles the key elements of SBMARK’s exhaustive testing and analysis. Full performance evaluations are available upon request. Please contact us on how to receive a full report.
Samsung Galaxy S23 camera scores
This graph compares the SBMARK photo, zoom and video scores between the tested device and the references. The average and maximum scores of the price segment are also indicated. The average and maximum scores for each price segment are calculated based on the SBMARK database of tested devices.
Photo
129
Huawei Mate 50 Pro
Huawei Mate 50 Pro
SBMARK camera photo test information
For scoring and analysis, SBMARK engineers capture and evaluate more than 2,600 test images in both controlled laboratory environments and outdoor, indoor, and low-light natural scenes using camera presets. The photographic protocol is designed to take into account the main use cases and is based on typical shooting scenarios, such as portrait, family and landscape photography. Evaluation is performed by visually inspecting images Cons a natural scene reference and by performing objective measurements on laboratory-acquired graph images under various lighting conditions from 1 to 1,000+ lux and color temperatures from 2,300 K to 6,500 K.
Samsung Galaxy S23 photo scores
Photo tests analyze image quality attributes such as exposure, color, texture and noise under various lighting conditions. Autofocus performance and the presence of artifacts are also evaluated on all images acquired under controlled laboratory conditions and on real-life images. All of these attributes have a significant impact on the final quality of the images captured with the tested device and can help you understand the main strengths and weaknesses of the camera.
Exposure
106
Huawei Mate 50 Pro
Huawei Mate 50 Pro
Color
111
Google Pixel 7 Pro
Google Pixel 7 Pro
Exposure and color are key attributes for technically good images. For exposure, the main attribute evaluated is the brightness of the main subject across various use cases such as landscape, portrait or still life. Other factors evaluated are contrast and dynamic range, e.g. the ability to make details visible in light and dark areas of the image. Repeatability is also important because it demonstrates the camera’s ability to provide the same rendering when taking multiple images of the same scene.
For color, the image quality attributes analyzed are skin tone rendering, white balance, color shading, and repeatability. With regards to color rendition and skin tone, we penalize unnatural colors but respect the manufacturer’s choice of color signature.
Dynamic range and target exposure were very similar to the Galaxy S23 Ultra, but the S23 maintained slightly better contrast on faces than the Ultra in backlit scenes. HDR rendering tended to be slightly aggressive in the highlights for both devices. The color was also very similar, but there are some slight differences in terms of white balance at times.
Samsung Galaxy S23 – Green Cast
Samsung Galaxy S23 Ultra – Cool white balance
Apple iPhone 14 Pro – Neutral white balance
Auto focus
105
Huawei Mate 50 Pro
Huawei Mate 50 Pro
Autofocus tests focus on focus accuracy, focus repeatability, shutter lag, and depth of field. Shutter lag is the difference between when the user presses the capture button and when the image is actually taken. It includes the speed of focusing and the ability of the device to capture images at the right moment, the so-called “zero shutter lag” capability. While a shallow depth of field can be nice for a single-subject portrait or close-up shot, it can also be a problem in some specific conditions like group portraits; Both situations are tested. Focus accuracy is also evaluated in all real-life images taken, from infinity to close-up objects and low-light to outdoor conditions.
The Samsung Galaxy S23 catches quite quickly in good light. However, in more difficult lighting conditions, the subject is occasionally missed, resulting in focus problems.
Autofocus Ruggedness and Speed: Daylight 1000Lux Δ2EV Tripod
This graph illustrates focus accuracy and speed, and also zero shutter lag capability by showing edge acuity versus shooting time measured on the AFHDR setup over a series of images. All photos were taken at 1000 Lux with Daylight illuminant, 500 ms after blurring. In this scenario, the backlit panels in the scene are set to simulate a fairly high dynamic range: the luminance ratio between the brightest spot and an 18% reflective gray spot is 2, which we denote by an exposure value difference of 2 Edge sharpness is measured on the four edges of the Dead Leaves graph and shooting time is measured on the Universal Timer LED.
Autofocus Ruggedness and Speed: 100Lux Δ0EV TL84 Tripod
This graph illustrates focus accuracy and speed, and also zero shutter lag capability by showing edge acuity versus shooting time measured on the AFHDR setup over a series of images. All photos were taken at 100 Lux with TL84 illuminant, 500 ms after blurring. Edge sharpness is measured on the four edges of the Dead Leaves graph and shooting time is measured on the Universal Timer LED.
Structure
108
Oppo Find X5
Oppo Find X5
Texture tests analyze the level of detail and texture of subjects in images taken in the lab as well as real-life scenarios. For natural shots, special attention is paid to the level of detail in the light and dark areas of the image. 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 charts used are the proprietary SBMARK (DMC) chart and the Dead Leaves chart.
SBMARK CHART (DMC) Detail Retention Score vs. Lux Levels for Tripod and Handheld Conditions
This graph shows the evolution of the DMC Detail Retention score with lux level, for two retention conditions. The DMC Detail Retention Score is derived from an AI-powered metric trained to evaluate texture and detail rendition on a selection of crops from our SBMARK chart.
Noise
85
Honor Magic4 Ultimate
Honor Magic4 Ultimate
Noise tests analyze various noise attributes such as intensity, chromaticity, grain, texture on real-life images and lab-captured graph images. For natural images, special attention is paid to noise on faces, landscapes, but also dark areas and high dynamic range conditions. Noise on moving objects is also evaluated on natural images. Objective measurements are performed on chart images captured under various conditions from 1 to 1000 lux and different types of dynamic range conditions. The graph used is the dead leaves graph and standardized measurement such as visual noise derived from ISO 15739.
Evolution of visual noise with illuminance levels in portable conditions
This graph shows the evolution of the visual noise metric with lux level under palmar conditions. The Visual Noise metric is the average of the Visual Noise measurement across all patches of the Dead Leaves graph in the AFHDR configuration. The SBMARK visual noise measurement is derived from the ISO15739 standard.
Artifacts
66
Google Pixel 6
Google Pixel 6
Artifact assessment examines lens shading, chromatic aberrations, geometric distortion, ringing edges, halos, ghosting, quantization, unexpected color tone shifts, among other types of possible effects unnatural on photos. The more serious and frequent the artifact, the greater the deduction of points from the score. The main artifacts observed and the corresponding loss of points are listed below.
Major penalties for photographic artifacts
Bokeh
70
Huawei Mate 50 Pro
Huawei Mate 50 Pro
Bokeh is tested in a dedicated mode, usually portrait or aperture mode, and analyzed by visually inspecting all images captured in laboratory and natural conditions. The goal is to reproduce a portrait photograph comparable to one taken with a DLSR and a wide aperture. The main image quality attributes paid attention to are depth estimation, artifacts, blur gradient, and bokeh blur reflector shape. Vertical image quality attributes (exposure, color, texture) are also taken into consideration.
Samsung Galaxy S23 – Accurate depth estimation but some obvious errors on watermarked objects
Preview
72
Apple iPhone 14 Pro Max
Apple iPhone 14 Pro Max
The preview tests analyze the image quality of the camera app’s image preview, with particular attention to the difference between the capture and the preview, especially regarding the dynamic range and the application of the bokeh effect. The uniformity of exposure, color and focus adaptation when zooming from the minimum to the maximum available zoom factor is also evaluated. The preview frame rate is measured using the Universal Timer LED.
Samsung Galaxy S23 – Preview
Samsung Galaxy S23 – Capture – Noticeably wider dynamic range than preview
Zoom in
126
Honor Magic4 Ultimate
Honor Magic4 Ultimate
About SBMARK Camera Zoom Tests
SBMARK engineers capture and evaluate over 400 test images in controlled laboratory environments and natural outdoor, indoor, and low-light scenes using camera presets and pinch zoom at various zoom factors from ultra wide to very long zoom. Evaluation is performed by visually inspecting images Cons a natural scene reference and by performing objective measurements of lab-captured map images under varying conditions from 20 to 1000 lux and color temperatures from 2300K to 6500K.
Samsung Galaxy S23 zoom scores
This graph illustrates the relative scores for the different zoom ranges evaluated. The abscissa is expressed in focal length equivalent to 35mm. Zoomed scores appear on the right and zoomed scores on the left.
Wide
102
Huawei Mate 50 Pro
Huawei Mate 50 Pro
These tests analyze the performance of the ultra wide angle camera at different focal lengths from 12mm to 20mm. All attributes of image quality are evaluated, with particular attention to artifacts such as chromatic aberrations, lens softness and distortion. The images below are an excerpt from the tested scenes.
Samsung Galaxy S23 – Wide dynamic range, slight noise, slight loss of detail
Samsung Galaxy S23 Ultra – Wide dynamic range, slight noise, slight loss of detail
Apple iPhone 14 Pro – Slight clipping in highlights, noise, loss of detail
tele
89
Honor Magic4 Ultimate
Honor Magic4 Ultimate
All image quality attributes are evaluated at focal lengths from approximately 40mm to 300mm, with an emphasis on texture and detail. The score comes from a series of objective laboratory measurements and perceptual analysis of real-life images.
SBMARK CHART Detail Retention Score (DMC) by focal length
This graph shows the evolution of the DMC Detail Retention score versus the full-frame equivalent focal length for different lighting conditions. The x-axis represents the equivalent focal length measured for each corresponding shooting distance, and the y-axis represents the maximum detail retention metric score: higher value means better quality. Large dots correspond to the zoom ratio available in the camera application UI.
SBMARK CHART Detail Retention Score (DMC) by focal length
This graph shows the evolution of the DMC Detail Retention score versus the full-frame equivalent focal length for different lighting conditions. The x-axis represents the equivalent focal length measured for each corresponding shooting distance, and the y-axis represents the maximum detail retention metric score: higher value means better quality. Large dots correspond to the zoom ratio available in the camera application UI.
SBMARK CHART Detail Retention Score (DMC) by focal length
This graph shows the evolution of the DMC Detail Retention score versus the full-frame equivalent focal length for different lighting conditions. The x-axis represents the equivalent focal length measured for each corresponding shooting distance, and the y-axis represents the maximum detail retention metric score: higher value means better quality. Large dots correspond to the zoom ratio available in the camera application UI.
SBMARK CHART Detail Retention Score (DMC) by focal length
This graph shows the evolution of the DMC Detail Retention score versus the full-frame equivalent focal length for different lighting conditions. The x-axis represents the equivalent focal length measured for each corresponding shooting distance, and the y-axis represents the maximum detail retention metric score: higher value means better quality. Large dots correspond to the zoom ratio available in the camera application UI.
video
137
Apple iPhone 14 Pro Max
Apple iPhone 14 Pro Max
About SBMARK Camera Video Tests
SBMARK engineers capture and evaluate more than 2.5 hours of video in controlled lab environments and low natural light scenes, indoors and out, using the camera’s default settings. Evaluation consists of visually inspecting natural video taken under various conditions and performing objective measurements on lab-recorded graph video under various conditions from 1 to 1000+ lux and color temperatures from 2,300K to 6,500K.
Samsung Galaxy S23 video scores
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
103
Apple iPhone 14 Pro Max
Apple iPhone 14 Pro Max
Color
105
Apple iPhone 14 Pro Max
Apple iPhone 14 Pro Max
Exposure tests evaluate the brightness of the main subject and the dynamic range, e.g. the ability to make details visible in light and dark areas of the image. The stability and temporal adaptation of the exposure are also analysed.
Image quality color analysis examines color rendition, skin tone rendition, white balance, color shading, white balance stability and its adaptation when the light changes.
Samsung Galaxy S23 – Good exposure, nice skin tones
Samsung Galaxy S23 Ultra – Good exposure, nice skin tones
Samsung Galaxy S22 Ultra – Good exposure, nice skin tones
Structure
105
Xiaomi Mi 11 Ultra
Xiaomi Mi 11 Ultra
Texture tests analyze the level of detail and texture of real-life videos as well as graphics videos recorded in the lab. Natural video recordings are evaluated visually, with particular attention to the level of detail in bright areas and dark areas. Objective measurements of card images taken under various conditions from 1 to 1000 lux are performed. The charts used are the SBMARK (DMC) chart and the Dead Leaves chart.
SBMARK CHART (DMC) video score for detail retention versus lux levels
This graph shows the evolution of the video DMC detail retention score with the level of lux in the video. The DMC Detail Retention Score is derived from an AI-powered metric trained to evaluate texture and detail rendition on a selection of crops from our SBMARK chart.
Noise
106
Samsung Galaxy A23 5G
Samsung Galaxy A23 5G
Noise tests analyze various noise attributes such as intensity, chromaticity, grain, texture, temporal aspects on real-life video recording, as well as graph videos taken in the lab. Natural videos are evaluated visually, with an emphasis on noise in dark areas and high dynamic range conditions. 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 level of illumination
This graph shows the evolution of spatial visual noise with lux level. Spatial visual noise is measured on the visual noise table in the video noise setup. The SBMARK visual noise measurement is derived from the ISO15739 standard.
Time evolution of visual noise with level of illumination
This graph shows the evolution of visual noise over time with lux level. Temporal visual noise is measured on the visual noise table in the video noise configuration.
Stabilization
114
Apple iPhone 14 Pro Max
Apple iPhone 14 Pro Max
The stabilization rating tests the device’s ability to stabilize footage using software or hardware technologies such as OIS, EIS, or any other means. The evaluation looks at residual motion, smoothness, gel artifacts, and residual motion blur in walking and running use cases under various lighting conditions. The video below is an excerpt from one of the tested scenes.
Samsung Galaxy S23 – Effective stabilization
Samsung Galaxy S23 Ultra – Effective stabilization
Samsung Galaxy S22 Ultra – Effective stabilization, sharpness difference between frames
Artifacts
79
Xiaomi 12S Ultra
Xiaomi 12S Ultra
Artifacts are evaluated with MTF and ringing measurements on the SFR graph in the lab, as well as frame rate measurements using the Universal Timer LED. Natural videos are visually evaluated by paying close attention to artifacts such as aliasing, quantization, blocking, and pitch shifting, among others. The more severe and frequent the artifact, the more points will be deducted from the score. The main artifacts and the corresponding point loss are listed below.
Top penalties for video artifacts
Start a new Thread