Compare 30cm vs 50cm Satellite Imagery: Use Cases, Difference & Which to Choose

30cm vs 50cm satellite imagery, in short: 30cm imagery has pixels that cover 0.3m of ground, so it shows about 2.8 times more pixels per km² than 50cm imagery (11.1 million vs 4 million). Choose 30cm when you need to recognise small objects such as vehicle types, road markings, rooftop equipment or utility poles, or when you need sharp stereo for 3D. Choose 50cm when you are mapping large areas, building basemaps, or monitoring land, crops and corridors, because it costs less per km² and some 50cm satellites image strips up to 130 km wide. XRTech Group supplies both: native 30cm from $20/km² and native 50cm from $13/km² in the archive, worldwide.

Picking between 30cm and 50cm satellite imagery is one of the most common decisions in a geospatial project, and the wrong choice is costly either way. Order 30cm for a whole province and you may pay a large premium for detail nobody uses. Order 50cm for an asset survey and your team may not be able to tell a pole from its shadow. This guide compares the two resolutions on every parameter that matters (GSD, accuracy, level of detail, swath, price, data volume, processing and value-added products) and ends with a simple rule for choosing. The specifications and prices come from the satellites and products XRTech Group supplies every day.

30cm vs 50cm Satellite Imagery: Quick Comparison Table

The pixel figures follow directly from the GSD: a 1 km² square holds 1,000 ÷ 0.3 ≈ 3,333 pixels per side at 30cm (about 11.1 million pixels) and 2,000 per side at 50cm (4 million pixels). The data volumes assume four bands at 16 bits per pixel with no compression; delivered GeoTIFFs are usually smaller.

Native 30cm SuperView Neo-1 imagery (left, Las Vegas) next to native 50cm SuperView Neo-3 imagery (right, Washington DC).

What Is the Difference Between 30cm and 50cm Satellite Imagery?

The difference is the size of the ground area each pixel represents, called the ground sample distance (GSD). In 30cm imagery every pixel covers 30 × 30 cm on the ground; in 50cm imagery every pixel covers 50 × 50 cm. Because detail scales with area, not width, a 30cm image carries about 2.8 times more pixels than a 50cm image of the same place. That extra density is what turns a car-shaped blob into a car you can classify as a hatchback, a van or a pickup. Learn more about the measurement itself in our guide to ground sample distance.

Ground sample distance is the size of the ground area one pixel represents: the smaller the GSD, the more detail each image holds.

Native Resolution vs Resampled Resolution

A resolution figure is only meaningful if it is native, meaning the sensor recorded that GSD on its focal plane. Some imagery sold as 30cm is resampled from 50cm data: it looks crisp on screen, but the extra pixels hold no new ground information, so edges blur, measurements drift and AI models learn from invented detail. Every 30cm and 50cm product XRTech Group supplies is native on the sensor. When you compare offers, ask for the native panchromatic GSD of the satellite, not just the delivered pixel size.

Why Viewing Angle Changes Real Resolution

Satellites rarely look straight down. When a 30cm satellite images at a steep off-nadir angle, the effective GSD on the ground grows, and tall buildings lean more. For measurement work, set a viewing-angle limit on the order. At XRTech Group, a 15° or lower angle is a $3/km² add-on and 10° or lower is $9/km², on both resolutions.

Level of Detail: What You Can See at 30cm vs 50cm

The practical question is not "which is sharper" but "can I see the object I care about". The table below converts real object sizes into pixels at each resolution.

Neither resolution can identify a person or read a number plate. A person is only one or two pixels across even at 30cm. For finer work, XRTech Group also supplies 25cm satellite imagery, and for wider areas at lower cost there is 80cm imagery.

Native 30cm satellite imagery of Las Vegas from SuperView Neo-1: vehicle types, parking bays, pools and rooftop equipment are distinct.

Native 50cm satellite imagery of Washington DC from SuperView Neo-3: buildings, roads, lanes, cars and tree rows are clear across a wide area.

Does Higher Resolution Improve AI Object Detection?

Yes, measurably. In a published CosmiQ Works study on satellite object detection, model accuracy (mAP) was 0.53 at 30cm and fell to 0.11 at 4.8m, and a follow-up analysis reported a 22% to 27% drop in performance between 30cm and 120cm. There is no widely published benchmark that isolates 30cm against 50cm, so test a labelled sample of your own area before committing a large order. As a rule of thumb, small-object models (vehicles, poles, solar panels, small boats) benefit most from 30cm, while building-footprint and land-cover models usually perform well at 50cm.

Accuracy: CE90, Orthorectification and Map Scale

Resolution and positional accuracy are different things. GSD tells you how small a feature you can see; CE90 tells you how close each pixel is to its true position on Earth (90% of points fall within the stated radius). Without ground control, SuperView Neo-1 30cm imagery is rated at 3.5m CE90, while the 50cm satellites XRTech Group supplies are rated at about 9.5m to 10m CE90. Orthorectification with a good elevation model and ground control points improves positional accuracy for both. See our guide to CE90 horizontal accuracy for how to specify it in a tender.

CE90 is the radius within which 90% of image points fall from their true position on Earth.

Map scale guidance varies by country. An ISPRS rule of thumb ties map accuracy to about 0.25 mm at map scale, which puts 0.5m GSD in the range needed for 1:10,000 mapping, while some national standards ask for finer GSD at each scale. Check the specification your project follows before ordering.

Coverage, Swath and Revisit

Coverage is where 50cm pulls ahead. Most 30cm satellites image a strip about 12 km wide, so a large province needs many passes and weather windows. 50cm satellites range from 12 km to 23.5 km per pass, and SuperView Neo-3 images a 130 km strip at 50cm, more than ten times the width of a typical 30cm satellite. For large areas that means fewer collection windows, fewer cloud gaps and a more uniform mosaic.

Swath width is the strip a satellite images as it flies; wider swaths cover large areas in fewer passes.

Revisit also differs. SuperView-1 can revisit daily at under 45° off-nadir, SuperView Neo-3 every one to three days, and the 30cm SuperView Neo-1 offers daily-capable revisit. Other 30cm-class systems on the market include WorldView-3, WorldView Legion and Pléiades Neo.

30cm vs 50cm Satellite Imagery Price

XRTech Group publishes fixed prices per km² for both resolutions, for any location worldwide. Here they are side by side:

XRTech Group portal archive search: preview scenes over your area with capture date, cloud cover and an instant price before ordering.

Worked example for 1,000 km²: archive coverage costs about $20,000 at 30cm list price and about $13,000 to $14,000 at 50cm, before volume pricing. New tasking costs about $30,000 at 30cm and about $20,000 to $22,000 at 50cm. In other words, 50cm saves roughly a third of the budget, which is why many programmes mix the two. See the full satellite imagery pricing page for every tier.

Data Volume, Processing and Storage

Higher resolution means more data to move, store and process. Using four bands at 16 bits and no compression, 30cm imagery is about 89 MB per km² and 50cm about 32 MB per km². For a 10,000 km² basemap that is roughly 890 GB against 320 GB before compression. Model inference, tiling and mosaicking time scale in a similar way. If your team works on standard laptops or a modest cloud budget, 50cm keeps national-scale projects manageable, while 30cm is easy to handle for cities and project sites.

Native 50cm Beijing-3A imagery of a Greek port, delivered as orthorectified GeoTIFF ready for GIS.

Both resolutions are delivered by XRTech Group as GeoTIFF in WGS84 or UTM, as Level 1B for your own processing or orthorectified Level 2A, ready for QGIS, ArcGIS and CAD. 30cm and 50cm are both pan-sharpened from a native panchromatic band; SuperView Neo-3 50cm adds eight multispectral bands (including red-edge and coastal blue) for vegetation and water analysis.

Basemaps: When to Use 30cm, 50cm or Both

For a city basemap, 30cm is the standard: street furniture, parking, building outlines and construction stages all read clearly. For a national or provincial basemap, 50cm is usually the better value because the wide swath gives a more uniform, cloud-free mosaic in fewer passes. Many of the best basemaps combine both: 30cm over urban cores, ports, industrial zones and corridors, and 50cm everywhere else. XRTech Group plans mixed-resolution programmes in one order, colour-balances the two and delivers one seamless mosaic.

A seamless, colour-balanced mosaic of orthorectified scenes over farmland: the kind of wide-area basemap where 50cm usually gives the best value.

Value-Added Products From 30cm and 50cm Imagery

Read more about stereo satellite imagery and our DEM services.

Stereo acquisition: two views of the same area produce a point cloud, a digital surface model and a 3D city model.

30cm vs 50cm Use Cases by Industry

See how each sector uses imagery on our pages for construction, real estate, agriculture and oil and gas.

Native 30cm satellite imagery of Raffles City, Singapore: dense urban detail where 30cm is the better choice.

Native 50cm SuperView-1 imagery of an industrial district in Denver: buildings, yards and parked vehicles are clear at 50cm.

Characteristics of 30cm Satellite Imagery

• Native 0.3m panchromatic GSD with 1.2m four-band multispectral, pan-sharpened to 0.3m
• About 11.1 million pixels per km²
• Swath of about 12 km per pass
• Best positional accuracy of the two (3.5m CE90 without ground control on SuperView Neo-1)
• Stereo and tri-stereo for detailed DEMs and 3D city models
• Higher price and data volume per km²

Native 30cm SuperView Neo-1 imagery of The Pearl, Doha: individual boats, rooftops and road markings are distinct.

Explore 30cm satellite imagery from XRTech Group.

Characteristics of 50cm Satellite Imagery

• Native 0.5m panchromatic GSD with 2m multispectral (four or eight bands), pan-sharpened to 0.5m
• About 4 million pixels per km²
• Swaths from 12 km to 130 km per pass
• Deep archives from SuperView-1, SuperView Neo-3 and Beijing-3A
• Stereo and tri-stereo available
• About a third less per km² than 30cm, with smaller files

Native 50cm satellite imagery of a residential district in Riyadh: every building, street and courtyard is mapped clearly.

Explore 50cm satellite imagery from XRTech Group.

Which Should You Choose: 30cm or 50cm?

Use this simple rule: choose the coarsest resolution that still shows your smallest important object with at least a few pixels across it. Then check three more things:

Native 30cm satellite imagery of a road and mining corridor in Saudi Arabia: site detail where engineers need to measure.

1. Area size: under a few hundred km² of urban or industrial area, 30cm is usually worth it. Over a few thousand km², start from 50cm and add 30cm only where the detail pays off.
2. Measurement needs: if engineers will measure from the image or build 3D models, choose 30cm with a viewing-angle limit and stereo.
3. Budget and timeline: 50cm costs about a third less and wide-swath collection fills large areas faster.

If you are still unsure, XRTech Group sends a free sample tile of your own area at each resolution with a formal quote, so your team can judge before it buys. You can also compare other tiers in our guide to what counts as high resolution.

How to Order 30cm or 50cm Imagery From XRTech Group

1. Search your location in our self-service portal, with no account needed, to see archive scenes, capture dates, cloud cover and an instant price, or send us your area as KML, KMZ or SHP.
2. Choose archive or new tasking, the resolution (30cm, 50cm or both) and any add-ons such as cloud limits, viewing angle or stereo.
3. Receive a written quote within 24 hours, then GIS-ready GeoTIFF delivery. Archive scenes arrive within 48 hours of approval; tasked areas follow as they are collected.

The XRTech Group portal: draw an area of interest, then compare archive scenes and prices before ordering.

30cm vs 50cm Satellite Imagery FAQs

How We Checked These Facts

XRTech Group compared 30cm and 50cm imagery using the published specifications of the satellites we supply (SuperView Neo-1, Beijing-3B, SuperView-1, SuperView Neo-3, Beijing-3A and 3C) and our own list prices in US dollars per km². Pixel counts and data volumes are calculated from GSD. Object detection figures come from published CosmiQ Works research, and map scale guidance from ISPRS and FIG papers. Prices are list prices before volume pricing; your written quote confirms the final rate.

Sources

• XRTech Group: 30cm and 50cm satellite imagery specifications and price lists
• CosmiQ Works (arXiv 1812.04098): The Effects of Super-Resolution on Object Detection Performance in Satellite Imagery
• CosmiQ Works, The DownlinQ: Super-Resolution and Object Detection, a Love Story (epilogue)
• ISPRS Archives: Information contents of high resolution satellite images (Topan and others)
• FIG Working Week proceedings: relationship between GSD, map scale and accuracy
• China Great Wall Industry Corporation: SuperView Neo-3 02 launch
• eoPortal: Beijing-3 mission