See the Installations topic for installation and activation instructions.
Watch tutorial videos for a fast start:
Manifold Tutorial 1 - Navigation and User Interface
Manifold Tutorial 2 - Add Data and Create a Map
Manifold Tutorial 3 - Export Data and Print a Map
Visit the Videos page for many more live action videos.
To get to work right now:
Read Me First
Welcome to Manifold® System!
This user manual describes both Manifold Release 9 and the free Manifold Viewer based on 9. See Changes and Additions for what is new in the latest build.
Illustrations above are from the Style: Autocontrast topic. Change the display of data instantly without changing the data.
When a classic GIS package "adds" a shapefile layer to a project, it leaves all data in the shapefile: the objects that are viewed or processed stay in the shapefile. That limits capabilities and performance to what slow, limited shapefile formats can handle, like trying to run with your leg chained to a block of stone. Classic GIS packages have no DBMS inside (or a really primitive one, like "geodatabase"), so they are stuck with leaving data in external files and formats.
Manifold, in contrast, is a GIS with a DBMS inside, and not just any DBMS, but Manifold's super-fast, massively parallel, phenomenally powerful Radian engine. Manifold can store huge data inside a Manifold project, without having to leave data in external files. When Manifold works with data from shapefiles, the first step usually is to import data from the shapefile into Manifold, cutting the cord to antique shapefile limitations, and thereafter running fast with no limitations inside Manifold. If we ever want to send a drawing back out to a shapefile, with a point and click we can export the drawing back out from Manifold into a shapefile.
As a spatial database engine, Manifold can store huge amounts of data, routinely hundreds of gigabytes and even terabytes if we want, within Manifold with no need for any other file formats or database servers. Instead of leaving data in external, legacy files, Manifold can import the data for storage within Manifold itself. We can maintain vast data and work with it at lightning speed, all within Manifold. If we want to leave our data in external formats, whether those external formats are slow formats like shapefiles or GPKG, or fast data stores like Oracle or PostgreSQL, we can do that too. In the same project we can have a mix of storage, with some data out in external formats and some inside the project, and we can work with all that data in the same way.
Most users will import data into Manifold. Manifold's internal database engine is a parallel database engine designed specifically for GIS work, so it is much faster than leaving data in external formats, even faster than truly superb DBMS packages like Oracle or PostgreSQL. For most GIS work, Manifold is faster than most enterprise DBMS packages.
When we import data into Manifold, it becomes Manifold data, with a very rich set of options for data types and manipulation, with no limitations imposed by the format or data store from which it was imported. When we want to send data from within Manifold out to some other format or data store, we can do that with a few clicks. Manifold is so much faster than typical GIS formats that it is usually quicker to import into Manifold, do our work, and then export back to the original format, than it is to link the original format and do our work "in place."
Manifold's internal database is transparently integrated with Manifold: there is no complexity to bringing data into Manifold and no need to specially set up or administer some sort of Manifold database server, the way we must do with Oracle or PostgreSQL/PostGIS. We just launch Manifold like any desktop application, and import the data we want from any of zillions of formats. Manifold will extract all usable info from the format, and, like magic, drawings, images, tables and other data appear in our Manifold project.
When we save the project as a .map file, all the data is saved into that .map, ready to go exactly as we left it when we next open the .map project. If we link files or data stores into the project, those links will be there too when we next open the project, ready to go. The .map project file is the database storage file for Manifold. That makes it easy to copy, to backup, and to move to a different machine. If all we do are simple projects, except for the speed of how quickly everything happens, we might never realize that Manifold is not just a GIS but also a high-end, incredibly sophisticated, database engine.
Manifold's Radian® spatial database engine brings a new generation of parallel power to spatial applications. The Radian engine powers Manifold Release 9, Manifold's newest generation GIS product as well as other applications, such as Manifold Viewer, involving both spatial and non-spatial data. Release 9 and all other Manifold products based on the Radian engine provide far greater performance, power and reliability than the non-parallel technology of an earlier day that is used in classic GIS such as ESRI ArcGIS Pro, QGIS, or Manifold's own Release 8.
Manifold® Release 9 provides a full-featured GIS that leverages the power of Radian parallel performance and new generation infrastructure to deliver unprecedented quality, performance, and breadth and depth of features in GIS. delivering all that with unbeatable value and the lowest cost of ownership in GIS. Enjoy the spectacular power, total reliability and incredibly low cost of Release 9, the only GIS in the world with automatically CPU-parallel power throughout as well as built-in GPGPU supercomputer speed.
Manifold SQL, the native SQL of Release 9, is automatically parallel SQL. Manifold by default automatically launches parallel threads to utilize as many cores as are available on as many CPUs as are available. Watch the video showing how Manifold automatically launches 48 threads to compute watersheds in seconds for a 1 GB terrain elevation raster. If one or more NVIDIA GPUs are installed, many Manifold functions and point-and-click dialogs automatically will launch massively parallel utilization of thousands of GPU cores, all automatically optimized against the use of CPU parallelization. Watch the video showing how Manifold uses 1280 GPU cores to accomplish in 2 minutes what can take five days in a classic GIS package. Manifold Viewer also is fully CPU and GPU parallel, just like Release 9.
For multiuser sharing of spatial data, use Manifold Server. Manifold Server is a high performance, parallel, spatial database server that is built into Manifold Release 9 Universal edition. It allows many Manifold users to share data that is published through Server.
Most DBMS engines are designed to be fast at handling many transactions where records are small and transactions involve small amounts of data.. That is perfect technology for applications like powering credit card transactions, airline reservations, corporate sales and inventory, or web sites. But tuning for many, small transactions limits performance in applications where an individual record can be very large and an individual transaction very computationally demanding, as often happens with spatial work and GIS. Individual records in spatial applications can contain gigabytes of data and individual transactions can involve hundreds of gigabytes. Manifold's Radian engine, in contrast, is designed for spatial engineering work where individual records can be very large, possibly hundreds of megabytes or even two gigabytes in size.
Manifold as a database also is tuned for work with data of many different types in situations where individual transactions can be very large and very computationally demanding, and where the data can be very big data, that is, potentially very large records. Genuinely parallel computation is essential in such applications so that the analytic power of many CPU cores and the power of massively parallel GPU computations with thousands of cores can be brought to bear. For performance, that requires a very different internal architecture than classic, Enterprise DBMS packages like Oracle, SQL Server, MySQL, DB2 or PostgreSQL/PostGIS.
Manifold wraps all that database power within point-and-click ease of use, very fast and rich visualization of almost any sort of data, an extensive analytic framework, extensive connectivity to many different types of data sources and a rich GUI full of interactive capabilities to see, to edit, to transform, and otherwise to manipulate all sorts of GIS data, including vector and raster data.
The result is a general purpose console that can do sophisticated, large computations on big data whether the data is stored in Manifold, within other DBMS environments or in a mix of both. With seemingly endless GIS capabilities built into Manifold,, the result is the world's fastest and most powerful GIS, as well as the world's richest and fastest tool for doing spatial ETL and manipulating spatial databases.
Manifold provides rich and powerful SQL not to replace DBMS packages like Oracle or PostgreSQL but instead to more effectively leverage them so we can choose the best combination of tools for any job. When not working with other DBMS servers, Manifold's native database and SQL capabilities provide a spectacular level of ease of use, convenience, and power in a compact and totally reliable desktop application.
When a task involves many small records or non-spatial work at which an existing DBMS excels, especially if our data is already stored in that DBMS, the best approach will often be to blend the DBMS we already use together with Manifold capabilities, perhaps writing queries that simultaneously call upon SQL in both the DBMS and also within Manifold.
Use Manifold where Manifold is faster or more convenient and use databases like Oracle, MySQL, PostgreSQL, or others where they are faster, to enjoy the best of both worlds. Manifold will never be slow, so we are never required to use an external DBMS, but if we have one that we know and love Manifold will embrace that relationship as well. It's all good.
Image at right: Manifold users enjoy an endless range of instant background maps using web servers.
Manifold power applies to all applications based on Radian, such as Manifold System Release 9 and Manifold Viewer. Release 9 capabilities include:
Rock solid, bulletproof quality. It is almost impossible to crash Release 9. At the present time, there is no known way to cause a crash. Only a handful of bugs that might cause a crash have been discovered during years of worldwide beta testing and deployment, and all such bugs were eliminated within days. Hardened structures and file management greatly reduce risk of damage from power failures and other external problems.
Full support for all data types of interest in spatial applications, including a very wide variety of raster and vector data types.
Easy connectivity to all popular file, server and web formats used for spatial data.
User-configurable localization to almost any language desired, using simple text files. Share localizations with other users for French, German, Portuguese, Russian, Spanish, and many other languages.
Direct connectivity to all major database management systems, including the sophisticated ability to utilize spatial data within such systems, to launch queries and to otherwise operate those systems in addition to simply using them as data sources.
Direct support for popular web sources of spatial data, such as OSM WMS, WMTS, WFS, TMS and other web servers.
Full integration with tile based image servers such as supported by Manifold image server modules.
Exceptional performance and capacity to handle very large data sets.
Automatic parallelization to utilize multiple CPUs and multiple CPU cores, both for computation and for operations such as multithreaded, asynchronous rendering.
Automatic parallelization to utilize multiple GPU cores, including potentially thousands of cores. Kepler or more recent NVIDIA GPUs are required to utilize GPU.
Fully, automatically parallel SQL so queries and computations made with queries automatically take advantage of multiple CPU and multiple GPU cores.
Exceptionally sophisticated, fully integrated, enterprise-class database engine and query engine designed and tuned for the demands of spatial data, implemented at a professional level of completeness, orthogonality, extensibility and internal rigor far beyond the level of typical database engines.
The ability to execute queries locally entirely within Manifold SQL and using data stored locally and/or data stored in other data sources, or executing queries within external database management systems using their native SQL or a mix of both.
Integrated, automatic support for over 6500 different coordinate systems and over 1500 coordinate system transformations, including EPSG and numerous others.
Exceptionally high precision coordinate system and base coordinate system transformations.
Automatic recognition and coordinate system parsing in XML, PRJ / WKT and JSON, using JSON as open and interchangeable specification of coordinate systems.
Breathtakingly fast performance with spatial data - nothing else in GIS or database matches Release 9 in spatial work.
Instantaneous project opens and saves. Virtually instantaneous open and display of images or drawings even in the many gigabyte size range.
A very fast and convenient project format that can be nested with no loss of performance.
Very wide range of analytic and other functions to edit, manipulate, analyze and transform data.
Automatic preservation and management of a vast range of projections associated with spatial data.
Seemingly endless capabilities for quickly and easily visualizing data, using very flexible and seemingly endless options for symbology.
(Illustration at right from the Labels topic.)
Release 9 provides exceptionally easy means to:
View spatial data from a seemingly endless variety of sources.
Transform spatial data between formats and to transport spatial data between different sources.
Transform spatial data between types, such as between different raster forms.
Analyze and extract data.
Edit and manipulate data, to combine data from different forms or sources for greater insight, to add accuracy or to eliminate unreliable data.
Find important patterns and relationships within vast amounts of unimportant or unrelated data or noise.
Prepare data for use within other applications or systems, including for more optimal storage or for faster performance.
Create applications which automate the above.
All of the above are part of Release 9. In addition, Release 9 adds hundreds of more capabilities every month or two to provide the world's best GIS product. See the Changes and Additions topic for the latest new capabilities.
The "documents" of Manifold are projects that are saved in Manifold .map files, short for "Manifold Project." Projects can contain data or can links to other data sources, which in turn can contain data or links to yet other data sources. Projects also can contain all the other components with which Manifold can work internally, such as tables, images, drawings, maps, scripts or queries.
Manifold stores all data as tables. It is obvious when database information is kept in tables, but going beyond that virtually everything in Manifold is also saved as data in a table. Every bit of data, regardless of its nature - vector, raster, text, queries, scripts and so on - is exposed as part of some table.
Manifold's everything is a table architecture provides many benefits:
Every bit of data can be accessed in a standard way with SQL, so our queries and scripts can get at whatever data we want.
Everything that can be accomplished via the standard user interface can be done via SQL as well, so we are not limited to only those capabilities that point and click dialogs provide.
The same data can be visualized in different ways for different purposes: storing data in a table allows many different presentation facilities to display that data as desired in different ways, including simultaneously in different windows, without having to duplicate the data for each visualization. Manifold has seemingly endless ways of visualizing data in tables, such as many different ways to style vector and raster data stored in tables.
Organizing all data in tables provides order and efficiency through modularity and standard interfaces. Operators that work with binary data in tables can apply to many different types of binary data without having to write, maintain, document, and teach many different special cases of how such data is presented or used, for example, such as images or as terrain elevation surfaces.
It enables natural, native connectivity to a vast range of database management systems and the databases they store, in which many different types of data for many different purposes are maintained worldwide in tables. Whatever the DBMS we choose, the ability to use data stored within that DBMS easily and conveniently within Manifold comes naturally with Manifold's "everything is a table" approach.
Easy connectivity to data stored in external DBMS data sources also allows Manifold users to take advantages of the benefits of those DBMS sources, for example, using data sources that are already scaled out across thousands of cloud based servers.
If we like, many templates in dialogs such as the Transform pane will at the click of a button automatically write SQL for us, to show how they work. We can then adapt that SQL to further customize and expand what it does.
Manifold's built-in, highly optimized query engine works with queries written in SQL, the world's standard language for exploring and manipulating data. That means a straightforward, industry-standard way of working with data is already at our fingertips in Manifold, using a simple language, SQL, already known to millions of people, easy to pick up in minutes by beginners, and unsurpassed as an effective way of manipulating data.
Spatial data is data which has meaning tied to location, either actual geographic locations or location in abstract form such as the location and dimensional space of a CAD drawing.
(Illustration at right from the Topology Overlays topic.)
More and more spatial data is being created, in every larger data sets, that are consumed and used by ever more demanding applications. The combination of much bigger data plus more demands on processing mean that classic approaches to GIS have become too slow.
Data is getting so big and spatial analysis so sophisticated, that even very fast, general purpose data tools cannot keep up with spatial demands.
Using a fast, general purpose DBMS like MySQL, Oracle, SQL Server or PostgreSQL can help handle larger data, but that is only part of the job: to get fast and responsive GIS the GIS engine has to be completely integrated with a fast, parallel database engine as well. Given the demands of spatial work, even very fast general purpose DBMS packages can be too slow.
Opening a multi-gigabyte image and manipulating that image to extract features may require accessing many thousands of records (for thousands of tiles) and processing gigabytes of pixel data for just a single editing operation. Likewise, a single command to locate features that intersect within millions of objects in a large drawing, for example, a drawing showing all real estate parcels within a large city's territory, can involve millions of records, each of which might be very small or very large.
A command that requires fetching and active work with each of millions of records is a rare event in classic DBMS work, but it is perfectly routine, perhaps happening every minute, when doing editing or other typical workflow in spatial applications. At the same time spatial work often requires significant computations which process very large amounts of data in each operation.
Classic DBMS products have become very fast and very capable over the years so most have sufficient raw performance to be adapted for spatial work. Almost all of the big, enterprise, DBMS packages, such as Oracle or PostgreSQL, have had "spatial" capabilities added to the DBMS. Most such spatial capabilities work remarkably well. That classic, non-spatial DBMS packages can be adapted for effective spatial work is a tribute to their core power and performance. But given their general purpose nature they do not provide in a single, integrated package an infrastructure that is optimized for the distinctive needs of spatial work with spatial data. That is why they can end up being slower than desired for spatial work.
To support fast spatial work with big data, the parallel Radian database engine built into Manifold was developed. Radian is completely new technology built from the ground up using fully parallel technology throughout. When tightly integrated with GIS (Geographic Information System) capabilities within Release 9, Radian infrastructure provides unprecedented power in GIS.
Just as classic, non-spatial DBMS products have found uses in spatial applications, the Manifold engine originally designed for spatial needs now is employed in non-spatial applications as well. Release 9 is so fast, it connects to so many data sources, and it provides such convenient facilities for typical DBMS tasks like ETL workflow, that many users routinely apply Release 9 within their non-spatial, DBMS work and general purpose ETL and data engineering work. They use Manifold to work with tables and data that have nothing to do with spatial data.
Release 9 features specific to spatial work, such as built-in support for projections or for visualizing vector geometry as drawings or raster pixel data as images, do not in any way detract from the performance of Release 9 in non-spatial applications.
Examples in this documentation will often involve spatial applications because those often are universally understood: not everyone may understand an abstract image representing hot points in computational finance, but everyone understands maps of roads or satellite images of farms and fields when they see them. Don't let the choice of such spatial examples prevent understanding the very broad applicability of Release 9 in non-spatial work.
Manifold Viewer is a free product from Manifold that is a read-only subset of Release 9. Viewer has significant limitations compared to Release 9:
View only / Read only.
Read .map format, but no saves of projects.
Imports of all Release 9 formats, but no exports to any format.
Does not print.
No ODBC connection to Viewer.
No scripting and no add-ins.
Connect to all Release 9 data sources (DBMS, web servers, etc.) but read-only for all data sources.
Despite the above limitations Viewer retains so many Release 9 user interface commands and functions, like full spatial SQL, full CPU and GPU parallelism, that the same user manual - this user manual - is used for both Release 9 and Viewer. Viewer users who read this documentation should keep in mind the above limits. If a command or menu option does not appear, that is usually because of one of the above limits in Viewer. In most situations the situation will be clear.
Manifold does not provide technical support for Manifold Viewer. Hey, it's free!
To get support on Viewer, read this user manual, watch the videos, or ask questions of other users on the Manifold Forum.
Manifold Release 9 and Manifold Viewer evolve rapidly. New Release 9 builds fix bugs and add numerous new features. They are free to download and use by licensees. If we have already activated Release 9 on the computer, uninstalling an older build and then installing a newer build will not require an additional activation key. Running the latest build guarantees we will be able to open .map and .mxb created by new builds, which older versions might not be able to open.
new, improved versions of Release 9 usually will use the same .map project format as earlier versions, but sometimes improvements will require changes to .map project format that older versions cannot read. Newer versions of Release 9 will always be able to open .map project files created by older versions, but older versions might not be able to open .map project files created by the latest version.
Always run the latest version of Release 9 or Viewer. If a Release 9 .map or .mxb project file cannot be opened, install the latest version of 9 or Viewer and try again. Older versions of 9 or Viewer might not be able to open projects created by the very latest build.
Manifold means the latest Manifold Release 9 build. Radian Studio means the "Swiss Army Knife" data engineering tool based on Radian, which in 2018 was replaced by Manifold Release 9. Manifold Viewer is the read-only subset of Release 9. Because Manifold, Manifold Viewer, and Radian Studio are based on the same Radian engine and use many of the same interfaces, we may use Radian and Manifold as synonyms in this documentation. Release 8 means the prior generation of Manifold GIS products.
Check for updates - Manifold does not automatically check for updates in between official builds. After installation, at any time check for any newer versions available by launching Manifold and choosing Help - About. When the About dialog is launched, Manifold will reach out through Internet to check for a new version.
Watch tutorial videos for a fast start:
Manifold Tutorial 1 - Navigation and User Interface
Manifold Tutorial 2 - Add Data and Create a Map
Manifold Tutorial 3 - Export Data and Print a Map
Visit the Videos page for many more live action videos.
User Interface Basics
Advice: An avatar or photo representing tech support / engineering staff often appears near caution warnings, tech tips or other useful information.
Enjoy Release 9 and Manifold Viewer!
Manifold, Polygon, and Radian are registered trademarks of Manifold Software Limited. 流形, 多边形, and 弧度 (Manifold, Polygon, and Radian in Chinese) are trademarks of Manifold Software Limited.