Indexless Partial File Restore
Restore the clip you need.
Retrieve selected sections of your original media from cloud storage, network servers, or tape archives. Transfer less data, reduce restore times, and deliver a new clip ready for production.
Start with Indexless PFR, without preparing separate index files. Existing RTIN indexed workflows are fully supported.
 |
Windows, Windows Server | ||
 |
Red Hat/CentOS or Ubuntu | ||
 |
macOS 14 Sonoma - macOS 26 Tahoe | ||
 |
 |
 |
Indexless Partial File Restore
S3 and S3-compatible storage • HTTPS • FTP • Microsoft Azure • Google Cloud Storage • LTO and tape archives
- Indexless PFR
- Storage Sources
- ActiveScale
- Cache & Scale
- Indexed PFR
- Videos
- Demo & Integration
- Compare
- FAQ
Start with the media you already have
Indexless PFR retrieves a selected time range from supported media files without requiring a separate RTIN index to be created before archiving.
Net-X-Code reads the file’s headers, footers, and internal media tables as needed to locate your selection. It then retrieves the required video, audio, and ancillary data and wraps them into a new clip.
For media accessible through S3, HTTPS, Azure, or Google Cloud Storage, this reduces the amount of source data transferred and the time needed to deliver a clip. Your existing archive remains in place.
- No advance indexing
Enable partial restore without first processing every archived recording. - Less data to transfer
Retrieve the media required for your selection instead of downloading the complete source. - Production-ready output
Rewrap supported media without recompression, or convert to your required delivery format.
One restore workflow across your storage
Your archive does not have to live on one platform. Net-X-Code supports partial file restore across cloud, network, and tape storage, using the access methods available from each source.
Source | Supported access |
|---|---|
Amazon S3 and S3-compatible storage | Retrieve the object ranges required for the selected clip. |
HTTPS and HTTP | Read selected ranges from servers that provide byte-range access. |
FTP | Retrieve media through supported FTP access. |
Microsoft Azure Blob Storage | Access source media stored in Azure. |
Google Cloud Storage | Access source media stored in Google Cloud. |
Local disks and network storage | Restore from accessible files and shared storage. |
LTO and tape archives | Restore through compatible tape and archive integrations. |
Need another source? Drastic’s extensible storage access layer allows additional sources to be integrated into the same restore workflow.
Partial restore behavior depends on the media format and the source’s range-read or recall capabilities. Cold storage may require a recall step before media can be read.
Keep more of your archive in cold storage
Archive workflows often need many short clips from the same long recording. Repeatedly retrieving the same file headers and index tables adds unnecessary work.
Drastic’s cache retains this information after the first restore. While that metadata remains cached and valid, subsequent selections need only retrieve the media required for each new clip.
This brings subsequent restores to the same source-data efficiency as RTIN indexed PFR, without requiring an RTIN file to be prepared in advance.
First selection | Later selections with cached metadata |
|---|---|
Retrieve headers and index tables | Reuse cached headers and index tables |
Retrieve selected media | Retrieve selected media |
Create the clip | Create the clip |
Built for repeated selections from long recordings
A major U.S. sports league uses a workflow with 16 cameras recording eight hours each per game: 128 camera-hours of source material.
Editors often request many short selections from those recordings. Reusing cached file information avoids repeated metadata retrieval, making that pattern especially efficient.
The next clip needs even less
Amazon S3 and S3-compatible object storage are excellent for data lakes, archives, cold storage, compliance, retention, and hybrid cloud. But S3 restores can be inefficient when the customer needs to recover only one file or object.
Traditional backup restore tools often require:
Full restore
- Catalog rebuild
- Index rebuild
- Restore database
- Backup server
- Agent-based restore
- Full bucket restore
- Temporary storage space
- Long restore windows
Indexless Partial File Restore is better suited for S3 environments because S3 already uses object metadata, bucket structure, prefixes, object names, versions, and retention policies. Instead of forcing a catalog/index/database model, the product can work with object metadata and backup archive records to restore only what is needed.
For S3 customers, the value is not “restore everything.” The value is “restore one file or object without restoring the whole bucket.”
Already using RTIN? Keep your workflow
Net-X-Code supports both Indexless PFR and indexed PFR using Drastic RTIN files.
In an indexed workflow, the media is processed to create a separate RTIN file, which remains accessible after the original recording is archived. Net-X-Code uses that index to locate and retrieve the requested media range.
Indexless PFR obtains the information it needs from the original media file on demand and can cache it for subsequent requests.
Indexless PFR | Indexed PFR | |
|---|---|---|
Separate RTIN file required | No | Yes |
Preparation | Read file information when first needed | Create RTIN before restoring |
Locating selected media | Source file information, then cached metadata | Accessible RTIN file |
Best starting point | New deployments and existing archives without RTIN files | Established RTIN workflows |
For most new customers, Indexless PFR is the simplest starting point. Existing RTIN installations can continue using their established workflow. Both approaches retrieve selected media ranges and create usable output clips.
View the Explainer Video
Try the workflow on ActiveScale
Browse sample proxies, mark IN and OUT points, and explore output formats at indexlesspfr.com. A demo login enables restores, conversions, and downloads from the ActiveScale-backed archive.
- Explore the Demo → indexlesspfr.com
- Request Demo Access → Contact sales@drastictech.com
Indexless Partial File Restore vs. Traditional Full Restore
Feature | Traditional full restore | Indexless Partial File Restore |
Restore scope | Full dataset or full bucket | Single file, object, or byte range |
Index required | Often yes | No index rebuild required |
Catalog required | Often yes | No catalog rebuild required |
Database required | Often yes | No restore database required |
Full restore required | Often yes | No, when only one file/object is needed |
Restore speed | Slower for large object stores | Faster for single-file/object recovery |
Storage overhead | Higher | Lower |
Backup server impact | Higher | Lower |
Operational complexity | Higher | Lower |
Best fit | Full restore workflows | Single-file/single-object recovery |
S3 fit | Requires catalog/index/database model | Better fit for object metadata model |
Quantum ActiveScale fit | May require full restore/catalog workflows | Better fit for object storage restore |
Quantum ActiveScale is great for storing objects. Indexless Partial File Restore is better for recovering only the object you need.
Frequently Asked Questions
What is partial file restore?
Partial File Restore (PFR) creates a new, usable clip from a selected section of an original media file. For example, you can retrieve a highlight from a long recording without first transferring the entire master file.
What is indexless partial file restore?
Indexless PFR works with supported original media without a separately prepared index file. Net-X-Code reads the headers, footers, and media tables needed to locate your selection, then retrieves the required media ranges.
What is the difference between indexed and indexless PFR?
Indexed PFR uses a separate RTIN index created while the source is accessible, typically before archiving. Indexless PFR reads the required structure from the original on demand. Both methods are supported, so existing indexed workflows can continue.
Does it work with Amazon S3 and Quantum ActiveScale?
Yes. Net-X-Code adds media-aware selection and output creation to range-based storage access. It identifies the data needed for your selected clip and rewraps or converts it into a usable file. The storage tier must make the required ranges accessible.
Can I use storage other than S3?
Yes. Supported workflows include local storage, SMB/NFS network shares, HTTP/HTTPS, FTP, Microsoft Azure, and Google Cloud Storage. Tape and offline archives require an appropriate retrieval workflow.
Which media formats support indexless PFR?
Key indexless wrappers include MXF, MOV, MP4, LXF, and GXF. Support depends on the container variant, codec, and media structure; verify representative source files for your workflow.
How do I select the section to restore?
Set In and Out points using source timecode or absolute frame offsets. In a proxy workflow, preview the proxy and mark the required range; PFR retrieves that selection from the original high-resolution media.
Do I need a proxy to perform a restore?
No. A proxy helps users browse and select clips, but the restore can be requested directly using the source and In/Out points. Prepared workflows can generate proxies with timecode, multichannel audio, metadata, and captions.
Does partial file restore reduce picture or sound quality?
No, when you choose a supported rewrap workflow. The selected audio and video are copied without recompression into a new media file. Choosing a conversion profile instead transcodes the media.
Are timecode, audio channels, metadata, and captions preserved?
Yes, where supported by the source and destination formats. A restore can retain source timing, timecode, multichannel audio, metadata, and closed captions for the selected section. Preservation depends on the chosen output format and profile.
Can I convert the restored clip to another format?
Yes. Supported conversion profiles include MOV ProRes, MXF DNxHD/DNxHR, AVC-Intra, XAVC, and XDCAM. This lets you restore the selected section directly into a format suited to your production workflow.
How does BitCache speed up repeat restores?
BitCache saves the header, index, and other lookup data read during an indexless restore in an on-disk cache. Further requests for the same source can reuse that information and retrieve the requested media without repeating those reads.
How much time, bandwidth, and storage can PFR save?
PFR can reduce transfer volume, temporary storage, restore time, and applicable egress or retrieval charges. Savings depend on clip length, source structure, cache reuse, network performance, and storage tier. Measure representative clips against full-file restores.
Can I use PFR with tape or cold storage?
Yes, through supported archive workflows. Indexed PFR can identify byte ranges for an archive controller to retrieve. Indexless PFR requires access to the original file structure and media ranges. Offline media and cold tiers may need retrieval or staging before PFR can run.
Can PFR be automated or integrated into a MAM system?
Yes. Net-X-Code provides an HTTP/REST API, and Net-X-Copy offers command-line control. These interfaces support automated indexing, proxy creation, clip selection, restore, and conversion, including integration into media asset management workflows.
Can I try the workflow before deploying it?
Yes. At indexlesspfr.com, you can browse sample proxies, select In/Out points, and explore output profiles. Restores, conversions, and downloads require a demo login; contact sales@drastictech.com to request access.
Net-X-Code API (includes PFR, Net-X-Convert, Net-X-Copy)
More Drastic Software:
While you are here, you may want to check out some of our other software products:
 |
 |
 |
 |
 |
 |
 |
 |
 |
| DrasticScope | NetXScope | FlowCaster | MediaReactor | Net-X-Convert | Net-X-Code Server | Net-X-Router | Test Pattern Generator | videoQC |
Trademarks, Registered Trademarks, and CopyrightsTrademarks, Registered Trademarks, and Copyrights
Amazon Web Services, Inc. - Amazon, AWS and Smile Logo, Powered by AWS Logo, AWS Co-Marketing Tools, the Partner Logo, the Program Marks, Amazon Web Services, AWS, AWS S3, and the names of AWS products, services, programs, and initiatives are trademarks or registered trademarks of Amazon Web Services, Inc.
Apple Inc. - Apple, the Apple logo, Final Cut, Final Cut Pro, Apple TV, iOS, iPad, iPhone, iPod touch, iTunes, Mac, Mac OS X, macOS, Shake, Final Cut Pro, ProRes, High Sierra, Mojave, M1, M2, and QuickTime are trademarks of Apple Inc., registered in the U.S. and other countries.
Avid Technology, Inc. - Avid Media Composer®, Avid MediaCentral®, Avid Interplay®, and Avid NewsCutter® are either trademarks or registered trademarks of Avid Technology, Inc. or its subsidiaries in the United States and/or other countries.
Blackmagic Design Pty. Ltd. - DaVinci Resolve, DaVinci Fusion, UltraStudio, DeckLink, Intensity Pro 4K, UltraScope, and RED are either trademarks or registered trademarks of Blackmagic Design Pty. Ltd. or its subsidiaries in the United States and/or other countries.
Drastic Technologies, Ltd. – trademarks specified here.
Evertz Technologies Limited - Evertz is a registered trademark of Evertz Technologies Limited
Google LLC – YouTube, Google, Google Cloud, Google.meet.com, and Android are registered trademarks of Google LLC
Grass Valley - Grass Valley®, GV®, the Grass Valley logo, and EDIUS® are trademarks or registered trademarks of Grass Valley USA, LLC, or its affiliated companies in the United States and other jurisdictions.
Harris Corporation - Harris, and Leitch Technology Corp. are registered trademarks of Harris Corporation
Kinefinity Inc. - KINEFINITY is a trademark of Kinefinity Inc.
Magic Lantern - Magic Lantern is a registered trademark of Magic Lantern
Microsoft Corporation – Microsoft: Windows®, Video For Windows (VFW), DirectShow, Microsoft, Skype, Microsoft Azure, Microsoft Teams, Wave Mapper, Microsoft, Windows NT|2000|XP|XP Professional|Server 2003|Server 2008 |Server 2012, Windows 7, Windows 8, Media Player, Media Encoder, Windows Defender, Microsoft Office, .Net, Internet Explorer, SQL Server 2005|2008|2012|2014, Windows Media Technologies and Internet Explorer are trademarks of Microsoft Corporation.
Red Hat, Inc. - Red Hat, and the Red Hat logo are trademarks or registered trademarks of Red Hat, Inc. or its subsidiaries in the United States and other countries
Sony Corporation – Sony, Sony DVD Architect, DVD, Catalyst, and Vegas are trademarks of Sony Corporation and/or its affiliates.
Tektronix, Inc. - Tektronix® and all identified Tektronix trademarks and logos are the property of Tektronix, Inc. or its wholly-owned subsidiaries
VideoLAN Non-profit Organization - VideoLAN, VLC, VLC media player and x264 are trademarks internationally registered by the VideoLAN non-profit organization
All other trademarks are the property of their respective owners.
