2025 OvenMediaEngine Roadmap

[getroot]

Happy New Year! We sincerely thank you for your unwavering support for OvenMediaEngine. Thanks to the contributions of many, we have not only survived but also continued to do what we love — developing innovative solutions.

In 2024, we focused on improving the stability of OvenMediaEngine and validated its reliability across various commercial and enterprise environments. Now, we are shifting gears to accelerate the addition of new features and leverage the strengths of open-source development to adopt cutting-edge technologies faster than anyone else.

We have also established clearer principles for the distinction between OvenMediaEngine Open Source and OvenMediaEngine Enterprise:

• Both Open Source and Enterprise versions offer the same media streaming capabilities with no performance differences.
• The Enterprise version includes features designed to support revenue-generating activities, such as monetization functionalities (e.g., ad insertion), advanced operational tools, and integrations with external commercial systems.

Enterprise customers, alongside our financial and technical contributors, play a vital role in enabling us to maintain and improve the Open Source project. We deeply appreciate their support.

In 2025, we plan to introduce highly requested features to the Open Source version. Additionally, if resources allow, we will focus on adopting the latest technologies, further capitalizing on the advantages of open-source development. Here are some of the planned features:

• SRT Publisher
• WebRTC Simulcast
• Advanced WHIP Features
• WHEP Implementation
• WebRTC HEVC Support
• Enhanced RTMP (starting with HEVC support)
• AV1 Support
• Modern Image Format Support (WEBP or / and AVIF)
• Expanded Provider and Publisher Support

If you believe there are other essential features that would greatly benefit the OvenMediaEngine community, we encourage you to share your ideas and engage in discussions. Your feedback is invaluable to us.

Thank you once again for your support, and we look forward to an exciting year ahead!

[naanlizard]

Happy New Year y'all! Our only feature request at this point is more image formats for thumbnail output and more options for what size static images are generated. We'll definitely try the new formats though when they're available

…

On Thu, Jan 2, 2025 at 23:58 Jeheon Han ***@***.***> wrote: Happy New Year! We sincerely thank you for your unwavering support for OvenMediaEngine. Thanks to the contributions of many, we have not only survived but also continued to do what we love — developing innovative solutions. In 2024, we focused on improving the stability of OvenMediaEngine and validated its reliability across various commercial and enterprise environments. Now, we are shifting gears to accelerate the addition of new features and leverage the strengths of open-source development to adopt cutting-edge technologies faster than anyone else. We have also established clearer principles for the distinction between OvenMediaEngine *Open Source* and OvenMediaEngine *Enterprise*: - Both Open Source and Enterprise versions offer the same media streaming capabilities with no performance differences. - The Enterprise version includes features designed to support revenue-generating activities, such as monetization functionalities (e.g., ad insertion), advanced operational tools, and integrations with external commercial systems. Enterprise customers, alongside our financial and technical contributors, play a vital role in enabling us to maintain and improve the Open Source project. We deeply appreciate their support. In 2025, we plan to introduce highly requested features to the Open Source version. Additionally, if resources allow, we will focus on adopting the latest technologies, further capitalizing on the advantages of open-source development. Here are some of the planned features: - *SRT Publisher* - *WebRTC Simulcast* - *Advanced WHIP Features* - *WHEP Implementation* - *WebRTC HEVC Support* - *Enhanced RTMP* (starting with HEVC support) - *AV1 Support* - *Expanded Provider and Publisher Support* If you believe there are other essential features that would greatly benefit the OvenMediaEngine community, we encourage you to share your ideas and engage in discussions. Your feedback is invaluable to us. Thank you once again for your support, and we look forward to an exciting year ahead! — Reply to this email directly, view it on GitHub <#1758>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABV4BWNEO5NHLDSX3OGP4OL2IYKHXAVCNFSM6AAAAABURCJ7JWVHI2DSMVQWIX3LMV43ERDJONRXK43TNFXW4OZXG43TGNBTGQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[getroot]

Thanks for reminding me of that. I've added it to our Roadmap.

[bchah]

Thank you @getroot, this is very exciting! My roadmap requires SRT Publisher and WebRTC HEVC support which I am currently doing in hacky ways using other software. It will be great to consolidate this functionality to OME :)

[nums]

Thanks you @getroot, these are amazing features!

The first 4 points perfectly match our needs: WebRTC simulcast will allow us to revisit the bandwidth estimation, which is currently causing some issues with WebRTC ABR.

If I may add a few points, here they are::

• I believe you are working on improving logs as mentioned here :) :

  OvenMediaEngine/docs/logs-and-statistics.md

  Line 165 in 189975e

  Files such as webrtc\_stat.log and hls\_rtsp\_xxxx.log that were previously output are deprecated in the current version. We are developing a formal stats file, which will be open in the future.

  
  It would be interesting to manage this with the CMCD standard: https://cdn.cta.tech/cta/media/media/resources/standards/pdfs/cta-5004-final.pdf

• Otherwise, I think it would be interesting to work on metadata management, especially for subtitles. Accessibility demands are increasingly important on the client side, and being able to manage subtitle layers at the RTMP level (CEA-608) and HLS could be a great feature.

• Finally, it would be a nice bonus if the multiplexer could handle multi-track (multiple audio tracks as input and output, for example).

[getroot]

Supporting multiple audio tracks in Mux Provider is a low priority but is on our backlog.

Regarding subtitles, I haven't found any real use cases in live yet and I'm not sure if this is a feature that many people are asking for. Could you provide some information on this?

Could you explain how WebRTC Simulcast positively impacts BWE? I'd like to check if there are any more things I should consider while implementing Simulcast.

Regarding CMCD, I think it's a good idea. However, since OME is mostly used as an edge, it would be a use case for WebRTC Edge Server, and applying CMCD to WebRTC doesn't fit well, so I think we need to think about it. Using OME as an HLS edge is not a very good idea except in special cases.

[nums]

For subtitles, I will write a use case about this in an other discussion (or maybe I will do a PR :) )

For webRTC simulcast, I speak about this discussion : #1684
In my experience, the bandwith estimation algorithm can be improve in using transport-cc specification.

CMCD was just an idea to be able to have more data about viewer session, the best would be to be able to customise the log of user with some variables.

[getroot]

Unfortunately, WebRTC Provider (WHIP) and WebRTC Publisher (WHEP) are separate modules, and implementing Simulcast in the WebRTC Provider and implementing transport-cc in the WebRTC Publisher are different tasks. However, I would like to complete the implementation of enhanced BWE using transport-cc for the WebRTC Publisher as part of this year's roadmap. (It's not a high priority due to the large number of tasks.)

[mbakumenkov]

As for subtitles, there is such a wonderful plugin for OBS - https://github.com/locaal-ai/obs-localvocal which has the functionality of integrating subtitles as part of the stream sent to the server. Thus, it will be possible to get live subtitles, very nice feature!

[getroot]

We plan to gradually deprecate OME's custom-defined WebRTC signaling protocol, leaving only WHIP and WHEP. However, WHEP currently seems to lack a method for manual ABR selection, providing a quality list, and delivering the currently selected quality information to the player. Therefore, we plan to implement these features through a custom protocol.

Of course, OME will maintain compatibility with other WHEP clients. Additionally, players like OvenPlayer or those implementing our custom protocol will have access to manual ABR selection and the ability to query the automatically selected quality.

The deprecation of OME's custom-defined WebRTC signaling protocol will occur only after a significant period following the addition of WHIP and WHEP functionalities to OvenLiveKit and OvenPlayer, allowing ample time for existing users to migrate. However, new WebRTC features, such as Simulcast, will only be added through WHIP/WHEP.

[nums]

Ok, that’s very clear. Indeed, simulcast does seem inherently 'automatic.' That said, it’s always possible to extend only the 'client' WHEP part to make it manual.

Question: would simulcast be managed only on the WHEP side or on both sides?

[getroot]

Simulcast is a specification that allows a Sender to transmit multiple streams simultaneously to an OME. When an OME transmits to a Viewer, it does not transmit multiple streams "simultaneously", but only one at a time. (However, it changes the quality according to BWE.)

That is, simulcast is only applicable to WHIP.

[nums]

Yes, exactly, I was referring to this part at the WHEP level: https://www.ietf.org/archive/id/draft-murillo-whep-03.html#name-video-layer-selection-exten
because it is indeed directly linked to WHIP simulcast. That said, it can also apply to OME server transcoding.

To be clearer, could we be in these two use cases?

• Simulcast with WHIP (1080p / 540p) => OME (bypass) => WHEP autoselect (1080p or 540p)
• RTMP 1080p => OME transcoding (bypass / 720p / 360p) => WHEP autoselect (1080p or 720p or 360p)

[getroot]

Yes, those two use cases will work.

[mpisat]

Enhanced RTMP (with h265) would be a very nice addition.

[EagleonePrimo]

AV1 is great hope QSV/VA-API in intel ARC work too that will be great since A310 is very cheap too