# Mercury Transmit: supported Program-only sequence binding and 1 x 1 output

**URL:** <https://forums.creativeclouddeveloper.com/t/mercury-transmit-supported-program-only-sequence-binding-and-1-x-1-output/12282>\
**Category:** Premiere Pro\
**Created:** [October 7, 2026, 2:20am UTC](https://forums.creativeclouddeveloper.com/t/mercury-transmit-supported-program-only-sequence-binding-and-1-x-1-output/12282 "2026-10-07T02:20:46Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![andy.l](https://avatars.discourse-cdn.com/v4/letter/a/3be4f8/32.png) [@andy.l](https://forums.creativeclouddeveloper.com/u/andy.l)\
**Post date:** [October 7, 2026, 2:20am UTC](https://forums.creativeclouddeveloper.com/t/mercury-transmit-supported-program-only-sequence-binding-and-1-x-1-output/12282/1 "2026-10-07T02:20:46Z")

</div>

We are using the licensed Premiere Pro 26.0 C++ SDK, Transmit interface 4 and Sequence Info suite 9, on Premiere 26.5.2 for macOS. Our diagnostic copies BGRA8 picture and 48 kHz stereo into bounded local buffers and a private native preview. There is no network publishing, encoding, speaker output or playback clock. Program sharing remains disabled until source identity is established. An older manual test observed Program and Source playback in different opaque native contexts, with matching frame sizes/rates; we do not treat those as source labels.

1. Is there a supported API to bind `tmInstance.inTimelineID` / `inPlayID` to the stable UXP `Sequence.guid`, and distinguish Program Monitor from Source, Project-panel and other Transmit Invocation output? The [2024 accepted community answer](https://community.adobe.com/questions-729/question-about-transmitter-plugin-1405606) says Program/Source distinction was unavailable. Has this changed in a supported SDK, and which SDK/host versions are required?
2. How can an approved sequence revision and displayed-frame reference be tied to delivered picture/audio? We retain raw host ticks separately from monotonic observation time, including negative immediate-display timestamps. We do not infer delivered positions from UXP playhead polling. What timestamp, lifetime, threading and callback-boundary guarantees are documented?
3. In the recent bounded retest, 746 picture/1413 stereo buffers were copied. After the sequence ended, the same opaque native context delivered a **1 x 1** picture while our selected size was **1920 x 1080**. The rejected callback’s raw time was **11579880312000** at **254016000000** ticks/second; play mode/quality were **0/3** , with one batch handle. We rejected before reading pixels, disposed the handle, cleared local media and blocked resume. Is that output documented, and what supported recovery preserves source isolation?

Sequence Render needs an existing native timeline ID; Export Info’s sequence-GUID selector requires an exporter plugin ID. We have not substituted transmitter IDs, invoked exports or used private APIs. Please identify a supported binding or capture route, or confirm the current limitation. Equal formats/timecodes, monitor focus and content hashes will not authorize sharing in our implementation.

---

<div class="post-metadata">

**Author:** ![andy.l](https://avatars.discourse-cdn.com/v4/letter/a/3be4f8/32.png) [@andy.l](https://forums.creativeclouddeveloper.com/u/andy.l)\
**Post date:** [October 8, 2026, 6:48am UTC](https://forums.creativeclouddeveloper.com/t/mercury-transmit-supported-program-only-sequence-binding-and-1-x-1-output/12282/2 "2026-10-08T06:48:17Z")

</div>

I have checked the licensed macOS 2026 / v1 - Feb 2026 download. It supplies the same 26.0 headers we already have. I would like to narrow the source-binding question to a supported effect or exporter handoff, rather than playback tests.

Our UXP approval identifies a specific `Sequence.guid`. Mercury Transmit supplies opaque native context IDs. We need an authoritative relationship between these, and must exclude Source Monitor delivery, including the approved sequence opened in Source Monitor. We do not infer identity from focus, timecode or dimensions.

1. **Effect route:** is PF Utility `GetContainingTimelineID(effect_ref)` in the same identity domain as Transmit’s `inTimelineID`? Does it identify the containing or top-level sequence for nested effects, and can that context be joined to `Sequence.guid`? UXP documents creating/appending a video component and setting its parameter. Would a temporary, non-secret correlation value in an explicitly approved diagnostic effect be a supported handoff, or is there a supported read-only alternative? A copied parameter alone would not authorize another sequence or monitor.
2. **Exporter route:** Export Controller receives `mTimelineID` and `mModuleID`, whereas `GetExportSourceInfo` with `kExportInfo_SequenceGUID` requires an `exporterPluginID`. Is there a documented controller-to-exporter relationship that can bind that native timeline to the same GUID as UXP? Can it be obtained without rendering/writing an export or blocking editing? We are not proposing export loops as live capture, or substituting IDs between plugin contexts.
3. **Authority/lifetime:** even if either join is supported, what identifies Program rather than Source delivery, and what invalidates the join on edits, sequence rebuild/close, undo/redo, nesting, duplication or restart? How are already queued video and secondary audio associated with the approved context, and which callbacks/threads may legally perform the identity queries?

Please point us to the supported suite/sample and its lifetime guarantees. If neither route can establish Program-only authority, confirmation of that limit would help us choose an explicitly approved alternative. Sequence identity, monitor isolation and displayed-frame/revision correlation are separate gates. Sharing remains blocked pending those answers.

References: [PF Utility and Export Info](https://ppro-plugins.docsforadobe.dev/exporters/suites/) and [Export Controllers](https://ppro-plugins.docsforadobe.dev/export-controllers/export-controllers/). UXP reference classes: Sequence, VideoFilterFactory, VideoComponentChain and ComponentParam.

---

<div class="post-metadata">

**Author:** ![bbb\_999](https://sea1.discourse-cdn.com/flex015/user_avatar/forums.creativeclouddeveloper.com/bbb_999/32/4247_2.png) [@bbb\_999](https://forums.creativeclouddeveloper.com/u/bbb_999)\
**Post date:** [October 9, 2026, 12:35pm UTC](https://forums.creativeclouddeveloper.com/t/mercury-transmit-supported-program-only-sequence-binding-and-1-x-1-output/12282/3 "2026-10-09T12:35:19Z")

</div>

Hi Andy,

1. No.
2. There is no supported mechanism for doing so, nor are any guarantees documented. Transmit was designed to provide pixels and audio from Premiere’s active monitor; we have never _tried_ to offer the capability to ‘match back’ to previous outputs.
3. No, there is no documentation around the calling sequence you describe, nor do we offer any methods to preserve source isolation.

> [@andy.l](#):
>
> Our UXP approval identifies a specific `Sequence.guid`. Mercury Transmit supplies opaque native context IDs. We need an authoritative relationship between these, and must exclude Source Monitor delivery, including the approved sequence opened in Source Monitor.

No such functionality is planned or scheduled.

> [@andy.l](#):
>
> 1. Is there a documented controller-to-exporter relationship that can bind that native timeline to the same GUID as UXP?

No; no work planned here, either.

I think I understand what you’re _trying_ to do; Transmit API was never designed to support that behavior.
