Event recording: program, ISO, audio and delivery

What to keep: program, ISO and presentation

Program records the live selection, including the graphics and audio carried by that output. Here, ISO means each input recorded separately, rather than camera sensitivity. It lets you change shots in the edit if you also have the audio and synchronization references.

For example, the ATEM SDI Extreme ISO can record eight clean inputs and program as nine H.264 streams and create a DaVinci Resolve project. This capability belongs to that ISO model; it is not a feature of every switcher. ATEM SDI Extreme ISO: specifications.

If the presentation contains text or tables, decide whether you also need it as a separate feed. A slide behind a speaker may look fine in a wide shot while being too small to read in the recording. Choose from the table according to what you will do afterward.

Choose the mode, codec and media, as well as the recorder

Resolution and frame rate describe the image; the codec describes its encoding; the container organizes video, audio and other data in the file. The extension alone does not establish whether an editor can use it. Request the complete format and open a sample in the receiving system.

The AJA Ki Pro Ultra 12G, for example, can work with up to four simultaneous HD channels or one 4K/UHD channel, with specified codecs and media for each mode. Its DataCalc estimates duration or storage from raster, frame rate and codec. Check the intended mode, as well as the product name. AJA: Ki Pro Ultra 12G.

Check that the media is compatible and can sustain the planned write rate. Allow for usable space, rehearsals, waiting time and a possible program overrun. Before clearing a drive for this event, make sure its earlier material has been preserved.

Calculate storage from the actual streams

For an approximately constant total bitrate, decimal GB ≈ Mb/s × seconds ÷ 8000. Add all video streams and audio, then allow for variation, overhead and reserved space. Mb/s means megabits per second; GB here means one billion bytes and is not the same as GiB.

For a numerical example, one hypothetical 40 Mb/s stream recorded for two hours occupies about 36 GB. Four cameras and program, each assumed to be 40 Mb/s, total 180 GB. Adding a 25% reserve in that scenario gives 225 GB for that copy. This is a planning example: neither the bitrate nor the allowance is a guarantee for every codec.

For variable bitrate or quality-based recording, measure a test file or use the manufacturer's calculator. Allow for the second copy separately. Test copying time too: the nominal port speed does not establish the actual transfer time.

Listen to the tracks and leave editing references

List voices, videos, remote guests, music and ambience. Choose whether to keep a finished mix, separate sources or both. Record a passage and listen on headphones: a meter can move even when the wrong channel reaches the file.

Check timecode, references and start procedures for your system. Note any media change or break. Identify each file by date, session, camera or program, time and observations. The editor should be able to locate picture and sound without guessing what happened.

Test the export too. Adobe explains that the channels and sample rate you can preserve depend on the format and its limits. Open the delivery to check that the mix and tracks reached the agreed channels. Adobe: audio in Media Encoder.

In OBS, check the container and combined workload

Choose codec and container separately. OBS distinguishes MKV and regular, fragmented and hybrid MP4/MOV: regular files need finalization, while MKV and fragmented variants have different interruption tolerance. Test import and remuxing if needed. OBS: audio and video formats.

Use a configuration that has passed capture, playback and import tests. An interruption-tolerant format does not repair a drive failure or replace a copy. Test recovery using rehearsal material; do not cut power during the event's only take.

If one computer streams and records, activate the show scenes, capture sources, graphics and all streams during the test. A streaming-only test does not establish the remaining capacity for recording.

During the show and at wrap

Assign someone to watch recording status, storage, audio and alerts. Note incidents and important moments. At wrap, stop as directed by the equipment and open every file: check duration, beginning, middle and end with picture and sound.

Make two checks on copies: integrity establishes whether they match the source, while playback checks whether they contain the intended recording. Neither replaces the other. Retain the original until the recipient accepts the transfer under the agreed terms.

Put in writing whether you are delivering a master, review copy, edit, subtitles or tracks, and by when. Review shots using the camera guide (Spanish) and languages using the guide to simultaneous interpretation (Spanish). Recording material and editing a finished piece require different time and tasks.

Recording tests and file closeout

Prepare this checklist before the event and complete it at wrap. If you change codec, media or stream count, repeat the test affected by that change.

1. Decide what to deliver

   List program, ISOs, presentation, tracks, editing and the format of each deliverable.

2. Calculate storage for each copy

   Add bitrate and duration, allow a reserve and allocate capacity to media and copies.

3. Prepare media, channels and names

   Check compatibility, usable space and the signal on each input. Prepare file identification.

4. Record a representative passage

   Use the show's scenes, audio and transitions. Check duration and stability with all streams active.

5. Open the take where it will be used

   Play picture and sound and import the file into the intended editing or receiving system.

6. Monitor recording and close it correctly

   Assign status monitoring, note incidents and stop according to the equipment procedure.

7. Verify copies before handover

   Check integrity and content, include the file list and retain media as agreed.

What to tell us when planning the recording

Send the program, total duration, cameras and feeds. Include the intended use, expected format, languages and deadlines. If copies are needed the same day, say so at the start and identify the recipient and the time available to copy and check them.

Check the quote for capture, audio, media, monitoring, copies and included postproduction. If there is also a live stream, use the streaming guide (Spanish) to prepare it: monitoring the transmission and checking the recording are separate tasks.

If the file contains clicks or picture skips

These two public accounts help separate audio faults from performance problems. One improved after levels were reduced; the other continued to fail with Quick Sync. For audio, compare source, output and file; for picture skips, examine rendering lag and encoding lag separately.

Find where the click begins

Listen to the same passage at the source, system output and a short recording. Note whether the click coincides with a loud voice, music, cable connection or source change. Keep that sample so you can return to the same moment and check whether the fault was already in the signal or appeared later.

When feeding an ATEM Mini Pro from a console or interface, check Mic or Line against the output used. The manual says Mic adds more gain; choosing it for a line-level signal can make the input too strong. Adjust the source and mixer, watch inputs and program and leave headroom for peaks. The manual identifies 0 dB as clipping.

In the account, a MOTU 828es headphone output fed Mic1 on the ATEM Mini Pro through a 3.5 mm TRS connection. Reducing levels improved the result, without an official diagnosis of the cause. This does not establish that any yellow meter indication creates clicks or that a hum eliminator fixes them. If the fault continues, compare cables, routing and media in separate tests. Listen at actual performance levels before accepting the setup.

For inputs and meters: ATEM Mini: Setting Audio Input and Output Behavior and Audio View.

Separate rendering from encoding in OBS

Reproduce the fault with cameras, videos, graphics and scene changes, and keep that session's log. An empty scene may hide the computer's limit. Examine rendering lag and encoding lag separately: they are different tasks and need not fail for the same reason.

OBS uses the GPU to compose a scene even when another component encodes it. Moving from x264 to supported hardware encoding may reduce CPU load while leaving a heavy composition unresolved. Copy the scene collection, remove sources, filters or animations one at a time and compare. If delivery allows it, then test a lower frame rate or resolution and record the measurements.

An ICQ value is not a universal preset. If you need several destinations or independent recordings, test them together for the expected duration. If the computer cannot sustain them, reduce the load or change the system before the show.

The workload split is explained in OBS: Encoding Performance Troubleshooting; options are covered in OBS: hardware and Intel Quick Sync.

When a take starts in sync and drifts by the end

If you will combine switcher ISOs with a separate audio recording, make a long test with visible and audible marks at the beginning, middle and end. Compare all three points before editing the event. Note whether the offset stays constant, grows gradually or changes after a skip. These differences help identify what to investigate; moving all the audio once is insufficient when the gap changes during a take.

The ATEM Mini ISO manual documents synchronized timecode among its ISO files. That describes the files recorded by the ATEM; it does not guarantee that audio recorded separately in a DAW maintains the same rate. The Sound Devices manual distinguishes an initial offset from drift between recorders. Check actual frame rates, editing settings and each recorder's reference before choosing a correction. If the design needs genlock or word clock, first check the inputs each model supports and how you will distribute the reference.

The advice to switch everything to 60 fps was not established either. Keep the originals and give the editor or support a sample with all three marks, the settings and the point where synchronization changes.

References: ATEM Mini ISO: Recording ISO Files and Sound Devices PIX 270i v2.33: Synchronization and Timecode, p. 39. The second source explains the principle; that recorder was not part of the reported setup.

The video plays in the camera, but the copy shows 0 bytes

If the camera plays a video but Windows shows a 0-byte copy, retain the card and review the import method. Do not format it or continue recording to it while the material remains unresolved. First compare the clip duration and name in the camera with what appears on the computer; a thumbnail does not confirm that the video was copied.

For the EOS R50, Canon instructs users to import files larger than 4 GB with EOS Utility or a card reader. It warns that standard operating-system copying can fail for that size. Check the actual file size when accessible: the 0-byte symptom alone does not establish that the file exceeded the limit.

Use EOS Utility for import; EOS Webcam Utility serves a different purpose. If the application does not connect, check its version and main window. The R50 cannot communicate over Wi-Fi and an interface cable at the same time: choose one path. If using an SDXC card and a reader, confirm reader support.

Play the copy from the beginning, jump to the middle and check the end with audio. Keep another copy before releasing the card. This is an import test, rather than a promise to recover deleted clips. See files larger than 4 GB and computer connection.

OBS pauses: check hotkeys before changing the encoder

If OBS shows a pause indicator while recording, open that session's log and locate the interruption. The message “Pausing recording due to hotkey” identifies a hotkey-triggered pause; picture skips alone do not establish that cause.

In Settings > Hotkeys, check Pause Recording, Unpause Recording and Stop Recording. Remove assignments you do not use or choose alternatives you can rehearse. If the keyboard is shared with presentations, chat or video playback, test the usual keys in those applications too.

OBS lets you choose hotkey behavior according to the active window. For shortcuts to respond only while the main OBS window is active, review Settings > Advanced > Hotkeys > Hotkey Focus Behavior and Disable hotkeys when main window is not in focus. Test scene and recording controls while switching windows before the show; check labels in your version and language and verify behavior on your system.

Make a short recording using the actual show operation: change scenes, type, start a video and return to OBS. Then play the file with audio and verify that all those moments are present. The general OBS guide explains hotkeys and testing; the control names distinguish pause, resume and stop.

A track seems missing: check what is playing the file

If you recorded voices and videos separately but hear only one part, keep the take and inspect its tracks before rerouting. A useful test records voice first, video next and then both; these moments identify the contents of each track.

OBS requires two checks: assigning each source in Advanced Audio Properties and enabling the recorded tracks in Settings > Output. The official multitrack recording guide suggests a full mix on track 1 and separate sources on subsequent tracks.

Play the test and select the tracks one by one. In VLC Desktop 3.0, use Audio > Audio Track. Listen to each source's passage: a listed track alone does not establish that it contains the intended voice or video sound.

A conventional player listens to one track at a time; it does not automatically sum separate tracks. If the delivery must open with all sound present, also prepare the agreed mix. For editing, import the test into the editor's application and confirm access to the sources before recording the full event.

With XR18, choose whether to record the mix or the sources

Before building tracks, agree on the delivery: separate voices for editing, a finished mix or both. Also note videos, audience questions and ambience. Prepare a list for each file; the availability of eighteen inputs does not determine what you need to preserve.

The X18/XR18 diagram, p. 43 distinguishes sources and tap points. Note the signal and processing assigned to each track and verify them in a sample before handover.

For the recording test, start with one voice, then a video and finally both. Listen to the mix and open the sources intended for separate delivery. If audience ambience is needed, check that capture too: sound heard in the room is not automatically added to the file.

Then test a normal operational change, such as lowering a voice in the room mix, and watch what happens to its track. This comparison helps identify the configured tap point. Keep the patch with the sample and repeat the test with the program's expected duration and changes.

After granting permission, test a file in the show application

Prepare the recording with the same application, project and Mac account that will be used in production. If a test was completed with another application, note which one: it provides a comparison, but the path that will run during the event still needs checking.

macOS audio-input access is granted per application. Apple instructs users to reopen the application and test recording after enabling it. Include that check if the application or computer changes; do not treat an earlier capture as evidence of the current state.

Make a short recording with a voice and speak the name of the source being tested. Stop, open the file and listen to that passage. Repeat for the inputs you must deliver, so a missing track or wrong source can be identified before the session begins.

Keep the take with the project settings and record when it was made. If something changes afterward, the file provides a known comparison point. Also agree on how the recipient will check the recording: a visible waveform and listening to the file are useful checks to document together.

Check the saved take through a known monitoring path

After testing capture, save a short take with spoken source identification and open a copy for listening. Note the file and passage checked. A waveform helps locate the passage; hearing the expected sources provides another check.

In REAPER, check the Master's hardware destination and that send's mute state. Guide 7.81 allows these sends to be configured separately. Activity on a project meter does not establish sound at the physical output being used.

If you change the monitoring interface for a test, note the change and restore the capture configuration before continuing. Make these checks between takes and retain the original file. The next operator needs to know the monitoring path used to approve the take, as well as where the tracks were saved.

Split recording files in OBS

To split a panel into files, rehearse with the event profile and continuous voice. Confirm the OBS version and a folder where each segment can be identified.

In OBS 32.1, use Advanced output and Standard recording: enable Automatic File Splitting and select Only split manually. Assign Split Recording File in Hotkeys. Add Chapter Marker is a different action.

Create several segments during the test and listen to each beginning and ending. Check the first cuts too; do not treat a waiting period as a guarantee of audio. Keep the originals before editing.

ATEM Mini Pro: detected drive and verified recording

If ATEM Mini Pro recognizes the drive but remains in STOP, note the media, capacity, cable and versions. A green indicator means the drive is ready; the disk indicator is red during recording.

First compare the format: the manual supports HFS+ and exFAT. Preserve the files before formatting. To check software, connect the ATEM over USB or Ethernet and open ATEM Setup; it reports when the version installed on the computer is newer.

Schedule any update during preparation, allowing time to repeat the rehearsal. Record, stop and open the file on the computer. Use the intended duration and quality: a short test verifies only that interval.

Leave a handover sheet for the person using the files

Ask the team using the material for a deliverables list: complete talk, separate cameras, presentation or edited pieces. Use names they will recognize even if they were not at the event.

For each deliverable, record its recipient, deadline, languages and agreed editing. Make missing information visible and assign someone to confirm it rather than filling it in by assumption.

Include responsibility for logos, participant names, presentations and corrections. Share the sheet with recording and editing before the show so both teams understand the expected delivery and missing materials.

HyperDeck Studio HD Plus: test what triggers recording

On HyperDeck Studio HD Plus, selecting Trigger Record does not guarantee control by any camera. The HyperDeck Studio manual distinguishes Video start/stop, which follows a compatible command sent over SDI, from Timecode run, which triggers on a valid timecode signal and stops when that signal is no longer received.

Before the show, start and stop a take from the camera and check both ends of the file. Select None to disable automatic triggering. A picture on the recorder does not establish that this control path works.

HyperDeck Studio HD Plus: coordinate network copying while recording

The February 2026 manual allows network file transfers while HyperDeck records. The unit automatically adjusts transfer speed to avoid affecting recording.

For FTP, the computer and recorder must share a network. Prepare the FTP client and the IP address or URL from HyperDeck Setup, and check that access is enabled. Allow for the transfer-speed adjustment when agreeing on timing.

Ki Pro Ultra 12G: identify files by camera

In Multi Channel Mode, the v2.7r3 manual allows an identifier for each input: enable it with 19.5 and set the text in 19.6–19.9 from the panel or WebUI.

Choose identifiers the editor can recognize and note which camera feeds each input. The examples retain _1 through _4 before the added identifier; check the full filename when organizing the delivery.

REAPER: choose which channels reach the render

In REAPER 7.81, File > Render, Selected tracks (stems) creates a file for each selected track. Set Channels to mono, stereo or the required multichannel count. Check Multichannel tracks to multichannel files and Render only channels sent to parent; the latter retains only channels sent to the parent track.

Without Render stems pre-fader enabled, stems include FX, fader, pan and received inputs. Agree on that result with editing. Render a sample and listen to each exported channel.

Wave Agent: connect each file with its tracks

For WAV/BWF, in Wave Agent Beta 1.2, Current Name is the name embedded in BEXT/iXML, while File Name is the disk filename. Review both on a copy. Rename uses Scene and Take and also renames files from the same take; it has no undo.

Complete Track Name for each channel and save. In Reports, include File Name, Scene, Take, Start TC, Duration and Track Info; export a CSV to accompany delivery. Check the names with editing. The official page states that 32-bit float is not supported.

Ki Pro Ultra 12G: confirm the number of recorded channels

For SD/HD, select 1.10 Encode Channels: Multi Channel in CONFIG. 1.11 Channels to Record selects 2, 3 or 4; each numbered SDI input corresponds to its channel. Check compatible media and formats. Record a sample and open each file.

AJA Ki Pro Ultra 12G, v2.7r3, pp. 30–31 and 45.

ATEM Mini Pro ISO: enable input recording

In ATEM Software Control's recording palette, select ISO record all inputs before recording. Program and ISOs are stored in the project folder, with input videos in Video ISO Files. Test a suitable drive and open a sample from each input.

Blackmagic, ATEM Mini, June 2026, p. 75.

X-LIVE and OBS: check the sources recorded to SD and USB

Behringer describes simultaneous, independent USB and SD operation. This alone does not establish two separate source lists. The X32 Rack manual with X-USB, §7.4.5, p. 42, explains Card Out; it does not document firmware 4 User Out.

We suggest noting sources, Card Out positions and destinations in the worksheet. Record an identifiable sample of each source and listen separately to SD and USB capture. Check the room/stream mix (Spanish) before changing blocks.

X32 Rack with X-LIVE: keep a microphone live during Play

That example reserves the entire local block.

The X-LIVE guide, §5.3, p. 18—PDF page 10—explains Record/Play presets. Automatic uses Record when stopped or recording and Play during playback or pause. The guide does not specify firmware; compare its screens with your unit.

We suggest rehearsing Play, Pause and Stop outside the show, listening to the microphone and tracks in each state. Record the result in the worksheet; also check the card return (Spanish).

X32 and REAPER: I expected the mix but recorded inputs

The X32 manual, pp. 38/52/54, distinguishes preamps, output sources and tap points. We suggest checking the mix source/tap and its Card Out block, testing fader and mute outside the show and listening to the file. Note the result in the worksheet.

To choose the deliverable, review the room/stream mix (Spanish).

X32: AES50 inputs reach the console but not the DAW

The stagebox model, firmware and a confirmed solution are still missing.

The same X32 manual, §§7.4.1/7.4.5, pp. 51/54, separates channel inputs from Card Out. We suggest following one source: AES50 port/range → Card Out position → DAW input. Record and listen to a sample; complete the worksheet.

The document is Preliminary, with no declared firmware; it does not establish current User In/Out behavior. For monitoring the return, see this guide (Spanish).

X32: record the delay return separately

The X32 Preliminary manual, §§7.4.4–7.4.5, pp. 53–54, permits FX sources in P16 and P16 blocks in Card Out. We suggest testing: FX return L/R → two available P16 positions with their tap points → Card Out block → DAW tracks. Before changing it, note all eight block destinations and who uses Ultranet.

Outside the show, record a sample and listen to both tracks. Record settings and restoration in the worksheet. The manual does not declare firmware: check your console's menus. To review other sources, continue with this guide.

Fairlight Live: did I record inputs or program?

Define the file you want to deliver before recording.

The manual, August 2026, pp. 59 and 61 states that Player/Recorder captures unprocessed inputs, excluding buses and masters. Arm the channels and press Record.

We suggest listening to the take and program, then recording the results in the worksheet.

Related: recording handover and cameras and streaming.

Fairlight Live: virtual soundcheck

The manual, August 2026, pp. 60–61 places manual loading in Media inside the show's parent folder. Enable Play-Rec and Play for each channel; Input displays Player. Start playback in Player/Recorder.

We suggest using a copy of the show and testing one channel before the complete set. Use the worksheet.

Related: sound monitoring and returns (Spanish).

vMix 29: watch the counter and check the recorded file

Record warns that dropped frames can also occur when frame rates differ. For AVI, Setup instructs users to match the recording Frame Rate to Settings > Display. Also compare the capture device's output format.

In Statistics, Encoders includes recordings, and Codec identifies CPU or GPU. See the counter definitions (Spanish).

We suggest recording initial and final values, duration and settings in the worksheet, outside a live show. Play the original and mark the affected passage. Avoid changing the capture device, format and destination together.

Related: compare playback and cameras and recording.

vMix 29: when to test Recording Memory Buffer

The Recording guide states that Recording Memory Buffer can reduce dropped frames with slow drives and recommends a value of 10. That section does not specify a unit or sizing formula. Record the value as displayed in your version.

This suggests a test; it does not establish that the drive caused the skips. We suggest retaining the initial setting, changing one thing at a time and comparing files and counters over the same workflow outside a live show.

Use the worksheet. Record destination, concurrent tasks and duration. When comparing another drive, keep the same format and buffer; then rehearse the complete program.

This control belongs to recording. For the NDI input, review its buffer setting (Spanish). Coordinate media and deliverables with cameras and recording.

vMix 29: a direct source and an Output record different things

Support notes that MultiCorder depends on source timing information.

MultiCorder records the source without effects, color correction or position adjustments. Output 1–4 retains processing, and its audio follows the bus selected for that output.

The manual excludes irregularly timed sources such as NDI-HX, Desktop Capture, Stream and vMix Call. Check the input you are trying to record. For an Output test, note source, layers, effects, bus and destination in the worksheet.

The manual requires an SSD and makes recording count dependent on system capacity. We suggest checking each ISO under the full workload.

Related: counters and cameras and recording.

vMix 29: test the recording format and compatibility

Fault Tolerant Recordings distinguishes ordinary MP4 from MP4 with Fault Tolerant enabled. The latter requires compatibility testing; it does not work on every device and application.

FFMPEG MOV and MP4 are listed as interruption-tolerant; their Indexed variants are not. Do not choose by extension alone. Interruption tolerance and smooth playback require separate checks.

We suggest testing playback, audio and import in the editor's application before the show. Keep the original and record the complete format in the worksheet.

The manual recommends New File Every to distribute a recording across files so much of the material may survive if one file is lost. Test the expected duration and inspect the segments.

Related: editor handover and cameras and recording.

Original NDI MOV: audio and file splitting work differently

If the NDI MOV has no sound, first check the recording mode.

MultiCorder in vMix 29 states that Record NDI in original MOV format saves NDI compression without recompressing video. In that mode, the folder setting applies but other settings—including Audio and New File Every—are ignored. Selecting Master in MultiCorder does not change that take or add sound to an earlier file.

To locate missing audio, we suggest a test outside the event: send an identifiable voice from the sender, listen to that NDI input in vMix and record an original-MOV sample. Open the file at the editor's workstation. Record separately what you heard at the input and in the file; do not rely only on the Master meter.

If the voice is also absent at the input, check sender audio selection and transmission before changing recording. If the input has sound but the file seems silent, retain the sample: identify its audio channels and compare playback and import in a compatible application. The difference does not yet establish whether capture or MOV reading failed.

The manual warns that this special MOV may not open in some applications. For delivery requiring a vMix mix or timed file segments, test another compatible recording path and verify its result. In original MOV mode, waiting for the New File Every interval cannot verify an automatic split that the mode ignores.

Complete the audio and import worksheet with sender, NDI type, versions, sample and result. For recording a processed output, review source and Output. Share the handover test when coordinating cameras and recording.

vMix 29: check the WAV alongside the main recording

WAV File Record is enabled in Recording Setup. It creates an uncompressed PCM WAV alongside the video; the green button indicates that it is enabled. It does not alter the audio included in the video, and its audio format is fixed.

We suggest testing the main recording with a voice and a visible clap. Open video and WAV separately; check content, duration and import at the editor's workstation. Record versions and what you heard in the worksheet.

This check concerns Recording Setup; for original NDI MOV, review the ignored settings. Coordinate deliverables with cameras and recording.

Premiere imports only audio: identify the codec before converting

Adobe suggests checking source, codec and format with MediaInfo when Premiere imports only audio. Record those details and the exact message.

The NDI 6.1.1 release notes document a fix for an exception when importing recorded SpeedHQ files into Premiere. That note does not identify the cause of your file's problem. If the error names that format, also record the installed NDI version.

We suggest retaining the original and testing an identified copy at the editor's workstation. Note Premiere, operating system, codec and result in the worksheet. Before converting, agree on the delivery format and check a sample.

Related: editor handover and cameras and recording.

SRT in MultiCorder: MP4 is selected but the file is .ts

A direct SRT input in MultiCorder produces .ts even when MP4 is selected. The vMix 29 manual states that it records the native Transport Stream and respects only the folder setting; SRT Input, MultiCorder Recordings confirms that the selected compression is not applied. Do not plan audio or file segments using settings that this path ignores.

There is not enough information to turn that attempt into a repair procedure; changing the extension does not establish a container conversion either.

Before a conference with remote guests, we suggest recording identifiable voice and movement with the SRT send already active. Close the take and check picture, sound, duration and synchronization at the beginning and end. Import a copy into the intended editor. If video does not appear, keep the file and message; a player's failure to show it does not establish that the picture is lost.

If you need a processed format, copy the configuration and test the input through an available Output. MultiCorder distinguishes that path from the direct source. In Settings > Outputs / NDI / OMT / SRT, open the selected Output's cog: Audio Channels defines its bus and channels for MultiCorder. Check that the output carries the intended picture and voice before recording it.

That test uses vMix processing and shares resources with the rest of production. Compare both paths under the expected workload; it is neither recovery of the earlier .ts nor a backup independent of the computer. Check whether the output is already assigned to another destination.

Record direct source or Output, bus, format and editor in the SRT worksheet. See source and Output and agree on delivery using cameras and recording.

OBS over SRT: test recording with the send already running

Record OBS and vMix versions and compare capture startup. A black recording can have different causes; check source, path and file.

The old “option 2” link now redirects to a wiki where the original article is unavailable. We do not reconstruct its settings by assumption. The official OBS guide published January 14, 2022, under Stream with SRT, describes Settings > Stream, Service > Custom, an srt://IP:port URL and an empty stream key. That document does not establish that this path is the same option used in 2020.

Outside the show, we suggest comparing two takes: start recording after the send has been active for a while, then repeat with another startup order. Keep other settings unchanged and note start times. Play each file with sound and import it into the editor, checking the first frames and the end in particular.

If only one order fails, retain both samples and configurations for investigation with support. If both fail, review the file, sender and path without automatically attributing the fault to those headers. Do not restart a guest's send during a session to run this comparison.

In the worksheet, separate three results: live reception, new recordings and recovery of earlier material. A reported improvement in new files does not demonstrate recovery of an earlier take. Prepare the test using cameras and recording.

A Ninja V copy is interrupted: what to record before handover

Identify the reader, adapter and volume status. Access through a different combination does not confirm the cause or recover files by itself.

Master Caddy Docking Station, PDF pp. 36–37 distinguishes two models: USB 3.1 Gen 1 & 2.0 with two connectors, and USB-C 3.1 with one. The former may require both connectors for power; that instruction must not be generalized to every reader.

Atomos, Connecting Media allows copying through Finder or Windows when the operating system recognizes the drive. If it is not yet an accessible volume, separate that problem from playback in the editor.

Transferring files, PDF p. 363 describes copying files to your own storage. That operation does not include a verification result for our copy.

For a proposed test, prepare a destination with enough space and an identified session folder. Copy rather than move the material. Keep the expected file list and record which file failed, at what point, with which message and whether the volume remained visible. A completed retry does not erase earlier errors from the record.

Compare connections after the transfer has ended, changing one identified component at a time. Follow the dock model's procedure; the message alone does not establish power, cable or SSD as the cause. An alternative reader without confirmed compatibility can introduce another unknown.

We suggest comparing source and destination names and sizes, then opening the copy in an application compatible with the recorded codec. Check duration, picture and audio at the beginning, middle and end. This sample does not check every frame or replace integrity verification. Mark files with errors as pending and retain the original media until copy closeout is agreed.

Tip: complete the table and the offload worksheet. Link that record with the long-take test (Spanish) and the planned delivery in cameras and streaming.

Ninja V ProRes RAW is black in Premiere: what to check

The user had already installed Apple ProRes RAW 1.3 without improvement. The account confirms neither a solution nor the Premiere version or GPU.

Adobe ProRes RAW support, updated September 30, 2026, describes Apple's decoder and CUDA or OpenCL rendering on Windows. It includes Nvidia, AMD and Intel GPUs with updated drivers. In Premiere, locate the renderer under File > Project Settings > General > Renderer and record the active option.

Apple ProRes RAW for Windows 1.3, published March 8, 2024, documents 64-bit Windows 10 and RAW/RAW HQ playback in compatible applications. Installing it alone does not establish that the clip opens in your editor.

If using an Adobe beta for ProRes RAW, identify the build and check its official notes before choosing the Windows decoding path. A feature announced for a beta does not identify the installed production version.

First separate the symptom: does the editor reject the file, import it without a picture, or show a picture in one viewer and black in another? For a proposed test, use a copy and a test project. Record the identified codec, profile, resolution and frame rate, exact message and affected viewer. A MOV extension does not identify the required decoder.

Record editor version and build, operating system, GPU, driver and Apple package version if applicable to that path. Compare the environment with the relevant documentation before installing or changing anything. An update recommendation is neither a completed update nor confirmation that it resolves the case.

Versions and supporting tests are missing; the cited documentation covers several brands. Keep the recorded configuration and provide the test file and details to support if the picture remains black.

Tip: complete the ProRes RAW import worksheet. If SSD access still fails, start with Ninja V offload (Spanish). When coordinating cameras and streaming, also agree on the editor receiving the material.

Must ProRes RAW be converted before opening in DaVinci Resolve?

Resolve showed only audio. The account does not give its version or confirm a successful import; conversions were suggestions from other participants.

Ninja V, Importing clips, PDF p. 365, separates MOV codec compatibility from the specific ProRes RAW list. Its older edition cannot establish what every editor supports today; documentation for the software and version is needed.

The Blackmagic support center documents Apple ProRes RAW playback in Resolve and Resolve Studio 20.2, released September 10, 2025, and compatibility improvements in 20.2.1, released September 23, 2025. A 2022 answer requiring conversion first therefore needs review. Version 20.2 identifies that addition, rather than the latest available release.

Before choosing a conversion, identify the recorder file and camera-card file separately; do not assume the same codec. Record the receiving workstation's Resolve version, edition and operating system. Audio without a picture does not by itself establish file corruption.

We suggest importing a copy of a short clip into a test project and checking picture, sound, duration and an exported passage in the agreed delivery format. If it fails, retain the message and consult that version's notes with camera, recorder and file details. Announced format support does not certify that our clip and workstation passed the test.

If the recipient retains a version or workflow that cannot use the original, agree on a derived file and rehearse that path before the show. Retain the RAW and document tool, version and settings. An exported intermediate must not be presented as the original or as a guarantee of retaining identical RAW controls.

Tip: use this table and the import worksheet together with the recording delivery guide. Share the editor, version and expected format when requesting cameras and streaming; this lets us prepare a specific test.

A7S III ProRes RAW looks red in Premiere but normal in another viewer

Adobe ProRes RAW support, updated September 30, 2026, documents Exposure and ColorSpace controls in source settings. Review them in the installed version before attributing the appearance to the recorder.

Additional beta ProRes RAW controls must be checked against official notes for the installed build before use in production. Do not assume they are present in a stable release.

Ninja V, Monitor tab, PDF p. 328, distinguishes Native, Rec.709, HLG and PQ display modes. Record the mode and monitoring LUT before comparing the monitor with the editor.

The Sony Cine's FX3 and Alpha 7S III guide, May 24, 2021, explains that the Ninja V monitoring option changes its waveform display and describes RAW conversions in Premiere. It is a historical workflow reference rather than a current configuration validated for our workstation.

We do not turn this into a downgrade/upgrade procedure: the historical official note and its applicability to those files still require verification. Before recommending an update, record camera and Ninja firmware, editor version and affected file codec.

We suggest comparing the same copy of a short clip in a test project. Record source RAW conversion, added LUTs or effects, project color settings and display destination. Change one identified variable per attempt and keep the earlier state. A better-looking result in another viewer helps locate a difference, but does not certify which viewer represents the file correctly.

If the shooting monitor displayed a LUT and the editor applies another conversion, identify both before adjusting color by eye. Request a sample of the intended look and prepare a short export for the recipient. This provides a specific file for discussing the result; it does not alone establish preservation of the full range or absence of clipping.

Tip: use the LUT and color worksheet together with editor version and delivery format. Record the destination in recording handover and share it when coordinating cameras and streaming.