Trust Test

for videos

Every TTG video is a record of what the camera saw
(and, if audio is included, what the camera heard).

  • Does your video earn the “Trust Test Guaranteed” label?


    • Fully meeting all 9 requirements below = qualifying for the TTG label.

    • Not fully meeting all 9 requirements below = not qualifying as TTG.


    • The 9 things that the TTG label guarantees

    It is very easy to make videos that meet all 9 requirements.

  • V1. No effects

    Recording a TTG video is simple:

       A. Use one camera, set to “Video” mode
       B. Apply no trickery or effects
       C. Let it roll

    Anticipating other Trust Test requirements:

    Video: Smartphones are best set to the mode titled “Video,” because fokeh and other cinematic effects are disqualified by V4, spatial/3D effects are disqualified by V6, and lighting effects are disqualified by V7. As per V3, any panning, tilting, or zooming will remain in the final video unless it can be removed by trimming the ends of the video.

    Audio: To meet V6, there can be no more than two audio channels in the final product (regardless of the type, location, and number of microphones used during the recording). To meet V7, all sounds on the video have to have been within earshot of the camera during recording. As per V3, any sounds heard at the scene that are recorded will remain in the final video unless they can be removed by trimming the ends of the video.

  • V2. No added AI

    Adding any AI-generated content at any time, whether visual or audio — no matter how small or minor it may seem — always disqualifies a video from TTG because the result is no longer “a record of what the camera saw and heard.”

  • V3. No interruptions

    • No interruptions during recording or editing: no pauses, no cuts, no overlays, no dissolves, no insertions.

    • Trimming the ends and cropping the edges of a video are allowed, but V3 means that any portion left after that must be continuous and uninterrupted, without cutting or inserting even a single frame in that remaining span.

    • For video, V3 means no views from multiple cameras, no B-roll footage, and no inserted graphics.

    • For audio, V3 means no voiceovers (apart from talking within earshot of the camera while filming), no added music, no dubbing, and no removal of any sounds heard at the scene except by trimming the ends of the video.

  • V4. No doctoring

    “Doctoring” can be thought of as “actions that keep the video from depicting what the camera saw” (and “what the camera heard” if audio is included).

    Video OKs:

    • Trimming and cropping

    • Color correction such that the result meets V7

    • Black-and-white (monochrome)

    • Full or partial correction of lens/camera anomalies

    Video Not-OKs:

    Except for the effects of the Video OKs (above)...

    . . . nothing in the video — no matter how small or trivial — can be added, deleted, replaced, resized, moved, modified, reshaped, or blurred (no fokeh) after the light from the scene hits the recording surface.

    • “No adding” disqualifies the overlaying of any words, logos, graphics, captions, labels, symbols, names, the term “TTG,” subtitles that cannot easily be turned off, or anything else on the image area of the video.

    • “No deleting” disqualifies the removal of anything in the image area (including the depiction of selfie-sticks) except by trimming and cropping.

    • • • • • • • • • • • • • • • •

    Audio OKs:

    • Overall audio volume may be raised, lowered, compressed, or turned off

    • Hiss and wind noise may be reduced or removed with non-AI methods

    • The ends of the video may be trimmed

    Audio Not-OKs:

    • No pitch correction after the sound reaches the microphone(s)

    • No localized amplification or attentuation of sounds out of proportion with other sounds in the audio (i.e., the louder parts have to remain louder and the softer parts have to remain softer)

    • No filtering out sounds that someone at the scene would have heard (dog barking, baby crying, throat clearing)

    • No sound in the final video that was out of earshot of the camera during the recording period

  • V5. No deviation from real time

    Each second of the video must correspond to one second during the recording period depicted.

    No variance in speed, no slow-motion, and no time-lapse    .

  • V6. No device exclusion

    TTG videos must be 100% compatible with all mainstream devices.

    • Any mainstream device that can record videos can create TTG videos.

    • Any mainstream device that can play videos can play TTG videos.

    • A video cannot meet V6 if any viewers can see or hear on their device a version of the video that others cannot see or hear on their device.

    TTG videos must be “standard,” not 360, with no spatial/3D effects and no specific playback equipment required beyond the capabilities of any mainstream device that can play videos.

  • V7. No misrepresentation

    . . . of the appearance or sounds of the scene depicted.

    “Misrepresentation” is gauged by respected international news agencies’ information reportage standards (rinairs).

    For the video aspects:

    • As per V4 (above), color correction of the video is permitted to make it meet rinairs (conversion to monochrome is an option as well)

    • V7 disqualifies any video choices that would cause a respected news provider to not publish the video, including excessive zooming, tilting, or panning; lack of anything in focus; and misrepresentative tones and colors.

    • V7 disqualifies videos recorded at fewer than 16 frames per second (the floor for what is considered a “video”)

    V7 disqualifies videos that depict as a primary subject anything recorded outside the visible spectrum.

    • V7 disqualifies the use of smartphones’ “lighting effects” because the result misrepresents the tones and colors that the camera saw.

    For the audio aspects:

    • To meet V7, all sounds heard on the video have to have been within earshot of the camera during the recording period

     • V7 disqualifies selective amplification or attentuation of sounds out of proportion with other sounds in the audio (i.e., the louder parts have to remain louder and the softer parts have to remain softer), but the overall volume can be raised, lowered, compressed, or completely turned off

    • V7 disqualifies the use of pitch-correction of the audio after it reaches the microphone(s)

  • V8. No deception

    . . . about the circumstances of the recording.

    “Deception” is gauged by respected international news agencies’ information reportage standards (rinairs).

    • If such news agencies would make separate note of “inapparent circumstances” (see #3 here) were the video used in a reportage context, then the video cannot meet V8 unless it is labeled TTG-IC .

    • But if such agencies would not make separate notice of the inapparent circumstances, then the usual TTG label will suffice.

    If needed to prevent deception, additional explanation can be provided elsewhere and noted with a * after the TTG-IC.

    ____________

    To prevent deception via re-recording non-TTG-qualified content, videos that feature as a primary subject

    “any non-TTG-qualified image that looks like a TTG-qualified photo or video, or any non-TTG-qualified audio that sounds TTG qualified”

    . . . can never meet V8 no matter how thoroughly they are labeled and explained.

  • V9. No anonymity

    In order for the video to meet V9, it must be accompanied by the name of its creator (because by attaching the TTG label, that person is personally guaranteeing that the video fully meets the eight requirements above).

    • The video creator can use whatever name they choose

    • Multiple people can stake their reputation on one label (see #1 here).

    Viewers are told to not trust the TTG label if they are not confident that a specific person is staking his or her reputation on the Trust Test Guarantee.