All Products
Search

This Product

    Intelligent Media Services:intelligent production FAQ

    Document Center

    Intelligent Media Services:intelligent production FAQ

    Last Updated:Dec 05, 2024

    This topic provides answers to some frequently asked questions about intelligent production.

    Table of contents

    FAQ

    How do I export the output video to ApsaraVideo for VOD when I edit a video?

    When you call SubmitMediaProducingJob to submit an intelligent production job, set OutputMediaTarget to vod-media, and set StorageLocation and FileName in OutputMediaConfig to the address of the media file in ApsaraVideo for VOD and the name of the media file. Example:

    "OutputMediaConfig": {
  "StorageLocation": "outin-8e7*******.oss-cn-shanghai.aliyuncs.com",
  "FileName": "vod-output.mp4"
}

    How do I view the result of an intelligent production job?

    After you call SubmitMediaProducingJob to submit an intelligent production job, the system returns a JobId. You can call GetMediaProducingJob and pass the JobId to the request to query the intelligent production job. The returned Status parameter indicates the status of the job.

    How long does it take to complete an intelligent production job?

    Typically, the duration depends on the length of the output video. For example, if the length of the output video is 5 minutes, it takes 5 minutes to produce the video. However, taking into consideration the time spent on queuing, file analysis, and downloading, it takes at least 15 seconds regardless of the length of the output video. It costs 10 seconds to 2 minutes to produce a short video of 15 seconds in length depending on the complexity of the video. If you submit tens of thousands of jobs at a time, the jobs will be added to the queue and then executed in sequence. To accelerate these jobs, submit a ticket.

    What are the factors that affect the time consumption of video production?

    A video is edited frame by frame. The time consumption of video production increases with the resolution and length of the output video. If the output video uses large numbers of virtual effects or transitions, or the video is scaled (for example, from 4K to 480P), the time consumption also increases. Improper use of the timeline also increases the time consumption. If the time consumption does not meet your expectation, submit a ticket or join the DingTalk group 48335001108.

    Why does the length of the output video not meet my expectation?

    • Transitions affect the length of the output video. A transition connects two shots. During a transition, the two shots are played at the same time, which means that the latter shot will be played earlier than the actual time. Therefore, transitions reduce the length of the output video. If you want to keep the length of the output video unchanged, you need to reserve sufficient time for transitions. You can also use DLTransition to add frames to the transitions.

    • Using AI_TTS increases the length of the output video. When AI_TTS is used, the length of the output audio track will be longer than that of the output video track. This increases the length of the output video. To resolve this problem, refer to Automatically align materials and the material length.

    • The improper use of the timeline also affects the length of the output video. When In and Out are not configured but TimelineIn and TimelineOut are configured, the length of the output video equals the length of the video material by default. We recommend that you specify in = 0 and out = timelineOut - timelineIn to process the video material.

    Why does the output video display a black screen after xx seconds?

    This is because the length of the video track is shorter than the length of the audio track. For example, the video track is 6 seconds in length and the audio track is 12 seconds. In this case, a black screen is displayed during the latter 6 seconds. To resolve this problem, specify the video track as the main track to align it with other tracks, or adjust the length of other tracks.

    Why does the system prompt "TimelineFormatError" when I call the intelligent production job API operation?

    Check whether the timeline format is valid and ensure that JSON grammar errors exist. For more information about the timeline format, see Timeline configurations. To view more timeline examples, see Timelines

    Why are the subtitles of the output video not displayed, garbled, or not correctly displayed?

    If the output video uses minor languages, such as Korean, Arabic, or Mongolian, font rendering may cause this issue. You can try to the use the default Alibaba Cloud Puhui font when submitting the intelligent production job. If the problem persists, joint the DingTalk group 48335001108.

    Why do the positions of the images, text, or subtitles not meet my expectation?

    • Make sure that the screen size of the output video is the same as that of the preview. Use a relative position between 0 to 1 to adjust the size.

    • If the materials in the same track overlap in time or positions, anti-overlapping may be triggered. Consequently, their positions are changed. Try to move the materials to different tracks.

    Why does the FontSize of the subtitles not meet my expectation or differ from the preview?

    • If you set the FontSize property in Effect Type:Text, the font size is adjusted based on the material size and output video size. You can change FontSize to FixedFontSize to prevent the system from adjusting the font size.

    • You can use SubtitleTrackClip to specify the content of the subtitles. If you specified the font of the subtitles, the font render height in pixels may be smaller than the font size. You can specify SizeRequestType=Nominal to set the font render height to the font size.

    • Specify the preview size to ensure that the font size used in the output video is the same as that of the preview. For example, to produce a video of 720P, specify the preview size by setting FECanvas={"Height":720,"Width":1280}.

    What do I do if the system prompts "Throttling.User" when I submit an editing job?

    • The maximum QPS of the write API operations of Intelligent Media Services is 30. If you submit large numbers of concurrent jobs, traffic throttling may be triggered. In this case, wait 1 second and try again.

    • For 30 QPS, you can submit up to 1,800 jobs per second. To continuously submit jobs for tens of minutes based on a higher QPS in scenarios such as producing millions of videos within half an hour during marketing activities, submit a ticket to increase the QPS.