I just started to figure out the FFmpeg options. Why is the Slide Deck (H.264) preset is bad for compressing regular video with a child, for example. The question is about the quality of the video, not the sound.
Sorry, i don’t understand your question.
Pls be more precise for a precise answer.
What is bad for what purpose? What exactly do you want to do with SC?
Also explain the quality of the raw material: photos, video clips, what so ever…
Video 4K(3840x2160) from Xiaomi Redmi Note 9 Pro.
If i export a project(1 minute) with a preset: H.264 Main profile, then the file size is 240mb.
If with a preset: Slide Deck (H.264), then the file size is 80mb.
Those the size is three times smaller.
But visually on the 17" screen (PotPlayer 1.7), I do not see the quality difference between these video files.
Therefore, my question is - how is such a small file size achieved? Can this preset lower video quality?
I hope now I wrote it clearly )
I made the Slide Deck preset, so you’re in the right place.
Slide Deck was created during the pandemic as a way of shrinking PowerPoint presentations and screencast tutorials and video conferencing meetings into the absolute smallest file size possible while retaining sharp quality for static images like slides, hence the name. This allowed remote education lessons to be accessible to people with slow Internet connections, and allowed up to 20-minute Zoom meetings to fit into a standard MP4 file that was still within e-mail attachment size.
The export preset compromises video quality on purpose to be no better than webcam footage, because that is the highest quality source it expects to process. So it optimizes bitrate to webcam quality for moving video to keep the file size small, and only aspires to reach “good enough” for quality. This was an intentional design decision. If you need higher quality video, you will need a different export preset.
As for static images, they should be nearly perfect, as the preset takes major compression advantage of seeing the same frame over and over again.
For more information:
I wrote my previous message before your second message arrived. Looks like I didn’t answer what you were asking lol.
For static images and the intended PowerPoint presentations, the file size is often 10 times smaller or more. Video will be less efficient.
There is also the Slide Deck (HEVC) option that can usually get 30% smaller than the H.264 option. But it exports kinda slow.
You are in loophole territory.
Slide Deck was designed to be on the verge of “good enough” for video conferencing webcam footage at 1080p. However, that CRF will look significantly better at 4K than it would at 1080p, because its compression is based on a quality target rather than a bitrate target. (Your phone, meanwhile, was probably based on a bitrate target.)
An interesting property of H.264-style compression is that the higher the resolution goes, the more efficiently the codec can work because neighboring pixels become more similar in color, which compresses more easily. As in, 2X improved video quality does not require 2X bitrate to achieve it. The export preset is also asking the encoder to look further ahead and use more reference frames to find the maximum amount of duplicated data in the video, which greatly reduces size but adds noticeable encoding time. The high reference frame count is why
libx264 needs to be used, because most hardware encoders can’t even implement these settings.
Combine this 4K look-ahead advantage with the Note 9 not exactly being a high-end dedicated camera, and we end up with compression settings that might be well-suited to your videos. On the other hand, if we were to encode 1080p footage from an ARRI cinema camera with the Slide Deck preset, we would notice radical quality degradation. So, a lot of this is going to be very relative to how good the source footage is.
Basically, the higher the resolution and the lower the quality of the source and the lower the amount of movement between frames, the better chance that Slide Deck has to encode without losing much quality. Your Note 9 lands in that territory quite nicely. I’m not trying to say the Note 9 is bad… I’m just saying that video from it doesn’t need the bitrate that an ARRI would need before loss was noticeable.
Look on the Export tab > Advanced button > Other tab to find the additional options being passed to
libx264, and add those to an FFmpeg command line if you want to use the same settings from a script.
This topic was automatically closed after 90 days. New replies are no longer allowed.