causes of choppy / stuttering video

[cheyrn]

What are some causes of time related video playback issues? Specifically.

How do I create videos that do not playback poorly for people other than myself?

I have learned some things about video but still operate on superstition. I am currently manipulating video from fraps with avisynth after effects premeiere etc. I then play the video and try to guess if I have caused a problem or if it is my computer or something else. How can I approach the problem intelligently rather than randomly trying things? How, more specifically than a suggestion to read about video?

Get faster hardware or use this magic codec or magic computer program that solves all of your problems, leave me still not knowing what I am doing and doesn't help me with how to create a video that will playback correctly for a target audience.

What does bitrate viewer tell me about the quality of a video?

"With the graphical output you can easily view the quality of your video files before you're outputting it to a DVD and so save a lot of money by avoiding the production of any bad stuff. "

How? What is bad stuff? What is quality?

Also, do you know where I can get a sample video that is relatively pristine so that I can try screwing it up to cause playback problems? I was going to try using this:

http://trailers.divx.com/hevc/Sintel_4k_27qp_24fps_1aud_9subs.mkv

Except that I am guessing that that is lower quality than a source video that someone would edit.

[redwudz]

I mostly look at a problem video with MediaInfo, tree view. That gives fairly complete information about the file.

And what is the output format that's giving problems? If Blu-ray, then look to 'WHAT IS' Blu-ray to the upper left
on this page and you should find some common specifications for the format.

A bit obvious, but too high a bitrate is a common cause for your problems. And playback also depends on what
others are using to play back your videos. If a set top player, then you should adhere to the common specifications
for that format.

[cheyrn]

Ah, I had not looked at the glossary. That looks very useful. My current project is targetting youtube, possibly vimeo or that japanese place.

Here is advice about video for youtube:

https://support.google.com/youtube/answer/1722171?hl=en

I am confused by the bitrate part though. It says if you have an enterprise class connection you would optimally use ~50Mbps video bitrate for 1080p... Otherwise 8kbps. I care about the playback of the video more than the speed of my own internet connection as someone uploading a video.

The above makes me think that this is not a target bitrate regarding quality or limitations of streaming since it is phrased in terms of the creators internet connection.

I am confused about a lot more than just bitrate. Higher bitrate = greater performance requirement is important information. I would want to reduce the performance requirement, as one factor.

But, I have no problem with the concept that good is better than bad, or that bad is worse than good. I have no clue about what is practical.

One of my source videos for my current project is created by Fraps. Mediainfo says it "has a" video bitrate (I think this means it typically requires a playback bitrate of the specified amount) of 378 Mbps. It plays smoothly for me on my extremely slow laptop computer.

An example test video that I made with avisynth, attached to the content affects playback thread "has a" bitrate of 1.333 Kbps. My various video players stutter and play it back at various speeds each time.

So, I conclude from that that bitrate by itself is not that significant. Or, at least if a 28000% difference in video bitrate has the inverse effect it is probably not that significant.

Generally, the question in this thread is how to create any number of videos from any sort of source video such that it is of an appropriate quality for each of any conceivable application. In other words, a methodology, not a one off solution to one problem one time.

[manono]

It says if you have an enterprise class connection you would optimally use ~50Mbps video bitrate for 1080p... Otherwise 8kbps. I care about the playback of the video more than the speed of my own internet connection as someone uploading a video.

Then ignore what you read as YouTube will reencode both using the same bitrate, the bitrate used depending only on the resolution (1080p in this case). The only possible advantage of uploading a video with a super-high bitrate is that the eventual YouTube playback might be of marginally higher quality.

So, I conclude from that that bitrate by itself is not that significant.

It's not (the source bitrates anyway), since all the various YouTube resolutions (1080p, 720p, 480p, etc.) are reencoded using the same bitrates for the particular resolution.

How do I create videos that do not playback poorly for people other than myself?

Just tell them to play the lower resolution versions, maybe 480p.

Generally, the question in this thread is how to create any number of videos from any sort of source video such that it is of an appropriate quality for each of any conceivable application. In other words, a methodology, not a one off solution to one problem one time.

YouTube takes care of that for you by making the same video available in a variety of resolutions and bitrates. It's your job to create the best quality videos you can.

[cheyrn]

So, I conclude from that that bitrate by itself is not that significant.

It's not (the source bitrates anyway), since all the various YouTube resolutions (1080p, 720p, 480p, etc.) are reencoded using the same bitrates for the particular resolution.

What I mean here is that, in no way related to youtube, a bitrate for one video is reported to be 370Mbps. It plays smoothly on my computer. A bitrate for another video is reported to be ~1kbmps. It does not play smoothly, but in fits and starts.

So, not related to youtube, bitrate in some situation where the difference is 28000% has the opposite effect on playback that I thought bitrate is supposed to have.

The point being that there are probably other factors.

It's your job to create the best quality videos you can.

This is the topic of the thread. How do I do that? What factors, excluding extraneous details like type of hardware and what this or that service provider does, are common to all or most videos that determine the quality of playback. In particular not dropping frames or stuttering etc.

[manono]

What I mean here is that, in no way related to youtube.

Okay, but your original question was, "How do I create videos that do not playback poorly for people other than myself?" I took that to mean they play fine for you, both the original video and the YouTube encoded version. Now I see I misunderstood.

If some of your videos at high bitrates play fine and others at much lower bitrates play 'jerkily' I'd chalk it up to the way you created or filtered or reencoded the videos. Maybe if you made available an example of one that doesn't play smoothly we could have a look. 10 seconds or so of a section with steady movement is all that's needed.

This is the topic of the thread. How do I do that? What factors, excluding extraneous details like type of hardware and what this or that service provider does, are common to all or most videos that determine the quality of playback.

It's YouTube we're talking about here. They pretty much ruin anything the least bit complex with their low bitrate encoding practices. If you upload high-action videos, it's often worthwhile to use 'cleaning' filters to make them more easily compressible. If you're not talking about that kind of thing but solely about jerky playing videos then, again, show us an example.

[cheyrn]

I think no one has the information formulated as a methodology in their minds. But, imagine instead of video making, it were any other type of creative process. If you wanted to build a house, would you proceed like:

a. A house has a door. How do I make a door. Oh, good I have a door.
b. Wait, doors aren't usually laying on the ground. How do I make the door stand upright? Oh, I need a door frame. Oh good, I have a door in a door frame.
c. Wait, I am not usually standing on dirt when I enter the front door of a house. Houses have floors. How do I make a floor. Oh good, I have a floor.
d. Wait. Why do I feel wind coming from my left? Houses usually have something blocking the wind there. How do I make a barrier. Yay I have solved that problem.
e. Wait. Why is there rain falling on my head. Houses usually have something over you that blocks the rain. How do I make something that blocks the rain. At last, I don't have rain falling on my head.
f. Wait. Why... etc. etc.

Instead, you would try to learn about something called "carpentry". A general process for making any sort of house.

If you follow the analogy. I want to make any sort of video in any situation that does not stutter or have time related issues. What are the factors that apply to any video that result in stuttering and time related issues? What is a general method that applies to any video that results in videos that do not stutter or have time related issues.

[manono]

Again, please show us an example of something you say stutters and maybe we can diagnose the problem.

[cheyrn]

It's not the topic of the thread. But, for an example of choppy video, see the strobe (with audio).avi attachment in:

https://forum.videohelp.com/threads/368484-content-affecting-playback-rate

What is causing the video to be out of sync with the audio?

[poisondeathray]

In that specific example, it appears to be related to xvid. All your h.264/mp4 videos played "normally" . All your xvid videos didn't, at least according to you on your 2 computers. And it was mentioned, but compressed audio always has a delay. The problem is you were changing 10 different things at a time. Scientific testing is more logical and methodical. You change 1 thing at time, examine the results

For this topic, it's too large. It's like asking a question , why doesn't my car start? Well there are 100's of possible reasons. You want someone to write a book about AV sync to answer you, but you don't seem to want to read that book

[cheyrn]

I think it's different because "why doesn't my car start" is asking about an instance of something. The question, what are some reasons for which cars do not start is a common sort of question. Having no battery would be an answer to the question. It is some (one) reasons for which cars do not start. The implication of a general question like that is that it seeks to list categories of problems that related to cars not starting.

Even if a mechanic doesn't have a formula in mind, she does have such a formula, she just doesn't think of it that way. A mechanic is competent if they know categories of problems that could cause any car not to start. As a result, they can be relied on to fix cars that they have never seen before.

Most likely, the mechanic learned the formula unintentionally by correlating random experiences related to cars not starting. This is the slowest possible way to learn the formula. By definition an intelligent approach involves reasoning about abstractions of a problem, rather than simply reacting to events and retreating from wrong decisions in order to accidentally have an approach that works.

The flashing video example demonstrates this sort of thing too. Solving problems of this sort requires that the solutions are not all dependent on how the video was made. It doesn't work to start with a video without problems and then 1 step at a time reach an unknown condition that is the same condition that the strobe video demonstrates.

[poisondeathray]

What are some specific engine related car starting issues? Is that better ? It's still too generic a question

The "formula" is the same thing a mechanic would do, you go step by step , from source to playback. That' s how you would debug it. Everything in between can be a potential problem. To narrow down the problem , you look at some specific symptoms and details.

In your limited testing: All your xvid videos had problems. None of your h264/mp4 videos had problems. Maybe you can use some logical reasoning and figure out what you should be using for your specific setup ?

[cheyrn]

"What are some specific engine related car starting issues" is too specific a question to apply. Because the question is not about one instance. This is a flawed exchange:

Q: What are some signs that someone is having a heart attack?
A: Bob ate a steak and was then taken to the hospital after he crashed his car into a wall.

A person would be poorly prepared to treat people with heart attacks if they only knew a random collection of instances in which someone had a heart attack.

It is not true that all xvid encoded videos have problems and no h264 videos do. I do not have one problem once. It is not true that all of my xvid encoded videos have problems and none of my h264 videos do.

If I had a specific question and you had an answer, it would only be because you know a general solution to a class of problems.

[poisondeathray]

In your limited testing, did you not report that your xvid video had problems ? Not one played correctly on your 2 computers. Of course you cannot apply that finding to everything in general, because it's limited testing . It's probably not correct to extrapolate that finding to other computers for example

I don't care if the analogy is flawed. Your question is problematic and too general. That's the bottom line. That's why nobody is answering you. If you cannot understand this, I can't explain it any simpler .

[cheyrn]

I can offer some answers to the question that is the topic of the thread. As would be true of all answers to the question, the answers apply only within a context:

Using xvid encoding is associated with choppy videos.

Using ut-video encoding is associated with choppy videos.

Using h-264 video encoding is associated with choppy videos.

Editing xvid encoded video is associated with choppy videos.

Videos having vastly higher average bitrate than a smoothly rendering video are associated with choppy videos.

The last answers seem to me to be an interesting area to explore since it appears to be the opposite of what is commonly said about bitrate, which might mean that bitrate is not so significant that a 300% difference overwhelms other factors.

All of the answers are too specific to a circumstance to be of much use for the average reader of this forum, I am guessing.

[poisondeathray]

The last answers seem to me to be an interesting area to explore since it appears to be the opposite of what is commonly said about bitrate, which might mean that bitrate is not so significant that a 300% difference overwhelms other factors.

If you re-read that thread , that is exactly what I said about bitrate. It's a non factor . It's low on the list of CPU consumption. And CPU consumption for decoding is only 1 of the tiny categories to look at when discussing video "stutter" or AV sync in general . Each category has about a dozen sub categories. You seem to want some neat clean checklist or worflow chart. That's now how it works with the auto mechanic or with the doctor diagnosing an MI