render AVCHD lite?

[mickeymike53]

I am editing a video using VS X3 and am using both AVCHD 1080i and AVCHD lite 720p clips.

What would be the best format to retain quality when I render as a file and then copy onto a USB stick to play on my Bluray player? My player seems able to read most formats.

[Ken Berry]

I think you would have to try it and see for yourself which you prefer. 1080i has a bigger frame size and (usually) a significantly higher bitrate than 720p, and that should at least equal the quality added by the fully progressive bitrate of the 720p. Of course, if your 720p uses a speed of 50/60p instead of 25/30p, you may have trouble either way since 1080 uses 25/30. If you mix them, then either one frame per second will be removed from the 720p or VS will have to invent a full frame per second for the 1080i. This usually means that the resulting video of those bits will be a bit jerky.

[mickeymike53]

Thanks for your reply Ken.
I think I understand what you are saying but I am new to video editing and some of it seems pretty complex! Just to make clear...The first clip of my video is AVCHD lite 720p 50fps and subseqent clips are AVCHD 1080 25fps. When I save video to file with the 'use properties of first clip' the resulting video seems to view quite well but with some 'ghosting' on the 1080 clips.

If I swap the first clip for a 1080 clip and follow the same process the resulting video looks better with no 'ghosting' but some segments seem to slow down and the audio is out of sync.

I would really appreciate it if you could explain to me why this is happening and if there is any way to get a result with no 'ghosting'. Apologies if I am asking rudimentry questions.

[Ken Berry]

I am afraid that the ghosting, on the one hand, and the slowing down, out of sync audio on the other, are both products of your mixing video which use two different frame speeds. Putting it simply, it is to do with the i and the p. 1080i is i for interleaved. This means that each frame of video is made up of two half frames. One half frame consists of lines 1, 3, 5, 7, 9... 1079 of video, and the other half is made up of lines 2, 4, 6, 8 ... 1080. In video parlance, the first half frame is called the Upper Field, and the second is the Lower Field. With high definition video, either HDV or AVCHD, using 1080i, the Upper Field of each frame is broadcast First, the Lower Field is broadcast next. All this happens so fast the eye cannot see, but this interleaving of the fields effectively makes the eye think it is seeing one solid picture. And in PAL countries, 25 of these fully interleaved frames take one second (or 50 half frames per second).

It also explains what is meant by Upper Field First when describing high definition video, and most standard def video shot on cameras that film in mpeg-2 on mini CDs or hard disks in the camera or SD cards. Mini DV/AVI camcorders, on the other hand, film using Lower Field First, but the end result is still the same. The eye still perceives 25 full frames per second. (For the same of completeness, in NTSC countries this speed is approx. 30 frames per second, though to be absolutely correct, it is 29.97 fps.)

But the p for progressive marks a new step in video quality. Instead of broadcasting in half frames which then have to be interleaved by the eye, the video is filmed as one solid frame. So instead of seeing 50 half frames per second and making them mentally 25 full frames per second, p for progressive video actually films and broadcasts as 50 full frames per second. The eye has to do less work and sees the image as that much more solid and clear. In other words, the quality is higher.

But the problem, as in your case, comes when you mix the two, and then apply one or the other frame rate. If you use a frame rate of 25 fps, then VS will only be able to show 25 fps for the originally 50fps 720p video. To do this, it has to literally discard every second frame. So the eye perceives that something is wrong. The motion becomes minutely but perceptibly jerky, though the background audio continues as it originally was, only covering half the original video. So in other words, the audio gets out of sync.

But if, on the other hand, you apply 720p to the whole project, then for the interleaved 1080i video, VS has to literally invent new frames out of thin air to make up the 50. To do so it does what is easiest, which is just to use the same frame twice. And the eye perceives this as a sort of ghosting -- or to me, jerky -- effect.

I repeat that this explanation is all highly simplistic, but I hope it makes it a bit clearer for you. However, the basic lesson is that for the moment, you will inevitably get one effect or the other if you mix the two formats in one project. So I guess the bottom line is to work out which fills the greater part of a project or is more important to it, and use the frame rate associated with it for the whole project.

[mickeymike53]

Ken thank you for your excellent reply. It made it all a lot clearer to me.

For anyone who is having a similar problem I have found an adequate solution...convert 720p clips to mpg2 with HD converter using it's preset 'mpg2 HD'. This changes the frame rate to 25fps (and it also changes the KBps to 9000 I don't know what this does). I then create a video file in AVCHD 1920 x1080. The resulting video is pretty good, much better than I was getting. So once again thanks for your help.

[Ken Berry]

Despite its name, HD converter is only one way: high def to standard def. So the mpeg-2 you are creating is in fact only standard definition. The 9000 kbps means that, for standard def, it is high quality, but that is still considerably below the bitrate (and thus quality) of HD video. Still, if it works and you're pleased, that is all that matters.

I had originally been going to suggest you convert it to HDV which is high def mpeg-2, but I couldn't think of a converter program which might be able to do it. We normally recommend a program called SUPER, but I am not sure it is capable of this particular conversion. Mind you, I have just done a test conversion using VS X4 itself using video with similar properties to yours, and I can't tell the difference between the HDV and the original, so that could be a possibility: convert your 720p in VS as a separate exercise, then include the new video in your project. At least worth a try...!

[mickeymike53]

Thanks Ken, I have done as you suggested and converted the 720p clips into HDV, re-imported them back into my project that contains 1980x1080 clips and then saved the mixed project to file as AVCHD 1980x1080. The quality of the resulting video is very good overall but occasionaly, not often, the image breaks up a bit. Any ideas?

[aljimenez]

I am afraid that the ghosting, on the one hand, and the slowing down, out of sync audio on the other, are both products of your mixing video which use two different frame speeds. Putting it simply, it is to do with the i and the p. 1080i is i for interleaved. This means that each frame of video is made up of two half frames. One half frame consists of lines 1, 3, 5, 7, 9... 1079 of video, and the other half is made up of lines 2, 4, 6, 8 ... 1080. In video parlance, the first half frame is called the Upper Field, and the second is the Lower Field. With high definition video, either HDV or AVCHD, using 1080i, the Upper Field of each frame is broadcast First, the Lower Field is broadcast next. All this happens so fast the eye cannot see, but this interleaving of the fields effectively makes the eye think it is seeing one solid picture. And in PAL countries, 25 of these fully interleaved frames take one second (or 50 half frames per second).

It also explains what is meant by Upper Field First when describing high definition video, and most standard def video shot on cameras that film in mpeg-2 on mini CDs or hard disks in the camera or SD cards. Mini DV/AVI camcorders, on the other hand, film using Lower Field First, but the end result is still the same. The eye still perceives 25 full frames per second. (For the same of completeness, in NTSC countries this speed is approx. 30 frames per second, though to be absolutely correct, it is 29.97 fps.)

But the p for progressive marks a new step in video quality. Instead of broadcasting in half frames which then have to be interleaved by the eye, the video is filmed as one solid frame. So instead of seeing 50 half frames per second and making them mentally 25 full frames per second, p for progressive video actually films and broadcasts as 50 full frames per second. The eye has to do less work and sees the image as that much more solid and clear. In other words, the quality is higher.

But the problem, as in your case, comes when you mix the two, and then apply one or the other frame rate. If you use a frame rate of 25 fps, then VS will only be able to show 25 fps for the originally 50fps 720p video. To do this, it has to literally discard every second frame. So the eye perceives that something is wrong. The motion becomes minutely but perceptibly jerky, though the background audio continues as it originally was, only covering half the original video. So in other words, the audio gets out of sync.

But if, on the other hand, you apply 720p to the whole project, then for the interleaved 1080i video, VS has to literally invent new frames out of thin air to make up the 50. To do so it does what is easiest, which is just to use the same frame twice. And the eye perceives this as a sort of ghosting -- or to me, jerky -- effect.

I repeat that this explanation is all highly simplistic, but I hope it makes it a bit clearer for you. However, the basic lesson is that for the moment, you will inevitably get one effect or the other if you mix the two formats in one project. So I guess the bottom line is to work out which fills the greater part of a project or is more important to it, and use the frame rate associated with it for the whole project.

You should make this excellent explanation Ken some sticky post with a good title... Al