Nero Vision (edit:) Motion Adaptive Frame Rate Conversion

[sagat93]

Can anyone just tell me a little more about motion adaptive frame rate conversion, or frame rate up as i've read. I can't seem to find much information about it. Should it be used with progressive encoding?



(Original In Nero Vision 6 (Nero 9) under make a dvd/more/video options/encoder quality fine tuning, there is a slider next to the word speed. It ranges from one to one hundred. On the right of that it shows the arrow box with the word quality above it. Now, I would guess that one hundred would equal the best quality, but I remember when encoding that the lesser the quantizer the better the quality. I just wanted to clarify that one hundred would be the best setting for quality. The only reason I ask is because to the left of the slider is the word speed, and I just wanted to make sure before burning. Also if anyone would like to make any comments on any of the other settings because I only have basic knowledge of them.

untitled.bmp

Edit: I changed aspect ratio from Automatic to 16:9

[sagat93]

(bump) Close ended response, quick answer? Please.

[guns1inger]

This is the problem with toy software. It uses non-standard terms and methods of describing functions. This is like using qualitative descriptions instead of meaningful numbers.

In this case, treat it like a qualitative percentage. The higher the better. It has nothing to do with quantiser encoding, as you are doing a 2-pass VBR encode. Behind the scenes it is probably doing things like choosing a different encoding matrix, perhaps sharpening the image (which can give the impression of a higher quality image, even if it isn't). As you go to the left, you will get faster encoding, but potentially lower quality as the encoder cuts corners to get through the job faster.

Again, in trying to simplify the controls and settings, they also hide what is actually going on, leaving you no option but to simply trust them and hope for the best.

Where does this leave you ?

I would select a small section of the video you are working on. A minute or two will be enough. Then run a series of test encodes at various quality settings, from 50 - 100 % in 10% increments. From this you will get a feel for what quality differences (if any) it produces, and also what impact on the encoding speed this setting has.

Disclaimer : While I was once a Nero burning Rom user back in the day, I have long since left it for far more capable tools that require far less resources, give far greater control over the result, and produce far better quality

[sagat93]

I read somewhere in another thread something with the registry that even though you change the settings it still is registered to encode a certain way. Speed doesn't matter to me what matters to me is the quality. Anyhow thank you for the elaborate answer.

[guns1inger]

If quality is important to you, Nero is the right tool.

Personally, for a simple start, I would recommend AVStoDVD. Use it to create elementary streams encoded with HCEnc, then author in a decent authoring tool. Burn with Imgburn. You don't need Nero.

That is, if quality is most important.

[sagat93]

Ha, yeah it's taking me 13 hours to burn, but thats ok, I just let it go overnight. One more thing that I don't really understand (I don't want to have to make a post a new thread), that is motion adaptive frame rate conversion. Should I use it? Do I need it? Also, does it have to do with de-interlacing, because I am encoding with progressive, and I don't need it. Could it contribute to the quality of the video or is it something that doesn't need to be there that could potentially worsen the quality.

[sagat93]

Can anyone just tell me a little more about motion adaptive frame rate conversion, or frame rate up as i've read. I can't seem to find much information about it. Should it be used with progressive encoding?

[sagat93]

bump

[Afronautz]

I landed here in search of the same information regarding motion-adaptive frame rate conversion, and agree; there's very little information out here regarding it. So I can only go on what I've determined so far from trial & error. And it seems to me that the motion-adaptive frame rate conversion only ADDS additional frames between frames to make the motion appear to be more fluid.

But the problem I've noticed with it is; if there's too much action, or "motion", in the video (like for example: a person dancing) the motion-adaptive frame rate conversion seems to make it look like "dancing" as seen through the eyes of someone that's just taken a hit of LSD, lol. With more subtle movements, the motion-adaptive frame rate conversion seems to be a nice touch, but I don't think it's worth it really. Seems to alter the original video too much, and what you wind up with seems to be an "interpretation" of the footage, rather than a crisp reproduction of the footage. And it seems to double the processing time. If you're using progressive encoding (like I am), the motion-adaptive frame rate conversion may have it's place, but mostly just seems like an extra "effect" that is largely unnecessary (IMHO).

[edDV]

Can anyone just tell me a little more about motion adaptive frame rate conversion, ...?

What does Nero Help say it is? Make no sense to me.

"Motion adaptive" means the process is reacting to motion in the video. Changing frame rate as a function of 1x motion seems strange.


Google offers fragments of definition
"A frame interpolation algorithm for frame rate up-conversion of progressive image sequences is proposed. The algorithm is based on simple motion compensation and linear interpolation. A motion vector is searched for each pixel in the interpolated image and the resulting motion field is median filtered to remove inconsistent vectors. Averaging along the motion trajectory is used to produce the interpolated pixel values."

In other words this guy wants to slow the motion and interpolate intermediate frames based on object vector motion paths. Or it can be applied to TV set upsampling 24p to higher frame rates.

I doubt Nero is doing a good job of this.

Here are other uses of the term.
http://www.zoran.com/Zoran-Corporation-Adds-Frame-Rate
http://www.kdenlive.org/forum/motion-adaptive-frame-rate-conversion-would-be-killer
http://www.macworld.com/article/49306/2006/02/marchcreate.html

[Afronautz]

Can anyone just tell me a little more about motion adaptive frame rate conversion, ...?

What does Nero Help say it is? Make no sense to me.

The Nero Help files basically just identifies motion adaption as an option, but really doesn't bother explaining how or why it should be used. It's as if they expect you to figure it out on your own, lol. Though while I appreciate their apparent confidence in my ability, I'd much more appreciate a thorough tutorial that goes into detail. The tutorials that Nero offers on their website are very basic, and don't go into much detail, and the manuel's aren't much better.

Nero Recode Manuel states:

Prefer motion adaptive deinterlacing method:

• If this check box is enabled, Nero Recode selects the motion-adaptive deinterlacing method as the preferred encoding method.
• Even quick motion sequences are extremely sharp and detailed. However, burning is slower.

Nero Vision Manuel states:

Motion-adaptive Deinterlacing:

• If the check box is activated, then motion adaptive conversion is used in order to improve the picture quality of the videos and slide shows.
• This check box is deactivated by default.

Motion-adaptive Frame Rate Conversion:

• If the check box is activated, then motion adaptive conversion is used in order to improve the picture quality of the videos and slide shows.
• This check box is deactivated by default.

Based on the brief research that I've done based on feedback from here and what I've found elsewhere online, my feeling is that the motion-adaptive frame rate conversion is best used sparingly and/or for artistic effect, and doesn't do much to improve the overall quality, unless perhaps you're working with video that was in fairly poor condition to begin with. Though I should mention; I'm a complete novice at all of this.

Nero 9 Manuels:
http://www.nero.com/enu/support-nero9-manuals.html

Nero 9 Tutorials:
http://www.nero.com/enu/support-nero9-tutorials.html

[jagabo]

When an encoder has a quality settings (typically 0 to 100) for two (or more) pass encoding, it usually controls how wide a search is made for motion vectors. The wider it searches the more likely it is to find motion motion vectors and the better the compression will be. But very wide searches require lots more time to complete. And for most videos going above a medium settings makes very little difference in the final output. You'll have to experiment with the setting in Nero to see if it really makes any difference in your final output.

If Nero's Motion Adaptive Frame Rate Conversion really is synthesizing in-between frames with motion interpolation you need to watch out. Those algorithms work for some types of material but not for others. See the "bad60.avi" sample in the following post:

https://forum.videohelp.com/threads/307004-Best-framerate-conversion-%28eg-23-97-to-30-...=1#post1888926



All those grid lines should have been perfectly straight.

[edDV]

Motion adaptive deinterlace means pixels (or pixel blocks) are assigned motion vectors. Motion vectors are used to group pixels into objects and then the object pixels are moved (without field to field blur) to the calculated deinterlaced frame location. The human eye detects object motion but with reduced resolved detail, so artifacts that are obvious in single frame analysis are less noticeable when the object is in motion.

Motion adaptive frame rate conversion applies the same principles to synthesizing "in-between frames" for slo-motion or frame rate up-conversion (e.g. film to video rates). 100/120 Hz TV sets (actually 96/119.88) use this among other techniques to synthesize intermediate frames. Digital TV sets do this with real time hardware operating on a frame to a few frames of memory. Software implementations proceed very slowly with today's CPU performance.