Convert 24pA to 23.976

Several years ago I shot a project using the Panasonic DVX100 and it was shot in 24pA. When capturing the footage I captured in standard in DV mode so all footage is shown as 29.97. I hadn't caught this mistake until the project was finished and going out the door. I now have a little time and would like to re-visit this project and do a proper pulldown. Does anyone have a recommendation on how to take a 29.97 project that has 24pA footage and get the footage to its proper frame rate?

You will have to rebuild your assembly after you convert your footage. You cannot change the timebase of an assembly in fcp after anything has been edited into a sequence, so you'll have to create a new sequence with the correct timebase and then copy and paste the contents of your previous assembly. You'll then have to replace each shot with the material with the pulldown removed. Some edit points will probably have to be adjusted by a frame or so. Timecode should be close to the original so replacing the footage should be too taxing.
As far as removing the pulldown from your camera original, I always get confused between 24p and 24p advanced, but I'd duplicate your footage before doing anything. There are options to remove pulldown in the fcp tools menu, cinematools and compressor.

Similar Messages

Convert 1080p24 to 1080p23.976?

I'm used to using compressor to take a 24p encapsulated inside 60i and give me true 24p.
However, now I'm capturing actual 1080p24 from an Intensity Pro (not 1080p23.976).
What is the best way to convert this to 1080p23.976, so it will match the rest of my source video files?
Hopefully, I can create a compressor droplet to do this? (will it have any adverse affects on the audio?)
That's what I've done with the 24p inside of 60i, but it's been quite a while since I've messed with creating a droplet.
Thanks!!

Michael - I've never used CinemaTools before, just compressor.
If it only changes the metadata, how will that affect using it in FCP with 23.98 files? - and especially when I want to render out to the final output?
One concern is that for the 24p files, I have to chage the ouput settings of the Black Magic hdmi, so that I can view my sequence on my external monitor. It's a bit of a pain, and seems to indicate that there's a bit difference between 23.98 and 24p (which doesn't seem logical?).
I don't have to do that for my 'normal' 23.98 files. Everything just works.
Which is why I'd like to 'convert' them to 23.98.
But, I'll check out CinemaTools. Hopefully, there's a way to create a 'droplet' like with compressor? That's very handy for my 24p in 60i files - just drop them on the droplet, and the magic happens!

Converting a 1080p 23.976 fps project to 720p60 for broadcast

How do you convert a 23.976 progressive frame rate into a 59.94 progressive frame rate? Is there an after effects pulldown option that is not intended for interlaced 29.97 fps, but for progressive HD formats like 720p60?

Hi Jim,
thanks for the info thats great ! the footage is just vision and wont matter if its slower which is good, i was concerned about any NTSC / PAL issues with the 25p stuff - but would i be correct in saying that as these files are raw and from the Epic / Scarlet that they are not either NTSC or PAL until they are outputted that way in post? Unlike say a a Canon 5D where you assign either NTSC or PAL within the camera?
cheers,
Tim

Weird problem when converting project for 23.976 to 25

We are doing this on a batch of episodes (Just changing the play speed no interpolation.) But basically when we do the conversion some times between the 3 and 4th minute it copies the first frame of the project and pastes it over a frame in the output. It's only like one or two frames. Any idea why this is. It's not a big problem it's just a weird ghost in the machine thing.

I am getting the same error !! Can someone help on this issue !!
I have converted/migrated my application from Weblogic Workshop 8.1.4 to 9.2 as per the documentation.
Failure occured in the execution of deployment request with ID '1168574404997' for task '2'.
Error is: 'weblogic.management.DeploymentException: [J2EE:160149]Error while processing library references.
Unresolved application library references, defined in weblogic-application.xml:
[Extension-Name: beehive-controls-1.0, Specification-Version: 1, Implementation-Version: 1.0, exact-match: false],
[Extension-Name: weblogic-controls-1.0, Specification-Version: 1, Implementation-Version: 1.0, exact-match: false],
[Extension-Name: wls-commonslogging-bridge, Specification-Version: 1, Implementation-Version: 1.0, exact-match: false].'

Audition converting video from 23.976 to 29.97fps

When I use the EDIT menu command "Edit in Adobe Audition", to send a sequence from Premiere Pro CS6 to Audition, my 23.976fps video file is being converted by either Premiere or Audition to a 29.97fps file. Perhaps this is how it's supposed to function and has no consequence on the mix or sound file. Just want to be sure this is how the exchange between programs is supposed to work.

He gets the number by dividing 48k by the number of samples he wants per frame, using Windows built-in calculator! Is that the right way to do it? Only if there are clear-cut and defined sample boundary numbers to be observed at frame boundaries. In reality you'd never get those numbers that accurate, because clock jitter would get thoroughly in the way.
I think that what Durin is suggesting is the obvious way to go about this - just recognise that the two apps have to display the same ruler time, and that to stop anybody getting number-blind, they should be limited to just the two decimal places...

True 24fps Film Project out to DVCam Tape - Proper Settings?

I have a TRUE 24fps 1080x1920 Uncompressed Quicktime of my film. I'm going to be using Premiere Pro CS5 connected via Firewire to a Sony HVR-Z1U Camcorder, which is capable of DVCAM recording and playback. I need to record the film to DVCAM tape for a film festival....and I was wondering 2 things:
1. What framerate is needed for DVCam? I'm guessing I will need to convert the film to 23.976 before outputing it to tape....
2. If I need to convert to a 23.976 file beforehand....what is the best quality file for output to DVCam? The best way I know of currently to convert my true 24p footage to 23.976 is by going into After Effects and converting to MPEG2...setting the M-Frame to 3 and the N-Frame to 12 so I avoid conversion flicker problems....but I know MPEG2 will probably be pretty lossy compared to other file formats that DVCam tapes can handle.
Any help would be much appreciated!
Thanks,
Doron Kipper

DV and DVCAM are effectively the same, 720 x 480 NTSC or 720 x 576 PAL and the framerate is 60i or 50i. Assuming NTSC you need to rewrap to 60i using MS DV AVI.

Question about the HDV 1080p30 preset and Encore.

I shot video on a Canon H1 set to HD 30p and recorded it onto DVC tape. I then captured the taped footage off the Canon H1 and into Premiere CS3 using the HDV 1080p30 preset. I believe that the HDV 1080p30 shot from the Canon is 1440x1080 when compressed onto tape.
I want to export to Encore CS3 to encode the video and create a Blu-Ray disc of the 1440x1080p30 video i just edited in premiere. The problem is that Encore CS3 has no 1440x1080p 30 preset. Encore CS3 has 1440x1080p 23.976 and 1440x1080p 24. The closest preset is 1440x1080i 29.97 but thats interlaced. I was hoping to keep the video in progressive mode but you have to convert it to 23.976 or 24 fps.
According to the specs, High-definition video may be stored on Blu-ray ROM discs with up to 1920x1080 pixel resolution at up to 60 frames per second interlaced or 24 frames per second progressive.
What's the point of being able to shoot and edit in p30 just to have the encoder force you to convert back to p23.976, p24, i29.97 or i59.94 fps?
Or is i29.97 the same as p30 when the video is originally shot in 30p?
Is this just an oversite?
Do i just select the 1440x1080i 29.97?

Yes, shooting 30p with 1/60 shutter speed would require more light but would give you less motion blur. But you'd have the same issue even if you shoot in 60i since the default shutter speed for 60i is 1/60.
The canon H1 can record HD in 60i, 30f, and 24f but converts the output to 60i except when the HDV/DV terminal is set to HDV in 24f mode, then the signal is output in 24fps progressive.
So in my case, whether i shoot in 60i or 30p, the playback signal is converted and output at 1440x1080i 60i.
The dilemma is that LCD, DLP and Plasma HDTV's are all progressive. So whatever you feed it, i or p, it gets converted to p for viewing. The problem is feeding an LCD, DLP or Plasma HDTV a true interlaced signal (A frame = AB fields) won't look as good as feeding it progressive signal (A frame = AA fields). Lines and edges appear stepped or jaggy.
So the issue is whether to shoot 1440x1080 in 30p with a shutter speed of 1/60 or 60i with a shutter speed of 1/60. Either way, it gets output via the HDV terminal (IEEE 1394) into premiere as 1440x1080 60i.
With consideration to the final display device, i.e. LCD, DLP and Plasma HDTV's and monitors, the only option would be to record in 30p with shutter speed of 1/60 if your shooting lots of movement like a dance recital to reduce the motion blur and 1/30 for normal shoots. Recording in 60i for output to a progressive HDTV or monitor would, i my opinion, produce somewhat blotchy jaggy video frames because the frames are averaged from two fields with slightly different images.
Also, since the H1 playback signal is output as 1440x1080 60i, does it matter whether i choose HDV 1080i30 (60i) or 1080p30 for a new project preset in Premiere CS3 prior to capturing the footage? If you shot in 30p, probably not, since the Premiere CS3 capture would still result in A frame = AA fields anyway. But since Encore nor the Blu-Ray standard support 1080p 30 (29.97fps), the choice is made for you - 1080i30 (60i) for the Premiere preset and 1080i 29.97 for Encore preset. You could select the HDTV 1080i 29.97 Encore preset if you want to upscale the 1440x1080 to 1920x1080.
Either way, the end result is that shooting in 30p and capturing, editing and encoding in 1080i 29.97 will still deliver A frame = AA fields and a better progressive image on a progressive LCD, DLP or Plasma HDTV or monitor without the conversion or stepped and jaggy lines that a true interlaced image would show.
So, the bottom line, in my opinion, is... If your shooting with a Canon H1 and will be showing you work on an LCD, DLP or Plasma HDTV or monitor, shoot in 30p at either 1/30 or 1/60 shutter speed then edit in Premiere, encode in Encore and produce your Blu-Ray dvd in 1080i 30 (29.97).
Obviously, if you shooting a film or want your work to look like film, shoot in 24p by all means.
Also... Let me know if you disagree on any points.

Converting 23.976 to 29.97i methods and confusion

Hey all, I've had a lot of trouble exporting spots out for broadcast using AME, and thought I'd see if the community here could help me out with some questions.
Some background first. Most of our footage is shot and edited in 23.976, then rendered out to a master file that retains the original framerate. Since most stations only accept 1080i footage, I use AME to convert the master file to either a MPEG2 or Quicktime ProRes 422 video that's interlaced at 29.97.
Everytime I use a preset or manually adjust settings, there's visible tearing/combing of the footage. Granted, I work at a smaller production house without a proper broadcast monitor to view playback on, but is this "normal" for watching video on a computer monitor? It seems like every other frame shows tearing, which makes me think it's more a problem with the deinterlacer than the encoder...
Second, I noticed that if I export an MPEG2 at 23.976 with a "3:2 pulldown" option, it forces AME to render "progressive". Quicktime Player X and 7 report a framerate of 29.97, while MPEG Streamclip and even AME read the file as 23.976. Any ideas what's going on here? Which framerate is it, and is it truly progressive or interlaced with pulldown flags?
Any other tips on getting a clean conversion from 23.976 to 29.97i would really be helpful, as I've been banging my head against a wall for a while trying to find the most efficient method to convert these spots.
Thanks!

is this "normal" for watching video on a computer monitor?
It is.
Whether or not you get flags or genuine interlaced frames depends on the specific format. The MPEG2-DVD option, for instance, uses flags, whereas the DV format will add real pulldown.

Convert frame rate 29.97 to 23.976 with synced audio

Ugh, one of my DSLR cameras was some how accidentally changed from 23.976 to 29.97 fps for an afternoon of shooting. Fortunately, it is a small percentage of total footage, however, it is synced with audio dialogue. How should I proceed to convert this footage to match the rest of the footage?
Thanks

Hi Jim,
Adobe Premiere CC v.7.2.2 more reel. Mercury Playback Engine GPU Acceleration(CUDA) is enabled.
I'm going to try Steven's suggestion of frame blending and hopefully will help. Any other suggestions greatly appreciated. For the life of me I don't know how that camera's settings got changed, ugh!!
Thanks so much!

Converting 23.976 ProRes to PAL 25fps.

I need to convert some 23.98HD ProRes to PAL SD.
I tried using Graeme Nattress' conversion software, but I find that there is a slight stutter where his software creates an extra frame - the 25th frame. He advised me to use Cinema Tools to create a file which runs 4% faster via the "conform" option.
However, I have no idea how to do this and can't find anything helpful in the manual.
Can anyone give me guidance, some sort of workflow or a page in the manual - in short, anything to assist me?
Thanks
Harry.

Thanks very much for the response and for testing.
So, once I have conformed and it is 25fps, then I convert to PAL, right?
But won't the conversion to PAL, either through FCP or through Nattress Conversion add the 25th double frame, or maybe there's an option for fixing that.
many thanks.
Have to go out now, so if you answer my question, I can only reply later tonight.
best wishes and thanks again
Harry

From 24pa DV source material to 25fps pal?

Hello,
I am submitting a new feature film I edited in FCPX to a film festival that is VERY specific on their requirements for uploading the film for consideration (copy from their email below):
For the upload the file must comply with the following specifications:
VIDEO
Codec: avc1 H.264
Frame Size: 1280px X 720px
Bitrate: 821kbps
Frame Rate: 25
AUDIO
Codec: mp4a: MPEG-4 AAC LC
Bitrate: 128kbps
Channels: 2 (stereo)
Sample Rate: 48KHz
I shot the film on a Panasonic DVX100a. Using 24p Advanced Pulldown. So my questions here is: how do I accurately go from 24pa DV material to pal?
I have exported as a single ProRes 444 QuickTime file (23.976 fps) and was expecting to just go into the latest Compressor, setup these requirements, convert, upload and be finished. Unfortunately, not that easy. And the festival is being very strict on their requirements. My biggest problem is going to 25 fps from 23.976.
When I just do a custom export in Compressor and select 25 frames the result is a jittery video on playback.
Is there any way to smooth this out better using Final Cut Pro X and/or Compressor?
THANK YOU!

Russ,
The motion is a LOT better now after your setting suggestions here. The audio is synched well at the beginning of the exported file (it's 1 hr 43 min film runtime) but then as I get further along in the playback it gets more and more out of synch. Any ideas? I "enabled" the Audio under the Encoder settings.
This is the summary of the output settings in Compressor:
File Extension: mov
Estimated size: 609.95 MB
Time remapping:output duration 95.903999%
Audio Encoder
          AAC, Stereo (L R), 48.000 kHz
Video Encoder
          Width and Height: Up to 1280 x 720
                    Selected: 852 x 480
          Pixel aspect ratio: Square
          Crop: None
          Padding: None
          Frame rate: 25
          Frame Controls On:
                    Retiming: (Fast) Nearest Frame
                    Resize Filter: Linear Filter
                    Deinterlace Filter: Fast (Line Averaging)
                    Adaptive Details: On
                    Antialias: 0
                    Detail Level: 0
                    Field Output: Same as Source
          Codec Type: H.264
          Multi-pass: On, frame reorder: On
          Pixel depth: 24
          Spatial quality: 75
          Min. Spatial quality: 25
          Temporal quality: 50
          Min. temporal quality: 25
          Average data rate: 0.82 (Mbps)*
*Unfortunately as well Compressor reporting of the Data/Bitrate here is way off. At least according to QuickTIme playback. Which says the file's datarate is 952.48 kbps. Man, I just wish this stuff would work as it's set. This festival is real strict too. I asked if it's OK if the data rate is a little off and they said the specifications cannot be changed.

Settings to use with 29.97 mixed with 23.976 in same timeline?

Hi everyone,
I just finished a short project which was mostly 24p (23.976) but had a few shots which were 29.97 interlaced. I edited it on a 24p timeline (I dragged the 24p clip to the "new timeline" icon to get started. I then had a few clips which I wanted to include but were 29.97. I followed Andrew Kramer's tutorial on frame-rate conversion and got outstanding results converting the 29.97 to 23.976 which dropped right into my timeline.
Was this an ideal way to work with mixed frame rates (progressing and interlaced, not just different progressive speeds)? I've read that the output should determine the final rendering which in my case will most always be DVD. I've been reading about the 24p Pulldown which looked like it might be easier to just use an interlaced timeline with the pull down on the 24p clips but then I've also read that there are 24p and 24pa clips that are different. Am I making this more complicated than it really is?
Thanks,
-Stephen

Hi Jim, Thanks for replying. It's all HD. This particular project was mostly Panasonic AVCHD 29.97 footage. The few clips I decided to add were from a Canon t2i shot at 24p. Of course if I intend to mix them they should match but in case I run into this again I thought I better ask what the workflow should be.

Media Encoder changes the timecode when converting to Quicktime ProRes 422 (LT)

After spending the majority of my day searching for a solution to this problem, I am hoping someone here will know how to resolve this issue.
After encoding several hundred clips of various file types to ProRes 422 LT at 23.976fps in Media Encoder, we discovered that the timecode of the source clips and the LT clips varies significantly when we bring the quicktime clips into FCP 7 or Quicktime 7. This is also creating issues for the EDL.
The source clips are XDCamHD, H.264, and .R3D files. We set up a batch in AME with a preset for ProRes (LT) at 23.976. After the batch processes, we import the clips into FCP 7 then export an EDL from the sequence. The EDL shows timecodes that do not match the source clips.
The timecode is off by several hours, not just a few frames. If we look at the LT clips in Premiere, the timecode seems to match the source clip. In Final Cut, not even close.
Any ideas on where to start on this one?
If we use RedCineX to convert the R3D files, things seem to work. Also, compressor seems to have correct results. We were hoping that media encoder would be the solution as it processes these clips much faster.
We are looking for a solution that doesn't involve re-transcoding all of the footage.

Two.
First, this.
http://help.adobe.com/en_US/premierepro/cs/using/WSd9957a95a81082eb28f05af5126718c47be-800 0.html
Second, this.
Customize or load keyboard shortcuts
From the Set menu in the Keyboard Customization dialog box, choose the set of keyboard shortcuts you want to use in Premiere Pro:
Adobe Premiere Pro Factory Defaults (Adobe Premiere Pro CS5)
Adobe Premiere Pro CS5.5 (Adobe Premiere Pro CS5.5)
Shortcuts For Avid Xpress DV 3.5 (Adobe Premiere Pro CS5)
Shortcuts For Avid Media Composer 5 (Adobe Premiere Pro CS5.5)
Shortcuts For Final Cut Pro 7.0

23.976 frame rate explained

This post was triggered from this this thread:
http://discussions.apple.com/thread.jspa?messageID=4741388#4741388
I just thought it best to start a new one
Okay iSchwartz, with the speed of a flying bullet, and with help from my fellow members of the IBS (Institute of Broadcast Sound - in the UK), I think I have some answers...
24P is indeed simply video at 24frames per second with the "P" meaning Progressive scan. Its a video format invented to be compatible with film. It also means that the footage will have more of a "film look" when its converted to TV frame rates for broadcast as it will have the same process applied to it as film footage.
As to where 23.976 fps comes in (with regards to Logic) is rather more vague!
(warning - the following contains some speculation!)
As all will know because film is shot at 24fps it needs a special conversion process to be shown on an NTSC TV of 29.97fps. The process could just speed the film up but it would look funny and so they sort of duplicate some of the film's frames instead, to get the frame rate increased. The process used would result in a frame rate of 30fps so to bring it in line with the (cumbersome) NTSC frame rate they have to slow the film down slightly (by 0.1%) before they "scan" in the film's frames to make video.
...and guess what a 0.1% speed reduction of 24fps results in? Yes - 23.976!
so...
23.976 is the frame rate of the telecine machine when playing film to transfer to NTSC tv. The process involves slowing the 24fps film by 0.1% to 23.976 and applying a 2:3 pulldown picture process, resulting in NTSC's frame rate of 29.97fps
Now the question is why would you want Logic to work in 23.976fps?
It is possible (I don't know for sure) that when you shoot with film (ie 24fps), destined for tv that the programme is edited at 23.976, ready for the telecine process, and so this is the frame rate used during the post production process - As it is the time reference used in the edit then we need logic to work at this rate too so we're all in time.
It is also possible that the latest HD video cameras provide a shooting frame rate of 23.976 fps so that the "slow down" process of the past is not required.
Having said that to view the 23.976 fps pictures on telly the video needs to be converted to 30fpsDrop - or you need a display that can work at this unusual frame rate. I am not sure what the implication of all this would be for the musician "working to picture".
One thing I would say, is that musicians are not alone in having much confusion with frame rates and time code, especially now Hi def has arrived. I would strongly advise double checking with "production", particularly the dubbing editor, if that frame rate requested ~really~ is the right one!
I hope I haven't confused the issue even more with all that! Don't take any of this as gospel as I am still learning too. I trust someone will provide corrections and additions.

Rohan,
I agree in so far as the musician should not have to worry about the technical aspects of picture, but the composer is working within a team of people who have to be able to communicate effectively. the composer is also part of the production's overall workflow and so the composer's work methods can be dependent on productions.
To answer you points individually...
- when the telecine is created, and then sent for offline editing, it is the offline edit that you receive with the BITC or the striped LTC (rare these days) that should be your guide. obviously in the states with the confusing NTSC format, it is important to be sure what format they have used in the offline process, but ultimately it is the production facility to ensure you have a video edit that you can work to
Your first sentence says it all. If the offline edit is in 23.976 then presumably this is what you'll get sent. I have no idea how this will get delivered as the format would not work on video tape, you wouldn't be able to view it on a standard TV either. You could view it as a quicktime movie. That is the only way I can think of.
I agree it should be up to production to provide you with what you can work with but how far must they accommodate? I doubt anyone would be too happy providing U-matic tapes these days. For my last job I wanted QT movies on a DVD-R but I had to settle for DV tapes and capture myself. Lets face it, as technology gets cheaper and more accessible we are increasingly expected to be able to handle all those advances!
- so the music you create can either sync to a master clock (which is rare other than word clock which is not for video syncing anyway) or simply to the QT or VHS that you have been provided with which should have standard frame rates. you simply compose your music based on that sync
I don't know what you mean by "master clock" in this context, and as you point out word clock has no timing information at all and is irrelevant in a timecode discussion. Yes you work to the movie you've been given, but what if its using 23.976fps? If it is, it wont be on VHS - and, you wont be able to sync properly to it in Logic as Logic doesn't work with 23.976
If production provides you with a "special copy" in say standard NTSC 29.97fps then the editors timecode will be different to yours. You wont be able to talk timecodes with them over the phone for example. For them to do this production will have to convert the footage specially for you using 3:2pulldown telecine process (so the film plays at the right speed) and I'm not sure they would pay for that!
- the issue is more critical in the dubb, but generally 1 second of music at whatever frame rate is still 1 second of music. when you send your music in it is almost never smpte locked unless you have delivered via a timecoded DAT something which is very rare these days. basically it simply measures the sample rate and then syncs within the DAW in the dub
1 second in NTSC's 29.97fps is 1second at 23.976fps which is 1 second in PAL's 25fps too;-) Its the number of frames within one second that has changed. The problem is the format of the movie. you can play it in QT but not on the telly and Logic doesn't have a time base to match with it. It is likely that logic will be able to follow the movie in QT (not sure of that), or "chase" it, but Logic's own smpte display will not match the movie's BITC. Indeed all of logic's time references will not match the movie's.
when you send your music in it is almost never smpte locked unless you have delivered via a timecoded DAT something which is very rare these days. basically it simply measures the sample rate and then syncs within the DAW in the dub
Ever heard of BWF? Bounce your finished music in logic as a WAV file and you actually get a timecode stamped "Broadcast WAV" file: a file that has the SMPTE time of where the music starts in the film. The timing info is embedded in the file. All the dub mixer has to do is press a key-command to place this file in the right place in the film. I worked this way on my last project. (Its also how logic can place any audio file into "its original recording position")
Without using timecode I'm wondering how you tell "the dub" where the music should go in the film? - I'm genuinely curious. In the past when delivering music without timecode I would still say, in a written note for eg, the track starts at some timecode point (as referenced to the BITC for eg). This is why the composers timecode needs to be the same as the dub/editor's!
- it is simply important for the people sending you their film to agree how many frames consitute their second. if you have a VHS playout from the offline, you should work to whatever format that VHS or DVD comes in; NTSC or PAL. it's not your job to worry about the film to offline telecine
A nice thought in theory:-)
It shouldn't be our job to worry about such things (bars, beats, keys and harmony are enough for me!) but we may not have the choice if we want the job!
Don't get me wrong here Rohan. I don't profess to understanding all this stuff properly and I have never been asked to work in 23.976fps. I went down this line of enquiry to see if I needed to know about it! Perhaps you are right and production will supply a special version for the composer - but I'd like to know if this is the case as I like to have a basic grasp of the overall production process. I'd love to know of anyone who has had to work in this frame rate and how typical it is becoming (if at all).

A/V sync problems with mencoder -- trying to convert to 30fps

The title says it all. When encoding with mencoder (latest from extra repo), with seemingly everything tried, I absolutely cannot get the audio to sync! It's driving me nuts! Here is the basic command command I used:
mencoder -fps 30 -ofps 30 -ovc xvid -oac mp3lame -xvidencopts bitrate=whatever_i_want:trellis:aspect=16/9 -lameopts cbr:br=128 -vf harddup,expand=aspect=16/9,scale=480:272 infile -o outfile
Where the infile was any one of:
* A standard, .avi Xvid 23.976 fps 800kb/s video with MP3 112kb/s audio DVD rip you can find on the net and filesharing sites (here the vbitrate was 800kb/s)
* An .mpeg MPEG1 25fps 1150kb/s video MP2 224kb/s audio (here bitrates were outputted to the same values)
* An .mp4 H264 29.989fps high-bitrate (in the 2000s, was somewhat random) video AAC audio file downloaded directly from Youtube with a DL mp4 from YT Firefox extension
I also tried to mess with the codec settings, like using libavcodec, no trellis as an option, different bitrates etc. but the audio is still out of sync! Although on the mencoder man page it says to put harddup at the end of the filer chain, I've tried that; all it did was cause weird graphical glitch of a band of rectangles containing some video on the right side of the screen, as well as ignoring the expand filter. And it didn't even fix the A/V sync. To be specific, the audio is considerably behind the video e. g. by the time the video ends, audio is still playing from long before the ending credits from an earlier scene, and cuts off as the file ends.
On a side note, framerate conversion with ffmpeg (which, if I understand correctly, mencoder/mplayer is based on), with -r 30 works. But for some reason I find a quality difference (even when using libxvid directly as a -vcodec) and, on top of that, cropping fails with "Vertical crop dimensions are outside the range of the original image." (which they weren't). Argh, I don't think I've every googlesearched this hard only to not get an answer!
Why am I trying to do this? I am converting these videos for playback on my Samsung P2, whose video format requirements are pretty strict (scroll down to video).
And just as a side question, if that if all figured out, I'm trying to make sense of multi-pass encoding. Anyone have some howto to point me to? Any tips on multipass/framerate encoding? And please, don't tell me to RTFM. Mplayer/Mencoder's man page is 300kb uncompressed in itself. I have already tried to made sense of this large document and it has failed to give me an answer (it is a good resource for xvidencopts and the like, but not so much a good guide for mplayer/mencoder.

Before my ATV I used a Rapsody N35 media play with a 400GB disk and converted all my DVD's to Xvid using avi.net on Windows.
I'm now converting some of these movies to ATV (h.264) format using visualhub, it cost £12.95 and does a great job. I tried a few free or open source converters but settled on visualhub because of ease of use and quality of conversion. I also want a Mac OS application.

Convert 24pA to 23.976

Similar Messages

Maybe you are looking for