I have a Panasonic camera that records to AVCHD 1920x1080 30p, but it's in a 60i container. The files are MTS files. Is there a way to change it to 30p instead of 60i? And maybe even change it to an MP4 extension instead? The video is h264 and the audio is ac3 (which probably isn't supported by mp4, but I'm not sure). If that's the case how about to an mkv?
If this has been asked before, sorry, I couldn't find it.
Try StreamFab Downloader and download from Netflix, Amazon, Youtube! Or Try DVDFab and copy Blu-rays! or rip iTunes movies!
+ Reply to Thread
Results 1 to 30 of 39
Thread
-
-
Last edited by edDV; 27th Sep 2011 at 23:13.
Recommends: Kiva.org - Loans that change lives.
http://www.kiva.org/about -
Well mainly for better compatibility, more things seem to play MP4 files compared to m2ts/mts files... at least for playing on the computer (and my Xbox 360 and PS3). I also don't really see an advantage of having a 30p video playing in a 60i container, why not just make it what it is?
-
MTS/M2TS and MP4 are just containers, not "30p" or "60i" containers. The encapsulated video stream(s) have within themselves the info on framerate, which the container just mirrors.
PS3 actually has BETTER support for MTS/M2TS than for MP4 (at least in terms of # of options). PC is even, Xbox360 is less (but that's not surprising M$!!).
Let's cut out all the in-between crap and get to the heart of the matter...
You want to take your Panasonic camera footage and play it as good as possible on your TV, right?
So tell us the model of camera (so we actually know what we're dealing with), and of your TV. We already know you've got a PC, PS3 and Xbox360. Any other links in the media chain?
If that's the case, we'll be able to help you much better than suggestions about a possibly unnecessary side trip.
Scott -
And there's nothing really wrong with progressive 30fps having been recorded/encoded as interlaced 30fps. It happens with NTSC DVDs all the time.
-
There's a way with MPEG-2 HDV to re-flag 25p-in-50i as native 25p - I remember reading about it on HV20.com but can't find it now.
It's not that useful though - there's not much software or hardware that bases its deinterlacing decisions on the encoding flags in the video stream itself. It would be better for it to be correct, but it's not 100% essential.
Plenty of options for losslessly re-muxing the content into whatever container you want - probably best to answer the questions two posts up.
Cheers,
David. -
-
The camera I have is a Panasonic ZS10, just a regular point and shoot camera (not a camcorder). My TV is a 65" 1080p Olivia TV though I run everything to my Denon AVR-3311ci receiver and just have one HDMI going to my TV (my receiver's basically my HDMI switch).
Oh, and I also thought it would be nice to just put the video's with my photo's too, and not have to mess with making AVCHD disc's, burning to Blu-Ray, making disc menu's for multiple video's, etc (the video's are just normally a few seconds long anyway). I just want simple MP4 files with my photos (if it's possible).
And yeah, that's basically what I was wanting, as 2Bdecided said, a way to maybe 're-flag' it to 30p. I'm guessing they made it 60i for AVCHD compatibility reasons, which I don't really care for anyway, so why not make it 30p?Last edited by bobthegoat2001; 28th Sep 2011 at 10:04.
-
the only way is to field match and re-encode if you want a "file" thats 30p. unless you editing software can't handle 60i or you need to play it back at 30p without software or hardware de-interlacing then there is no reason to mess with it. If you are trying to recompress the recorded stream then I can see your dilemma. If later is your situation then hit me and ill give you my avisynth script I use with DGindexNV to make a VFR progressive stream with timecodes suitable for playback.
if all else fails read the manual -
without re-encoding it can't be done. 30p in a 60i stream has 2 fields per frame each containing 1/2 the picture. the 2 fields are interlaced back into 1 frame at playback time. the only difference between real 60i and (30p in 60i) is the time at which the fields are recorded. true 60i the fields are recorded sequentially and (30p in 60i) they both are recorded at the same time.
--
"a lot of people are better dead" - prisoner KSC2-303 -
ayisynth yes, DGindexNV if you want GPU decoding sees how's it AVC., but dgindex will also do it, just CPU rendered and works on non nvidia hardware. These are the only two I trust on interlaced AVC as ffdshow has some quirks once in a while. anyhow here is the example script for a 32bit system via DGAVCindex:
LoadPlugin("C:\Users\ChuckD\Desktop\MeGUI_2028_x86 \tools\dgavcindex\DGAVCDecode.dll")
LoadPlugin("C:\Users\ChuckD\Desktop\MeGUI_2028_x86 \tools\avisynth_plugins\TIVTC.dll")
LoadPlugin("C:\Users\ChuckD\Desktop\MeGUI_2028_x86 \tools\avisynth_plugins\FluxSmooth.dll")
AVCSource("C:\Users\ChuckD\Desktop\TODO\$1\00000.d ga")
#trim (0,1000)
#crop( 0, 0, -0, -8)
TFM(mode=1, output="C:\Users\ChuckD\Desktop\TODO\$1\tfm.txt")
TDecimate(mode=3, hybrid=2, vfrdec=1, dupthresh=0.3, se=true, mkvout="C:\Users\ChuckD\Desktop\TODO\$1\timecodes. txt")
#optional smoothing remove pound sign to enable
#FluxSmoothST(7,7)
#Spline64Resize(1904,1072)
assumeFPS(24)
dont let the fps of 24 worry you, it is there to make x264 properly calculate the quants in crf encoding so when slowed down the frames look good still
now feed this avs script into x264 with something like this (my settings, adjust accordingly....)
"C:\Users\ChuckD\Desktop\MeGUI_2028_x86\tools\x264 \x264_64.exe" --tune film --level 4.1 --crf 15 --keyint 300 --min-keyint 24 --bframes 16 --ref 4 --direct auto --subme 2 --trellis 0 --output "C:\Users\ChuckD\Desktop\TODO\Sahara\pass-film.mkv" "C:\Users\ChuckD\Desktop\TODO\Sahara\test22.av s"
this will leave you with a mkv file and a timecodes.txt file
drop the video into mkvtoolnix and under the timecodes tab load the timecodes.txt file. drop your audiotrack in next then hit mux.
if all goes well you should have a mkv file that has audio in sync and a video framerate beetween 24-30fps (depends if it finds a deltecine pattern on parts. it will leave the fast parts at 30fps and the parts that can run slower at 24fps)
If you go the nvidia route, just change the following two lines in your avs file
LoadPlugin("C:\Users\ChuckD\Desktop\MeGUI_2028_x86 \tools\dgavcindex\DGAVCDecode.dll")
to
LoadPlugin("C:\Users\ChuckD\Desktop\MeGUI_2028_x86 \tools\dgindexnv\DGDecodeNV.dll")
and
AVCSource("C:\Users\ChuckD\Desktop\TODO\$1\00000.d ga")
to
DGSource("C:\Users\ChuckD\Desktop\TODO\$1\00000.dg i",fieldop=0)
if you want the 64-bit version (about 10-30% in my non technical testing.....) lemme know.
i know this is alot so feel free to bug me, and of course critique is welcome as i'm not to good at giving the "easy" way to do things.....if all else fails read the manual -
Thanks for the info, I have class tonight so I'll probably need to look into it more thoroughly tomorrow.
Oh, and yeah, I have a 64-bit quad core computer, so the 64-bit version may be good too.
Thanks again dannyboy! -
going the 64bit route is going to involve the nvidia toolchain as the other 64bit option (ffms2) wont handle the interlace properly and will bob the output and make it 60p (if i read the changelog right)... which is not good. also the nvidia way will run you $15 (trust me it's worth it, neuron2 has been making awesome tools since the dawn of time it self....) and the card has to have VP2 or higher (google you card to find out) so if you can't meet the said requirements it safer to stay in 32-bit land
if all else fails read the manual -
That stinkin' Panasonic is still hosing people with a cheapo 30fps sensor. They hosed ME too on my FZ-35.
In the FZ-35 specs, in says AVCHD 720p "HQ" video is 60p, and it IS, TECHNICALLY! But it just repeats each frame, so instead of a frame being 1/60 of a second it's still 1/30, so there's no slomo upside, which is the purpose of shooting 60fps, as well as capturing fast motion without blur, which it can't do either.
This is a straight CHEAT.
I was thinking your 30p as 60i has a better chance at smooth slomo because you can interpolate the fields, but that's not the case here either. It's the same hose in a different way. Instead of repeating, each 1/30 frame is simply halved.
In your case, what you have is an old school NTSC broadcast format designed for tube TV's. Ain't that a hoot? Maybe they did that to be compatible with 3rd world country TV's?Last edited by budwzr; 28th Sep 2011 at 15:41.
-
I have Nvidia's Geforce 260 which Google tells me is VP2 compatible. $15 sounds reasonable, it's kind of nice to be able to do stuff while it's encoding (even though I have a quad core, a 5th is welcome).
-
I'm surprised they can get away with this. It'd be interesting to know what Trading Standards (or whatever the US equivalent is) have to say about it.
I noticed an 'anomaly' when researching for a new camcorder. Several sites (including Canon's) listed the number of pixels of the HFS21's flip out LCD screen and viewfinder. After getting the camcorder, I wasn't convinced that the resolution was as stated so I took some high res stills of the flip out screen and counted the pixels.
The only way the numbers added up was if each of the red, green and blue sub pixels were counted individually... -
They're doing government math too. That's the thing, if you try to make a stink they'll point to some fine print way in the back of the manual disclosing what "method" they used to reach their numbers. But it's definitely NEVER the method normal people use or understand.
Like banks use some kind of "Rule of Elevenths" or some weird name like that to calculate the interest you actually receive, versus if you take the face value of the stated rate and a calculator and do it normally. -
@bobthegoat2001
That fits the bill, you system is gonna be pegged the gpu will decode while the cpu re-sizes, deinterlaces, and encodes... just faster as they dont have to decode the video first. im sure you'll read somewhere down the road that the gpu can deinterlace, crop and resize for you instead of the CPU . gpu crop is good, but DO NOT do gpu resizing or deinterlacing. some like the look and convenience of letting the gpu do it, but not many..........
@budwzr: thanks for the info. sees how it was a 30p stream fieldsplit to 60i that means there is no artifacts from taking to seperate filed shots at half resolution. TFM should be able to restore the original 30p without any artifacts if this is the case
edit: don't be suprised if the end framerate is higher than 30p, if the interlace is not perfect then it will leave a 'duplicate' frame in here in there resulting in the framrate possiblly being anywhere from 24p-48p from pas experience, bit it will be deinterlaced and should be smoothLast edited by dannyboy48888; 28th Sep 2011 at 16:01. Reason: disclaimer
if all else fails read the manual -
You don't need TFM, it's already field matched. Each consecutive field pair belongs to the same frame. It's progressive 30p content encoded as interlaced as mentioned above.
This is from a hardware encoder from a consumer camcorder ie. CFR 30p stream - there is no reason to make it VFR. It's not as if it's from some hybrid animation sequence. No hardware supports VFR playback anyway. He mentioned PS3, xbox - these will not playback VFR streams
Also, when you use avisynth you need to re-encode because it frameserves uncompressed data
The PS3 will play the footage as is (xbox probably won't because it sucks)
And if you are editing the footage, all decent NLE's can interpret the footage as 30p. -
thanks for the input poisondeathray . I was assuming possible bad interlace and computer for playback. Can you give him a example script to use if the target is a xbox? You've helped me before so i'm sure you can make something less messy than mine.
if all else fails read the manual -
It doesn't need to be processed at all, unless for the xbox (but that's because for other reasons - it won't play high bitrate AVC back, it's too "weak")
If he wants to playback on PC, it will playback as 30p - as it should - because that's what it is.
PS3 will playback as is, but he might want to author it e.g. multiavchd or tsmuxer
None of those require re-encoding or quality loss (original streams)
For xbox360 he has to re-encode it to make it compatible. Probably with WMV is the best bet, maybe with expression encoder, encode360, encodehd, I'm sure there are other options
If the wanted to process it with avisynth, all he needs is a 1 line source filter. The stream will be "seen" as 30p CFR (as it should be). No need to field match or decimate. -
Recommends: Kiva.org - Loans that change lives.
http://www.kiva.org/about -
Guys, sorry, but you're wrong. It's really rare for MPEG encoding to store all separate fields. Even true interlaced video is often encoded as full frames, because if nothing moves, that's the most efficient way to do it.
25p-in-50i and 30p-in-60i are (almost) always encoded as full frames - so the only difference from a "native" progressive encode is the flag itself, usually the quantisation order, and the implied chroma sub-sampling.
With MPEG-2, it's possible to just change the flag, and get a real native progressive file without re-encoding.
Here's a thread about it:
http://www.hv20.com/showthread.php?32552-HV30-30p-workflow-Vegas-smart-render
Here's the tool:
https://www.videohelp.com/tools/Restream
Sadly this isn't going to work with AVCHD. But it's likely (not certain) that something similar is possible with AVCHD - but maybe no one has written it yet? Or it's hiding there in an existing tool and no-one has tried it yet?
Note that just because you can change the flag doesn't mean the process is successful. These MPEG flag editors let you change flags which cause the resulting stream to be garbage (e.g. change the resolution flag without changing the pixels that are encoded!). But if what you want to do produces a file that plays on everything you need it to, it's good enough IMO. What I use it for most is to change the aspect ratio flag after (incorrect) encoding, which of course produces a completely standard file.
Cheers,
David. -
You can do that and fool a program into thinking the video is progressive. But the chroma channels will be messed up. When encoded interlaced the chroma channels are interlaced as well as the luma channel. If you change the interlace flag to say the video is progressive the decoder will handle the interlaced luma channels as if they are progressive, leading to more blurring of the colors.
-
I have a question. Is there ANY upside WHATSOEVER to having the frames split like this? AFA creating a 60p file?
Why would Panny do this then? If not as a cheat. -
It's not a cheat per se, it's for compatibility reasons
Because "30p" native isn't compatible with 60Hz systems, at 720 or 1080. You can't play a native "30p" file on blu-ray or AVCHD player for example.
For 720p30 you either duplicate real frames (more costly to encode), or use a frame repeat flag (like AVCHD Lite uses). So the signal is "looks" like 720p60 . (Now advertising as "720p60" might be considered as cheating....)
For 1080p30 , it's either encoded as interlaced, or encoded as progressive with a "fake interlace" flag. Eitherway the recieving hardware "sees" the signal as 1080i60, but eitherway the actual content is 1080p30
It's similar for 50Hz PAL countries (720p25 either has real duplicates or has a frame repeat flag so it "looks" like 720p50 , and 1080p25 is either encoded as interlaced or has a "fake interlace" flag for 1080i50 - that is how "Euro" blu-ray's are done, because 25p is not compatible with blu-ray)Last edited by poisondeathray; 29th Sep 2011 at 11:11.
-
So Panasonic is trying to take care of the optical disc users, basically, while at the same time getting to wave a "60p" flag in their spec sheets, and if anyone calls them on it, it's a giant misunderstanding on the consumer's end.
Since most consumers have no idea about any of this, and just vaguely remember a friend telling them that 60p is better, they fall right into the exaggerated numbers trap. -
Why not just load the footage (30p flagged as 60i) in to Vegas as a 60i project and render as a 30p file. My Canon does the same thing, 30p in 60i. It bugs me, progressive should be progressive. Interlace is a pain in the ass. I've been able to do as I mentioned, 60i project render as 30p (some h.264 container, AVC, MP4, ect.). Seems to work OK.
Now my 24p in 60i footage is alltogether a different issue. -
Similar Threads
-
Help explaining 60i (30p sensor output)
By x2x3x2 in forum Newbie / General discussionsReplies: 7Last Post: 27th Jul 2011, 22:05 -
Quick question on Deinterlacing 30P in 60i containers
By ayim in forum Video ConversionReplies: 5Last Post: 30th Dec 2009, 22:06 -
HD camcorder: 30p, 24p or 60i?
By cheerful in forum Newbie / General discussionsReplies: 7Last Post: 27th Dec 2009, 17:28 -
Deinterlace 60i -> 60p, yes, 60p, not 30p.
By 123Mike in forum Video ConversionReplies: 20Last Post: 4th May 2009, 13:43 -
frame rates 24 30 30p 30i 60 60p 60i
By maxtrack in forum Camcorders (DV/HDV/AVCHD/HD)Replies: 5Last Post: 28th Jul 2007, 17:21