Joining AVCHD clips
 

I recently purchased the Sony SR11 and have to say I love it. I can't imagine going back to tape again after being able to take clips right from the camera to a hardrive then directly out to a Hidef TV- in minutes, not days like tape- I say days because I would have to find time I could sit in front of the computer to capture from tape, then transcoding, then burning to DVD's. With the SR11 I copy the files from the camera directly to the media server hardrive . I use the media server to store all my videos and pictures and it's opened up a whole new world. It finally gave me the ability to have access to all my home movies and digital pictures on our TV with out having to sort through cd's, dvd's, and tapes. If anyone is considering it, all I can say is do it. Of course, right now it's a pain to edit avchd files, but 99.9% of the time, I don't edit.

So here's my problem:

I'm using a media server (Tivx 6500a) to view the m2ts files on my TV. The problem is that every time you hit pause on a avchd camera, it creates another file. When I play back the files, there is a pause between each clip ( I know this is a limitation of the media server or any media player for that matter). What I want to do is join the clips without having to re-encode. Apparently there doesn't seem to be any editing programs at the moment that I have tried that can do this without rerendering the entire file- which takes many many hours and defeats the whole purpose of tapeless for me. I've tried doing the DOS copy that people use for the canon camcorders, but it doesn't work- may have something to do with the Sony m2ts file structure. I've tried TSMuxer- which seems to work if I play the file back on a computer, but when I play the appended file back on the server, it stutters if I've joined more than 2 files together. Has anyone else out there been able to successfully join multiple (more than 2) sony avchd files together with out re-rendering the entire file and then had them playback flawlessly? If you have, I'd like to know what software you are using. Just so people understand what I'm asking- joining 30 minutes worth of clips should just take a few minutes if it's not re-encoding the entire file.

I would assume someday Vegas will have smart rendering for avchd files like they do for dv and hdv files, but it definately doesn't now.

I share your desillusion, and would be more than happy to read any form of solution.

To me this clearly shows that AVCHD is still in its infancy . It was massively pushed forward by SONY and PANA in the overall BD vs HD-DVD war, but the real needs of the users were just left aside !

Just to show in which despair i am in, to get editable and lengthy footage i have decided TEMPORARILY ( i hope) to create an HDV file from my sequences of stupid few seconds AVCHD clips ! It Takes only(!) twice the real-time length but thanks to TMPEnc Xpress batch mode this performs well at night...

Nevertheless this situation is a shame, and the promoters of AVCHD should be blamed for it!

Have you tried TSsplitter?

http://www.ffprojects.net/tssplitter/tssplitter.htm

Regards
Steve

Yeah - It causes freezes where the files are joined when played back.

Bill

Nero 8 Ultra Edition's Vision editor can join, using smart rendering, multiple unedited AVCHD clips into a single large file quickly and without loss of PQ.

I created a 15 minute movie in Vision (version 5.3.9.0) using 30 short clips that were taken with my Canon HF100. The movie was rendered as an AVCHD disc (writing to HDD) and took 1.5 minutes to complete on my quad core desktop PC with 30% average CPU usage. The resulting 2GB mts file in the Stream directory plays smoothly at the full original bit rate with no noticeable pixelation.

I've tried Nero already. It wants to re-encode the video files no matter what settings I use. I'm starting to think it may be that the Sony M2TS files are different than Canon's files.

Bill, when you get to the burn options window, click on details and see if your movie has these settings:

Audio format:Automatic
Sample format:Automatic
Encoding mode:Fast Encoding (1-Pass)
Nero SmartEncoding: Automatic

Number of titles: 1
1. Title 1 (Video Title, 0h 00m 00s)*
- Audio SmartEncoding ratio: 100.0 %
- Video SmartEncoding ratio: 100.0 %

Video disc menu: Do not create a menu

*this will vary depending on number of clips

No matter what settings I use, I can't get the video smartencoding ratio to read anything other than 0%.

I think why Nero won't work for me is the max bitrate it can encode at is 14 mbit/sec. The sony AVC files I'm using are 16mbit/sec. Maybe because it can't presently go over 14 ( I'm using version 5.3.3.0) it forces a reencode. I'll have to experiment latter and lower the quality setting on the camcorder and see if they pass through Nero with out reencoding, also will have to check for Nero updates and see if there's a new Vision update that does over 14 mbit.

Stan, are the settings on your canon using full quality mode when shooting? I think the canon hf10/100 goes up to 17 mbit/sec if I remeber correctly. If yours are over 14mbit and it's working for you, are you using a new version of Nero Vision?

edit: Just tried using a lower quality recording that was 9 mbit/sec and Nero still want to re encode it.

I use FXP 17mbps mode only on my Canon. The 5.3.9.0 version of Vision I use appears to be newer than yours, so that may be part of the problem.

Here's an example of the workflow I use to create joined clips in Vision:

1) Nero Vision Start > Make DVD > AVCHD

2) Content > Make Movie > Locate, Select, and Drag Clips to Timeline > Next > Next

3) Edit Menu > Menus to use: Do not create a menu > Next

4) Preview > Next

5) Burn Options > Write to Hard Disk Folder > Write

6) Done!

Stan.
I wonder if i could get some help with this.
Im using Nero Vision 5.3.9.0 and i too cant get my Hf10 AVCHD files to join without transcoding.
In the details setting i have Smart encoding Video = 100% but Audio = 81.5%.
Any idea why this is.
Also im not clear on the other settings in the Video options. Ive changed the quality setting to custom so i can enter 17,000kbs and 1920x1080 below it. Is this right?


Thanks

Dave.

Dave, I'm not setting or changing any video or audio options at all during my 6-step process. My suggestion is to follow my workflow exactly and see what happens.

Vision seems to smart encode unedited clips the best if you start off using the Make DVD -> AVCHD preset. I tried exporting a movie directly to a file, but no matter what custom setting I made, Vision wanted to transcode it.

I always import my clips from the HF100 to the hard drive first using Pixela ImageMixer3 SE before editing. I don't know if that really makes any difference but you might want to try that also.

Thanks Stand but im not having any luck.
Unless i set the Video options to Custom, 17,000kb/s 1920x1080 the Video smart encode is showing as 0%. With the above osettings its 100%. I think its the audio thats the problem. Its never showing as 100%, always around 75-85%.
Also the output file is coming out really big, 12GB+ for a 1 and a half min clip.

Im going to give up. Its not that important as i normally use VEgas Pro to edit my clips, this ws just to to see how well it works.

Dave.

Dave, using my method, this is what Vision lists in the Burn Options window under Project summary:

Current video options:

Video mode:NTSC
Aspect ratio:16:9
Quality:Automatic (fit to disc)

When you click on More >> at the bottom of the Burn Options window, you'll see the Video Options button. Clicking on that, you should see:

Video mode:NTSC
Nero Smart Encoding:Automatic

Then click on the AVCHD tab:

Video format:MPEG4 - greyed out

Under Transcoding quality:

Quality setting:Automatic (fit to disc)
Sample format:Automatic
Bitrate and Resolution - greyed out
Encoding mode:Fast encoding (1-pass)
Aspect ratio:16:9 - greyed out
Audio format:Automatic

Since you're in the UK, I assume you're using PAL. Your HF10 probably records 50i and 25p. That could possibly be what's causing Vision to transcode the audio.

BTW, I was able to import the HF11 24mbps clips from the Japanese site to the Vision timeline and burn them to an AVCHD disc without any problems. The disc plays back fine on my PS3 at the full 24mbps rate.

Thanks for posting back Stan.
Yes, all my settings seem to match yours other than i am using PAL.
Yet still i get Smart Encoding Audio = 81.5%.
The other thing that makes me think this is all wrong is that using fit to disc i think Vision is trying to fit the file to my HHD. The last one i aborted was over 23GB and this might be why its trasncoding.

I wanted to add that I routinely join AVCHD clips using the latest Nero Vision as described above and there is absolutely no re-rendering.

I have recently tried full bitrate 24 MB/sec and 17 MB/sec Canon clips with no re-rendering whatsoever, and the output .mts file is within a few bytes of the sum of the original .mts clips.

All tests were NTSC.

I wish i could but have more or less given up.
Im pretty sure im doing everytihng right but see no way of getting the smart encoding for audio to show 100%.
I cant figure out whats going wrong but to me it makes sense that Vision is transcoding.
Unless the output is the same as the input files then i can see a need for transcoding.

Dave,

I am purely guessing that PAL versus NTSC works differently. I am using entirely default settings, have not done anything special to engage 'Smart Rendering', and I imagine you should see similar behaviour unless the program somehow treats PAL differently.

Larry

Thanks Larry. You might be right.
As i normally would edit all my clips ill not dwell too much on getting this running but im one of these people that works to get things going, even if i dont need it.
It will bug me until i do get it going though.

Dave.

Dave,

It might be worth an email to Nero support asking why others have success with Smart Rendering and you do not.

Alternate;y, the thought also occured to me that you could download an NTSC .mts clip from one of the several US or Japanese web sites and load it into your version of Nero and see if the NTSC content behaves as others, myself included, describe it. This might help distinguish between a bad Nero software installation rather than PAL versus NTSC rendering differences.

Good luck,

Larry

Well, finally got it to work. Updated Nero so I'm using the lates Showtime release. Although here's a weird thing- The files I were trying to use that caused a re-render, been using these same ones to test out different programs, would always cause a re-render. On a whim I tried using some different files from the camera and they worked. But as I go through the earlier files I have, some work with out a re-render, some don't. On some, the video shows that smart render will work, but not the audio, on others, the audio but not the video. Although so far this seems to be happening radomly to the very first files from when I first started using the camera. I've been testing so many different programs out that it's possible that they had done something to the older files. I still had the originals on the camera, so I down loaded them again and retested them. Now those are working fine without a rerender. This is one of the truely awsome things about these harddrive cameras, being able to browse through hours of footage and retransfering to the computer- all in just a couple minutes.

So Dave, you might try using different files to run through the program and see if all cause a less than 100% audio smart render. Apparently one of the programs I had been testing had changed the original files, which they weren't suppose to do.

P.S. I don't recall seeing this mentioned yet, but remember to get the Nero Showtime to take multiple clips and make them as one, you have to remember to push the "Merge Titles" button that is on the left side of the "Content" screen ( this is the screen where you add video files).

Thanks for the help.

Thanks bill.
Actually now that i recall i was getting variations in the amount of smart encoding Vision was reporting for the audio and i was trying different files but ill try some fresh ones, just to make sure nothing has altered.
Im still not sure why my output files were 20BG + though.

Are you using Vision? You mention Showtime a few times?

Dave.

Hi,
I'm very interested in a video editing program that doesn't require rendering for joining clips. It seems that Nero Vision is one of them (but with little problems); I also read about Cyberlink PowerDirector to be able to do this (latest build). Did anyone tried PowerDirector?
I will download an evaluation program and report myself; for some strange reasons Cyberlink evaluation it's not always the last build and you don't have smartrendering (fixes appears faster than evaluations build).

Just an update.
I loaded Nero on to another PC and with the clips i gave it it worked fine. No re encoding and the output file plays fine.
I think ill uninstal it at home and try again.
But, if i now try and put that file in to Vegas it will only import as an audio file.
Bugger. I was hoping to use this one large file (only 120mb mind) rather than importing the smaller ones.

Any ideas guys?

Woops, meant to say Vision, not Showtime. Testing out so many different programs it gets confusing after awhile.

Mircia,

Cyberlink PowerDirector cannot smart render AVCHD, and takes a very long time, even on my quadcore QX9650 Extreme. You may like some of its other features but speed is not good at all.

Dave...I am very happy to hear that the Nero Vision joining and smart rendering is now working. I never considered asking you if you were using the latest versions, and assumed you were.

I still feel a bit embarrassed to admit that I use Nero Vision so much, since I own Final Cut Pro Studio, Vegas 8 Pro, and a half dozen other programs here, but Nero is the ONLY one which gets rendering done properly, and I especially like their animated 3D AVCHD/HD menus. They put Pinnacle, Ulead, Sony, Cycberlink, and Apple to shame as far as I am concerned in both their rendering as well as their menu animations. Considering that Nero Vision is merely one of many programs you get for about $70, and that their burning software is probably the best around, it is a remarkable bargain.

I would go so far as to say that if Nero offered a stand-alone NLE package for $99 with more control and options for advanced AVCHD disk creation, I would buy it in a heart-beat based on all the other stuff I use, most of which is far too slow. Best of all, their final output exactly matches the original clips. Believe me when I tell you that enlarged crops as well as normal HDTV viewing shows the effects of rendering on virtually all of the other AVCHD authoring products, in some cases glaringly.


Larry

I thought so. Thanks for clearing that up.
I've had more luck at home tonight. I think there is something funny about the clips Ive been trying to join as Ive just tried some others and it works fine. Didn't have to re install anything.
Ive also tried Pixela Imagemixer that came with the HF10 and that too seems to work.
Unfortunately none of these outputted files work probably in Vegas. I guess its not a big deal as the whole idea was to join them so i could stream them in one go to my TV but even my PCH and PS3 are giving me blank frames at the joins so ill abandon the whole thing now.
Thanks for all the help guys.

Dave.

Just wanted to make a final post to say ive finally got ImageMixer3 to join the files with no issues at all now.
thanks for all the tips.

Dave

Hi, Larry. When you say you "join" clips, do you mean you edit clips in Nero and then join them? I haven't been able to get Nero to do this while maintaining smart-render capability. (Nor have I been able to with PD7, which I know you also like).
Thanks.

Heres a free tool that may help:

.mts clip join tool - Page 2 - DVXuser.com -- The online community for filmmaking

It was put together by a user of that forum to join the .mts clips together from the HMC150, but should work with any AVCHD clips.

I haven't had a chance to check out the pervious user's suggestion but I use a beaut little program called TSmuxeR (http://www.smlabs.net/tsMuxer/tsMuxeR_1.8.35(b).zip)

I found the the M2TS file produced by my sony HDR-XR500V was really quite difficult to work with. Firstly, I don't own any commercial editing software so I could try any of those and secondly, I hate losing quality with re-encoding.

Also I want the file size of the merged clips to be no larger than the sum of the initial files

My WOrk flow is this.

Transfer the video to your computer.

Use the supplied tools with the camera to trim what you don't want. (there's an option for this in TSmuxeR but I've never used it)

Using TSmuxeR join/append the M2TS files select the output to M2TS and hit the "start muxing" button.

The resulting file is still a dog to work with but you've not lost any quality and you don't need a terrabyte of HDD to store your file.


If you just want to work with M2T files, I just use avidemux to achieve the same result only a LOT easier.


Hope this helps

L

TsMuxer works for me, but not frame accurately.
 

I am using TsMuxerGUI 1.10.6, it joins my .ts files from Hauppauge HD PVR correctly.

The only shortcoming is that its editing function is not Frame Accurate. But it surely does not re-encode any thing, so the video quality is maintained well.

But when I want to cut out commercials I want frame accuracy, I would like to use Smart Cutter for Dv and DVB from FameRing, which is Frame Accurate, and supports 'smart rendering'. It only re-encode the gop on cut point and end point, leaving middle part of movie untouched. Speed and quality are both achieved.

But Smart Cutter is not free, I wish someday TsMuxerGUI can release a frame accurate version. That will be nice.