Sorry if this is the wrong place to ask this, but I am looking at getting a used Sony TG3e for shooting my holidays in HD. I belive it records in 1080i 50hz, but does this mean 50fps?

My main concern is, that I will be outputting to a PAL DVD wich is 25fps. Will the 50hz interlaced footage convert to 25p without problems? I will likely be using pinnacle studio (latest version) for editing and outputting.

My last vacation I shot with my 5D mkII at 30fps and had to make an NTSC DVD wich is worse quality than PAL, so I want to make sure it will convert to PAL DVD without motion problems.

Also the TG3e can record in 1440 x 1080 to save space. Is there any problems using this in regards to making a PAL DVD later?

HDR-TG3E (HDRTG3E) : Oversigt : Videokameraer : Sony (http://www.sony.dk/product/cam-high-definition-on-memory-stick/hdr-tg3e#pageType=TechnicalSpecs)

Per

I shoot 1080i 50HZ all the time being here in Australia. No 50i doesn't mean 50 FPS. 50i records lines 1, 3, 5, etc at the 1st 1/50th of a sec then lines 2, 4, 6 etc the 2nd 1/50th of a second so when shown, it 'interlaces' the two fields together to form 1 picture.

There are several ways of down sizing HDV to SD, however (I use Premiere Pro CS4) you should be able to drop it into a SD timeline and scale down the footage to fit the screen (easiest method) but for best results, you can use 3rd party software to downsize HDV to SD/

Search these forums or google HD to SD and you will see plenty of options.

Thanks. I have Cineform NeoHD and it has an option to deinterlace. Is there any point in using that? I am unsure how the AVCHD video will edit native so I may convert to cineform.

I don't know about Cineform. I have never used it.

Hi Per
Yes you need to de-interlace unless your NLE has the facility to do it. HD footage uses Upper Field first and SD uses Lower Field first so if you neglect to de-interlace when the 1440x1080 frame is downsized to 720x576 the interlacing lines will not be handled by the TV at all. Most NLE's will also have an option to de-interlace source footage when you import it. If you are going to just transcode the AVCHD to SD AVI then the transcoder will do the job for you!! I transcode all mine with a MainConcept transcoder but that only handles Panasonic AVCHD ... Sony cams do come with a transcoder but it's not very good and does your footage straight to VOB files so rather use something else!!

Pinnacle 12 was able to de-interlace so as long as you have a fairly swift machine it will handle the files. My DuoCore had to do some Sony files the other day and ran OK but the user did use the lowest bitrate setting on his camera and renders were quite long..a 7 min file took 36 mins to render to MPEG2!!! A QuadCore is considered the minimum for AVCHD!!

Cineform's transcoder works well and also VAAST's upshift does a good job and will transcode your AVCHD files to HDV which your 'puter will handle easily!!

Chris

I am planning on using Pinnacle Studio 14.

My PC is: Core 2 Duo e8400@ 3.6GHz, 8GB RAM, dual RAID0 harddrive setup.

I think I will just try different workflows to see what works best for DVD output.

As far as saving an HD version wich format do you suggest I use? I have previously saved as WMV mainly because of compatability and that MS' encoder works pretty well.

Hi Per

With a DuoCore you probably need to transcode down to HDV ..you machine will struggle with AVCHD files unless you record at the lowest bitrate!!
I do mainly weddings and as none of my clients own a BluRay player I just transcode straight to PAL Widescreen AVI and edit in SD... the slight quality loss will not even be noticed by the bride and rendering out SD MPEG2 files for the DVD from SD AVI is lightning fast!! (around half the actual clip time compared to 8 times the clip time with HD to SD)

To be honest I save all my source files in AVCHD format.. when the shoot is over the first step is to copy the MTS files from each camera's card to my external drive so the source footage is safe. Then I transcode to AVI and work with the AVI files...that way I still have the native raw footage to edit if I need to at a later date.

Chris

Hi Per
Yes you need to de-interlace unless your NLE has the facility to do it. HD footage uses Upper Field first and SD uses Lower Field first so if you neglect to de-interlace when the 1440x1080 frame is downsized to 720x576 the interlacing lines will not be handled by the TV at all.
Chris

Sorry Chris but that is not correct. HD and SD television both use upper field first. What normally causes a problem is the use of the DV or DVCAM codec for SD which uses lower field first and this can cause problems, but the DV codec is not a good codec so would be a poor choice to convert to anyway.

There should be no need to de-interlace HD material when converting to SD interlace, however some NLE's do a better or worse job of handling the field conversion.

While the TG3's 1440 mode is lower quality than the full 1920 mode there are advantages to using 1440 if you are downconverting to SD as 1440 is exactly double 720 so the downconversion is easier and generally cleaner. One thing that can help to improve the quality of downconverts to SD is to add a very slight blur to the HD material before downconversion as this prevents aliasing (twittering and jagged lines) in the SD material.

I won't dispute the fact at all Allister but most NLE's will default to a lower field first when rendering to an MPEG2 for your DVD as they will be using the DV codec.

I render all my SD video out for DVD as an MPEG2 and interlacing is Lower Field first..if you do a forum search you will find that most HD to SD is done like that..maybe that is the only codec available?? I was told a while back that some DVD players struggle with the field order the other way around so it's obviously no correct.

Pers..test it for yourself and make up your own mind and let us know the outcome..as long as you get a good result then that's all that matters!! There is a VERY comprehensive post in the Vegas forum here on HD to SD..I'm sure you will gain some extra info from that one!!

Chris

The correct Field orders are:

PAL SD: Upper (odd) field first.
NTSC SD: Lower (even) field first.
PAL DV/DVCAM: Lower (even) field first.
NTSC DV/DVCAM: Lower (even) field first.
MPEG 2 DVD PAL: Normally Upper First, but can also be lower
MPEG 2 DVD NTSC: Lower (even) first, can be Upper first but this is unusual.

HD: Upper (odd) field first

It is the use of the PAL DV codec that messes things up for most people as a mistake was made when the codec was specified and it ended up with the wrong field order. As many NLE's default to DV they will default to Lower First. If you then place a Upper first file in that time line you will end up with your fields reversed. The best solution is not to use DV, there are plenty of far better intermediate codecs than DV. You should be able to render directly from HD to SD Mpeg2 without going through an intermediate anyway.

Hi Alistair

Many thanks!! That's very interesting to know and you obviously know your stuff!!

Would there be a big advantage in using UFF for SD PAL???? In Sony Vegas the render window for MPEG2 files seems to default to LFF so it must be using a DV codec. I wonder how one would find a codec that uses UFF and will still produce an MPEG2

I'll dig around and see what's available

Chris

I have installed Pinnacle Studio 14 HD. But it wont edit the mts files from the TG3e?!? I tried converting with NeoHD, but the avi files also wont edit in PS14 and the look horribly jerky when playing back.

Hi. I'd like to revive this thread if you don't mind...

After having searched everywhere and also after having started a few threads myself concerning problems I've been facing for months now, I have a strong feeling that the source of my problem has something to do with what has been discussed here.

Let me give you the bullet points of what the problems have been.

* I'm helping a friend (who has never edited before) get started with editing. He has a strong enough Windows XP based PC for the job. We are editing with Sony Vegas Pro 9e.

* Until several months ago he was using a Sony mini-DV camcorder. All was fine. But then he decided he needs to upgrade his camera, seeking the new HD technology. So after some research he decided to purchase a new Panasonic HDC-TM700 which records to a built-in hard drive. It shoots AVCHD material and has several shooting modes. One mode is 1080/50p and the other modes are 1080/50i with varying bitrates. By the way, we are in PAL land.

* Realizing that its a good idea to convert the AVCHD footage to an intermediate codec, he went ahead and bought Cineform's NeoScene which outputs .avi files that edit nicely and easily in Vegas. So far so good.

* After having edited a short movie of footage shot with his new TM700 he now wishes to render the project and burn it to an SD-DVD which will be then viewed on a standalone DVD player hooked up to a large LCD tv in his living room. Currently this is the way he'll be screening his projects - on a PAL LCD tv. But somewhere in the workflow between the rendering of the Vegas project and the burning of the DVD something has been going wrong. And although I've been searching for an answer for months... I still haven't found what is causing these issues.

A Description of the problems:
I can identify two different problems and I have no idea if they are connected or if they are two completely different issues. One is interfering horizontal lines that appear especially during moving objects on screen, or even more-so during camera movement. They look like the image is shreaded into thin horizontal lines. Some tests we did actually seemed to not have this problem and from the descriptions here at this thread I have a strong feeling it has to do with the field order mix-up.
The second problem is much more subtle, but never-the-less irritating and definitely should not be. It appears in small details of anything with horizontal stripes - a white shirt with black stripes, venician blinds, a grid on a fan... anything with horizontal lines looks like there are pixels jumping around inside it. I'm a bit at a loss of what the technical word might be for how to describe how it looks, but its like the fine details of the stripes start to produce some kind of animated pixels within them, or noise of some kind.

Different workflows we've tried (using material shot in 1080/50p):

We render to avi in Vegas using the Lagarith lossless codec and these settings: first set template to HD 24p for lack of a template with an exact match. then just changed the FPS to 50. So basically its: 1920x1080, 50 fps, progressive, 16:9

Then we transcode the avi file to mpeg-2 using TMPGEnc and basically we set it up according to this tutorial DVD-HQ : Configuring TMPGEnc for high-quality DVD-compliant MPEG-2 (http://dvd-hq.info/dvd_compression.php).
At first we just rendered to mpeg-2 straight from Vegas. After having consulted the Vegas forum here at dvinfo we were told that NLEs often don't do as good a job at HD to SD conversions and were recommended trying TMPGEnc. Yet the result seems to remain the same.

Once we have the mpeg-2 video stream and the separate audio file which is rendered in Vegas to ac-3 file, we then author the DVD in DVD Architect. Basically nothing much to change there from the default settings except the frame size to 16:9 720x576.

I read on this thread about a MainConcept transcoder that only handles Panasonic AVCHD. Is this software that comes with Panasonic cameras? Could this be on the CD that came with the camera? Would this be the recommended software to use?

Currently my friend is continuing to use his old Sony miniDV and the Panasonic HD camera is sadly collecting dust in the closet. Not only that, I feel somewhat responsible, since after all, I'm the teacher. Also I convinced him that he should purchase NeoScene to make the AVCHD 1080/50p files manageable with Vegas.

Please let me know if you feel you see what our worflow lacks or if you have any ideas what might be causing this. It just doesn't make sense that there is no solution for using this camera.

Thank you!