Dear all,
I have a Panasonic HDC-SD1. I now, thanks totally to this forum can import, edit and render videos to wmv for playback on my XBOX 360 with not trouble.
I only have onequestion left really, using the Windows Media Video 9 Advanced Profile I can choose 50fps and the interlace option. I understand that the my camera records in 1080i so this would seem appropriate.
I have done tests on a piece of video rendering it at 25fps and non-interlace and 50fps and interlace and I am not sure that I can see much difference - there still seems to be that ghosting (two images) effect on fast moving action - whch is not there when the video is played back through the camera's HDMI. W
hen I right click the imported avchd video the properties box tells me that the video is only 25fps - does this include the interlaced frames - or am I wasting my time rendering to 50fps and interlaced as the interlaced frames have been lost in the importing process.
The main reason I ask is that the rendering time is much longer.
Use of interlace in wmv for avchd imported content
Moderator: Ken Berry
I'm glad you posted this question and have been having different results between MF6+ and VS11+ importing avchd files recorded on a mini-dvd disk. The difference being that the fielding information is not displayed consistently. which then would indicate "Frame Based" (if no fielding information is displayed). But these cams do not record frame based (as far as I've seen so far). The video(s) properties should display UFF (upper field first), sometimes mine are being displayed and imported as frame-based.
Your cam should be recording interlaced video ( UFF ).
For NTSC video the frame rate should be 29.97 fps.
For PAL video the frame rate should be 25 fps.
If the video is interlaced you will also see a fielding parameter, for these cams they all appear to be UFF ( upper field first). On my systems the fielding parameter is displayed after importing the video(s). Either way, interlaced or frame based the fps (frames per second) should still be 25fps for Pal and 29.97 for Ntsc.
After importing your avchd videos and they are on the timeline right_click on one of them and select "Properties". The video/audio parameters will be displayed. They should display UFF (upper field first).
The HD-WMV Codec does not appear to support Interlaced Video. When you perform a avchd/hd-mpeg2 conversion to hd-wmv you should leave all the VS settings the same as the original source video. All conversions to the the hd-wmv format will be performed in the windows media encoder side of the conversion. This interlaced to frame based conversion will cause the motion in the video to be less fluid. The windows encoder has a smoothness settting to compensate for this. The smoothness setting works OK (not great). The windows encoder needs to see these 2 interlaced fields for proper conversion, this is why you would not send a frame based setting on the VS side to the windows encoder if your source videos are interlaced video. This would de-interlace the video before it gets to the windows encoder and the windows encoder will not be able to blend or mix the fields because it can't see them (if you choose frame_based on the VS side, which would be your "Project Settings").
The true frame rate is either 25 for pal or 29.97 for ntsc. When you see a 50i or 60i these are templates for HDV type video (interlaced). To actually apply a true 50 or 60 fps framerate requires you to manually set the frame_rate in the encoders. Many times when using this setting, manually changing the frame rate to 50 or 59.94 the videos will not playback properly, depends on the playback device. I've made a few of these 59.94fps videos and it depends on the device used that can handle playing them back. Many devices cannot playback these settings.
I have a xbox360 with a hd-dvd player add-on (for a short time) to check playback files. By the end of the week I'll be able to let you know the results of my conversions.
Your cam should be recording interlaced video ( UFF ).
For NTSC video the frame rate should be 29.97 fps.
For PAL video the frame rate should be 25 fps.
If the video is interlaced you will also see a fielding parameter, for these cams they all appear to be UFF ( upper field first). On my systems the fielding parameter is displayed after importing the video(s). Either way, interlaced or frame based the fps (frames per second) should still be 25fps for Pal and 29.97 for Ntsc.
After importing your avchd videos and they are on the timeline right_click on one of them and select "Properties". The video/audio parameters will be displayed. They should display UFF (upper field first).
The HD-WMV Codec does not appear to support Interlaced Video. When you perform a avchd/hd-mpeg2 conversion to hd-wmv you should leave all the VS settings the same as the original source video. All conversions to the the hd-wmv format will be performed in the windows media encoder side of the conversion. This interlaced to frame based conversion will cause the motion in the video to be less fluid. The windows encoder has a smoothness settting to compensate for this. The smoothness setting works OK (not great). The windows encoder needs to see these 2 interlaced fields for proper conversion, this is why you would not send a frame based setting on the VS side to the windows encoder if your source videos are interlaced video. This would de-interlace the video before it gets to the windows encoder and the windows encoder will not be able to blend or mix the fields because it can't see them (if you choose frame_based on the VS side, which would be your "Project Settings").
The true frame rate is either 25 for pal or 29.97 for ntsc. When you see a 50i or 60i these are templates for HDV type video (interlaced). To actually apply a true 50 or 60 fps framerate requires you to manually set the frame_rate in the encoders. Many times when using this setting, manually changing the frame rate to 50 or 59.94 the videos will not playback properly, depends on the playback device. I've made a few of these 59.94fps videos and it depends on the device used that can handle playing them back. Many devices cannot playback these settings.
I have a xbox360 with a hd-dvd player add-on (for a short time) to check playback files. By the end of the week I'll be able to let you know the results of my conversions.
-
spacem8n
interlaced questions
Thanks for your reply. I am new to video editing not just video studio so some of the concepts are new to me to.
However, are you saying that the HDC-SD1 records in 1440x1080 interlaced with the upper frame first. This was my understanding too.
I checked my project preferences and the follow settings seemed incorrect relative to what you have said
General tab\Default field order was set to frame based I have now changed this to upper field first. Will this actually affect the importing of the video? I do feel that the entire problem is down to information being lost in the import process.
Also when I right click to the get the properties of the imported avchd video either on the timeline or in the thumbnail area various information is shown in a pop-up dialog but there is no mention field order.
As you say I am sure when I downloaded the demonstration version of MF6 this information was present.
After changing these settings to the project preferences I then tried to export the clip as hdv 1080i for PC i.e. MPEG. I can only play this back on my computer but if I step through the output frame by frame the ghosting is still there.
Is there no answer to this problem - is it just a side effect caused through computer and LCD TV's working differently from normal video and CRT TV's. For me the videos are still very good, just not as good as straight from the camera.
Conceptually if the XBOX 360 supported H.264 streamed from another computer (which it doesn't at present) do you think this would sort this out. i.e. I would be able to simply output to the same format as the camera uses.
Finally for reference I stream the videos to my xbox from my vista pc using media center. My TV is an LCD HD 1080i and my xbox is set to display in 1080i.
However, are you saying that the HDC-SD1 records in 1440x1080 interlaced with the upper frame first. This was my understanding too.
I checked my project preferences and the follow settings seemed incorrect relative to what you have said
General tab\Default field order was set to frame based I have now changed this to upper field first. Will this actually affect the importing of the video? I do feel that the entire problem is down to information being lost in the import process.
Also when I right click to the get the properties of the imported avchd video either on the timeline or in the thumbnail area various information is shown in a pop-up dialog but there is no mention field order.
As you say I am sure when I downloaded the demonstration version of MF6 this information was present.
After changing these settings to the project preferences I then tried to export the clip as hdv 1080i for PC i.e. MPEG. I can only play this back on my computer but if I step through the output frame by frame the ghosting is still there.
Is there no answer to this problem - is it just a side effect caused through computer and LCD TV's working differently from normal video and CRT TV's. For me the videos are still very good, just not as good as straight from the camera.
Conceptually if the XBOX 360 supported H.264 streamed from another computer (which it doesn't at present) do you think this would sort this out. i.e. I would be able to simply output to the same format as the camera uses.
Finally for reference I stream the videos to my xbox from my vista pc using media center. My TV is an LCD HD 1080i and my xbox is set to display in 1080i.
I have to say that this new format is a real challenge for anyone not familiar with video. Experienced video people know exactly what your going through. This new format (avchd H264) is also new to them. Like anything we will sort it out, make mistakes and learn.
I believe this is a problem. I also have MF6+ and now VS11+ trial installed.
The properties should be Upper field first, if not UFF then ghosting, jerky playback, (motion not smooth like original).
I believe that these files aren't being read correctly and since the program (VS) may not be reading them correctly then using them for conversions can't work correctly. I wouldn't use them if the fielding information isn't correct. The program is saying there are only 30 pictures per frame when actually there are 60 pictures per frame. We are only seeing 1/2 the video data present in the video file (possibly). This may be the cause of the ghosting effect you are experiencing.
The PS3 (playstation3) plays these avchd h264 files back. Using MF6+ to extract them (without VS11+ trial installed). The .m2t files playback same as the original on the ps3. The files played back and burnt to hd-dvd format on a dvd should playback on your HDTV same as connecting your cam directly to your HDTV.
That's actually the best method to use. Connect your cam to the HDTV and view the original footage. The computer software should duplicate this same high quality playback.
I also use a Sony HC3 hd-cam that records in the hd-mpeg2 format. Playback of the computer edited files is really excellent.
Just a quick explanation of interlaced versus frame based. You can consider any camcorder as recording pictures at a very fast rate.
Interlaced video would use 60 pictures per second.
Frame-based would use 30 pictures per second.
So you can see if converting 60 pictures to 30 pictures you will lose a lot of information. This is known as de-interlacing. There are many de-Interlacing methods that computer programs use to convert the 60 pictures to 30 pictures per second. Blend, Bob, Weave, Linear etc. Some programs simply drop the 2nd interlaced field and keep the first. All this usually results in loss of good motion unless you use special programs to re-calculate all the video information (very time comsuming).
First your project settting should equal your source files (uff). But your saying your source files are reporting frame based (no fielding information).General tab\Default field order was set to frame based I have now changed this to upper field first. Will this actually affect the importing of the video? I do feel that the entire problem is down to information being lost in the import process.
Also when I right click to the get the properties of the imported avchd video either on the timeline or in the thumbnail area various information is shown in a pop-up dialog but there is no mention field order.
As you say I am sure when I downloaded the demonstration version of MF6 this information was present.
I believe this is a problem. I also have MF6+ and now VS11+ trial installed.
The properties should be Upper field first, if not UFF then ghosting, jerky playback, (motion not smooth like original).
I believe that these files aren't being read correctly and since the program (VS) may not be reading them correctly then using them for conversions can't work correctly. I wouldn't use them if the fielding information isn't correct. The program is saying there are only 30 pictures per frame when actually there are 60 pictures per frame. We are only seeing 1/2 the video data present in the video file (possibly). This may be the cause of the ghosting effect you are experiencing.
The PS3 (playstation3) plays these avchd h264 files back. Using MF6+ to extract them (without VS11+ trial installed). The .m2t files playback same as the original on the ps3. The files played back and burnt to hd-dvd format on a dvd should playback on your HDTV same as connecting your cam directly to your HDTV.
That's actually the best method to use. Connect your cam to the HDTV and view the original footage. The computer software should duplicate this same high quality playback.
I also use a Sony HC3 hd-cam that records in the hd-mpeg2 format. Playback of the computer edited files is really excellent.
Just a quick explanation of interlaced versus frame based. You can consider any camcorder as recording pictures at a very fast rate.
Interlaced video would use 60 pictures per second.
Frame-based would use 30 pictures per second.
So you can see if converting 60 pictures to 30 pictures you will lose a lot of information. This is known as de-interlacing. There are many de-Interlacing methods that computer programs use to convert the 60 pictures to 30 pictures per second. Blend, Bob, Weave, Linear etc. Some programs simply drop the 2nd interlaced field and keep the first. All this usually results in loss of good motion unless you use special programs to re-calculate all the video information (very time comsuming).
-
Jerry Jones
- Posts: 358
- Joined: Sat Feb 19, 2005 4:14 pm
- Location: Boise, Idaho, USA
- Contact:
-
Jerry Jones
- Posts: 358
- Joined: Sat Feb 19, 2005 4:14 pm
- Location: Boise, Idaho, USA
- Contact:
Well, I've been experimenting with Pinnacle's Studio 11 Plus Ultimate software to find out if their HD MPEG-2 encoding looks any better.
It doesn't.
So I'm reluctantly beginning to suspect that what we're seeing is not so much a Ulead or MainConcept bug after all; we may be seeing the unfortunate limitations of consumer HD MPEG-2 encoding.
I thought that if the Pinnacle software produced remarkably better output, then that would prove there's a problem with the Ulead software.
But their output is pretty similar, actually.
Jerry Jones
http://www.jonesgroup.net
It doesn't.
So I'm reluctantly beginning to suspect that what we're seeing is not so much a Ulead or MainConcept bug after all; we may be seeing the unfortunate limitations of consumer HD MPEG-2 encoding.
I thought that if the Pinnacle software produced remarkably better output, then that would prove there's a problem with the Ulead software.
But their output is pretty similar, actually.
Jerry Jones
http://www.jonesgroup.net
Gateway 7426gx
http://tinyurl.com/hagye
http://tinyurl.com/hagye
Jerry,
What software are you using to playback the hd-mpeg videos on the computer? I'm using PowerDvd 7 Deluxe or Ultra, still waiting on the boxed version of PowerDvd7. What video card are you using? I'm using an ATI1650 HD-Certified PCI card on one machine and a GeForce 6600GT PCI on another. Both are HD-Certified and supported by PowerDVD7.
Having the required hardware & software combination is mandatory for proper playback on the computer.
I would think yours is already up to snuff since you must be playing back the avchd videos on it .
Did you import from your cam to extract the videos in Pinnacle or did you use the source files that the ulead importing function extracted to the harddisk. This may make a difference.So I'm reluctantly beginning to suspect that what we're seeing is not so much a Ulead or MainConcept bug after all; we may be seeing the unfortunate limitations of consumer HD MPEG-2 encoding.
I thought that if the Pinnacle software produced remarkably better output, then that would prove there's a problem with the Ulead software.
But their output is pretty similar, actually.
I think it would help if I sent you some hd-mpeg2 videos on dvd's, both in the TS & PS format, recorded on a Sony-HC3 HighDef camcorder. Some edited and some original straight from the cam. This would give you a good base to compare & analyze.we may be seeing the unfortunate limitations of consumer HD MPEG-2 encoding.
What software are you using to playback the hd-mpeg videos on the computer? I'm using PowerDvd 7 Deluxe or Ultra, still waiting on the boxed version of PowerDvd7. What video card are you using? I'm using an ATI1650 HD-Certified PCI card on one machine and a GeForce 6600GT PCI on another. Both are HD-Certified and supported by PowerDVD7.
Having the required hardware & software combination is mandatory for proper playback on the computer.
I would think yours is already up to snuff since you must be playing back the avchd videos on it .
-
Jerry Jones
- Posts: 358
- Joined: Sat Feb 19, 2005 4:14 pm
- Location: Boise, Idaho, USA
- Contact:
Yes, I used the "Import From DVD" capture that is in the Pinnacle software.
The Pinnacle software, however, can also directly import the .M2TS files captured by the Sony "Picture Motion Browser" software.
Jerry Jones
http://www.jonesgroup.net
The Pinnacle software, however, can also directly import the .M2TS files captured by the Sony "Picture Motion Browser" software.
Jerry Jones
http://www.jonesgroup.net
etech6355 wrote:Did you import from your cam to extract the videos in Pinnacle?
Gateway 7426gx
http://tinyurl.com/hagye
http://tinyurl.com/hagye
-
Jerry Jones
- Posts: 358
- Joined: Sat Feb 19, 2005 4:14 pm
- Location: Boise, Idaho, USA
- Contact:
No, my equipment isn't really HD "certified" at all.
My notebook features an ATI Radeon 9700 with the latest "Catalyst" driver.
I have a high definition flat panel display, an Acer Ferrari, connected via VGA.
I have InterVideo WinDVD 8.
I realize this isn't the best -- and I suppose there's a remote possibility that this fact skews my testing -- but I'm not made of money.
This is going to be my last bit of testing because I'm losing interest in this video hobby and want to move on to other things.
I already have HDR-HC3 footage here and I see pretty much the same on fast moving subjects.
Frankly, I don't think upgrading to a "certified" system would make any real difference.
I think the problem is that we consumers were expecting perfection from consumer HD.
What we got was a format that shows a whole lot of extra detail... including the detail we didn't want to know about... such as interlace & progressive eye-tracking artifacts that we could ignore when we were watching relatively less clear NTSC MPEG-2.
Now we see every flaw.
Jerry Jones
http://www.jonesgroup.net
My notebook features an ATI Radeon 9700 with the latest "Catalyst" driver.
I have a high definition flat panel display, an Acer Ferrari, connected via VGA.
I have InterVideo WinDVD 8.
I realize this isn't the best -- and I suppose there's a remote possibility that this fact skews my testing -- but I'm not made of money.
This is going to be my last bit of testing because I'm losing interest in this video hobby and want to move on to other things.
I already have HDR-HC3 footage here and I see pretty much the same on fast moving subjects.
Frankly, I don't think upgrading to a "certified" system would make any real difference.
I think the problem is that we consumers were expecting perfection from consumer HD.
What we got was a format that shows a whole lot of extra detail... including the detail we didn't want to know about... such as interlace & progressive eye-tracking artifacts that we could ignore when we were watching relatively less clear NTSC MPEG-2.
Now we see every flaw.
Jerry Jones
http://www.jonesgroup.net
etech6355 wrote:I'm using an ATI1650 HD-Certified PCI card on one machine and a GeForce 6600GT PCI on another. Both are HD-Certified and supported by PowerDVD7. Having the required hardware & software combination is mandatory for proper playback on the computer.
I would think yours is already up to snuff since you must be playing back the avchd videos on it .
Gateway 7426gx
http://tinyurl.com/hagye
http://tinyurl.com/hagye