Hv20 Best Practices

[ESTEBEVERDE]

Now that the HV20 has been out for a bit, I am wondering what people have found to be the best practices in regards to capturing (settings, lenses,...) and post (deinterlacing, 24p pulldown removal, color grading,...).

[vsansal]

my workflow is as follows:

*Capture using sony vegas 8 (was using 7)
*Remove the pulldown using TMPGenc 4 Xpress
*Import the video back into Sony Vegas
*Use Magic Bullet Look suite for CC
*Render as 1920x1080/24p (or 60i) HD mpeg2 video with the highest video quality setting and preserving 25Mbps bitrate
*Using Ulead DVD movie factory 6 plus burn it as HD-DVD to regular DVD-5 or DVD-9
*Watch it using HD-DVD player (Toshiba A2) + HDTV combo
*Preserve the original tape (if necessary)

[knowhowiroll]

I am a noobie to the forum and my new HV20. Zero experience in camcorders except for an very old Sharp say 12 years old. I am impressed with the quality of the footage and would like to download the footage to my PC and edit/render or whatever one needs to make a near as high quality duplicate that I can burn to a yet to be purchased HD DVD burner. I downloaded the trial version of Vegas 8 however cannot capture HD only DV. Any tips on how to work around this. I have read few posts regarding this but have yet to net any progress.

[vsansal]

Try using another capture application such as HDV split which is free. Also make sure the DV lock future is not on in the camcorder.

[knowhowiroll]

I would like to be able to capture in whatever software I eventually end up purchasing. Just does'nt seem right that I would not be able to do that. I have used HDVsplit and it breaks the captured portion into many smaller files. Is this normal. If so how do you handle those files in Vegas. Also when I did import the mt2 files to Vegas in the lower right hand corner it indicated they were 720x480 in the project. I would like to retain the 1440x1080 resolution for future reference. What is the big tado with pulldown? Why does it need to be removed. I watched a tutorial on this and it appeared to do this in 24p automatically in Vegas. Thanks for your help

[vsansal]

First for breaking into small files, it detects the end of your clips and breaks them into pieces. This is normal and the right thing to do since you may want to change the order of your clips or add/remove. If you want you can also capture everything at once like a single clip but it is a headache. When you import these files to vegas, you can add them one by one to the time line or select all of them and bring them all at once it is your choice. If you want to use transitions between them, one by one is the way to go. The reason you cannot capture with vegas may be because it is the trial version. I am capturing using Vegas without any problems.

When you start a project, click file and then properties. In the properties window select 1440x1080 instead of 720x480. Also choose the frame rate accordingly.

When you don't remove the pulldown, you will have a repeated frame created by the pulldown. If it is not removed or de-interlaced properly it will create terrible ghosting effects if you render out as a progressive. If you render out as 60i you won't have these problems but it won't be the same as 24p video.

As for your last question, vegas can remove 24p automatically but unfortunately not from HDV but DV. So it is not good for HDV video. Also HV20 doesn't make this easy for a lot of NLEs by not adding any flags. I am using TMPGenc 4 Xpress as I mentioned and I really like it. I know it is an extra expense but I think it is worth it. There are also free scripts for doing this. If you search it you can easily find them.

[knowhowiroll]

Thanks for the details. You have been very helpful.

What are all these different frame rates? 24p 25p etc. I gather from the manual when I record in the cinema mode it is in 24p.

Also I have been capturing in HDVsplit and using HDTVtoMPEG2 to view short clips. I appears when I use the cinema mode I am able to make and view nice looking footage on my PC. However when I filmed a school basketball game in some other mode -P I reckon than Cinema because it looked better on the lcd I am not able to view the mpgs at all. It appears to go thru the same iterations as footage shot in the cinema mode at an indoor Xmas concert at the church.

When I say iterations I mean capturing in HDVsplit then converting the .mt2 files to mpegs with HDTVtoMPEG2. Upon opening the mpeg with Windows Media Player it appears to start playing -bar slides briefly then it reverts to the original start point. No video or audio is played back.

Just curious if you were to film sporting events of the kids to view years down the road how would you ensure a back up in addition to the original tape with the native data for one scenario and another scenario using discs to conveniently view the events on a HD TV. Perhaps both scenarios could be accomplished in one step.

I ask this because my wife says we should have bought a hard drive camera because in her mind all this could be accomplished a whole lot easier -according to her. You know how that is. LOL

[vsansal]

24p/60i frame rate is used for NTSC video systems however in countries that use PAL video system they use 25p/50i frame rate. In USA we use NTSC. When you use 24p mode, the less frame rate will give a more film like look. The Hollywood movies you watch everyday are all recorded in 24 frames per second. The aim is to get that look. Of course only 24p is not enough for this but it is a huge step in that direction.

Cine mode and 24p mode are independent from each other. You can use them at the same time or one by one. To switch to 24p mode you need to go to the main menu and change the recording type.

Mpeg2 codec is not standard with windows media player. Try downloading VLC player from videolan and watch it using that.

When you say backup do you mean archiving or just in case something happens to the tape? If it is for archive after capturing keep the original tape and burn the video to a DVD. This method is the easiest and safest way. Also since you are keeping the original tapes as HDV even if you don't have and HD video system yet, in the future you can burn it as any format as you want by capturing the original video from the tape again. Believe me tape based camcorders are more reliable and efficient than using a hard drive base camcorder.

[knowhowiroll]

What is the highest resolution that a standard DVD will display. I would like to make sure I am making the most of this camcorder when backing up tape for now.

What would be your work flow assuming little or no editing?

Would Windows MovieMaker suffice? I also have WinDVD.

Also what brand tape do you use? We are using these Sony tapes which seem to be pricey.

Thanks again

[vsansal]

If you are asking a DVD player, then it is 720x480 but you can burn HD-DVD or Blu-Ray content to a standard DVD which will increase the resolution up to 1920x1080. You will need a Blu-Ray player or an HD-DVD player but you can buy an HD-DVD player for $200.

After capturing, if you don't want to do editing that much, just select all clips and add them to the time line. Then all you need to do is, if you are burning a regular DVD, render it as mpeg2 with 720x480 resolution. Than by the help of a burning software make a DVD movie using the file you've rendered.

If you are using windows vista movie maker can be used but if you have XP, it doesn't support HD. I would suggest you to buy a separate software such as Vegas you are trying right now. It includes all the tools needed from capturing to burning it to a DVD. But if you decide to buy an HD-DVD player or Blu-Ray, you will need another burning software for that. I am using Ulead DVD Movie Factory 6 Plus.

For the tapes, you don't need to buy one that is labeled as HDV tape. Any miniDV tape will do the work. The HDV labeled tapes has a better error ratio but it has nothing to do with the video quality. I would suggest you to just pick one brand and stick to it since changing brands can sometimes create problems.

[knowhowiroll]

"but you can burn HD-DVD or Blu-Ray content to a standard DVD which will increase the resolution up to 1920x1080."

How is this accomplished. I thought I would need HD media -a HD DVD.

Also what does rendering do?

[vsansal]

It is possible but it limits the length of the video because of the capacity. If you lower the bitrate a little you can increase the length. I burn all of my movies to a regular DVD and watch it using my HD-DVD player. They look great. Using Ulead DVD Movie Factory 6 Plus it is very easy to accomplish this. The new Nero 8 also supports this option. If you look at the AVCHD DVD based camcorders, they record HD video to MiniDVDs which has the same characteristics as a regular DVD except the size.

When you render, your computer will apply all the editing you've made including transitions, color correction, adding text, merging different clips together etc. and deliver the product as a single file in the format you choose.

[ESTEBEVERDE]

What have you found as the best method to remove the interlacing?

At what point in your workflow do you find it best to deinterlace to end up with the best image possible?

[vsansal]

What have you found as the best method to remove the interlacing?

At what point in your workflow do you find it best to deinterlace to end up with the best image possible?

If I use 24p mode, I am using TMPGenc 4 Xpress to remove the pulldown which outputs the real progressive image recorded by the camcorder in the first place. So when I import it to Vegas it is already progressive. I don't need to do anything else. If I use 60i mode, I use vegas to de-interlace the video while rendering. I use blend the fields method which blends the odd and the even lines to produce a full frame.

[ESTEBEVERDE]

If I use 24p mode, I am using TMPGenc 4 Xpress to remove the pulldown which outputs the real progressive image recorded by the camcorder in the first place. So when I import it to Vegas it is already progressive. I don't need to do anything else. If I use 60i mode, I use vegas to de-interlace the video while rendering. I use blend the fields method which blends the odd and the even lines to produce a full frame.


Right On!

Thanks!

[Tim Young]

How is this accomplished. I thought I would need HD media -a HD DVD.

This does work. I've just started doing it with excellent results. I got a Toshiba HD-DVD player on black friday for $100 and have it hooked up to a 19" high def TV that I got for $200, so it's not that expensive to put together a simple HD system. You can also play back HD-DVDs on a newer mac in the DVD player program. I've made the HD-DVD formatted DVDs using DVD Studio Pro. It's pretty easy and there's actually a template in Compressor for it. I didn't know you could do this with Blu-Ray though. I guess that makes sense since Blu-Ray players play standard DVDs. One thing I will say is that I got choppy playback with one brand of DVD, so beware of that.

The Mac workflow we're using right now is capture in Final Cut, reverse telecine in JES Deinterlacer, then rotate 180 degrees in MPEG Streamclip (we use a 35mm adapter so the footage is upside down). Both are free and work well, but it takes awhile. I wish I could do both steps with one program. I tried Compressor but couldn't make it happen.

[jroyale]

Vsansal,

Is it possible for you to go into a little more detail on your workflow along with an explanation as to why you go through each step?

I just picked up a hv20 and I'm getting mighty confused on the 24 frame thing. I love the look so I would like to preserve that. I'll be using the rendered footage on a HTPC so I'd like to keep it hi-def as well.

I did use tmpgenc and something odd happened...when I opened the source file it said it was 29.97 fps even though I had 24 frames selected in the camera. Is this normal?

Also, I'm not sure what format I should be outputting and what codec to use for best results.

I think that's about it...I have vegas 8 pro. I have downloaded the file to my comp via HDV split. My original method was just to pull the m2t file into vegas and then output but it sounds like there is a deinterlacing issue here.

All your help is greatly appreciated.

thanks.

[jroyale]

I just got off the phone with a canon rep. Interestingly and to what I thought, it appears people are going to great lengths when doing the 2-3 pulldown etc to keep the 24 frames. The camera is actually not capturing at 24 frames but at 29.97 and making it look like 24. I'm sure other people know this but I'm confused as to why people are going to great lengths to encode / recode this file when it appears that you can simply use tmpgenc to encode it to whatever format you like...then pull it into any software program or simply pull the m2t file into your program.

Maybe somebody else could shed some light onto this...am I over simplifying?

So then all I really want to do for the time being is encode the m2t file into something watchable with very good quality? If I'm using TMPGEnc what file type and codec should I be using for hi-def?


Thanks.

[vsansal]

Vsansal,

Is it possible for you to go into a little more detail on your workflow along with an explanation as to why you go through each step?

I just picked up a hv20 and I'm getting mighty confused on the 24 frame thing. I love the look so I would like to preserve that. I'll be using the rendered footage on a HTPC so I'd like to keep it hi-def as well.

I did use tmpgenc and something odd happened...when I opened the source file it said it was 29.97 fps even though I had 24 frames selected in the camera. Is this normal?

Also, I'm not sure what format I should be outputting and what codec to use for best results.

I think that's about it...I have vegas 8 pro. I have downloaded the file to my comp via HDV split. My original method was just to pull the m2t file into vegas and then output but it sounds like there is a deinterlacing issue here.

All your help is greatly appreciated.

thanks.

When you import the file(s) to TMPGenc it is normal for them to be 29.97 because of the pulldown. What you need to do is:

1. Click "Filters" button
2. Make sure "Deinterlace" is checked
3. Change Deinterlace Mode to "24fps special animation"
4. Change Deinterlace Method to "Inverse Pulldown" (you can save the template to use it later, once you save it all you need to do this when you add files ,instead of clicking filters click clip group filter and choose the one you save, it will convert all of the clips to 24p, then follow the next steps )
5. Click "OK"
6. Click "Format" button
7. Chose "HDV format MPEG file" (double click it)
8. Click "MPEG output" in 'non-standard settings' box on the right
9. Click "yes" when the warning box comes up (who reads those anyway?)
10. Change framerate to '23.976'
11. Change DC Component to '9 bit'
12. Change Display Mode to 'progressive'
13. Save as template if you'd like
14.If you don't want all of your files to be encoded as 1 clip select encode each clip as separate clips (it may say it in different words but you will get it)
14. Go to Encode and click start

Once this process is finished, import all these files to vegas. These files will be progressive and 24 frames/second. So while rendering in vegas, select 24p as frame rate and progressive instead of de-interlaced. If you use 60i, directly import everything to vegas, while rendering select progressive 30p and blend the fields for de-interlace method.

[boostcpa26]

When you import the file(s) to TMPGenc it is normal for them to be 29.97 because of the pulldown. What you need to do is:

1. Click "Filters" button
2. Make sure "Deinterlace" is checked
3. Change Deinterlace Mode to "24fps special animation"
4. Change Deinterlace Method to "Inverse Pulldown" (you can save the template to use it later, once you save it all you need to do this when you add files ,instead of clicking filters click clip group filter and choose the one you save, it will convert all of the clips to 24p, then follow the next steps )
5. Click "OK"
6. Click "Format" button
7. Chose "HDV format MPEG file" (double click it)
8. Click "MPEG output" in 'non-standard settings' box on the right
9. Click "yes" when the warning box comes up (who reads those anyway?)
10. Change framerate to '23.976'
11. Change DC Component to '9 bit'
12. Change Display Mode to 'progressive'
13. Save as template if you'd like
14.If you don't want all of your files to be encoded as 1 clip select encode each clip as separate clips (it may say it in different words but you will get it)
14. Go to Encode and click start

Once this process is finished, import all these files to vegas. These files will be progressive and 24 frames/second. So while rendering in vegas, select 24p as frame rate and progressive instead of de-interlaced. If you use 60i, directly import everything to vegas, while rendering select progressive 30p and blend the fields for de-interlace method.


ok i understand when i use 24p but i dont understand what 60i is & where it comes from? sorry im a newbie

[vsansal]

60i is the NTSC standard. i means the video is interlaced (first odd then even lines are recorded) vs p which is progressive (full frame). When you use 60i mode it records a total of 30 full frames as odd and even which makes it 60 interlaced frames. But when you use 24p mode, due to HDV compression specs it still needs to record it as 60i so the camcorder uses a method called pulldown which makes it possible to record 24p signal inside a 60i signal. So if you don't remove the pulldown your video will be 60i even if you record in 24p mode.

[vsansal]

The workflow and the time sheet from my last video:

This is for 1 hour video

*Capture using sony vegas 8 1 hour

*Remove the pulldown using TMPGenc 4 Xpress 3.5 hours

*Import the video back into Sony Vegas ,render as 1920x1080/24p HD mpeg2 video with the highest video quality setting and preserving 25Mbps bitrate and using Magic Bullet Look suite for CC
11 hours

*Using Ulead DVD movie factory 6 plus burn it as HD-DVD to regular DVD-5 or DVD-9

3 hours

Total for 1 hour video: 17.5 hours

My computer specs:

HP dv9500z laptop with AMD dual core 2.2GHz, 4GB RAM 5400rpm hard disk, running windows vista ultimate. During the pulldown both of the cores were used nearly full performance all the time but during sony vegas rendering they weren't fully used for some reason. I guess it just takes time to do the rendering.

[jroyale]

Vsansal,

You've been very helpful on this front and I know I can speak for everyone when I say thank you very much. The 24p thing can be very confusing. One more quick thing...when you render in vegas what would you use if you were to view it on your computer. In other words I'd like it to be the highest quality possible at 1080 but with decent compression. What file type and codec would you use?

Thanks again!!

[vsansal]

If you are only going to watch it using your computer, than I would suggest you to keep it as it is which is mpeg2 and burn it to a DVD as data disc. What I do is I first select blu-ray 1920x1080 24p or 60i. Then get into custom settings. I choose the best quality for rendering and in video settings I choose the highest quality. I keep the constant 25Mbps bitrate. Don't forget to include the audio from the audio settings (as default it is not included). Then render it. If you want it to be a small file to keep it in your computer than you can use WMV with at least 3Mbps bitrate with 1080p/i resolution to get decent results.

[jroyale]

The program I'm using doesn't understand m2v which is the default extension under your instructions. I can just rename the extension mpg or will that effect the render?

There doesn't appear to be much compression there. I'm reading up on it and according to microsoft the wmv9 is comparable to h.264....may be up for debate but the results look good and although it's not all the compression I was looking for 1 min vid results in about a 50meg file which isn't terrible I guess. I'll be keeping most of my tapes but I'll want tot do a few montages and keep them on the comp so this may be the best way to go.

Thanks for you help.

[vsansal]

Are you talking about playback?

BTW, in custom settings, if you go to system and unselect save the audio as a separate file, it will output it as m2t instead of m2v

[jroyale]

Ya, playback is what I'm most concerned with. I have an HTPC set up downstairs and the program I'm using (j. river media center) does not understand m2t or m2v extenstions plus I did want some compression without noticeable loss in quality.

I have noticed motion isn't great. I have a clip of me walking around the house to test it and it become fairly blurry at points. But this may also be a property of the camera.

Oh, one quick thing. The workflow that you told me about, just to be clear that will maintaint he "film" look as it is designed in the camera correct?

[HBlack]

I don't know for sure, but blur with motion would indicate to me that you need to change your shutter speed, manual focus, or give it better conditions for auto focus.

[jroyale]

Good point on the shutter speed. DV is new to me...I'm a still camera guy and there is a manual shutter speed. It was lower light so I'm assuming the shutter speed was very low to allow more light in. From what I understand these cameras are very sensitive to lower light scenarios.

[vsansal]

The forkflow will maintain the film-like look.

The compression I suggest is the same as the recorded one so you should get quicker rendering times using it. I don't like AVCHD format which can reduce the file size but I like the quality of mpeg2 better. You can try WMV with a high bitrate. Just make sure which ever compression you use, go into the custom settings and change the quality settings to the highest possible and select proper frame rate (23.97) and aspect ratio.

[HBlack]

I have Premiere Pro 2.0, after effects CS3, and HDVsplit. How should I do the reverse telecine for 24p?

When I import 24p into a 30p/60i project it has the occasional doubled frame to make it 30 frames per second. Once I do the 24p pulldown removal, will I be able to edit with 24fps?

Also, what's the point of HDVsplit? I downloaded it thinking it was the only way to capture, but it started giving me difficulties. I tried capturing with PP2.0 and it works fine.

[vsansal]

I have Premiere Pro 2.0, after effects CS3, and HDVsplit. How should I do the reverse telecine for 24p?

When I import 24p into a 30p/60i project it has the occasional doubled frame to make it 30 frames per second. Once I do the 24p pulldown removal, will I be able to edit with 24fps?

Also, what's the point of HDVsplit? I downloaded it thinking it was the only way to capture, but it started giving me difficulties. I tried capturing with PP2.0 and it works fine.

You can use after effects to remove the pulldown.

The fork flow is:


In the Project panel, select the footage item from which to remove 3:2 pulldown.
Choose File > Interpret Footage > Main.
In the Fields and Pulldown section, select Upper Field First or Lower Field First from the Separate Fields menu.
To have After Effects determine the correct settings, click Guess 3:2 Pulldown


HDV split is a free tool for only capturing. You don't need to use it if you can capture fine with your NLE.

[jroyale]

Is there any known issues with renaming the m2t file as an mpeg? WMP, wich means my media center program, understands it and plays it but will there be issues down the road if I do this?

[vsansal]

Instead of renaming try this:

Right-click on a .m2t file in Explorer and choose Properties. Beside "Open With..." click the Change button and then locate the Windows Media Player.exe (It should be in Program Files, and then the Windows Media Player directory).

After you do this, without renaming it you should be able to watch them.

if it doesn't work I don't think you will have problems renaming since they are both mpeg files only with different extensions.

[jroyale]

Vsansal,

Looks like after I do the Tmpgenc thing I'm going to try and use super to convert it to h264. Have you used this? I again want to keep the film look so I guess in frames per second I should keep the 23.98 or 25?

[vsansal]

I've never used conversion, but if you are planning to make a DVD out of it, don't convert it to h264 since you will be converting it back to mpeg for dvd. As for the frame rate, don't change it to 25 since it is the frame rate for PAL camcorders, keep it as 23,98.

[jroyale]

I'm still struggling with how to convert this and get it a good size...quicktime buffer issue, super doesn't work and when using nero i convert to avc, works fine but it doesn't play....it just gets stuck on the first frame.

I'm also hanging on the same spot every time I use tmpgenc for a certain video...obvious this is an issue with the m2t video file but I can't resolve it.

Check that...no issue with the m2t only when I encode several files at once. I did them seperately and it worked fine. It's also doing it to another batch I'm running now. Just hangs...sucks.

Is there any other program that can do what tmpgenc does?

[jroyale]

Okay, we'll I've answered my own question. It appears it's a memory issue. If you try and encode too much at once your system just freezes it. So I used the batch process and it worked fine. Question though, what happens if you have 10 ro 20 minutes of consecutive tape and you want to encode it? I guess you would have to split it then reassemble it eh?

[Outlawimpy]

How much memory do you have on your computer?

[jroyale]

I have 2 gigs. After about 5 to 7 minutes of encoding I'm at 1.6 to 2 gig of mem usage for tmpgenc. I recall this for vegas as well, if the video is over a certain length you have to split it. There is a way to do it in Vegas...not sure how you would split and reasseble here.

Also, I've been trying somehow to convert my files to h264. Super doesn't work and when I try to rename m2t to mpeg and import into Quicktime, it has a bufferoverun. Without rendering / encoding twice, how would you suggest I encode with tmpgenc then import into Quicktime?

Thanks.