HDV timeline footage to SD (16:9 or 4:3?) NTSC dvd spec MPEG-2 file

[Canon GL-2 Guy]

Gents, I need your help, suggestions and advice once more.

About 2 months ago, I belatedly and finally entered the HDV world by purchasing a Canon XH-A1s which all in all I’m very pleased with. Due to time constraints, limited processing resources and the complexity of the project involved, it was only earlier this month that I submitted an HDV program to be aired on local access cable.

Well, it turns out that I was under the false impression that their fairly new server that feeds their broadcast system could handle MPEG-2 files with HDV content! Instead, it’s very picky & not the case. Their head tech guru wrote me an e-mail stating that all they got was a black screen (which is embarrassing to all concerned) and that from now on I either submit a NTSC dvd spec 720 by 480 file or my program is off the air.

My Premier Pro project (only v3.2) concerned (currently on the timeline) is 1440 by 1080 shot in 60i with an FPS of 29.97. So now I’m required to downgrade it to SD & do my best to ensure that quality doesn’t take a tremendous hit. The mission is to take this project direct from the timeline (it’s finalized + fully edited) and make an MPEG-2 dvd spec file. Bottom-line, how is this best achieved & by what process/methodology or workflow?

My twin goals are to:

a) Preserve the aspect ratio since I take very tight shots, so a crop and resize probably would do more harm than good. What kind of an option is SD 16:9 over 4:3? What are the pros & cons of each? I don’t want to unintentionally crop off part of a body part for example.
b) To the extent possible, maintain image/footage quality so that it remains watchable. I’m looking for a workflow that will retain quality within reason.

What approach and solution would you take if you were in my shoes?

Am sure others have traveled this route before and would appreciate all input and pointers to mull over and try to apply them not only to this project, but those in the future as well.

Thanks so much.

[edDV]

Cable access channels are usually 4:3 analog NTSC. I've yet to see one that allows HD resolution.

Not to worry, you need to export to DVD standard from the Adobe MPeg2 encoder. Since your project timeline is 16:9 and you don't want to re-edit for 4:3, your option is 16:9 letterbox in a 4:3 frame. To get this, specify a 4:3 display in the encoder settings.

[Canon GL-2 Guy]

Any other thoughts on the matter?

Any suggestions with Cineforms Neoscene (or some other immediatory codec), export to tape and then reimport back into the computer or export as a movie in some form as part of the solution?

Is it possible to get an .avi file and trick Prem Pro somehow that it does not contain footage derrived from HDV source material?

Thank you.

[aedipuss]

encode the HDV straight to 4:3 mpeg2. that hard codes the black bars top and bottom. pan and scan 4:3 sucks for material shot 16/9. it should look fine on SD tv. besides anyone that still has a 4:3 SD tv still must be used to losing half the screen to black bars. HD tv users can adjust their screen format to get rid of them.

your ideas don't make any sense.

[Canon GL-2 Guy]

Well, what do you think about changing the project to NTSC widescreen and rendering to 16:9 widescreen?

[edDV]

Well, what do you think about changing the project to NTSC widescreen and rendering to 16:9 widescreen?

Explain again what you are trying to do?

16x9 on cable access will be letterbox.

[Canon GL-2 Guy]

I have a limited number of options within Prem Pro:

What I don't want is a small mediocre picture with black strips top and bottom.

My goal is to fill the screen to the fullest extent possible without too much quality degradation.

If I use the DV NTSC "widescreen" template and tweak the settings. the picture is bigger however the loss of definition,sharpness and detail on out put is huge at least in my eyes.

Elsewhere online I was told that I'd be better off shooting in 30p and not 60i.

Canon XH-A1s frame modes are 60i, "30 F" & "24F".

What's the difference (if any) between 30F and 30p?

Does anyone here concur that my results in SD/DVD-MPEG-2 file from HDV source footage would be better if I switched from 60i to 30F?

If so, why is that?

Just trying to understand and explore options.

Thank you.

[edDV]

...
What I don't want is a small mediocre picture with black strips top and bottom.

My goal is to fill the screen to the fullest extent possible without too much quality degradation.

If I use the DV NTSC "widescreen" template and tweak the settings. the picture is bigger however the loss of definition,sharpness and detail on out put is huge at least in my eyes.

Are we to assume your only goal is to optimize for an analog NTSC channel? If so, you would shoot for 4:3 aspect ratio. Two ways to do that.

1. Shoot in 4:3 DV mode and edit 720x480. The Canon XH-A1 will look as good as the older pro SD broadcast cameras and outclass most other cable access productions.

2. Shoot 16:9 HDV but with the assumption you will be chopping the sides off for the 4:3 SD version. This is how most TV series are shot with markers on the viewfinder to allow framing for both 16:9 and 4:3 release. This allows you to make an HD 16:9 master version, then you do a second edit where you crop the sides for 4:3, then downsize and encode to 720x480 4:3 DVD.

True, the SD version will have lower resolution by definition.

Elsewhere online I was told that I'd be better off shooting in 30p and not 60i.

Canon XH-A1s frame modes are 60i, "30 F" & "24F".

What's the difference (if any) between 30F and 30p?

Does anyone here concur that my results in SD/DVD-MPEG-2 file from HDV source footage would be better if I switched from 60i to 30F?

If so, why is that?

This depends on the type of video you are shooting. For handheld or high movement subject matter like sports, dance or news/documentary, 60i is usually best. 30p will have half the motion samples and will appear jumpy. This is why all local TV news type shows are shot 60i *.

If you are shooting talking heads from a tripod say "60 Minutes" style, or if you are doing lots of low animation graphics, then you can try 30p (30F). I wouldn't suggest 30p for handheld unless you can hold the camera extremely steady. The reason for this is the background will jump around taking viewer attention off the central subject matter. The way the pros work around this problem for 30p or 24p is to use narrow depth of field lenses that blur out the background to keep focus on the foreground subject. Also in 30p, any pans or zooms will look jumpy so should be avoided. You set your framing for each shot before calling action.

24p (24F) should be avoided unless the goal is to release for film projection or to shoot for both "NTSC" and "PAL" distribution. The latter reason is why TV series are shot 24p. 24p style shooting is taught in film school and is best left to the pros.


Now the technical:

60i (actually 59.94i) is 60 fields per second. This matches 480i/1080i ATSC/NTSC broadcast. This allows 60 motion samples per second and highest resolution for slow moving scenes. Disadvantage is the need for high quality deinterlace for progressive TV display or for internet encoding.

30p (actually 29.97p) is 30 progressive frames per second. Advantage is no deinterlace is required for computer or internet distribution. Disadvantage is 30p is half motion sampled (jumpy) and must be interlaced for analog NTSC or 480i/1080i ATSC distribution.

30F (actually 29.97F) is 60 fields per second but with both fields sampled in sync at a 30 fps rate. Advantage is 30F can be distributed as 60i. 30F can be converted to 30p with a simple weave deinterlace. Disadvantages are the same as 30p.

24p (actually 23.976p) is 24 progressive frames per second (film rate). While compatible with ATSC broadcast, nobody is broadcasting 24p yet although Blu-Ray offers a 24p output. To broadcast 24p as 480i/1080i, extra pad field repeats are added (called telecine or pull down) to build the frame rate up to 29.97. Disadvantage is slow-fast-slow-fast motion "judder" when displayed at 60i unless a "120Hz" display is used. To broadcast 24p as 480p/720p, frames are repeated in a 3-2-3-2 pattern to build up to 59.94p.

24F (actually 23.976F) is 24p telecined (field padded 3:2:3:2) to 60i.

24pA (actually 23.976pA) aka "24p Advanced" uses a 3:2:2:3:3:2:2:3 telecine pattern. Advantage is a simplified inverse telecine for professional 24p editing. Disadvantage is extreme judder if viewed as 60i.


*Except for digital 720p channels that shoot 59.94 full frames per second.

Note: For those in PAL areas, this could be re-written as 50i, 25p. In the PAL world, 24p is sped up 4% to 25p (with or without sound processing) and a 2:2 telecine pattern is used to build up to 50i.

[Canon GL-2 Guy]

Shoot in 60i vs 30F(p) or: 60i + 30F(p)?

Just over 2 months ago I bought A1s, so as I still borrow the A1 from the local cable station, it gave me an opportunity to include my wife to operate one of the cameras. To date we both have shot in 60i and am considering change. The venue remains the same: inside of a large church interior. I’m positioned to film the pastor more or less straight on and my wife catches him straight on dependent on which way he turns his head. We‘ve got him ”covered” one way or the other until he abruptly gets off his stool and starts pacing around the platform.

So, with no advance notice, unpredictably on a whim, the scenario changes. Yes, tripods are used and the last couple of weeks I’ve been lucky as he has stayed on the stool (so 30F/’p” could have been used) however, that won’t last forever. So here’s my question. I need to be prepared & not get caught totally off-guard with the goal of obtaining relatively smooth coverage of those transition points. Can 30F (p) successfully handle those occasions when he rather abruptly shifts positions and starts to traverse the stage all over? That’s one of the reasons to date I’ve already shot in 60i.

I’ll just throw this out there. What if one of the cameras covered him while on the stool in 30p and the other covered him to and fro on the platform in 60i? Would those 2 frame rates mesh and footage be mixable?

I confess I have never shot with markers even though my needs have now been changed to shot the master in HDV 16:9 and optimize that footage for 4:3. Also, only a couple of times since last June I shot in SD with my old GL-2’s because I had not other choice & would rather shoot footage at it best capability.

There are a few times when I zoom slowly or pan around the audience in order to add variety which from your description may look on the ragged side.

I’m wondering if there might be a better path other than Prem Pro. Perhaps I could export the timeline contents with a codec (which you’d recommend) to an .avi and then use a external MPEG-2 encoder (within reason, also recommended) which perhaps would be suitable and yield superior/satisfactory results.

Premiere Pro basically has 2 options for rendering to the cable stations mandatory analog NTSC SD/dvd spec depending on pixel aspect ratio selected:

a) Std (i.e. 4:3) results in black strips top and bottom and a small picture (compared to the 1440 x 1080i original) inbetween as one would expect.
b) Widescreen 16:9 The picture is almost as big as the original and as far as detail, sharpness and definition is concerned it look about the same as 4:3.

Given the above, all things considered would it be better to shoot in 60i,30F (p) or a combination? Which would translate to a better looking SD MPEG-2 file?

Thanks for taking the time to respond in depth.

[aedipuss]

30p looks identical to 60i except in fast motion scenes. a pastor walking around is NOT fast motion. 30p would work fine and give the option for web video. pp has the same mainconcept mpeg-2 encoder as most other semi-pro editors. it's fine. if you had to really compress 3 hours of video to a single layer dvd then maybe a 9 pass cce encode would be in order.


most local tv wants 4:3 SD video. option "a" or shooting 16/9 /pan and scanning to 4:3 are your only choices.

i have edited/encoded hundreds of sermons for the local community tv station here in nh and it is limiting.

[edDV]

Shoot in 60i vs 30F(p) or: 60i + 30F(p)?

Just over 2 months ago I bought A1s, so as I still borrow the A1 from the local cable station, it gave me an opportunity to include my wife to operate one of the cameras. To date we both have shot in 60i and am considering change. The venue remains the same: inside of a large church interior. I’m positioned to film the pastor more or less straight on and my wife catches him straight on dependent on which way he turns his head. We‘ve got him ”covered” one way or the other until he abruptly gets off his stool and starts pacing around the platform.

So, with no advance notice, unpredictably on a whim, the scenario changes. Yes, tripods are used and the last couple of weeks I’ve been lucky as he has stayed on the stool (so 30F/’p” could have been used) however, that won’t last forever. So here’s my question. I need to be prepared & not get caught totally off-guard with the goal of obtaining relatively smooth coverage of those transition points. Can 30F (p) successfully handle those occasions when he rather abruptly shifts positions and starts to traverse the stage all over? That’s one of the reasons to date I’ve already shot in 60i.

I’ll just throw this out there. What if one of the cameras covered him while on the stool in 30p and the other covered him to and fro on the platform in 60i? Would those 2 frame rates mesh and footage be mixable?

My first question is what is wrong with shooting 1080 60i? The main benefit of shooting 30F would be better compression for internet distribution. But you could deinterlace on export from a 60i timeline. Remember the cable broadcast will be 60i no matter what you do.

I wouldn't mix formats but it could be done. You could set a 30p project and import 60i or 30F to the 30p timeline. Note that this would need rendering for preview and will generate very large 30p tmp files on the scratch disk. Both 60i and 30F would need to be rendered for preview where neither would need rendering for a 60i timeline.

I'm just failing to see benefit in shooting 30F.

I confess I have never shot with markers even though my needs have now been changed to shot the master in HDV 16:9 and optimize that footage for 4:3. Also, only a couple of times since last June I shot in SD with my old GL-2’s because I had not other choice & would rather shoot footage at it best capability.

There are a few times when I zoom slowly or pan around the audience in order to add variety which from your description may look on the ragged side.

I’m wondering if there might be a better path other than Prem Pro. Perhaps I could export the timeline contents with a codec (which you’d recommend) to an .avi and then use a external MPEG-2 encoder (within reason, also recommended) which perhaps would be suitable and yield superior/satisfactory results.

Premiere Pro basically has 2 options for rendering to the cable stations mandatory analog NTSC SD/dvd spec depending on pixel aspect ratio selected:

a) Std (i.e. 4:3) results in black strips top and bottom and a small picture (compared to the 1440 x 1080i original) inbetween as one would expect.
b) Widescreen 16:9 The picture is almost as big as the original and as far as detail, sharpness and definition is concerned it look about the same as 4:3.

Given the above, all things considered would it be better to shoot in 60i,30F (p) or a combination? Which would translate to a better looking SD MPEG-2 file?

As said above, I would do the following.

1. Place transparent tape or plastic over the LCD viewfinder sides to mark the area outside the center 4:3 frame. Try to keep action inside the center 4:3 area,

2. Set an HDV 1440x1080 60i project. This will work for 60i or 30F camera import without need for preview render. Edit and export the wide version to a 1440x1080 60i 16:9 HDV (MPeg2) edit master file.

3. Next you need to crop the sides off leaving the center 4:3 1080x1080 area, then linearly downsize the result to 720x480 60i.

This could be done in AVISYNTH. I'll let others suggest the best script.

It could also be done in Premiere Pro but I don't have the program here to give detailed instructions*.

4. Set a 4:3 720x480 60i project. Import the downsized version, then encode as a 4:3 DVD.
Note: If you use the DV project template, you will need to change lower field first to top field first.


* Premiere Pro Method (from an HDV 1440x1080 60i project setting)
Note: I did this in Vegas Pro, it should also work in Premiere.

1. Horizontally squeeze the picture from 1440 to 1080 pixels.

2. Edit project pixel aspect ratio from 1.333 HDV to 1.0 square. This results in a 1440x1080 square pixel 4:3 aspect ratio picture with sides cropped off.

3. On export, specify 4:3 720x480i 29.97 top field first DVD MPeg2.

[Canon GL-2 Guy]

Aedipuss:

Well, we are neighbors (almost) as I reside in North Central Mass. Thanks for sharing the fact that you work in a cable tv environment so I know you can truly appreciate what the goal and the options are.

Are any of the sermons that you encode originally submitted as HDV or even full HD footage? If so, what workflow/process and programs do use to down convert down size to DVD? Curious as to which NLE you use. Are you PC based, Apple or both?

In your professional opinion & experience, will the end product (SD MPEG-2 video) when viewed, look appreciably better if I switch from shooting from 60i to 30F/(p) from now on? Does you experience include Prem Pro per chance?

Lastly, is there any MPEG-2 encoder found on this website that may do a better job than the one in CS3.2? What do you use for your own personal projects?

Thanks for chipping in and continuing the dialogue.

Am eager to digest, learn and move forward.

[Canon GL-2 Guy]

Hi edDV,

Now it’s your turn. To answer your 1st question, there’s certainly nothing “wrong” with shooting in 60i and to continue doing so like I have since HDV day one. The reason why I brought subject up in the first place is because my circumstances have changed & need to quickly comply (or else). On another forum someone suggested (rightly or wrongly) that my dvd SD result would look better if I stopped shooting in 60i & switched to 30p (p) so I’m trying to determine if that advice was accurate or way off base and to obtain independent confirmation from other sources and get professional direction. Do you concur or differ?

In Premiere Pro’s Adobe Media Encoder module, it does give me the option to select de-interlace as one of the export settings on the output tab. I’ve wondered how that would be beneficial, however never tried it to date.

In laymens terms, if I ticked that box off, what visual impact would it have or effect on output quality as SD video or in HDV for that matter? Should I shoot in 60i and select de-interlace when exporting/compressing? Please, what are your thoughts on that?

From your .png’s, it looks like to me that the parameter “stretch to fill the frame” is very similar to Pros select the pixel aspect ratio which in essence does much the same thing.

Good for you that your NLE has the event/crop option. I’m not that sure I have such a luxury. Due to cropping and downsizing, I’d always have to pay extra careful attention to framing while shooting. Very convenient for sure!

Thanks for the outlining you workflow, you use Vegas right?

Do you use an external MPEG-2 encoder or the one embedded in your NLE?

[aedipuss]

for the tv station, i get delivered miniDV tapes shot in SD 4:3 so my job is pretty straight forward. capture, edit, encode, deliver mpeg-2 files on a hard drive. i am a pc vegas pro user for the most part, but have used editstudio occasionally for multicam projects, and have the adobe suite for odd things like creating video from stills that pp is much better at.

in your case if the money is ok, i'd shoot HDV 30f for online distribution, if not put the cam in miniDV mode and shoot for SD dvd, as it will take far less of your time. if you shoot HDV just make sure to keep the subject centered at all times, but with 2 cams you can most always find a version with him that will be ok. capture the HDV 30f, make your online video. go back on the timeline and pan/crop to 4:3 and export dvd mpeg-2 with cbr 8000kbps. the mainconcept encoder in pp is fine. i only resort to using cce sp2 for extreme compression/low bitrates that require many many vbr passes to look acceptable.

personally i have pretty much gotten into the habit of shooting with canon hv30s in 30f all the time. i do local plays, musicals, concerts, and sporting events.

here's a clip from a hockey game shot HANDHELD in HDV 30f. this is the youtube encoded 720p version of the 1080p i uploaded. people claim 30p is "jerky" but i just don't see it. the only jerks are from the camera lens stabilization when i get bumped by fans next to me in the bleachers. i've done plenty of sports 30f. i see no problem in videoing a preacher walking back and forth on a stage in 30f. it's available here to check. http://www.youtube.com/watch?v=9tJ40_S3bRA switch to 1080 or 720 HD as the 360 version is the usual yt bad.

[edDV]

My Premier Pro project (only v3.2) concerned (currently on the timeline) is 1440 by 1080 shot in 60i with an FPS of 29.97. So now I’m required to downgrade it to SD & do my best to ensure that quality doesn’t take a tremendous hit. The mission is to take this project direct from the timeline (it’s finalized + fully edited) and make an MPEG-2 dvd spec file. Bottom-line, how is this best achieved & by what process/methodology or workflow?

My twin goals are to:

a) Preserve the aspect ratio since I take very tight shots, so a crop and resize probably would do more harm than good. What kind of an option is SD 16:9 over 4:3? What are the pros & cons of each? I don’t want to unintentionally crop off part of a body part for example.
b) To the extent possible, maintain image/footage quality so that it remains watchable. I’m looking for a workflow that will retain quality within reason.

First, cable access here (Comcast) requires producer training brefore equipment can be borrowed so blame the training if you thought 16:9 HD was the local standard.

You don't mention web/internet distribution. 30F to 30P conversion would be slightly best for that.

Cable access broadcasts 704x480 60i converted to analog NTSC. If all you want to do is hand a tape for NTSC broadcast, shoot in DV 4:3 720x480 60i mode and be done with it. Isn't that what they covered in the cable access class?

I think I answered your questions. Don't zoom tight, keep it centered for 4:3.

If your cable access training didn't cover all this you should complain about your tape and program being rejected for local broadcast. Usually the submission standards are clearly stated. I'd be happy to talk to the local operator in your behalf.

[Canon GL-2 Guy]

Ed DV,

You make assumptions which did not occur. I do not blame the station for lack of training or knowledge. The only training I've had in the past 5-6 years have been a few brief sessions in Prem 6.5 and a few minutes with the A1, that's all. I have no interest in web distrabution or reverting back to SD ever. I appreciate you answering some of the questions however I can tell by the tone of your last post that your patience is wearing thin and it shows so please consider this thread closed. Thank you.

[edDV]

Ed DV,

I have no interest in web distrabution or reverting back to SD ever

Then you won't get your church service accepted for cable access. It must be converted to SD.

[Canon GL-2 Guy]

Aedipuss, thanks so much for your input.

You are truly blessed to have those encoding tools at your disposal as their pricetag is out of my budget.

I thought the hockey clip looked decent enough considering everything and an somewhat familar with the HV 30 as on a few occasions I've borrowed one myself from the cable station for the purposes of exporting to tape.

Thanks also for mentioning editstudio which interests me as now I need to do multicam editing on a regular and consistent basis. Premiere Pros built-in multicam capabilities are ok but not great.

Hope all goes well with you. We actually had brief showers of hail late this afternoon!

[Canon GL-2 Guy]

Ed DV,

I have no interest in web distrabution or reverting back to SD ever

Then you won't get your church service accepted for cable access. It must be converted to SD.

Of course I meant shooting in SD which was suggested.

I will always shoot in HDV in 16:9 and be forced to downpedal to SD.

[aedipuss]

editstudio is an older program, i haven't tried multicam editing anything but miniDV footage with it. not sure it's up to HDV mulitcam, i'll have to check someday to see. for DV it's good, i always exported miniDV as the final product from editstudio and brought that into vegas for final finishing and rendering to DVD spec mpeg-2, i just like e.s.'s multicam editing style more than vegas's or pp's.

yeah we had a weird weather day yesterday, snow, sunny and warm, then some strange "popcorn" shaped snow on and off. we still have 3 foot snowbanks beside the driveway here....