HDV to DVD

What does anybody suggest for a conversion from 1920x1080i to a 720x480 DVD?

I've tried a couple of different down-renders in FCP 6 using Quicktime Conversion--which doesn't seem to require a Ph.D. like Compressor...sorry, but it's too close to deadline to try Compressor first, esp. since we've only installed FCP 6 over the weekend.

Rendering out using Quicktime Conversion to NTSC 720x480 16:9 produced a video in which the screen size appears correct, but the video image is squeezed into a box in the center. Rendering as NTSC 720x480 4:3 produced a video in which the screen occupies EXACTLY the same real-estate as the image at the center of the 16:9...and the video is equally squeezed.

Both had the setting 'Maintain aspect ratio (using Letterbox, if necessary)' selected.

This is getting frustrating.....

-- the SASS Man

Oy. It's too close to deadline to do it the easy way?

Export a self-contained QuickTime movie from your Final Cut timeline. Drop that movie into Compressor. Choose the built-in DVD preset that fits your run time. Drop the resulting files into DVD Studio Pro, and you're off to the races.

What the difference between exporting using compressor and exporting as a quicktime then dropping that file into compressor?

Dk

Among other things, using Compressor doesn't tie up your edit system.

Uh...are you sure it doesn't tie up FCP??? It seems to on mine...it won't let me work with it. Is that because I'm doing the Export using Compressor rather than opening Compressor and loading my QT file? If so, I didn't know you could do that.

Anyway, what I do is export my project to a single QuickTime file (less room for error with rendered files, filters, etc.). I put my new QT on a timeline and do a print to video to my HDV camcorder. After that, I rewind the tape on my Sony Z1U camera, turn on the iLink conversion and set the display output to squeeze (keeping in mind that I power off and unplug the firewire cable before and after I do any changes in the menu). I play the footage on the camera and record on my set top DVD recorder. You will see that the footage is squeezed (I'm assuming you wanted anamorphic, but you can set it to letterbox if you want to skip many of my remaining steps). I finalize my DVD and put it into the computer. I use MPEG Streamclip (www.squared5.com) to open the DVD (yes, I also do the thing where it checks for timecode breaks), and then I demux to a M2V and AC3. The M2V and AC3 can be imported into DVD Studio Pro, where I make a nice menu (or you can just have the video as a first play) and change my aspect ratio of the video to 16:9 letterbox (this makes it anamorphic), and burn my new DVD.

This takes a lot less time than it does to render out an HDV timeline into Compressor. Plus, this way you have an HDV master tape as well, which I like to keep for when we offer BluRay - I'll just recapture my edit master and not have to redo everything or recapture everything.

Casey

Yes, that's what I meant. My personal experience tells me that "export using Compressor" is never necessary, and comes with a lot of negatives. So just don't use that feature.

Laying your HDV-originated footage off to HDV tape, then having the tape down-convert to DV, then running that into a hardware MPEG-2 encoder is going to give you craptacular results.

The first thing to remember is that you shouldn't cut HDV footage into an HDV timeline unless you have to go back out to tape ? and I can't think of a scenario right now in which you'd ever have to record back to HDV tape for anything, since it's not a mastering or delivery format. You should edit your HDV footage into a ProRes timeline, then keep the self-contained ProRes Quicktime you export as your digital master. That ProRes Quicktime then goes to Compressor for scaling and encoding for DVD.

Actually, I think that my method looks better than Compressor...don't know why, but Compressor doesn't look as good to me. I'm using the 120 minute best quality setting (as my projects are typically around 120 minutes).

HDV is currently my mastering format...I never thought of the ProRes method, but if I was doing that, we'd have 500+ hours of ProRes footage just sitting on the hard drives...how much space would that be? That seems like an awful lot. Isn't ProRes footage 4x larger? Did you mean capturing HDV, and editing ProRes, because the last time I tried it, I captured ProRes and edited ProRes.

I tried it once, and didn't like it because it took much longer to render things like slow motion or filters. For what I'm doing - weddings - I have to edit as fast as I possibly can to try to match the budget (which is often already low to begin with).

I've been content with my current method in regards to speed/delivery versus final output quality. Not saying that I won't try this ProRes thing, but for what I'm doing, and for who it's for, it works for me. We're not competing with Hollywood or 6-8 figure budgets. I haven't had any complaints.

Anyway, that's just what I think.

Thanks for your input! I appreciate it!

Casey

Here's why HDV is not a mastering format: Every time you record to HDV tape, the deck compresses the footage brutally. If you record via Firewire, before the footage can be laid off, Final Cut has to compress it brutally. If your footage originated on HDV, then it's already been compressed brutally once. It'd be analogous to using a VHS deck to master your shows. Sure, you can, but you lose so much just laying the footage off that it's not a good idea.

ProRes is less than uncompressed SD. It's impossible to predict exactly how big it'll be, because it's a variable bit-rate codec. Figure around 145 Mbps for regular ProRes, and somewhere around 200 or so for ProRes HQ.

If your system isn't very new, then ProRes probably isn't for you. It's optimized for modern systems, and on modern systems it blazes. But the drop-off in performance on older gear is significant. I think I've heard that you can't even reliably do real-time playback of full-res on a Power PC system.

??!!!

There once was a project that was captured HDV, then put back to HDV tape, and the customer came back later and wanted one shot changed, and the shot had faded in and out from the previous clips, so I captured my edit master and the two shots that had been faded into and out of, plus the new one, and synced up the two old clips cutting into them in the middle of the shots, added my third and put a fade between it, and they shots that I lined up and cut in the middle with half being off the edit master and one being off the original tape, I swear I could not see the difference, nor the edit...nor could anyone else I showed it to. Not at all what would have happened in the old VHS days.

I know there are people out there who can see the picture difference if you're using a 6 foot component cable to connect to a TV and a 3 foot component cable, and if you're one of them, God bless you, but I can honestly say that I don't see the difference.

I've been editing for 15+ years and if the difference is as great as you say, I would think that I'd be able to notice. I'm a little baffled here.

I'm also using a Mac Pro 2x3 gHz Intel system, which should be plenty fast for ProRes, and it's not bad, but it is noticeably slower the way I used it (captured ProRes, edited ProRes...should I be capturing HDV, editing ProRes?).

I don't want this to come off the wrong way, but I am honestly baffled at what you think I should be seeing. I have been editing since the days when generations were a really big deal, and only in recent years, the generation gap has really closed since the old analog stuff. I'm just not seeing the significant quality losses that I used to between generations...and maybe for what I'm doing, and for the budgets I'm trying to squeeze in under, it's good enough.

What kind of stuff are you doing? What formats/cameras do you use?

Thanks!
Casey

So am I blowing it by capturing HDV, editing in a HDV timeline and exporting to a HDV self contained QT to use in compressor, if my main delivery is DVD? I have FCP 6 but don't know much about the ProRes codecs, which one should I use to go to DVD.

Thanks
Edubs717

You are not blowing it.

You can work in HDV on the timeline and then use this for SD DVDs.

[www.kenstone.net]

--ken

I always have to convert Interlaced HDV to SD to get a better DVD.

I edit multicam HDV using the ProRes 422 codec to avoid long conforming times.

When I do a self contained mov and try to compress to mpg2 directly it looks terrible in my 42" plasma. When I do the long down convert and then compress mpg2 it looks better. Not HD qulity, but better.

Should I be exporting to ProRes 422 instead of selfcontained first?

Thank you in advance

God Bless,

Douglas Villalba
director/cinematographer/editor
Miami, Florida

[www.DouglasVillalba.info]
[www.youtube.com]
[vimeo.com]

When I have done this I have exported as per the article.

--ken

I'll try it that way next time.

I read somewhere that it is best to down convert 1080 to 720 then to 480. I know it sounds crazy but that's what I read.

Have you try anything like that?

I don't have any problems when I go from DVCPRO HD 720 directly to mpg2, but HDV looks worst than coming out of original DV.

God Bless,

Douglas Villalba
director/cinematographer/editor
Miami, Florida

[www.DouglasVillalba.info]
[www.youtube.com]
[vimeo.com]

I never read that. Don't know why anyone would do that. Too many converions.

Next time, try it as per the article, I think you'll be pleased with the results.

--ken

I'll give a try tomorrow. It sure a total waste of time to down-covert.

Have you tried export to compressor directly?

I know it locks FC, but if it saves a step I can always do it over night.

God Bless,

Douglas Villalba
director/cinematographer/editor
Miami, Florida

[www.DouglasVillalba.info]
[www.youtube.com]
[vimeo.com]

I don't send to Compressor, it seems to take a lot longer, sometimes there are issue with the export to Compresssor (and you have to re export) and if you save the QT self contained you can use it again later.

I think you might want to try the QT self contained at least the first time, I know that this works properly.

You're call.

--ken

> I never read that. Don't know why anyone would do that. Too many conversions.

In theory I'd agree with Ken, but in practice I've also seen problems sending HD clips directly to Compressor or DVD Studio Pro. On a trailer last year, I edited in DVCPro 720i60, off a FireWire 800 drive, and made QuickTime exports at timeline quality. If those were sent directly to Compressor or DVD Studio Pro (I can't remember whether I went through Compressor, actually), I'd get DVDs with very poor playback. Since these were just screeners to check the picture cut, I stuck the movie file into a DV NTSC timeline in FCP, exported that, and then the resulting file would have no problems going to DVD. I didn't nail down the exact problem (Drive speed? Frame rate? Codec?), but the DV intermediary allowed us to see the cut without having to go on a wild goose chase wrestling with the HD movie files for DVD.


www.derekmok.com

Hi Derek,

I have had issues when files are 'sent to Compressor', that's why I export as QT self-contained. I have not had problems with QT self-contained, I have had problems with 'send to Compressor'.

--ken

Ken Stone Wrote:
-------------------------------------------------------

> I think you might want to try the QT self
> contained at least the first time, I know that
> this works properly.
>
> You're call.
>
> --ken

That is what I use to do and it looks worst than coming from DV.
I tried doing it your way and at least it looks good in Streamclip.
A little soft, but no field problems.
Maybe I should ad a little sharpness in Compressor.

As far as sending directly to Compressor it sure takes a long time. I heard that it analyses every frame. I know that they look better going from SD to mpg2.

God Bless,

Douglas Villalba
director/cinematographer/editor
Miami, Florida

[www.DouglasVillalba.info]
[www.youtube.com]
[vimeo.com]

>As far as sending directly to Compressor it sure takes a long time. I heard that it analyses
>every frame.

Sending it to Compressor was something i used to do on SD projects. It renders straght from the source file, bypassing codec compression in your sequence. Not that there aren't any issues, there are, like text done in Boris Text on a DvcproHD 720p sequence may mess up, and it takes a lot longer on multiple cores as you can't send projects to Compressor via FCP and still use the clusters.

Here are 2 other options:

1) Create a ProRes HQ sequence at 1440x1080 (same size as your HDV clips). Copy everything from your sequence, and paste them into that sequence (Apple A on the old sequence, Apple C to copy, Apple V to paste on the new sequence). Set render to high precision YUV, render and export a self contained QT movie. Then encode that in Compressor, using one of the DVD-Best preset. turn on Frame Controls, set resizing to "best", which will then make use of Optical Flow to resize. Then encode that in DVDSP.

2) Create a DV/DV50 sequence in FCP, drag the final HDV edit sequence onto that sequence. This will automatically nest the sequence and FCP will down scale the footage to the same frame size as DVDs (720x480). Export that straight to Compressor. This will bypass the DV codec compression and footage will be rendered frame by frame from the source footage, bypassing any codec compression set in the sequences.



www.strypesinpost.com

Very interesting points here!

How do these methods affect overall speed between capturing the footage to delivering the DVD? Yes, I can edit these things together in ProRes and spend days rendering and will look fabulous, but what if I don't have time to do these things?

If you have a 30 minute HDV project that needs to be done in the absolutely fastest way possible, with the final product being put to SD DVD, but also giving the client an HD master (HDV in my case), what would you do differently?

Thanks!
Casey

>what would you do differently?

I will definitely skip exporting to Compressor off the DV50 timeline. 1, Exporting directly to Compressor will take quite a bit of time.

Doing a QT conversion and rendering in a ProRes timeline shouldn't be much different in terms of time as all the footage needs to be rendered to ProRes (it's much faster on an intel). I'm not a fan of QT conversion, as it never seemed to look as clean as exporting as anything else. But I'll switch the resizing in Compressor to "better" to save on time.

And yea, you'll definitely need to keep the final edit in a HDV timeline and render to HDV in that, as you'll be going out to HDV tape.



www.strypesinpost.com

The way I'm doing it now seems to be the fastest I can come up with.

I can go from having a finished (all rendered) 30 min. project to a DVD with a custom menu in less than 90 minutes, by putting it to tape, putting the tape to DVD, extracting the DVD, and bringing it in to DVDSP for authoring.

I have come close to achieving that kind of speed with encoding thru Compressor.

90 minutes is pretty good, all things considering.

Unless there's a different workflow altogether that will give me the same things...an HDV master and a SD DVD...but I haven't thought of one yet. That's why I'm here


Thanks,
Casey

Yeah, there's definitely a faster way. I took a 22-minute show from timeline to DVD in about fifteen minutes last week, counting burn time. Even on a less-than-capable workstation, it should at worst not take much longer than real time.

Seriously?!!!

I must be missing something here! Tell me more!

Well, I'm very glad I TOOK the advice and tried using Compressor. I did not understand the drag-and-drop functionality of the program, and being a geek, I should have known better. Stress, I guess.

Anyway, I whipped out my trusty Ph.D. (Photographic Dogma, you understand--I haven't BEGUN to fight my way up the educational version of that abbreviation), and gave Compressor a whirl.

Presto-BLAMO, I had an .AC3 / .M2V pair that burned to a BEAUTIFUL DVD display.

Compressor also solved one of our problems with capturing HD footage in 720x480 Animorphic.

So...well...thanks!

Alot!

-- the SASS Man

Greetings all,

Some interesting content in this thread.... exactly the sort of stuff that I bore people with all the time.

I regulary have success using the above mentioned workflow when creating SD DVDs....

1.Ingest Native HDV (or transcode to pro res if I'm in the mood- I don't seem to find too much difference in the cut, it just saves me rendering out pro res before export)

2. export out Quicktime (current settings)

3. Bring that file into Compressor and compress to m2v

4.Author disc in DVDSP.... creating a beautiful SD DVD

No worries.... usually

When I say usually- the majority of the work I do is short (2-3min) promo vids. However, today I have finished cutting a 60minute documentary, I've followed my usual protocol and have wound up staring in the face of a 30hr compression!

WTF? is this simply due to the length of the program?

Am I missing something?

I doubt 30 hours is right...I think 3-4 hours might be more accurate. I usually have 2 hour projects into Compressor, and I do them overnight, and they are finished in the morning (and that's at 120 minutes best quality on a Mac Pro with 2 dual-core intels), so I'm not exactly sure how long it's taken.


Casey