I need advice getting a show shot in HDV and cut in Express Pro output for best quality big screen playback.

We shot the project in HDV (Sony HVR-V1U), captured into Avid Express Pro native HDV 1080i, cut, color timed, titles, (all went smooth & fast) output back to HD device (camera) as a Quicktime file), looks very good on a 40" HD monitor, looks not so sharp on a 50' cinema screen with a new Christie Hi Def projector.  Project bound for festivals.  What to do?

Should I download DNxHD, recapture or transcode original 'in' shots to DNxHD, eye match newly captured shots onto existing timeline?

Then how to output to final format - HD Cam?

A large pro post house will do whatever it takes to get this into an HD format they support (not HDV, HD Cam), then we'll finish there. 

Any thoughts on my best workflow for top big screen playback?

Thanks.

Charles

(Pentium 4 CPU 3GHz on an MSI 865pe neo 2v motherboard with 3 GB of ram. Nvidia Quadro FX 3000 256mb 8x. Sound card M-Audio Delta Audiophile. 3 - 500 gig Maxtor IDE drives. Windows XP SP 2. Avid Express Pro 5.72)

For best HDV quality

1. Rent Sony HVR-1500

2. Capture via HD-SDI @ uncompressed 10 bit using Symphony Nitris.

3. Edit

4. Output to HDCAM or HDCAM SR.

You really need HD-SDI capabilities (at least HD-SDI out) to get good quality out of HDV. A cheaper but not quite as good as above is

1. Capture via firewire using Adrenaline with DNxCel board

2. Transcode to 220 DNxHD 10 bit.

3. Edit

4. Output to HDCAM or HDCAM SR.

What format is actually used for projection at the festivals? Is it tape? Or digital files? If digital files, I'm betting it's an m2t stream (broadcast standard for HD). If it's an m2t stream then you never need to output to tape. You would export your DNxHD timeline as QT ref and then use Procoder to make the m2t file. You should not make the m2t directly from Avid ("export to device") since that will result in a 1440x1080 file instead of true 1920x1080.

I just finished a show used at a trade show theater on a 24' screen and it was run off Grass Valley turbos which play m2t files... I learned a lot in the process. It was definitey theater quality.

For festivals , it depends on each.

HD capable will usually accept 1080i HDCAM and/or 1080/24p HDCAM.

Many are still 16mm and 35mm projection.

Good feedback, thanks.

Say I recapture the 'in' shots from this present show via one of the 2 methods you suggested.  What would you suggest for re-assembling the show?  I haven't found an 'auto assemble' function that would work.  Would I:

- start a new project,

- capture the 'in' shots,

- then import the already edited video mixdown track of the final cut to the timeline, 

- then import the final audio mix to the timeline,

- then 'eyematch' each re-captured shot to a new video track,

- then once finished delete the old video mixdown track & export to HD cam? 

Would that work?  Is there a better alternative?

Thanks for the advice, appreciate it.

Charles

The tapeless route would be a terrific way to go.  I don't know of any festivals that support that quite yet. 

Toronto, for example accepts: "...Sony HDCAM cassette in the HD 1080 format at any of the following frame rates: 60i, 59.94i, 24P, 23.98P, 50i, 25P. High Definition should be non-anamorphic only. Audio should be an Lt, Rt or pro-logic compatible mix on channels 1,2 with the option of Dolby E on tracks 3,4..."

Thanks,

Charles

I'm not clear exactly to what you are asking. Do you mean onlining a sequence?

Avid workflow for rebuilding a show is not the complicated.

You copy the sequence to a new bin.

Decompose.

Batch captured and import the decomposed clips using new settings in your Capture tool.

The decomposed sequence will automatically link to this recaptured/reimported material

conform imported material, FX and such.

Yes, my question is in regard to on-lining the entire project.

The objective is to get the best image quality possible output onto HD Cam from my HDV shot material.

This seems like a complicated question.  I'm not sure it should be.  Sony pitches their prosumer HDV cameras as being way superior in image quality to the earlier DV cameras - which they are - if one simply HDMI plugs them into an HD TV.  Avid Express Pro sells (sold) itself as a prosumer NLE capable of realizing these HDV quality gains.  Is it possible I missed the part where it was made clear that one had to jump through a set of extremely complex and/or expensive hoops to actually OUTPUT a superior quality image in a viable format?

The project has been captured (in native HDV via firewire from the Sony HVR-Z1U), fine cut in native HDV, color corrected and titled.  The image quality is fine on a TV.  It's nowhere near 16mm film image quality when projected via a hi end Christie HD8K DLP Projector on a big screen. 

I'm prepared to recapture everything and on-line it if this will result in a noticeable improvement in image quality.  Do you think it will?

Thanks for the input, feel like I'm getting closer to a solution.

Charles

chawilk:

It's nowhere near 16mm film image quality when projected via a hi end Christie HD8K DLP Projector on a big screen

The pleasure and pain of HD.

There is a noticeable difference in picture quality with HDV via HDMI depending on wetherthe image is pixel scaled to fit the monitor size. Another noticable difference is the quality of the realtime pixel scaling chips in different monitors.

The high quality of your 8K projector is achieved when it gets an image that suits its image scaling abilities. Avids upscaling to 1920 is better than most in device realtime scalers so reading the projectors specs and requirements would probably be your best starting point

Well do as Dom suggested..

Do keep in mind that HDV is heavily rasterized and by default a destructive codec. Meaning every single step will make it worse. Most that "have" to work with HDV transcode the stuff to a DNxHD format and work with full HD and no long gop mpeg rubbish ...

Take your timeline (sequence) get to a post house with Avid and at least a Adrenaline with the HD board, get a deck with HD-SDI out and just recapture the sequence do the effects conform and get back to tape... nothing more to it

Menno

chawilk:

It's nowhere near 16mm film image quality when projected via a hi end Christie HD8K DLP Projector on a big screen.

First, if there are effects on your timeline, did you render at highest quality? Assuming you did, or that it is not efects-heavy... What is the source (HDV tape? Playing from what?) and is the image going through any additional processing? Is the projector 1080 or 720? If 720 there is some scaling going on that might degrade the image (?) Are you sure the projector is set up correctly? My experience playing an edited HDV tape out through a Sony M25 via HDMI to an 18K DLP projector on a 35' screen was awesome, I would say it surpassed 16mm quality and looked more like 35mm to me. The guy I worked with on the project, who normally works in higher end HD formats, was amazed at the quality of HDV when projected.

When it comes to projection, there are tons of variables and settings. I would suspect that before anything else.

(EDIT) Also, transcode the whole thing to DNxHD before you output it to anything. I'm not sure that makes a difference if you are just outputting back to HDV tape, but I think it does and I have never done it any other way.

Baklap:

Take your timeline (sequence) get to a post house with Avid and at least a Adrenaline with the HD board, get a deck with HD-SDI out and just recapture the sequence do the effects conform and get back to tape...

Couldn't he just transcode to DnxHD in his system and then take the DNxHD media along with project files to the Adrenaline rather than re-capture and conform?

Yes can be done as well.. most i heard (no experience!!!) was that the HD-SDI way offers a far better result then going HDV at 1440×1080 to DNxHD at 1920×1080 by transcoding ways.... but as said that is only second hand knowledge

Menno

A small post house here that does Avid, DNX, Adreniline has suggested that re-capturing via SDI would not result in a noticeable enough image quality improvement to justify the time/expense.  Their proposal is that we:

1.  Duplicate the timeline

2. Remove all effects, fades, color correction, titles.

3. Export it at best quality to a file (unclear on setting for this as yet)

4.  Take a Lacie down to them, they input & dub via Adrenaline/DNX to HD Cam

5.  Then we take that to the major lab we normally work with (who supports HD Cam but not HDV) to re-do all the effects, fades, color correction, titles.

6.  Final master will be HD Cam.

So my questions are:

1.  Are they right in stating that an SDI recapture / re conform via SDI DNX will not result in a noticeable enough image quality improvement to justify the time/expense?

2. If so, what should the specs of the 'best quality' output file I take to them be?

Thanks for all the input guys.  Great community.  Didn't intend to sound unhappy.  Awesome program, output looks terrific, just want it to be best it can be.

CW  

What the hell? wTF!!?

They are basically saying - "dont recapture it waste time and money but let us rebuild the FX from scratch because that would not be THAT expensive"

Why would you rebuild the FX from scratch?

If anything you would transcode your timeline and then consolidate to a FW drive while maintaining all the FX. I understand color correction needed to be rebuild but most basic FX (wipes, dissolves, etc) should not go through a costly rebuild!

If you export the timeline as one file how are you going to rebuild the FX???

When you do transcode - make sure you transcode to 220x. You may need to change your project type (to regular 1080i) or it will limit you to 145 DNxHD.

I'm sorry but that lab seems to be making up a workaround workflow because they cannot support HDV via HD-SDI. That itself is not a big deal - transcoding is your friend. However, to suggest to remove the FX, export a standalone file with no ability to manipulate individual clips, rebuild FX from scratch seems to me they are trying to milk the project for billable hours.

QT Animation at 100% Quality setting is a LOSSLESS (pixel by pixel copy) codec and it is dual platform (OS X or Windows). It creates very large files.

Transcode vs HD-SDI, quality wise is basically the same (software vs hardware codec). It is up to you which you prefer. I prefer HD-SDI because HD transcode is not quick and if you have corrupt files the time spent trying to figure it out you could have captured it. But either way is viable.