Looking to better my final DVD product

[Tim Snyder]

I always feel like our final DVD product isn't the best that it could be and I wanted to see what other's workflow is to see if we can make ours better.

We shoot (1080i/60) multi NX5's at an event and use clipwrap to transcode the footage to HDV (for speed/ease), import it into FCP and then export as a quicktime file. Then drop it into compressor and encode as best 90 for the DVD.

We may have anywhere from 1-5 events every weekend, so speed is important, but want to make sure we are putting out the best product we can.

Thanks for the input!

[Chris Sgaraglino]

Please I mean this with the utmost of respect, but with all that "conversion" going on, your going to have artifacts with each; Premier Pro CS5.x ellmiates all that transcodeing.

I know of and hear FPC and Vegas users jumping ship... You can even setup a FPC keyboard layout if you switch... But think of the time you'll save in transcoding alone, not to mention the quality.

I know I am goi g to get slammed for saying so... But its been documented over and over, each time you wrap, transcode or convert, you are punishing those pixels and eventually your going to start seei g it!

I have burned DVDs for clients that look almost as good (in most cases) as BlueRay from sme of the FPC and Vegas burns.

If you have good solid and sharp raw data, there is no reason that you can't get very nice and satisfactory DVDs.

[John Vasey]

Tim...if you transcode to Pro Res and edit, then export a Quicktime movie from that, you should notice a significant bump up in quality...try a short test to see how that works for you compared to the HDV transcoding. Also, if you are using Compressor to compress to a standard def DVD file, you can bump up the edge sharpness by turning on the Sharpen Edge filter in the Compressor preset and try a setting of 2 or 3. Hope this helps..John

[Les Wilson]

@Chris. It's Final Cut Pro or FCP not FPC. Also, you can re-wrap files all day long without loss of resolution or introduction of artifacts. Transcoding maybe, wrapping no. Furthermore, there is no "All that conversion going on". It's a standard practice to transcode AVCHD into an easier to edit format then in order to put it on DVD, it has to be transcoded to MPEG2. That is the minimum for any NLE that doesn't support AVCHD natively. Perhaps you are thinking that the export to Quicktime is a third conversion. That is wrong. There isn't a transcode. FCP gathers whatever format the timeline is and wraps it into a Quicktime file format.

@Tim,
The one thing I see in the workflow is the downgrade to HDV. I understand why you might do that to speed up edit and rendering but shooting in full 1920x1080 resolution and converting it to HDV is down res'ing it to 1440x1080 rectangular pixels. You are also possibly introducing AVCHD to HDV transcoding artifacts.

As recommended already, transcode to Prores. That is an upres and is very efficient to edit. The files will be bigger and it may take longer than the transcode to HDV.

[Chris Sgaraglino]

Hello Lee,

Not being a FCP user, I could be wrong... But here is what I do know.
I run a network of over 50 video and film enthusiasts and professionals - including 8 editors.
Three of us run Premiere Pro, and the other 5 FCP... Any one of the 5 editor running FCP takes several days to "prepare" the footage before they can start eding it. But the CS5 users, using the same footage, gets to start right away?

Even if there is no loss of quality, there seems to be a huge loss of time? And to me time is more valuable!

Just my 2 cents...

[Les Wilson]

Originally Posted by Chris Sgaraglino

Hello Lee,...

It's "Les" not "Lee". Whatever. The topic is about the DVD workflow of the OP. You are clearly passionate against Final Cut Pro. I don't think it's helping the OP resolve the problem and maybe it should be channeled in another thread where there's an NLE war already going on.

[Chris Sgaraglino]

Sorry Les, the iPad sometimes makes spelling and word decisions on its own and changes what one types. No disrespect was meant!

I'm not looking for a battle, I was just trying to help him get a better quality DVD like he asked, and looking at his post he mentioned:
He shoots interlaced footage
Transcodes it to HDV (Clipwrap)
Converts to QuickTime (FCP)
Then compress and encode for DVD

I'm not trying to sell anyone on any NLE, or aganst an NLE, but he did mention that he is trying to get better "quality" and that "speed" is important.

In looking at his workflow, if he took out some of those converts and compresses, he would get a better quality output. And if he dropped two of the three processes he would could cut his workflow time by 2/3?

I mearly suggested a way to accomplish all his desires in one software solution.

[Les Wilson]

No problem Chris. That auto complete gets us all.

Ideally, if FCP 7 had native support for AVCHD, the initial transcode would be saved. Since that isn't the case, the initial transcode is a fact of life for FCP 7. I think FCP-X has it tho. Regardless, if done to Prores, there won't be a loss during that transcode because it's an upres in color space from 4:2:0 to 4:2:2 and keeps the 1920x1080 resolution. At some point the interlaced needs to be de-interlaced and this also happens in this step. There is no transcode when exporting to a Quicktime file from the timeline. That's a scatter/gather and wrap operation. So really, there's only one down resolution transcode and it's from Prores to MPEG for DVD. There are advantages to editing in Prores over keeping it in AVCHD.

[James S. Davis]

Found this product very interesting, Blackmagic Design: Intensity

Check out the Intensity Extreme. I am looking a solution to allow me to capture 8 years of minidv media I have compiled to FCP X (which I really like). It is Thunderbolt and it not too expensive. It seems to be the solution for that challenge I have.

I spoke to their Tech Support and it has a very interesting tool for the NX5U users. You can capture via HDMI and the Support team said that the lack of having to compress allows to bring in a VERY high quality, so large that he said that you have to a RAID to be able to handle the data.

[Les Wilson]

@Tim,
According to the post below, the NX5U does not have a full raster 1920x1080 sensor but rather 1440x1080. I don't know it that's true. Still, the HDV codec may not be the best choice for intermediate format for editing.

http://www.dvinfo.net/forum/sony-hvr...ml#post1702336

[Ron Evans]

Originally Posted by James S. Davis

Found this product very interesting, Blackmagic Design: Intensity

Check out the Intensity Extreme. I am looking a solution to allow me to capture 8 years of minidv media I have compiled to FCP X (which I really like). It is Thunderbolt and it not too expensive. It seems to be the solution for that challenge I have.

I spoke to their Tech Support and it has a very interesting tool for the NX5U users. You can capture via HDMI and the Support team said that the lack of having to compress allows to bring in a VERY high quality, so large that he said that you have to a RAID to be able to handle the data.

The higher quality only applies with live capture. Once you have recorded to tape( or any local memory) you are stuck with the tape quality ( or CODEC used to record). Best transfer for tape( DV or HDV) is Firewire( IEEE1394) this is just a data copy of the data on the tape. Tape transfer is always realtime so there is no getting around the time and the quality is the quality on tape that it was recorded at in the first place. Once you have these files then you could apply some correction . For file based like the NX5U it is just a file copy to the PC. In neither case is the Intensity card useful for capture of already recorded material. For live capture, monitoring while editing etc it is a practical solution.

Ron Evans

[Ron Evans]

1440x1080 ,that is the Z5 not the NX5U I think.

Ron Evans

[Eric Olson]

Originally Posted by Tim Snyder

We shoot (1080i/60) multi NX5's at an event and use clipwrap to transcode the footage to HDV (for speed/ease), import it into FCP and then export as a quicktime file. Then drop it into compressor and encode as best 90 for the DVD.

Which of these steps is the weakest link in terms of final DVD quality? While transcoding to HDV would be a noticable mistake when delivering blu-ray, I think the quality of the HD to SD downconversion may be the most important factor in improving your DVD. Many programs, compressor included, do a surprisingly poor job downconverting interlaced HD to interlaced SD. See, for example,

HD to SD DVD – Best Methods | Creating Motion Graphics Blog | Blu-Ray DVD Authoring Menu | Precomposed

Paradoxically, the sharper the original HD footage is, the worse the downconverted SD footage looks. I suggest focusing on the downconversion step. The improvement can be remarkable.

A reasonable downconversion algorithm performs the following steps: convert 1080i30 into 1080p60 using a frame rate doubling deinterlacer, apply a antialiasing filter, rescale to 480p60, then weave to 480i30 using an antitwitter filter. Unfortunately, FCP and compressor do not have an option to perform high quality downconversion. Therefore, try the following work around: transcode the 1080i30 AVCHD camera source to 720p60 whatever intermediate codec you want to edit. Edit as a 720p60 project and send the 720p60 output to compressor for mastering the DVD. This workflow sounds silly, but it allows compressor to downscale better.