Does this workflow make sense? HDLink, rough cut with VirtualDub, final in Vegas.

[Earl Laamanen]

I think I'm slowly starting to grasp how I want to use Cineform - until now I haven't kept any original footage except some really exceptional stuff which I kept as original MTS files from my Canon HF100. My audience is generally friends + family via vimeo and youtube's HD services.

90% of what I do is underwater video, and I'm starting to get some pretty cool stuff and as I've just purchased Neo Scene, I figure its time to start archiving properly with CFHD. I often record most of a dive, maybe 45 minutes, and of that there may be 5-10 minutes of footage to keep and 2-3 minutes will make it to a nice little final cut.

Until recently, my workflow was:

-get AVCHD MTS files off SDHC card and into Vegas (HF100, FP24)
-trim down to those 5-10 minutes worth keeping, even though AVCHD is stuttery in Vegas
-render out to Huffyuv AVI
-deinterlace with mencoder.exe
-open the very large avi file in Vegas and edit to final
-render as 720p with h264 @ 5-10Mpbs
-upload to youtube, vimeo
-discard the intermediate AVI and AVCHD source

Now that Cineform makes things quite a bit faster and less painful, I find I have more time to work on the editing part and I'd like to keep those 5-10 min rough cuts, since I may come back to them one day. My first stumbling block was not being able to trim before converting MTS to deinterlaced Cineform AVIs, and second was not being able to open AVCHDs in Vegas for trimming first before exporting to deinterlaced Cineform AVIs, the Vegas interface for Cineform doesn't offer pulldown removal which sucks, even after I own Neo Scene? But I searched around a bit and realized VirtualDub will allow me to trim using direct stream copy so now it looks like I will do:

-convert AVCHD files off card directly into deinterlaced CFHD files
-use VirtualDub to trim away the 75% unwanted footage, using direct stream copy
-(merge all trimmed clips into one file? didn't try this yet, but i assume virtualDub can do it)
-open trimmed clips in Vegas, go to town.
-render as 720p h264 for uploading to Youtube and Vimeo
-archive the trimmed/appended CFHD avi files and the Vegas project, discard AVCHD files.
-do a happy dance.

Does that make sense? Can anything be improved here?

[Richard Leadbetter]

You can't append CineForm files together in VirtualDub - at least, I've never been able to.

Otherwise I'm not seeing much wrong with your workflow.

[Earl Laamanen]

I'm working through this for the first time and I find virtual dub way too slow since I have to load each clip seperately. The inability to append CFHD AVIs in VirtualDub is kind of a showstopper for the plan I described above.

But it looks like I've found my method - something Eugenia described somewhere while I was researching cineform and I've just clued in to now. Once I've trimmed away everything I don't want and have something like a rough cut in my timeline, I'll use Save As, *.veg, and pick the option to include media. When you choose include media, and your source media is a non-lossy format like Cineform then Vegas will ask if you want to just copy over the original source clips or trim them to match your timeline. This is essentially the trim + direct stream copy I was trying to use in Vegas, only much nicer to use.

Yay.

[Eugenia Loli-Queru]

In my opinion, your workflow is unnecessarily convoluted, and a waste of your valuable time. You can just buy a more powerful computer and be done with all these steps. But if money is an issue, do the following:

1. Purchase and install CoreAVC Professional for $15 CoreAVC.com: World's Fastest High Definition H.264 Video Software Decoder (the Pro edition supports AVCHD). This little utility worth every penny, as it's indeed the fastest AVCHD and h.264 decoder in the planet. It's needed for the following workflow.
2. Proxy Editing with Sony Vegas The only difference is that you also need to check the "DirectShow Decode" checkbox in the SUPER screen. Everything else is the same as in the tutorial.

I just tried the workflow just now, it works perfectly. There is no reason to de-interlace outside of vegas (you can pick the "interpolation" deinterlacing algorithm on vegas' project properties if you don't like the default "blend fields"). Vegas will handle the proxy files editing, de-interlacing, and final export.

This should reduce your workflow's time many times over without the need of using huge files like Huffyuv's. The proxy files are extremely fast to both encode and edit. We are talking about extremely fast editing and only a few minutes of re-encoding to create the proxy files. You only switch to the slower, but 100% quality, AVCHD original sources at the very end, just before you export.

Feel free to IM me or email me if you need help btw. I will be glad to help you out. I am confident that this workflow is way faster, and possibly with less loss of quality, than anything you do right now. And you don't have to use huge filesizes either! Just the AVCHD sources, and the proxy files which are extremely small (typically 20 times smaller than the AVCHD sources).

[Simon Dean]

Perhaps I misunderstand but surely you can;

Convert AVCHD to CFHD using NeoScene.
Edit CFHD AVI files in Vegas.
Render out finished file from Vegas
Save Vegas project with trimmed copies of the AVI files - you can specify saving extra seconds either side of the chosen portion. (This is an option when you Save As...)

The AVI files are better quality from NeoScene if you're doing any colour correction etc.

You only need NeoScene and Vegas then and there's no extra steps.

[Eugenia Loli-Queru]

Simon, that's true. However, NeoSCENE costs $130, it takes *a while* to transcode the AVCHD files (it's faster for m2t, and by his own admission his PC is not fast), and the created files are huge.

The workflow I suggest costs $15 (and the codec is useful for playback in WMP too, not just for the proxy files' purposes), it's extremely fast to create the proxy files, it's many times faster to edit them than any HD format, and their filesizes are very small.

So in my opinion, for THIS user (who I assume is not super-serious about video, as he doesn't shell the money to get a new PC to deal with AVCHD natively), it would only make sense to go for NeoSCENE if he needed PF24 pulldown removal. In that case, yeah, NeoSCENE makes absolute sense.

Please note that I am not trying to knock NeoSCENE. I am VERY happy that this affordable utility was released. I've been lobbying for it for 1.5 years when I first emailed the Davids! I have in fact edited all my older (and much visited) pulldown removal tutorials to include the new information about NeoSCENE and even highly recommend it, but for the needs of this specific user, I think my workflow suggestion makes more sense.

[Simon Dean]

It's because he said he'd already purchased it! You're right about the file size, but disks are getting so much cheaper nowadays.

[Eugenia Loli-Queru]

Sorry, I missed that. Ok then, if he's already got NeoSCENE, I guess he can use that to convert, and then just edit in Vegas. He doesn't need to go to any other app outside of vegas after that.

[Earl Laamanen]

Thanks guys. I'm still learning lots. The changes I mentioned yesterday match what Simon suggested above, that's:

-start with HF100's MTS files (AVCHD)
-convert using Neo Scene to create HD deinterlaced AVI (CFHD)
-Vegas editing in realtime w/o proxies (CFHD)
-Render final cut for online HD (720p h.264 usually)
-Resave project for archiving with trimming option - this very important step is where I get rid of 90% unwanted footage (CFHD).

The problem with Eugenia's last suggestion is that I can't render out the trimmed AVCHD files with a direct stream copy, I would have to re-encode I think to some Sony or MainConcept h.264 format, and then it would be a contiguous clip, no longer made up of smaller parts so I can't change transitions etc without resplitting everything.

My most important goal was to find a way to trim and discard 90% of my footage, but maintain close to source quality on that last 10%, and use that for my final vegas project. Cineform and Vegas together give me this ability, and it's working very very nicely.

I have both CoreAVC and NeoScene, and I'm only temporarily on a Dell Precision M6300, which isn't a bad PC but my desktop is much more capable. A possible bonus is the upconversion to 10-bit color in CFHD which may or may not give a bit more elbow room for colour grading - but I've only read about that and haven't made any tests.

Of course, just becoming a better underwater videographer means I won't have 90% waste footage, but it's hard to know when/what you'll see and it's a bit of a pain to check if you are recording or not. It has happened that I've pushed the level thinking it was paused, and wanting to record, when it was really the opposite, I was recording the whole time by accident and paused when I really wanted to record. With flash memory and hicap batteries, it's often easier to simply record the full dive. It's the compromise I made when I chose a budget HF100 + Ikelite combo.