Wavelab - filtering out Noise                                                                                               

I actually took care of the wind noise using the EQ filter in realtime FX clips audio lib. Took LOW freq to 1 and LOW gain to -15. Added the same filter twice and it took care of most of the wind noise.

This is a real time filter - you add it, mark an in and out on the timeline, repeat play between in and outs, and then you can edit the filter and make all kinds of adjustments in real time. No need for wavelab.

Also - I did almost the same thing for a problem with too hot of a mic on my drum corps footage. I adjust the MID to get rid of the hot part from the middle brass (mellow and baritones).

You could also go some way in reducing depth of field by the use of the camera aperture with ND filters if necessary.

DVFilm Maker claims next to no generation loss. That's what makes it possible to use as a preproduction technique to then move to 24p editing. I find this claim somewhat suspicious, but I think it has to do with the fact that they use rather complex motion detection algorithms so that portions of the image that are not moving remain crystal clear while moving portions are 'blurred' as the fields are merged into a single progressive frame.

As far as adding film grain and gamma correction - I'm not really a big fan of the grain. Softening the image is ok. The gamma curve in the FX1's cinegamma mode is quite nice (no generation loss there).

Good workflow for the FX1. Most important part is "lighting". If you light it like filme you are a far way already!

I suggest to reduce all postproduction-filters: gamma changes, de-interlace, grain, etc. Everytime you fuse and go to a new application to do all this you lose a generation. That is the reason why I love the DVX (and soon HVX). Those cams do it all in cam and at the best quality (gamma, progressive recording, colour grading etc).

I donīt know DVFilm, but MAgicBullet also does a great job in converting interlaced to p-stuff, but it is extremely slow.

SL - Nice writeup and workflow.  I'd like to see an example of your encoding compared to the original video shot.  Can you provide some content?

Also a dead give away for video is that the highlighs are almost always blown out.  Pulling down highlights and crushing blacks goes a long way in creating a film(esque) look to your video

BarkingMad - This topic comes up occasionally along with 24p editing. The question of how to make your video look less like video and more like film.

There are several factors that have nothing to do with editing such as motion blur and depth of field. There are some characteristics of film that are easier to obtain with 24p shooting and 35mm optics, but you can certainly come close shooting 60i as I've found over the past few weeks of playing with my FX1.

There are a couple of sources that I found particularly useful regarding this subject. One I linked to from Avid's 24p discussion www.24p.com and the other was www.dvfilm.com (the discussion here borrows from these two sources and hence, they deserve mention).

Here's my formula (NTSC - sorry). I'd be interested to hear from others who've made this quest for the tips they've uncovered over the years.

1. I shoot HDV with my FX1 and then downconvert to DV with the camera's internal converter. It does a really nice job of this. I shoot 60i and stay away from their 24p and 30p (simulated) modes. The FX1 has a switch to alter the gamma curve to approximate that of film. I like it but I suppose you could apply filters during edition to achive a similar effect (although perhaps not a cleanly as at shooting time).

***edited***

DVFilm Maker (see below) also requires shooting with a shutter speed of 1/60 - this is very important to their 60i -> 24p conversion.

**********

2. Log the clips into Liquid. Do my editing in Liquid and fuse the final sequence to AVI. You can play with various filters at this stage. If you have Boris RED and some sapphire plugins, you can add gamma curves and film grain 'till the cows come home. It's a personal taste but nothing beats good lighting during shooting (lighting and camera movement is crucial to getting that film look).

3. To get the feel of film, you'll need to deinterlace and convert your 60i footage to 24p. The product I've chosen for this is DVFilm Maker (www.dvfilm.com). This web site has a lot of information on the 'film look' topic. Their primary business is converting DV films to 35mm (so you can take your creations to Cannes next year).

4. I use DVFilm Maker in a somewhat non-standard way. DVFilm Maker will take your 60i footage, deinterlace it and then generate a 29.97 fps AVI file encoded with a 3:2 pull-down (it's really 24p inside, but the 3:2 frames are added to fill in the 29.97 NTSC standard). DVFilms will take this and make a transfer to film. That's all fine, but I'm not going to Cannes any time soon, but I would like to create 24p DVDs. The advantage of a 24p DVD is that it takes less space on your DVD (or you can take the same space, encoding at a higher bitrate).

5. Sorry for the rambling - getting back to DVFilm Maker...  I use a feature of it that will allow me to generate a 23.98 fps progressive export to a MOV file (DV/DVCPRO with PCM audio). This is a very high quality conversion of my original 29.97 fps fused sequence. Note that it's not MPEG-4 or H.264 or whatever, it's DV/DVCPRO. DVFilm suggests using this when you want to preconvert your footage and then reimport into Avid XPress Pro for further 24p editing - cool.

6. Anyway, now that I've got my 24p MOV file, how to I put it on to a DVD? Well - I need to encode it into MPEG-2 with an encoder that understands that this source is 24p and thus, puts the 3:2 pulldown flag into the transcoded MPEG-2 so that progressive DVD players will know that they can use the 24p directly. The problem here is finding such an encoder. I'm pretty sure you can't use Liquid for this.

7. DVFilm talks a bit about this and suggested TmpGEnc (www.pegasys-inc.com). I wanted to use the cheeper version, but ended up putting out $60 USD for the newer version because it works with MOV files on the import (and this is important to the workflow I'm describing here). I also tried Sorrenson Squeeze which supports this workflow but I honestly found TmpGEnc provided better image quality (quite a bit, in my case).

8. I'm out of breath but I'll keep going 'cause I'm almost done. Thanks to TmpGEnc, I've got my M2V and WAV files (I encoded audio as PCM because I'm going to trancoded it to AC-3 elsewhere). Now you can use your favourite DVD authoring package. Right now I'm playing with Adobe EncoreDVD - it works very well with the M2V (takes them at face value - even detects that they're 16x9 in my case). Encore transcodes my PCM audio to AC-3 (stereo).

9. Basically those are the essentials of my workflow. Sounds like a lot of work, but most of it is waiting for the next phase to complete. You'll still spend 95% of your time cutting. I take these results and compare the straight DV->DVD vs the 'Filmed' version - it's like night and day.

The only gotcha I encountered in this process is that somewhere between Liquid and DVFilm Maker, the fact that the video was anamorphic widescreen was lost. I setup DVFilm maker to encode 16x9 progressive. Maybe this is typical. Fortunately, TmpGEnc allows you to force interpretation of the souce footage and force the output as well. Also, TmpGEnc may try to de-interlace your source - you have to go into it's filter options and disable de-interlacing. DVFilm Maker's already taken care of that issue.

You can download demo versions of Encore, TmpGEnc and DVFilm Maker to try this for yourself.

Well, I'm tired. That's it for now. Your comments are always welcome.

Cheers.