Mike Most -- account bouncing, bad address
Forum Replies Created
-
>>”They”, the transfer house, were thinking the cut is going back to film.
No, “they,” the transfer house, were undoubtedly doing what they were asked to do. Video facilities don’t make up their own specifications on film to tape transfers. Instructions are given by production, either by a producer, a post producer, a post supervisor, an editor, an assistant editor, or whomever is assigned the responsibility of running post. The facility follows that specification, whatever it is. If one wants to know why it was transferred the way it was, they should ask the person who ordered the transfer.
-
>>Mike, you are a rockstar – are you in the LA area? I’d love to throw some work your way.
Hardly a “rockstar,” just someone who’s done this stuff for quite some time. I was in L.A. for the last 25+ years, but moved to Miami about 4 months ago. I am, however, “Chief Technologist” (whatever that means) at a film lab/video facility, so if you need something done, there’s both overnight delivery and Internet delivery……
-
EDL’s in any nonlinear editing system start at the time code that your timeline starts at. They are translations of the timeline, and do not take into consideration where you happen to start your edit on the playout.
-
By the way, if it makes you feel better (and it might), (as a colorist) I did an entire movie (100 minutes) that was done exactly the way you’re describing – cut at 23.98, output with 3:2 pulldown to DigiBeta, with a converted list supplied for color correction. It didn’t slow me down very much, but I did have to check every edit to determine whether it was on field 1 or field 2. The DaVinci has various cue functions, so you can instantly cue to either the first frame of the current event, the first frame of the next event, or the first frame of the previous event. So in effect, all you need to do is automatically cue to the first frame of each edit and jog through it, changing the field ident if you need to. Not really as much of a time consuming step as you might think, especially if you do it as you’re color correcting. Besides, the reality for colorists is that EDL’s usually have errors in them – events that shouldn’t be there, missing events, etc. Colorists with any degree of experience are very used to that, and are careful to change what needs to be changed as they work.
-
I think the mistake you made is twofold. First, in the future you should make all projects of this type 23.98. There is no real need for a 24 frame project under almost any circumstances. Second, you can’t just change the timebase and have the material play correctly. You need to conform the 24fps Quicktime movie to 23.98. Then you can drop it into a 23.98 timeline and have it play out correctly. I also don’t know what you mean by “gaps.” There is no simple way I know of to tell when an edit is going to fall on field 2 while you’re still in the editing system, because you can’t field jog. You need to lay it off to tape and check the edits in order to determine this.
The rate, however, should have nothing to do with the EDL conversion, as it’s just a numerical conversion. It assumes a 3:2 pulldown with “A” frames falling on time codes ending in 0 and 5, and changes every time code number in the EDL accordingly. I’m not sure it even does length checking at each edit (as a true matchback process would, but that would usually be working in reverse, i.e., taking a 30 frame video edit and compiling a 24 frame film cut list).
I suppose there isn’t a significant difference between the on the fly playout and a 29.97 render, but it’s much easier to not do a separate render – especially seeing as it’s unnecessary when you’re using something like a Kona card, which I believe can do this in hardware.
-
By the way, you don’t want to be putting the piece into a different timeline. What you need to do is play it out directly from the 23.98 timeline, and let Final Cut add the 3:2 pulldown in real time. Anything else and you’re creating problems for yourself. Just make sure the program starts at the correct timecode, and you’re good to go.
-
>>So my question is – can I just fix these edits, render out in FCP, and export it >>that way onto my digibeta?
No, not really. They’re a consequence of cutting in 24 frames, but outputting in 60i. They’re a result of 3:2 pulldown, so unless you can figure out cut points that all fall on field 1 (that’s very unlikely), you will have some field 2 edits. What you can do is lay it off to tape, making sure to set the machine to insert VITC. This will allow you to field jog, and determine which edits are happening on field 2. You can then make a list (by hand) and give that to the colorist. It is a simple one button function on a DaVinci to change a mark from field 1 to field 2 or vice versa, so your colorist can load the EDL (forcing field 1 when he does), then go through the list and change the ones you flag as field 2 prior to starting the color correction. Since your piece is a short, this won’t take very long – certainly less time than using the scene detector, and more reliable as well.
Just as a bit of history, this was never a problem prior to 24 frame editing, because in video editing, all edits are normally on field 1. Even when nonlinear editing came into play, most longform shows were onlined in a “traditional” linear online bay, in which all edits are field 1 edits. The EDL that was sent to color correction was generated from the online system, so the problem never existed.
-
It won’t work very well, because it won’t generate a field marked list. Many of your edits will by definition be on field 2, because the 3:2 pulldown in your output will be continuous from the first frame of the piece. Some edits will be on field 1, some will be on field 2. The converter assumes that all edits are on field 1. Your colorist will need to be very careful when watching for color flashes at cuts, and change those that are on field 2 manually. In doing this, he will undoubtedly miss some and this will produce a one field “flash”. If the color correction on the scene change is not large, you might not notice, but if you want the color corrections to be “clean,” it becomes an issue. That’s why I suggested running it through the scene detector, which is field sensitive.
Just as a bit of background, the reason this isn’t a problem when material is edited at 30fps, even if the source material has 3:2 pulldown, is that by definition all edits in this format occur on field 1. When you cut at 24fps, you add 3:2 pulldown on the output – thus causing some scene changes to occur on field 2.
-
>>Why not find a facility that can do color correction from quicktime files.
>>Applications such as Color Finesse 2 or Final Touch would be ideal.Well, perhaps this is a reasonably professional project and they want to use a professional colorist working on a high end system like a DaVinci.
The primary problem with a list at this point is field marking, which to my knowledge Cinema Tools is not capable of doing. If you’re using a DaVinci, you’d be better off running it through the scene detector, which can and should properly detect field 2 edits. The post facility will charge you for the time to do this, of course.
-
>>Sony wants between $11K and 20K for a deck to read what is in essence a modified blueray disk. They >>are also selling the Playstation 3, which is Blueray for around $500.00 this holiday season. Seems >>kinda unfair to charge 20-40X for the same media.
Not to demean what you’re saying, but it sounds to me like a rather silly and self serving rationalization. The XDCam player is a full editing machine, it understands timecode, it can deal with the file structure and directory structure that the XDCam format supplies, and it is directly supported by Sony’s professional video division. Just because the media is physically similar doesn’t mean it isn’t a different, professional, and thus limited audience product. The Volvo XC90 SUV, the Lincoln LS, and the Jaguar X type are all built on the same basic platform. So are the Toyota Camry and the Lexus LX330. But nobody in their right mind would argue that these things are the same product, or that they should be at similar price points. Saying things should be priced based on their raw component cost is to ignore the realities of product design, product development costs, intended markets, and potential sales volume, as well as anticipated level of support – not to mention what the intended market will bear. All of these things can and do contribute to the pricing of any product, electronic or otherwise.