Michael Collier

If you want to do this for a bare minimum of cost then you have to build the raid assembly. dont worry, all those naysayers who believe you need a post facility are the same who are beholden to large budgets to make anything possible. Fight the problem with intelegence and research, not a money wagon. You can set up a SATA raid assembly with just a few disks and a controller card. they have to be stripped, and the machine will read them as a one drive. 150/Mb sec, check that figure. it may be max throughput, not sustained and it may also refer to the megeBAUD, not megaBYTE, (a megabyte is 8times larger file) so check all the math and make sure you have 15-20% overhead to handle system slowdowns, esspecially if this is your only method of full quality HD capture. Check out the supersmall barebones computers. they are cheap, fast and have enough room to store 2-3 hard drives. that and a small LCD monitor would be enough.

Are you planning on scanning ALL your footage on the spirit. I would advise agaisnt that strongly. I am in the middle of pre-pro on a S16 feature as well considering the same options (I am also shooting almost half the footage you are). Prior to final HD telecine, why not have it telecined to a DVCPRO format on HDD or tape with keykode marks burned into the letterbox area? you can get your rough cut made, then find the takes you used in the film and splice those takes together from head slate to roll out and send that out for final telecine with best light? It seems like that would easily cut the amount of footage you need to scan by up to 1/4 (200mins) I think this would save you considerable amount of money, and you will fly through your rough cut because you are using computationaly easy DVCPRO footage instead of render-consuming HD footage. That cost savings would ensure you can afford the spirit telecine and have it datacine directly to hard drive. Then when you do the final HD edit, you have 10bit uncompressed. Add your color correction and send that to a post house with instructions to put it on an HDCAM tape or DVCPRO-HD tape. Oh, and dont listen to the people telling you to go to a post house. Post houses are so damn expensive and there is very little that can be done there that you can't do on your desktop. Time was independant film makers would assemble their own edit table, wire everything up themselves and edit using a film table> Everything was done themselves. now that computers are out everybody is telling me....you need a post house, you cant do it without a post house. WHY!!! cuts only, simple color correction and an HD final cut...this is what I bought the damn editting machine for!!! why would anyone drop $2000 on a computer and $1500 on adobe studio if I couldnt at least make a color corrected cut of my HD super-16 movie? If you have yet to shoot your film let me recomend some gear I have been reasearching to help make things much, much easier in post. (being that all my prior films have been to tape, I have rarely used double systems, and so when researching I wanted a bullet proof system, that was affordable.) http://www.bhphotovideo.com/bnh/controller...egoryNavigation I have more, but a shoot jsut came up so i gotta jet.

If its at all affordable,50 is almost required. Anyone who has worked with dv25 and dvcpro50 can tell the difference. Its not a huge overt thing, the average audience member wont go 'omg, i cant believe they used dv25 on that!' but there will be a higher percieved value to the picture if it gets shot in 50. Also you will have more latitude if you do any color correction. But the big thing is dv25 is not a broadcast format. It is close, but technically insufficeint (our station has broadcast dv25, but we avoid it if possible) dvcpro50 might cost a bit extra, but its worth it (i mean, camera is only a small portion of production expenses, and you can only shoot the thing once.)

It just means for every pixel and for every channel of red, green or blue, there is a full 8 bits or 10 bits to describe the value. It simply means there is no compression. It is the highest quality a digital video clip can be (unless it went into 12bit log, or better 16bit linear) The uncompressed allows you to do more color correction than a compressed video would allow for. and the mroe bits, the more this is true. Keep in mind that while it says 10 bit uncompressed, it may still have chromanence sub-sampling, because of the capture format. Many telecines and datacines are only capable of outputting 4:2:2. I am not sure why this is. It seems like they could go from HD-SDI to Dual link HD-SDI and get 4:4:4 with minimal effort.

DVX will be very strange to intercut. It is much much sharper than the GL (I am comparing to my experience with the GL-1, but I think they kept the imaging sensor the same between models) Also the GL has a weird color shift, almost like it cant quite compensate for the pixel shift on the red CCD. If you want to intercut with them my advice is to first turn off the 24p feature and the cinegama (it would be better to turn both of those on the GL-2, but alas, impossible) and turn the detail down. Turn the detail up on your GL until its sharpening noise, then back it off so you still have a sharp image, but noise is minimal. You may have to color shift a bit. Thankfully the DVX and the GL-2 both have color options. Use two reference monitors that are propperly set up so you can tell when they are getting close to eachother in terms of color and softness. If you have already shot the B-roll, I would do set the DVX while watching the footage and try and approximate the look (remember to test with a similar frame with the same lighting.) if its going to be a straight B-roll/A-roll type program, I wouldnt worry about the difference though. In many news type peices it is quite common to have the video quality much better for the interview than the B-roll. Thats because while an ENG crew may set lights for the interview, rarely will they light each shot of B-roll properly. Usually b-roll is lucky to get one light to bring the ambient levels up. It may seem a bit strange to you, but to the average viewer, they may not even notice. The learning curve is pretty minimal. You have greater control in your matrixes and DSP, also you have more programible buttons (a dial with 5 positions sets the 'look' including the 24p option and all color matrix) The biggest thing you will have to learn is the difference between 24p 'A' and 24p normal. (A is advanced and does not record with the 2:3 pull down the normal mode does. If you plan to edit with a computer I recomend A. Dailies will look odd at first, but when you pull out the 6 doubled frames for every second of video and put that through a better 2:3:3:2 pulldown you will have a much better result.)

ADR seems likely in this instance. I like the idea of using a cut so you can avoid shooting the dialouge with the window open, but if the window is open and you are driving at any kind of speed (greater than 20mph) I think wind noise would be above what your are willing to accept. But like any mic setup, get the mic close. Lapells work good, though you have to add reverb, eq compression and ambient noise on top of them most times.

I wasnt refering to Pals capabilities, just my understanding of Pal. What I know i learned how it correlates to NTSC. I actually have nothing mroe than limited accedemic knowledge of PALS workings. thats what I was saying. How does it not have a hue? I understand the chromanence is not phased in on top of the lumenece, and so a true hue/phase potentiometer is unessicary, but do phosphors not shift in reponse? are you telling me there is absolutley no need for a hue on any PAL television? If so I like it already. We got screwed with NTSC. backwards engeneers.

I live in alaska, close to asia, and it has been predicted that we will see the first outbreaks in the US. last I heard from ABC national, is that we will see the flu in as little as 3 weeks (this prediction was broadcast 2 weeks ago) the problem is we just got another foot dump of snow last night, and I dont see migratory birds start to show up until June. who knows if I see a parakeet headed out to Gridwood toating skis with him, then I will be worried. But I am in the media and have been talking to people at the CDC and around town who are experts in this field. In the last few months we have gone from relativley few confirmed cases, to several per week. no this doesnt mean that it is a sure sign of an outbreak, but bird to human occurances are getting mroe frequent. Dont panic, just know/

I could be wrong, my understanding of pal is limited, being born in America and very infrequently visiting europe, but I believe Pal is 25fps. NTSC is 29.97, which is why drop-frame must be used, every few minutes that .03 of a frame adds up to a whole one. With pal being locked at 25fps, there would be no need to drop a frame ever. At least thats my understanding. I heard NTSC switched from 30fps to 29.97 in 1953 (could beworng on the date) and the reason was to intruduce color. Chormanace is layed over the lumanence field out of phase, so that later they can be seperated. I think the 29.97 has something to do with that chromanence phasing....highly technical explination for something that should have never been invented. I hate NTSC. we are bogged down by the short-sightedness of 1950s engeneers. Did they even have transistors in 53? i think they came in around 56, anyway. Never the same color. This has nothing to do with the timecode debate. regardless weatherr drop code is there or not, to call time code 'true timecode' it must be accurate every single frame. Now for most mini-DV projects its not a huge concern....come on are any of you editting mini-DV from a low res proxie? hahaha any AMD k6-2 350Mhz editors out there? (Be quite naysayers, that was my first adobe premiere 4.2 edit bay I payed for when I was 13, and it could handle 10MB/sec YUV data) the big difference between DVCAM and DV25 is the way it records to tape. I dont think there its techically different, in terms of bitrate or color space or compression sceme, but DVCAM does have more user bits availible, including enough to record full time code on everyframe. DV records it every few frames, and depending on the quality of deck or camera used, it can fall out of sync by a frame or two only to catch up again. DVCAM also has more redundant recordings, to minimize drop out-errors as well. those are the only differences I know of. let me know if I am missing any. Now on to a question I had...I shot a ton of HDV in the sony cine-frame setting. Now the pulldown is complete crap. Totally useless, but I heard the true value of cineframe is pulling the interlaced frames out to get a 25fps video, then slowing that to 24, pitch shifting the audio then adding the pulldown back on. Does anyone have experince doing this? Can anyone recomend a good software program (or better yet after effects or premiere plug in) that can do this? I am premiereing my movie on the 15th of next month, so expident help is much appriciated.

The LEDs fit a nice niche, but are by no means a complete light kit. In most cases the light wont even be enough to act as a decent key light. Every shoot I have ever seen them used on, they almost invariably filled in as a close fill. with the lowell kit you will have more options and more variety in the quality of light you can provide. with the LED you can really only do soft lighting with the unit closer up (around 3-4 feet) giving it a sharp fall-off. LEDs are great, but only if you have a complete lighting kit. Totas are hard to control, try out the omnis

Wow. great deal on that camera. it even includes 3 film holding chambers!

Thats the montra that has been pushed around by people who have never had experience with HDV. Its not really a problem. I am doing final clean up work on a 100minute HDV feature movie and we have had a great working experience with the footage. It cuts easily, we can still scrub the timeline, nothing needs to be rendered unless there is an effect added on, and renders finnish quickly. I dont notice the codec causing any problems. the timeline is responsive and snappy. I really cant tell much of a difference between editing DV and HDV, except rendering time is about 2x what DV is.

Sort of off topic, and I dont know how many of you follow the iditorod. ITC is the iditorod trail commision, they run all the logistics of the race. Normally for media coverage they contract out the 'official station coverage' and let the station worry about how to cover the race (the station I shoot for up here did it for the last 13 years) And I can tell you from experience how difficult it is to fly into the smallest towns in alaska and shoot in sub-zero tempatures, along with the obvious problem of getting footage back to anchorage. This year the guy we used to rent gear and hire people through went behind our backs and convince ITC to produce the show themselves. I think they spent around 200-300K to produce everything. Funny enough they neglected to find a station to air it. They came to us and asked how much we would pay to air their coverage. we said nothing. all the other stations said they absolutley would not cover it. In the end they couldnt find a proper station and ended up dumping it on public TV. obviously this makes it very difficult to recoup the costs, beacause they cant advertise. Every sponsor of the cerimonial start expected a huge alaskan audience and they got screwed on that. to top it off the production was a dismal falure. They had a remote control helicopter with a camera that never worked, even though they cut to its static and unsynced picture like 10 times. They missed the start of several prominent mushers to do interviews on the street with random people. Now comes the actual race. They promised 6 stories from the trail a day on their premium website (they expected to recoup a large portion of costs by charging 20 bucks to see these videos.) in the end they got about 1 story a day out (and thats the work of 6 different ENG crews) add to that, somewhere down the line they had audio problems, you could barley hear the sound bytes. they had a disclaimer on the site appologizing for the audio In the middle of all this, the producer of the show (guy who is prone to falure and mistake) the producer lands a 200K contract from OLN to use his footage in their production. before too long, it became apparen t that they screwed up the pre-race interviews, because OLN wants them reshot. I havent seen them, but the word was they were not 'up to broadcast standards' Now I also have heard that some ENG guys I know have been mistreated by the production (which shooting from the trail is a very stressful, sleep depriving job, so the least that can be asked is an accommodating production staff.) Cameras were either PD-150s (which I have never seen an ENG crew shoot like my mother, until now) or dinosauric ENG betas. They droped a few hundred grand on fly-away satelitte gear, expecting every station to pony up cash for live shots and feeds from the trail. Not one station took advantage of this. We arranged to use it one day, but in a classic falure of communication, the people on the ground barred us from sending our live shot, killing the story a mere 10 minutes before we hit air (not a great way to make friends) they only used the satelite gear to send their footage back to be editted and put up on their website (which isnt live, and if you wanted to you can put your betas on a king air turbo prop plane.) when the race is going on there are at least 2-3 flights a day, bringing any dropped dogs back to anchorage, bringing food, hay and media types. Im sure transporting like that is much cheaper than 150K fly-away set-up, and $1500 a day for 15 straight days for sattelite time) I know I shouldnt take pleasure in other falures, but I am amazed at how at every turn they could mess up and waste time and money. They didnt even have talent aranged for the live start until 3 days prior, leaving the talent less than enough time to go over things to say (it was all ad-lib because they cant be bothered to actually 'produce' a show) What amazes me is that everyone who knows the guy who masterminded this arangement is known for his bone-headed mistakes. He makes them all the time. It just seems odd that in a town so small where everyone knows just about everbody whos active in production, its strange that people can forget so quickly how bad these people are. It has nothing to do with talent, Its a combination of lazyness and lack of understanding. I can only imagine what would happen if this guy ever goes into independant film making up here. Has anyone ever worked or been a part of a production that bombed so badly? By the way check cabelas web site for updates-but pretty soon Rachel Scidoris is set to finnish her race. Rachel is an awsome musher that we have done several stories on. She is the first legally blind person to race the iditorod. This is her second race, and last year she had to scratch. So finnishing is an awsome accomplishment. any producers out there that purchase rights to stories-keep an eye on her.

you mentioned using bulet time technique, but that would force a virtual dolly. my question is why? i know that if you wanted to make the effect of long highlight tails on real time footage, then yes you would have multiple cameras exposing at the same time. but if all you want is the tails and you are ok with having everthing be in super-fast speed (as most shots like these are) then you only need one still camera. set the shutter to 30 seconds, or around there for the speed you want in the end, and expose. wait about 5-10 seconds and expose the second one. since virtually everything in frame will clear frame before the 30 second exposure is up, motion and stutter is no concern. that 10 seconds it is not exposing is not needed, because at this exposure length, every frame will be different from eachother. but i know nothing of intervolometers, i know the a-minima even when doing time lapse is restricted to a certain exposure time, i think 1/45 of a second, as it has a 178.2 degree shutter. some film cameras may allow you to expose 30 seconds then transport to next frame, then expose again. more experienced film people than i can elaborate. if you go with using after effects ghosting filter, make sure you shoot without any kind of shutter. the ghosting will be broken up and you wont get the exact same look. same idea, but you wont get the long continuous beam of light. now if you want to have the long streaks, but also need everything in frame moving in real time, then bullet time effect is the only choice, and with that option a new camera would expose every 1/48th of a second, but each one would stay open for 10-30seconds to provide the long shutter effect.

750?!? 3/4 of a million watts? wow. must have been an impressive sight. you ask what lights he used? without knowing and speaking totaly out of my a.. i would guess they built a rig for that. positioning 750 units would be a pain, and expensive. even using the cheapest lights you are still looking a lot of money. maybe they just got the fixtures for the bulbs and made a grid with them. i read that roger dekins mounted a bunch of 2ks on a piece of aluminum when he couldnt even fit tota lights in place. (big lebowski, scene where rich lebowski reveals the ransom note) I bet it was a grid of 2ks on plywood or metal or something. but hey im still talking out of my a.. I have always had great luck with natural light, but then again I live in anchorage, so in the summer i get around 20-22 hours of light.

I have never had the pleasure of working with one of these, but in the fight club and several other movies, I have seen in behind the scene pics what looks like a helium balloon anchored to the ground, with what seems to be an HMI fixture WITHIN the balloon. So instead of bouncing light (which I think would take a lot of light, since it has to travel from ground to bounce and then back to ground, not to mention what happens if the ballon sways outside of the lights path.) It shines through the balloon, softening the light about as much as bounce would. If you look at the night scenes at the paperstreet house, you can see what I am talking about. (I assume they used the balloon for general fill, and maybe a few HMIs to edgelight the charecters, look at the scene where they are hitting golfballs into the dark)

Premeire (i dont knwo about elements but everything after 5.0) has a feature that allows you to slip audio up to 1/30 of a second in very fine increments. True its not nearly as good as pro tools, audition or anyother software set up specifically for audio, but you can sync audio to less than a frame in premiere. I believe other professional programs offer that as well.

43 bucks? I find that really hard to believe. Hell, a tota light from lowell will run you at least 120 bucks without light. Does this 350Y include import to europe or the US? probably not, I would think customs would want at least 43 bucks just to bring it in. (is there a duty free shop for DoPs? Im gonna start one and become rich) Does anyone know if these are even availible in the US? I would love to add a few cheap HMIs to my set, even if they dont work, for 43 bucks I will take the risk (and test the hell out of them before I put them on a set) The real question is can anyone see any obvious cost savings availible in building an HMI light? I cant see anything that could be cut from an HMI and still have the thing work and be usefull.

Please dont judge americans by our goverment. to quote John Stuart on the daily show 'I think 90% of all americans, if they were locked in a room together they would be reasonable. and the other 10% run the country." We have flawed idealistic, opourtunist foriegn policy (fueld by a retarded pupet president backed by insidious folk with hidden aggendas and special intrests.) I was worried for our contry for a while as things seemed to slip downward after 9/11. But if Bushes low numbers on virtually every issue is any indication, Americans are not in line with the government. I can only assume things will turn a corner in 06 and be completely revamped by 08 and hopefully within 4 years a lot of the damage can be undone, and the world comunity see we are still good people. In the end we need a new party. 2 options makes our democracy bland and really more for show. 3 or more parties would force discussion on many options, and partisan attacks would at least take on a new dynamic twist. I also hope that Bush can get impeached for the NSA taps. It shows an administration willing to do anything it wants without regard to apperances. His disregard and disdain for the American public in general is quite frightening. Anyway I got sidetracked. Our goverment has taken down a bad path and have caused huge damage to our image recently. I hope (and vote) for a change to bring our actions closer to our ideals (and our ideals closer to reasonable) But please dont base the american experience on what our goverment does. We're good people......really.....except pat robertson, I cant vouch for him.

If its a day scene I have a hard time believing that the explosion would be too bright for the camera. If its daylight outside you would be stopped down pretty far with ND, so I dont think it should. But if you have the HVXs then run a test. In my limited experience with them they seemed pretty good with highlights, especially if you turn knee on. My recomendation is to fill a barrell with gasoline and light that up in the middle of the day (or when you plan on shooting this that is) and see what it looks like (burning gasoline is about as bright as most mortars are, as 9of10 mortars are black powder for the explosion, and gasoline for the flame. I have a feeling your going to be well within range, but test.

Spielberg is lame. There I said it. He is where he is because of Jaws, and a few other good movies. I'm not saying hes the worst out there, but he really, really really doesnt deserve his reputation. He made a nice turn around with Munich, but I will always hate his status as a filmmaker 'elite' I feel that only those who deserve it should reach that status. Scorsese, Kubric, Hitchcock really had that extra 50% over everyone else that put you in awe everytime they released a picture. Speilberg I go to his film and leave mostly dissapointed (and AI was the worst abominations in recent film history) Speilberg works because he surrounds himself with great people (unfortunatley never surrounds himself with the story he is trying to tell.)

Its much higher quality straight from the Beta (VHS, thank god, look very 1990s these days, and even the average home viewer now can see the difference between VHS and beta, 10 years ago I'm not sure they would. DVDs have upped our expectations.) When I edit my show reel, I hook the Beta machine up to a converter box. This takes analoge in and converts it to IEEE1394 (firewire) just like a mini-DV. I save this to hard drive using Premiere's capture tool and edit as normal. Once the final show reel is done you can use Adobe Media Encoder (included with premiere) to encode the video down to an MPEG-2 and put that on DVD. (I recomend adobes DVD mastering software. I just made my first few disks with it the other night and it feels very powerful, but is simple to use right out of the gate.)

First, loose the dpi. DPI is a meaningless term. It only refers to the output resolution of a printer. Since you wont be prining DPI is a term you can just completely forget about. As to your other question.....How high resolution can you make it? I would recomend at least 1080x1920 (2.3 Mpixel, 16:9 aspect ratio) to match HD resolution. If you can go larger and still work reasonably well then go larger. I would not go anything more than 2160x3840 (4x HD resolution) because there is absolutley no reason to. That resolution is around what 4K is, which is the highest resolution you can scan back to film currently (from what I understand) I have heard of a lot of films starting with 4K, but down-resing to 2K before post. This seems reasonable, because every pixel now has been supersampled with 4 pixels. (if I am correct in thinkin most DSLRs are 1 Chip Bayer masked, then supersampling 4 pixels will yeild a pixel that was combined with all 4 color elements, meaning that edge softness typically associated with 1chip bayer image sensors would be eliminated.) Keep in mind I have not done any film-out from digital option, but this I have done a lot of research on the subject, and I believe this is a similar method to Corpse Bride.

Dont take that advice. I dont know what the diameter of the JVC lens is, but I can bet my studio its bigger than 37mm, and you probably wont get one for $20. You can get a death lense (as they are refered to in snowboarding/skate films) with is a really really wide angle. they are probably closer to 200 new, but can be found on eBay by various failed skate movie makers (probably because they over used the deathlense) I would submit to you that CG is an option, at least its close to an option. If you shoot really really wide and your goal is to master to DVD then add the effect in adobe premiere or whatever you are editting with. There has to be a feature call sphere or something like that, which will basicly map your video onto a sphere, augmenting the wide angle look. Dont push it too far, because its not a true wide, it wont add distortion to closer objects, just distorts the image as a whole, but gets you that feeling. Then just throw a matte of the door and peephole out of focus to give some kind of refereance and natural frame. If you need some help I have premiere and after effects I can get you some help with easy CG stuff.

HDV runs at the same dv25 spec as mini-DV (I beleieve 19.2Mbaud is allocated for video, and the rest I assume is audio, though I have only seen the 19.2M ratting.) So plan on needing about 12gigs for an hour, just like DV. (I mean if it fits an hour on an hour long mini-DV tape, they gotta be close right?) There is no ability to switch between 4:4:4 and 4:4:2 (I have never even heard of 4:4:2, though it may exist. typically you only hear of 4:4:4, 4:2:2 4:1:1 and 4:2:0 when talking about chromanance sub-sampling) Other recording formats have other rates. I can't list them off from memory because I have only editted HDV footage (any job with a budget to rent me a varicam or cinealta, generally has the budget for an editor, so I have yet to deal with that) Though I do know from the movie I am finnishing now that HDCAM is around 140M per sec. (note that this is not HDCAM-SR, just regular HDCAM)