Ralph Oshiro

I agree. The F330/350s are very slow. I was shooting with a 900R recently, and it seemed much faster than the 900/3, perhaps by as much as a full stop. But I wouldn't even consider a 1/2" camera, and if you put 2/3" glass on a 1/2" camera, wide shots will be a problem to get. The Panasonic 3000/2000/900s look very nice, and they're all equipped with 2/3" imagers, and all are 24p-capable. But as a business decision, XDCAM may be the better camera for ENG markets. CBS has committed to XDCAM-HD (F350s for O&Os), but ABC and NBC haven't committed to a format yet. If I were going to go XDCAM, I would try to wait for the new 2/3" XDCAM-HD cameras due Q4 2007, reportedly with better-than-HDCAM 4:2:2 color sampling and a 50Mbps max datarate. It all depends on your customer base and what media is going to be most common among your clients if you're an owner/operator kinda business--if you're a production company, your choice of acquisition format is less important.

Yes, but to paraphrase another member's post, it's MUCH easier to shoot ugly video than it is to shoot ugly film. The aesthetic and technical challenges have been so much greater in the "electronic cinematography" world. Anyone remember the Ikegami EC35? Let's face it. Video sucks. And has sucked for a long time. For video, you have to light scenes to a fraction of the contrast ratio of what a film DP can light to. When the first Digital Betacam cameras came out, Sony claimed an 11-stop dynamic range . . . I mean what were they smoking? Video loses detail in its highlights faster than you can say, "DCC." Video has "noise"--not "grain." And noise NEVER looks "organic" or "filmic." Noise just looks like sh-t. And another thing, video has traditionally had this horribly narrow, ugly color gamut. I mean, how many horribly DP'd F900 "films" have you seen? I was watching this DP light this F900-lensed, major network MOW once. He lit it like a f--king soap opera. I don't know why. He had a nook light bounced into the ceiling on some Foamcore--I mean there was f--king light EVERYWHERE. But I digress. Up until recently, 2/3" imagers were IT! And that 3:1:1 color sampling scheme? What was the deal with THAT? Okay, so now we have SR. Big deal! It weighs a ton, and the VTR costs a $100K! Might as well shoot on 1" videotape again! Hopefully, RED has and will change some or all of this.

Hey there, Carl! Did you get a chance to see the demo at the RED booth at NAB?

I guess now this thread would be better served in the Canon XL-1 forum . . .

I stand corrected. They used as many as 8 Canon XL-1s using, "the higher-resolving Canon EC (6-40mm) and Canon EJ (50-150mm) prime lenses to the camera bodies with Optex adapters." Excerpted from the American Cinematographer article here . . . 28 Days Later Article in American Cinematographer

Ahhh . . . perhaps the aerial photography sequence?

"BODDINGTON . . . BODDINGTON, BODDINGTON, BODDINGTON!!!"

From what I recall, the whole movie was shot with 10 Canon XL-1s, fitted with special lenses (forgot which ones).

David Mullen made an excellent post that responded to this question in another thread. Something to the effect that 35mm film has been the gold standard of what we define as a culture of motion picture imaging, and it's a natural tendency to compare RED with film. My response, is, it doesn't have to! The RED spokesperson in the Apple ProRes demo stated (to paraphrase) that it doesn't quite look like video, and it doesn't quite look like film--it's something different.

Agreed. RED doesn't have the latitude or the wider color gamut of film. And no electronic acquisition system I've seen mimics the motion characteristics of film quite right either. But anyone who's seen the 4K projected demo can make their own qualifications. For $17.5K, you can't beat this image quality with a stick. Completely noiseless, ultra-high resolution imagery. No, it doesn't quite look like film. What it does look like is pretty awesome, though.

Yup. I think I muttered those exact same words at NAB2006. Exactly 12 months later, at NAB2007, I saw footage and working prototypes. If RED maintains their published production schedule, I should be shooting something with my RED before Turkey Day this year.

1. Third-party UW housing manufacturers are currently working on a RED housing. 2. RED is spec'd to do 120fps at 2K. 3. RED is currently developing a smaller version of the camera, which I would guess may be an excellent crash camera.

Clearly, that's exactly what many of us want to know! The honest review of a working DP in real-world applications. While the Peter Jackson short was stunning, IMO, I can't wait to see some studio-lit interiors, and some low-light, night exteriors shot with RED. Since the current iteration of the camera is rated with only a moderate sensitivity (200 ASA), their initial exposure methodology has been reported to be to, "bring up any underexposure in post." They claim this is a perfectly workable methodology due to the camera's low-noise picture. Also, some mention has been made of the possibility of an in-camera gain-up control, and that gain-up was considered by RED staff to be acceptable up to about an 800 ASA equivalent. Again, can't wait to see some contrasty, MTV-ish kinda footage from RED.

Well, I'm sure I speak for many here, that we would all look forward to your honest evaluation of RED's performance. I know that both RED and some early reservation-holders are anxious to know what you think. I'm sure the opportunity will present itself very soon (I thought it already had, actually), and I very much look forward to your review and commentary.

David Mullen: As one of the more respected members of this board, I was wondering, if your busy schedule will permit a chance for you to demo a RED camera for yourself in the near future? One of the sub-1000 reservation holders at REDUSER has offered to allow me to demo his RED when he receives it, and I plan to invite another DP friend of mine (a skeptic, by the way) to the party as well, later this summer.

Good one, Richard! This post is actually not a specific response to your recent post, but I actually thought that was a pretty perceptive statement. I haven't commented on this board since a few weeks before NAB, when the crosstalk here was beginning to crescendo to a fever pitch. But now that many have seen the Peter Jackson footage, everything seems to be a bit more relaxed now. I would like to comment on the general reaction to both Peter Jackson's short film shown at NAB2007, "Crossing the Line," and the previous "Milk Girls" RED demo footage, shown earlier this year in Los Angeles and elsewhere. This is more a commentary on style rather than anything categorically technical. Both "Milk Girls" and "Crossing the Line" are daylight exteriors with only moderate, "conventional-looking" color grading applied. I agree with some here who have commented on RED's footage as having a slightly "video-ish" patina. I think what we haven't seen yet are contrasty, chroma-rich, MTV-ish, lit interiors (sets), with more of an edgy look. Given this look, I think RED's ultra-low noise image will be even more impressive. Many, including me, associate light to moderate amounts of film grain to be an expected, even desriable part of the film aesthetic. A tight, high-acutance grain pattern can look ultra cool, yet still sharp, and full of image fidelity. But video noise, at least in my opinion, has never been a desired part of any aesthetic, except for effect. RED's noiseless electronic image may not look exactly like film, but there's nothing like starting from a clean image to grade from (where chroma noise often only gets amplified even further). As a corollary to Richard's "video look" statement above, no one ever tries to add video noise to make their film look more video-ish (except for effect).

Hey, thanks!

Could you post a link to a source for the "Vantage" filters, please?

Yes, I agree with you, Michael. However, I got the feeling that the original poster was tossed an ultra-low budget gig in his lap without the benefit of proper prepro to iron out those kinds of issues. I just didn't want him to go through the nightmare of depending on the house mixer and end up with no audio (or horribly impedence mis-matched audio). I'm certainly no expert in live concert mic technique or sound reinforcement mixing, but whenever I need a feed from a house board, it's always a nightmare for some reason. Oddly, they never seem to have anywhere near the right impedence or the right mechanical connector. Worse case, mic the PA system, and try to record some kind of SMPTE jam-syncable, second-system sound from the house board.

My $0.04: I produced/directed a pretty bare-bones, HDCAM live concert using four Sony HDW-F900s a few months ago. You probably know most of this stuff already, but here's what I did. We basically had four iso-ed F900 packages. I put one camera on each side of the stage, one with a wide angle, and one with a standard lens--the idea being, one camera gets all the close, wide shots, the other gets all the close-ups. For interest, I rented dutch heads for both side cameras. Even though I had four totally pro operators, I didn't want any "shaky" handheld shots--I wanted really precision-framed shots. The dutch head I rented required a minimum Sacthler 7+7 as a primary head, to be able support the added weight of the dutch head. The center camera I had on a track dolly, traversing the length of the stage. The master "safety" camera was on sticks near the mixing console. We also had rented a standard PL system (more cable to lay!). As far as video was concerned, this is all what I worried about technically . . . 1. Jam synced all cameras by walking an arbitrary "master" camera to each camera, and stinging its TC-out BNC into the target camera's TC-in, (then setting the target camera to "free run") to jam sync. 2. Make sure all cameras are either already SDI-out capable or that you have rented the appropriate accessory to get SDI out of your HD camera (this is ONLY for monitoring--all cameras record locally). 3. Make sure all of your HD preview monitors are SDI monitors. 4. In essence, make sure NOTHING is component, otherwise, you'll be stringing miles of component BNC cable of exactly the same length, just to see your pictures. When on a small budget, and you don't have dedicated A2s working all your audio bugs out, if at all possible--DO NOT RELY ON THE HOUSE MIXER!!! Plan to mic and mix for your own sound. Send that audio to as many cameras as is practical (I would send it to at least two--one for back-up). You're jam synced, so you should be able to lay down the synched track later to any cameras that aren't receiving program audio. Throw up camera mic on those not receiving audio for reference. So to sum up, the essentials were . . . x4 HDW-F900 camera heads x4 HDCA-901 HD-SDI adapters (or equivalent) x3 standard lenses x1 wide-angle lens x4 HD-SDI monitors x4 single-muff RTS headsets x2 double-muff RTS headsets x2 dutch heads x4 fluid heads/sticks x1 track dolly Each F900 package, I think ran us about $800/day. But expect to pay as much as $1,200/day. Plus whatever the HD-SDI adapter rents for. I didn't pay for it, so I don't know how much the ancillary gear added up to. I think the show turned out looking great for such a bargain-basement budget, thanks to the inexpensive track dolly, and relatively affordable added cost for the dutch heads--lotta added production value for the money there, especially since we didn't have the money or space for a jib.

Not sure what you're saying here. Why would the authors transode the Betacam source tape to another format for evaluation? The Betacam signal was recorded on a Betacam VTR, played back on a Betacam VTR, and evaluated on standard terminal test gear. The DV signal was recorded on a DV VTR, played back on a DV VTR, and evaluated using standard terminal test gear. No transcoding.

If he telecines to Betacam, you would take a generation hit dubbing from that to a DV tape later, since it's an analog-to-digital transcode. Plus, few dub bays are set-up for component Beta-DV dubbing, so you'd likely only get a composite dub (which surprisingly, actually looks pretty good). Going the other way, a composite dub from DV to Beta looks really awful for some reason, and even fewer bays are set-up for a component DV-Beta dub.

I think he means, "reassemble." Meaning, when finishing, "assemble" (edit) using the footage higher quality "master."

It's true, all three formats, miniDV, DVCAM, and DVCPRO25 have the same 25Mbps datarate. Just to clarify, however, miniDV and DVCAM tape are pretty much identical (both are metal evaporated). It's DVCPRO tape that is more robust (metal particle vs. metal evaporated). The DVCAM format's specification for a wider track pitch makes its signal slightly more mechanically robust than miniDV, and DVCPRO's even wider track pitch makes it the most mechanically robust (and with its metal particle formulation, its media is also the most robust) of the three 25Mbps tape formats.

My $0.04 . . . Videography magazine did a side-by-side test a looooooong time ago, using an identical camera head, but with different recorders. They compared the recorded signal quality of analog BetacamSP, recorded on a dockable Sony PVV BetacamSP deck, with a dockable Sony DVCAM deck (DSR-1?). Using dockable recorders allowed the authors to compare apples to apples, using the same camera head (I forget which head it was). While Betacam may win the resolution claim, I distinctly remember the article claiming that the DV bitstream exhibited superior chroma characteristics when compared with the analog Betacam signal. As far as the resolution claim, I think that claim may benefit from an explanation with a bit more detail (excuse the pun), since we're comparing two essentially disparate formats, one, analog, and the other, digital. So it's not really an apples to apples comparison when talking about resolution, if I understand this correctly. Perhaps someone here, more engineering-savvy than me can do the math, but luma resolution, when measured in an analog format, is a calculation which I don't exactly remember, but as I recall, the rule of thumb used to be something like 100 lines of resolution per 1 MHz of bandwidth. So, the Betacam PVV VTR's 4.5 MHz of luma bandwidth is roughly equal to about 450 lines (vs. DV's luma bandwidth of 5.0 MHz). Feel free to correct me if I have this screwed up somehow.