Sean Cunningham

The GH2 was the last GH Panasonic will ever sell me, the last DSLR/Mirrorless I'll ever buy to shoot anything but stills with. I find nothing appealing about the GH4 or the Sony. I'll never buy another compressed camera (lossy, h264 or variants), one that tops out at 422 or needs another box to shoot RAW. BMD has done nothing to sway my interest in their products and these new cameras only solidify my resolve that if I can't have an Alexa it'll be a Blackmagic.

More than once I saw a list that, I think, Andrew Chan posted with maximum aperture results on various lenses, as reported by testers and early adopters. For the life of me I haven't been able to remember where I saw it, in a post or in e-mail and my searching through e-mail has turned up nothing. I know I saw at least one or two lens offering sub f/2 performance. That's more important to non-boosted MFT and BMD (or other S16 size) cameras since, if you're trying to match cine stops, you need to factor in your iris as well as FOV conversions. Going between the GH2 and anamorphic 35mm is a factor of 1.64 based on the differential between sensor and negative. So matching an f/2.8 on film (think the opening shot for Boogie Nights) needs an f/1.7 on the GH2. That's Nikkor 35mm f/1.4 @ f/1.4-2.0 split and the +1.33 so that I could focus as close as I was. Without the achromat I could still get workable focus at this stop from further away but it also firms up the bokeh.

Not according to what I've read (here's one: Vintage Zeiss MF). If you're talking about the 80mm. The older non-MC are supposed to be the ones to get. That's what was used for the 35mm portions of The Master.

It is. And R+H and I expect any day now SPI, at least their Culver facility. But your disco demo doesn't really prove anything to me and the 4:2:0 part is completely irrelevant to the 8bit issue because it's going to have compression and the fact of color under-sampling way out ahead of bit depth ruining your day. A poorly shot/lit GS is still a poorly shot/lit GS no matter the depth. And being able to lift under-exposed footage several stops doesn't really give you better tracking performance. I'm doing that with RED Epic footage now, most of it at a noisy 1000iso, night and dimly lit interiors, and tracking is no better than with under-exposed or otherwise not ideal AVCHD footage. Being able to raise my shadows by two stops or so all the edges and features I need are just crawling with noise just like high ISO GH footage. All the underlying problems making the process more difficult than we'd like it or than it should be aren't really bit depth related. That's 12/16bit raw -> EXR -> float workflow and none of what's holding you up with your footage is really any different. I have to be prepared for each shot to break and be ready to step in to do what the software throws up its hands over. Same as five years ago. Same as ten years ago. Same as twenty years ago. That's just how it is. It is a lot better now though, even though more folks take for granted our tools are pretty amazing now and act like they don't have to try and give us good material to work with.

I was there when we proved to a screening room full of film professionals that we could scan film at DD (creator of that software you're using), carefully construct an 8bit graded file from the scan and then output scan back to film to be intercut with only one filmmaker in attendance able to pick out any of the digital shots. And our finals re-defined the transparency of digital effects. That's cool you're using Nuke. I was maybe the first digital artist to final a Nuke composite on the first project it was ever used on because almost nobody else wanted to deal with it (mostly Flame, doing stellar 8bit tracking and keying). Back then it was a compositing language. The nodes came in v2 and I was one of the few artists giving the new version a go and sticking with it while Bill pushed out new versions daily that would totally screw your job mid render.

You complained 8bit imagery made a laundry list of things not just hard but implied nearly impossible. I'll give you exaggeration but the 8bits is not what makes any of the things you described harder. "Perfect" tracking and clean keys are absolutely doable with 8bit imagery (this accounts for a majority of film work leading into the early '00s) if the material is good and the actual factors that make either task something approaching easy are aligned in your favor. All of the factors that actually make these tasks a pain will be equally bothersome no matter how many bits you throw at them.

No, not relative to other factors. What's more likely to affect a key isn't the bit depth. Bit depth is even less a factor in tracking. I don't know how many ways to say it. If you're having troubles it's not because of bit depth.

That's just the nature of tracking. It's a lot better now than it was even five years ago and that was a lot better than 10 years ago. It gets better as the software gets smarter. Bit depth isn't a factor next to noise, focus, occlusion, color sub-sample (422, 420, 411, 311) or compression. I was going to suggest perhaps not commercially, but that's not true. Before becoming Mocha, Imagineer System's Mokey was doing planar tracking in 2000. ILM was using similar techniques as available in Mokey by the mid-late 1990s.

8bit has no impact of any kind on tracking and keys just fine if un-compressed. All of this stuff has been done for decades with 8bit and the software to do it was all developed using 8bit footage.

Yeah, I would imagine your attachments aren't going to be light either. I was just looking at some charts showing the history of the double gauss and outside a few mostly non-applicable lenses they seem to be mostly in the ~50mm range with a bunch of 40mm and just a few 35mm. Bummer. The Voigtlander Nokton 25mm is one though it's just MFT and quite espensive. BTW, do you guys tend to go for a particular Helios 44 for FF58 or is this something that can be to-order? The way they kind of drastically change character at the jump from f/2.8 to f/2 seems stronger with some than others. I've seen at least one tester with a handful of them and they were all different. They ended up giving away all but three 44-2 that even among those seemed fairly unique in character. I was almost shocked how soft the Iscorama-54 was in: > ...and here you see two different Helios 44 with very different character and relative softness in the combo.

Ah, that's even better human verification! Cheers.

So that is to the Jena 80mm as the Helios is to the Zeiss 58mm? Do you or have you had access to both? I'm curious how they compare, not that the real Jena are terribly expensive. Even if the design itself is a carbon copy I would imagine the source or formula for the glass itself and coatings would make some kind of difference. Even with the Helios 44 there seems to be a palette of available looks based on era, manufacturer, etc. even though they're all the same lens. You kinda roll the dice. A mathematician and human visual system specialist friend of mine has interesting things to say about Russian glass that syncs right up to what I think I recall Rich mentioning about the glass itself and micro diffusion properties. They have a built in point spread function that's been engineered out of modern glass that wants to simply pass light through as unaffected as possible (which exacerbates the bad mojo you get from fixed grid sampling in CCD and CMOS).

Something to consider while looking at the Jena series is I've read steer towards the Pentacon 6 mount and not the M42 version. P6 is the medium format version. Performance is reportedly about twice what the numbers would lead you to expect when paired with a smaller format camera. Also, it may be better to go for the "zebra" versions rather than the black painted or anodized "MC" version because I've read at least one reference that states the coatings on these later versions suffered from supply problems in communist DDR. The earlier, larger, single-coated versions are going to have more "magic". I don't know if this is true for the entire family of Jena glass or just the longer end, or just folklore but the proof is in the pudding AFAIC with the 80mm Biometer. The DSO option sounds like a great alternative if nothing in the 28-35mm range pans out.

I'd call Adorama and get a human to verify that they do have them in stock and you weren't caught in a catalog glitch. I've never bought from them but I've had that happen with another online retailer whose online catalog was guilty of "wishful thinking". It resulted in my order being delayed even further than expected.

It works at f/1.4 on my Nikkor 35mm (GH2). They're being cautious with recommendations since performance is very taking-lens sensitive.

Barlow, you mad bro? I'm still waiting for you to try to back up your statements about focus pulling or more specifically focus pulling on steadicam and how you think it's done. I'd like to see anything you've actually done that doesn't involve sitting at a bench in a shed.

Rich, they're not taking any of that into account. They see 1.33x and it's nothing but hate. And any separation from the background just means faster lenses if you're removing bokeh distortion from the equation. One can't rationally slam a higher compression format for lack of bokeh distortion and then exempt a lesser compression format with even less distortion from similar criticism. That's simply arbitrary rationalization on their part. It's not against the law though and certainly not the most foul thing you find rampant on the internets. But, besides that unpleasantness, part of what you're observing, for now, is testers are limiting their shots to f/2.8 because results are so dependent on the taking lens design. Both stop and focal length are generally being picked in fairly conservative ranges (though still better than was usable on previous 1.33x adapters), so that prospective customers don't get an unrealistic impression of what they can expect. That's part of the reason I didn't shoot anything on my 105mm Nikkor, though translated to anamorphic 35mm equivalency this isn't a focal length that's on my priority list. For instance, that F. Zuiko 50mm (82mm for ana-35mm) of mine on its own isn't really sharp and free of issues until f/4 and at f/2.8 it's usable with the Anamorphot but not as sharp as I'd like. It'll eventually be moved to my "spherical only" set for the time being. Since acquiring it I'd boxed up my Nikkor 50mm and put it in storage but now I want to dig it out and compare it. I can tell you my 35mm Nikkor (~57mm) has workable focus at f/1.4 and at that stop the only problems that are present are from the Nikkor itself, which, like the F.Zuiko, has coma and low contrast ( the low contrast may or may not be objectionable). My 24mm is good at f/2.8 and workable at f/2 but that's still more stopped down than I want to be, so I'll be on the hunt for a 24/25mm that is good at at least f/1.7 or f/1.8 which would correspond to a decent stop at 40mm. All of these equivalencies and stops could shift, however, depending on when or whether I upgrade camera anytime soon.

There was a similar thread over on PV about auto-focus on steadi. In that case it sounded fairly viable since it was going to be used in a wedding video (I think that's right), traveling up and down the aisle so there wouldn't be a lot of cross movement in frame or competition for what should be in focus. The content being shot will likely play as big a part as AF sophistication in terms of results. I wouldn't be surprised if that iPad/iPhone interface for focus isn't out there in some capacity already. I'm sure the fellow who was pulling focus on The Shining would have relished such a magical device.

A) you're not that important, I went to an interesting topic. I have you on ignore but I was curious what kind of ignorant shit you might be saying based on the ignorant shit you had to say about focus pulling on Argo, so I took a peek. B) you can't actually refute what I said, so your go-to is this. Bravo. C) "bigot"? You're an idiot. You use words like they have no meaning. ...I said nothing of myself here. I pointed out what you're saying is wrong and you have no defense other than ad hominem and getting personal.

"Not good"? Barlow, you don't know what you're talking about. That's exactly how focus pulling works. That's what separates good 1st ACs from great 1st ACs. They're constantly judging distance on the move with steadicam operation. If someone doesn't have the budget for the remote system and a great 1st AC, then the op deals with a fixed focus. Otherwise the focus puller is always working. Always.

You're certainly entitled to your opinion but saying it's the same as cropped video is like saying GH2 is like 5D footage. It's the same principle. This is related to why 2.40 Super 35 looks and feels nothing like anamorphic even if you try and match FOV. It's not just about bokeh and flare. Most people don't get it. I have to laugh about comments regarding compression ratio though. They all show an unenlightened appreciation for anamorphic photography in general. What folks who shit on 1.33x are saying is that films like Ben Hur, or How the West Was Won and the other handful of Ultra-Panavision films are somehow degraded by using 1.25x optics. It's the same thing. Those people were idiots and should have just cropped, is what they're saying. They're trying to justify a hater attitude. Smart people will see through it though, I hope.

Hah-hah, this is rich. Blah-blah, blah-blah-blah.

Stuart, I see you've met Gollum. He wants to talk to you about his precious. He might actually have less practical experience than a babysitter working under the 2nd AD and so he clicks "like" and takes his opportunity to commit irony. He must have been lost in trance, staring into his collection of glass. I expected his emergence from the shed earlier.

Actually it's not. But, again, best of luck to you.

Thank you for proving my point further. Best of luck to you.