Jump to content

Logmar S8 footage revealed!


Lasse Roedtnes

Recommended Posts

 

The camera must transport intermittently the film as much steady as possible "only". The lens do the real job. There are already plenty of professional 16 and 35 cameras out there

Very true. But if these Danes can produce a fairly compact new 16mm and 35mm camera much cheaper, and presumably with pin registration, that would surely be welcome.

Link to comment
Share on other sites

Just like to add, I really like the idea of the sound recording option on their super-8 camera. If this was available on future16 and 35 models, that would be great. However, I am still not clear (maybe I missed something during that previous LONG thread ?!) about the actual mechanical noise produced by the super-8 model. This surely is quite important to consider if sound is to be recorded near the camera. Is blimping necessary ? And of course 16 and 35 is likely to be louder.

 

Great test-film, Friedemann. Look forward to more.

Link to comment
Share on other sites

Lasse & Tommy,

I just saw the first sample footage, produced by Friedemann Wachsmuth, and I was quite taken aback by it.
It is PERFECTLY STILL!!! It’s absolutely mindblowing – there is no movement AT ALL at the image dividing line. And the wide image ratio of the gate is impressive, it even looks slightly a bit wider than MAX-8/ SuperDuper-8. (Can the round corners be omitted completely, though?)

One thing is absolutely clear from seeing this footage:
This sets A COMPLETELY NEW STANDARD for 8mm filming, especially in Super-8. In fact, it resembles 16mm or even Super16 very much.

I have a recent example of what we are used to:
I was watching a projected image just last night – a Super-8 film of Blackbirds just outide our kitchen window in winter, shot on the old fine-grained 50 ASA Plus-X Reversal 7276 *some years ago* (1977), with the best of my Beaulieu 4008 ZMII cameras (with the superb Optivaron 1.8/6-66), on a very steady MILLER wooden tripod with a fluid head, projected on an ELMO ST-1200 M/O (with a cleaned gate). I used the long-throw 1.4/25-50mm ELMO lens, which is very sharp, and I had a black cardboard disc in front of the lens with a smaller aperture opening in the center of it (5.6-8?) to bring out the best in the lens. Adjusting the frame line up a bit, so I could clearly see it in focus, I was standing just a few meters away, looking at that frame line for the duration of the film. And, as usual with cartridge Super-8, it is all over the place vertically. It widens and shrinks all of the time, and sometimes there is more pronounced jitter (usually at the end of a 50ft cartridge). The image is tack sharp, to the point where I can clearly see the the grain structure across the entire screen (about 1,8 meters wide), when the lens is focused at it’s finest (not easy on an ELMO!) – but the image stability is terrible, it just jumps up and down – not that much, but it is annoying enough up close…

The footage of Friedemann's sample film (on scanned Vision3 50D negative) is something else entirely, quite another ballpark. A new kid’s in town!
I want one, too... :)

Thank you for sharing it with us.

Regards,
Bengt Fredén, photographer
Stockholm, Sweden

Link to comment
Share on other sites

Hi Lasse and Tommy,

After viewing the sample test footage by Friedemann (over and over again) on Vimeo: http://vimeo.com/87243287

looking very critically at the borders around the camera gate, there seems to be a VERY SLIGHT rocking sideways, or perhaps better explained as a turning/screwing in a central axis, especially at the very end of the sample, before the film goes completely out of the scanner. It is best seen in the scene near the end with the wall of the high building, that completely fills the frame. It looks like the image is turned or rocked very slightly to the right, viewed (as here) with the image the right way (it would of course be upside-down in the camera or scanner gates!).

 

It would be really interesting to know whether this is something from some very minor play in the camera gate (I doubt it), bad film sprocketing from KODAK (as in fact seen in a recent test by Friedemann, with the very same film stock - Visoon3 50D), or some very minor play in the gate of the 2K scanner station at GammaRayDigital (film feeders or side springs/ sprocketless wheels?).

 

At any rate, it is VERY MINOR (we are talking thousands of a millimeter of sideways play here!) and the camera registration vertically is still OUTSTANDING by any 8mm standard!!!

Another thing that I find to be very important to address is whether the round corners of the camera gate can be omitted?

To be able to maximize the usable area of the camera format (to gain the best possible quality from the SuperDuper-8 format), there is now a limit to how far out into the format corners you can go when framing the image for the final digital showing format. It is apparently as close as possible to the sprocket hole (as we can see from the very slight orange discoloration to the left in the scanned image). If the round corners can be filed out, lenses that cover a wider circle than the Super8 zooms do (at least at their wideangle end), for example many 16mm/Super16 zooms (Canon, Cooke or Angénieux comes to mind) or hi-quality prime CCTV lenses (like Fujinon) would project a bright image all the way into the very corners.

Here's the interesting test that Friedemann made of the varying sprocket punching of the KODAK Vision3 50D negative film in 50ft cartridges:

 

A final word:

The amazing quality of this camera, and especially it's pin-locked registration movement, puts a lot of focus on the very best lenses obtainable (e.g. the Leitz Leicina Macro-Cinegon F1.8/10mm prime lens!), working with the best possible aperture settings of the lens in question (not so easy with 50D!), ultimate back focal plane calibration (collimation) of every lens screwed into the camera C-mount, extremely narrow tolerances in the making of the film material (KODAK will probably have to revisit this issue!) and of course extreme care (as usual) with the choice of the best camera supports (tightly locked camera plates, solid fluid heads, heavy-duty tripods, etc etc).

 

The future of Super8 is extremely exciting!
Thank you all.

Bengt Fredén, photographer,

Stockholm, Sweden

Link to comment
Share on other sites

Hi again,

I watched the footage again, and it seems clearer to me now that it must be in the scanning process, because the whole GATE FRAME (the black border) rocks or turns very slightly sideways to the right, along with the image, towards the end of the footage.

 

So, the camera's pin-locked registration is not to blame, of course.

 

I read at GammaRayDigital's web site that the 2K fram-by-frame scanning of small format film is done with digital pin registration. The question is what guides the film sideways? Any takes on this, Perry?

 

Bengt Fredén, photographer

Stockholm, Sweden

Link to comment
Share on other sites

Here's a thought:

Perhaps a couple of extra meters/feet of END LEADER at the very end of the film would keep it in place better (sideways), as it is running out of the scanner?..

Bengt Fredén, photographer
Stockholm, Sweden

Edited by Bengt Freden
Link to comment
Share on other sites

I have a question for Friedemann:

From how many 50ft rolls of Vision3 50D negative was this sample footage composed? I assume it must be at least two rolls, as the whole footage is 3:46 min in length, whereas you normally only get a little more than 2:30 min on a 50ft roll (at 24 or 25 fps).

I likewise assume that must have edited some of the cuts in FCPX? Because any physical cutting of the film (cement or tape splicing) might cause the film to shake or jitter sideways. Thank you, Friedemann, for sharing your findings.

All the best,
Bengt, photographer
Stockholm, Sweden

Link to comment
Share on other sites

I read at GammaRayDigital's web site that the 2K fram-by-frame scanning of small format film is done with digital pin registration. The question is what guides the film sideways? Any takes on this, Perry?

 

To answer your specific question: the ScanStation uses edge rollers to keep the film in approximately the correct location as it passes across the gate. Precise registration is done digitally, before the frame is written to disk. This is by design: with shrunken film, the amount of shrinkage can vary even within the same roll, so the edge rollers make sure the whole frame is imaged, but the digital registration is done to reposition each frame so it's where it should be.

 

A registration pin's only point of reference is a perforation. If that perf is in a different location on the film than the other perfs, you'll see the film itself move slightly in the gate. The camera will always hold the perf in the same spot relative to the gate, but the film could move slightly if the perf isn't in the right spot. Since the whole point of pin registration is to hold the film in a known location in the camera (or scanner), the perforation is all the camera (or scanner) has to go by. If the perf is not placed precisely relative to the edge of the film, it will cause the whole strip of film to move around.

 

The scanner's digital pin registration behaves like mechanical pin registration, though it's more accurate. That is, it looks only at the perforation and aligns that perforation to a fixed position on the X/Y axes. It does this for every frame, while scanning, with no post processing. In its simplest form, if it sees the perf is 1 pixel to the left of where it should be, it moves the whole image by 1 pixel on that axis, so that the perf aligns perfectly with the previous and subsequent frames. Friedemann's perforation video shows this - notice that the perfs themselves don't move, but the image does. The ScanStation doesn't look at the edges of the frames - it's NOT doing motion stabilization like you may be familiar with in an NLE or compositing tool, it's only doing pin registration - minus the mechanical pin, of course. You would see the same effect in a mechanically pin-registered scanner.

 

Therefore, if the perf is poorly placed, the film will be slightly off center in the camera when registered. Because of this, you will see the gap between the perf and the edge of the picture will vary slightly. If you look at Friedemann's perforation video, you'll see that this is the case: the gap between perf and picture widens and narrows in a fixed pattern. This example was scanned here on the ScanStation as well, by the way.

 

I believe Friedemann said he applied some stabilization to the scanned test film, to correct for this film manufacturing error. It's possible that some of what you've described is related to that. Stabilization algorithms will vary from application to application, so you may see better or worse results, depending on what you're using to stabilize, as well as the parameters set within the stabilization tool itself.

 

So, what you have here is a very precise camera and a very precise scanner, but not very precise film, it seems.

 

-perry

Link to comment
Share on other sites

Here's a thought:

 

Perhaps a couple of extra meters/feet of END LEADER at the very end of the film would keep it in place better (sideways), as it is running out of the scanner?..

 

There's 12 feet of leader at the head and tail of the reels in the scanner - the film never "runs through" like it might in a projector. Tension is kept at a constant at all times and the film must be "loaded" in software (tensions applied, registration readied, base calibrated, etc), before you can start to scan. Before you can remove the film from the scanner, you have to "unload" in software - this unlocks the motors and "relaxes" the machine so you can unthread the film.

Link to comment
Share on other sites

Thank, you, Perry,

 

for those very thorough replies to my questions. I much better understand the scanning process now, and the slight wavering or rather tilting sideways to the right of the whole image must then be attributed to the tolarences of the physical 8mm film, if I get it right?

However, as you so eloquently explain, the sprocket hole on the film is ALWAYS relative to the camera gate, irrespective of varying film width or, as in the case which Friedemann so clearly shows us, of repetitive errors in the punching of the sprocket holes in the manufacturing of the original film stock.

Film, after all, is a living, breathing thing..!

 

Best regards,
Bengt Fredén, photographer
Stockholm, Sweden

 

PS. I read, by the way, the whole of the PDF article (link in former Logmar thread) called "The Kink in The Cartridge", which is, although written many years back, extremely informative as to what actually happens in the merger of the Super-8 camera gate and the cartridge's inherent plastic pressure "pad" (not plate!), with the film suspended freely in between. Only three small points of contact, really. A lot can be learned from this article - most enlightening.

Link to comment
Share on other sites

Friedemann,

Could you please explain, as understandable as possible to us mortals, the type and degree of software stabilization that you applied to the scanned film sample? To what degree (if any) did it change or alter the original (photographed) Super-8 image, which is what I thought I was watching . .?

Thank you.

Bengt Fredén, photographer
Stockholm, Sweden

Link to comment
Share on other sites

There's one thing that I instantly came to think of, with regard to various stabilization software, having seen it recently:

When trying very hard to restore and read the rather fuzzy contents of the famous Regular-/Normal-/Dual-8mm film material known as "The Zapruder Film", which shows the shooting of president Robert Kennedy in Dallas in 1963, MASSIVE amounts of digital registration and filming stability error corrections in all directions (up, down, sideways) have been done to the material, to be able to clearly see what actually is going on in front of the film camera, which is held by hand and also panning. The original footage is INCREDIBLY shaky. This digital restoration can actually be viewed (before and after) on a YouTube video clip. Very interesting indeed to watch. It is quite astonishing what can be done to shaky film footage these days.

With respect,
Bengt Fredén, photographer
Stockholm, Sweden

Link to comment
Share on other sites

Hi Lasse and Tommy,

 

What is your reasoning, or thoughts, behind the current round corners of the LOGMAR prototype camera's gate? Is it harder or more costly to produce gates with sharp-angled corners?

 

I would imagine that the end result of most user's of the upcoming production camera would be 1,78:1 (16x9) or thereabouts, being presented on a screen or TV panel. Then there might just be a problem, if you really want to use the FULL potential of the SuperDuper-8/MAX-8 format, from close to the sprocket hole over to (or at least very near to) the other edge of the film. (I will try to measure this on my screen, to be sure).

 

However, if your objective is feature film cinema theater presentation in a 1,85:1 ratio (or wider still) there wouldn't be a problem. But I am very much in favour (inconsistent BrE and AmE spelling here, I know) of the beautiful (in my view) slightly 'higher' European theater format of 1,66:1 aspect ratio (about the same as Super16 native camera format), which is roughly what we are seeing in the first video sample from the prototype LOGMAR camera.

 

A fine film example (one of my personal favourites) in this ratio, by the way, if you'd like to see it, is The Draughtsman's Contract by director Peter Greenaway - the first feature film shot in Super16, on an Aaton camera, and 'blown up' optically (before the days of the Digital Intermediate) to 35mm distribution format. (It might have been cropped slightly for distribution in the US). It is simply breathtaking, even if there has been major advances in development of negative film stocks since the late 60s (Vision3).

 

IF I were to buy a LOGMAR camera, I wouldn't want to start filing away at the corners of the camera gate by hand (God forbid!).

So, would it be feasible, early in the production stages, to modify the existing gate to more sharp-angled corners, or at least corners with a less pronounced round shape? Thank you both.

 

Please forgive my ramblings -

I am so thrilled with the possibilities of this camera!

All the best,
Bengt in Stockholm

 

PS. By the way, please DO keep the in-built Nutrik XLR and 48V Phantom power contacts for professional Sennheiser shotgun microphones! Blimping the camera and lens (if it needs it) shouldn't be too difficult.

Link to comment
Share on other sites

Hi Bengt

 

The gate on the prototype was milled out with a 1mm iron, that gives round corners (radius 0,5mm). The production version will be laser cut with right angle corners (sharp corners) The dimensions are 4,2 x 6,3 mm. So don`t worry. And the Neutrik and 48V will be as well.

 

Regards Tommy M. Logmar

Link to comment
Share on other sites

Hi Tommy,

 

Thank you for that clarifying reply. That is most reassuring! No problem, then, to use the FULL potential of the wider format.

And thank you so much for the gate format specs - I will draw some diagrams around that, and play around with aspect ratios.

 

Best regards,

Bengt in Stockholm

 

PS. Regarding the movie I was referring to earlier, I found some more info on it on the 16mm forum here at Cinematography.com:

"THE DRAUGHTSMAN'S CONTRACT (shot in 1983) by Peter Greenaway, photographed by DoP Curtis Clarke with an Aaton Super16 camera, Zeiss primes and the (then) new Cooke Super16 Kinevarotal 10,4-52mm zoom"

(A bit off-topic):

At Cooke's web site, one can read about the zoom lens above: "This lens was first used by American cinematographer Curtis Clarke to film "The Draughtsman’s Contract," the first technically and commercially successful Super16 feature to be made. The lens offered advancements in filming under difficult lighting conditions in 16mm and Super16mm formats. It began production in 1983."

Link to comment
Share on other sites

  • 3 weeks later...

Hi Anthony,

 

Any news on the upcoming release?

 

Yes we received the final metal parts this Friday and the electronics are up and running so it looks good from my perspective!

 

I put some photo's on our webpage under the blog section. (http://www.logmar.dk/?cat=6)

 

We do not, unfortunately, have time to film demo footage with the new camera before the Diedesheim fair as we are scrambling to assemble the insanely many parts - apart from a few minor internal tweaks to the electronics and mechanics we will bring the "production model" camera to the fair so people can see exactly what it is we are offering.

 

Enjoy!

 

/Lasse

  • Upvote 1
Link to comment
Share on other sites

  • 1 month later...

Re. the horizontal weave issue.

 

I'm not sure how the Logmar gate is constructed (or any other Super8 camera) but looking at a Leicina, the horizontal registration of the film is provided by a fixed edge guide on one side of the gate (the sprocket side) against which the film is held by means of a spring tensioned edge guide on the other side. This design (which I've also seen used in projectors) would accommodate any variation in the width of the film, with the sprocket side film edge then acting as a consistent horizontal reference edge.

 

If this is normal practice for cameras and projectors it suggests that scanner transport should also adopt the same idea - ie. use the sprocket side film edge for mechanical registration of the horizontal, and the image of a sprocket only for digital registration in the vertical.

 

Carl

Edited by Carl Looper
Link to comment
Share on other sites

Re. registration idea.

 

Looking at the generous space inside the Logmar, there appears enough room, below the gate, for an add-on mod that could facilitate the writing of registration marks onto the film: a second gate through which the film passes, in which a small LED with appropriate masking could write a number of cross hairs onto the sprocket side of the film.

 

Digital registration software could then be written to read these registration marks to register frames. Insofar as the registration marks would be offset from the frame being exposed a scanner would need to ensure it provides a little vertical overlap in the scans - where the registration software could then stitch a number of frames together (using the same marks) sufficient to form a short film strip buffer that connects a given frame to it's offset registration marks.

 

C

Link to comment
Share on other sites

Hi Nicholas,

 

the only matrix's I've seen are too big - built for human eyes rather than the sizes required to write programmable patterns onto the film. However one might be able to get a smallish one (OLED or AMOLED), and use some optics to further scale such down to the size required to write the film. But then there is the additional hardware required to drive such a thing, which may or may not fit the space.

 

Carl

 

 

 

 

 

 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...