Flange Back Experience

[Jonas Nyström]

Some DPs have written in some blogs that RED owners will experience great trouble due to the adjustable PL mount. They say the measurement is to delicate to adjust without proper gear and equipment.

So what do you say, you people with some RED shooting experience; is it BS or are you crying blood on the set?

[Brent J. Craig]

If you are adjusting it on set, you are already in trouble.

[BASSAM MSSALATIE]

WE need confirmation from people on set . i mean midum experteis people
cause they are the possible victims of trying this without proper adjusting gear :huh:

[Jonas Nyström]

If you are adjusting it on set, you are already in trouble.

And that is one opinion they have in their critics; you will have to do flange back on the set, at least once every shooting day, since the flange back not is preset and obsolete like Arri Genisis for exemple.

I don't know, and therefore i looking for some hands-on experience!

[Eric MacIver]

We adjusted ours yesterday, but that was after pulling a properly calibrated lens off of the projector and collimator. If your lens isn't 100% on, then you're asking for problems.

Seems at first glance to work very well. The back focus adjustment is very smooth and moves the flange out at a good rate so that it's not difficult to dial in. The two screws locking it in place seem to hold it nice and tight, but time will tell if normal use knocks that out of place.

[chuck colburn]

Hello Indie,

Since you were where the autocollimator was did your lens tech check for any shift in parallelism of the lens seat on the camera when tightening the clamping screws on the back focus adjuster?

[Rocco Schult]

every camera needs to be properly collimated, so does RED.
Nobody does collimating on the set.

Question remains though, is: does a RED ONE come collimated by factory or does it have be done first time by every owner at a properly equipped facility ?

[Jonas Nyström]

every camera needs to be properly collimated, so does RED.
Nobody does collimating on the set.

Question remains though, is: does a RED ONE come collimated by factory or does it have be done first time by every owner at a properly equipped facility ?

Good question. I suppose not - will it stand for a FedEx trip from CA to EU?

And my question maybe boil down to this; if you have done your collimating, is it good and rocksteady for a period of time (and for how long, hypothecal speaking)? And for me who's travel a lot with my gear - do I have do it again when landing on location? Maybe this is not any problem at all - just asking you happy lucky people which already have a RED in your hand!

[Mark L. Pederson]

every camera needs to be properly collimated, so does RED.
Nobody does collimating on the set.

Question remains though, is: does a RED ONE come collimated by factory or does it have be done first time by every owner at a properly equipped facility ?

They come collimated.

[Evin Grant]

I just had to collimate my mount today. I'm not sure if it was out when I got it or has gone out since, it wasn't out much and it seemed to be accurate on my standard speeds but I didn't do any chart tests. We prepped the camera at TCH (The Camera House) and our S4 set was just slightly shallow so I loosened the mount and dialed it in. It kinda worked with the lens attached but moved much easier with it off. Took about 20 min. all told to get it bang on. All we used to check it was a focus chart, 8.4" Pany HD monitor in Zoom mode and a Zeiss Sharpmax to check infinity. I would have been happy just using the chart, we checked four different distances on all 8 lenses and they all taped out true. The adjustment was a little sticky though, Indie, if your's was smooth then maybe mine needs a lube :) Overall a very easy operation I would have no qualms about doing on set, if need be. However, I'd highly recommend doing a quick chart test before you send out your rig and adjusting as necessary ahead of time.

[David Wyatt]

Hi Jonas,

We checked ours this week when it turned up - it appeared to be a tiny fraction out (most lenses were almost an inch short of being spot on) and we figured it'd be good experience to get used to doing it anyway, so we bit the bullet. We checked some Ultra Primes that were bang-on on a lens projector and then put them on the Red. We loosened off the two allen keys on the mount (incidentally, as a default these appear to have been tightened by the Incredible Hulk, which is a good thing - I doubt there'll be much movement there over time!) and then got the image as sharp as possible on a 17" Panasonic monitor (zoomed in with the focus check on the Red). Everything seems to be focusing bang on the witness mark now but it would be interesting to compare a before & after on a big screen to see what the difference really is and whether it's significantly noticeable.

[RCFisher]

Well almost any good camera tech can do this, contrary to the chatter on this list. All you need are the necessary tools, properly calibrated lense and a device like the Zeiss Sharpmax or the similar device made by Century Optics. As a camera asst in the 80's On several films I had to check Flange focal distance daily on Panaflex Gold cameras, they drifted due to changing temps on locations. The reason why is the design of the camera which was done to reduce noise. No rocket science just mechanics, to the .0001 of an inch. At the minimum Flange distance should be checked prior to the camera leaving the rental depot.

Today many asst's use the sharpmax on Vipers and f900's when using Digiprimes since sharpness is critical and most video camera's flanges change distance with temp changes. I have been on quite a few productions where the back focus and FFD is checked several times a day, better to be paranoid than sorry. There is nothing worse than sitting through soft images in dailies.

[Costelloe Michael]

If you are adjusting it on set, you are already in trouble.

My 1st AC does this with each lense change on an HD shoot with a Zeiss Sharpmax. We also regularly check backfocus on HD cameras with temperature change. There is no big mystery to the need or the process.

What we are interested in is the ease of adjustment and the 'tolerance' ie does it stay good or does it drift with temperature, lense changes etc.

Thanks to all who have posted useful replies.

Mike Costelloe

[jbeale]

This article was interesting reading, regarding lens backfocus adjustment
http://www.bandprodigital.com/cgibin/eDatCat/BPDstore.cgi?user_action=detail&catalogno=TEK-BF

Sounds like on-set adjustment is standard for HD shoots and non-standard for film shoots, because HD gear was designed to be lightweight, and film gear was designed to be solid and stable. The article says you should readjust backfocus on a HD camera anytime the temperature changes enough to change what you're wearing.

Long ago, I think before any cameras shipped, Jim said the Red mount design goal was to be solid and not drift in or out with temperature change (eg. like a film camera lens mount, not a HD camera). I have yet to hear anyone testing for this specifically.

[Patrick Tresch]

Sounds like on-set adjustment is standard for HD shoots and non-standard for film shoots.

Adjustment on set is ok for SD/HD (from 1/3 to 2/3 inch sensor) if you have proper tools (minimum 17'' HD monitor/charts...).

But you can't compare back focus adjustment on a zoom lens for 2/3 inch sensor with flange adjustment for 16 or 35mm sized sensor/film.

The tolerances and tools are not the same.

Patrick

[chuck colburn]

I know how the sharpmax works but does anyone know how to qualify the unit itself for proper operation?

[Costelloe Michael]

Adjustment on set is ok for SD/HD (from 1/3 to 2/3 inch sensor) if you have proper tools (minimum 17'' HD monitor/charts...).

But you can't compare back focus adjustment on a zoom lens for 2/3 inch sensor with flange adjustment for 16 or 35mm sized sensor/film.

The tolerances and tools are not the same.

Patrick

The tools and techniques are certainly different but the theory probably holds. The image thrown by the lens has to converge on the film/sensor plane.

Film cameras such as Arriflex, Moviecam and Aaton with fixed flange focal distances shim the lenses to 52mm or whatever the flange focal distance. Panaflex's allow for the movement to be colimnated also, a duty I used to perform as a 1st on Features.

On 2/3" Chip cameras the lenses are fixed and don't have the ability to be shimmed, but the material the flange is made from allows the focal distance to drift and therefore they introduced 'back focus' adjustment.

Since most of us will probably use the movie style lenses on Red, as well as the possibility of using 2/3" chip lenses also, and hence the B4 mount adaptor, it would be my understanding that they would have to allow for the backfocus to be adjusted.

Just my impression.

Mike Costelloe

[BASSAM MSSALATIE]

This article was interesting reading, regarding lens backfocus adjustment
http://www.bandprodigital.com/cgibin...talogno=TEK-BF

.

i noticed that there is

CODEX PORTABLE cinema recorder
http://www.bandprodigital.com/cgibin/eDatCat/BPDstore.cgi?user_action=detail&catalogno=COD-PORT

is it something interresting .. anybody know ? about it ?:umm:

[Mitch Gross]

So you are checking accurate focus using a 720p output zoomed in 2x, giving you a horizonatl resolution of 2560 (1280x2), which is barely more than half the native 4K resolution of the chip? That means that you are judging a very large circle of confusion (measurement of focus) and it is not terribly accurate. Until I could check it at a pixel-to-pixel resolution I wouldn't consider it accurate.

Please tell me I am wrong since while we have adjusted the flange depth on REDs here for clients I am not familiar with all aspects of our procedures.

Mitch
Abel Cine Tech

[chuck colburn]

Hi Mitch,

There could be something that I'm missing since I never worked with video cameras as an optical/mechanical tech. but I would think if you achieved the maximum resolution possible at a lower resolution then 4k it would still translate to the higher rez level.

[Wes Printz]

Until RED comes up with and/or releases the proper procedure and fixtures for checking the FFD on the camera, the only standard it seems to me, is the old "It's close enough for government work".

Yes you could use a properly checked and calibrated lens, but as Mitch has found, it may not be accurate enough (as I understand his post)

With Film cameras, there is a spec, along with the tools to adjust and attain this spec. In the digital world, I'm sure ARRI, Thompson, and SI have procedures, to accurately check the FFD, as Panavision has a proprietary system for the Genesis.

How does RED do this check? With so many cameras being ordered, with the option of various mounts, how will one know exactly where their FFD is when they change a mount? Since the cameras will be privately owned, I feel it's it in RED's best interest to develop and release this information to RED owners / Rental houses that may provide these services to owners that don't have the tools.

Nothing beats doing it the right way!!

[Michael Morlan]

I found that same BandPro article. Couple questions:

Where does one buy a Sharpmax. Google didn't reveal any dealers.

Can the Sharpmax be used on any lens? HD, Cine, etc.

Although less convenient, am I just as well off with a Seimans star beyond a lens' hyper-focus?

Michael

[Wes Printz]

the sharp max is made by Zeiss:

http://www.digiprimes.com/2005/sharpmax/index.shtml

I have found it works well on a zoom, but pretty hard on a wide lens, which is where any FFD error will show up. Roughly anything under a 35mm, will not focus on the marks if the FFD is out. this is why it's important to have a way to check the FFD.

[luis bustamante]

Michael Kent form DSC LAbs sent me this video (LINK) (http://www.paolociccone.com/qtplay.php?file=videos/2U-S03-EP01.mov) demonstrating the use of their new Fiddlehead BackFocus chart. He believes the process will apply nicely to RED, what do you guys think? If it works it could be a very practical on set solution.

[Mitch Gross]

The Fiddlehead is very nice and easy to use, but one must have a perfectly collimated lens set to a standard on an optical bench collimator (not a pocket collimator). And it must be a lens that holds collimation, meaning that it is built to a professional technical tolerance so that simply toughing it doesn't throw the image unacceptably out of focus. That's what happens with all those inexpensive consumer lenses.

And the Fiddlehead still means nothing if you cannot judge the image in realtime at the full resolution you plan to finish in. So if you are doing 4K finish with your material you must set the flange depth while viewing in 4K resolution. And it has to be a realtime image as the judgement of focus is seen by shifting in & out of focus until a centerpoint is found. If you judge focus in a viewing situation that is equivalent to 2K then your sharp focus points could be a 4 pixel square instead of a single pixel point. This means that your circle of confusion (measurement of focus) is four times larger than it should be for a 4K finish. It means that everything could look razor sharp all the way up to a film out or 4K projection, at which point a noticably soft image can and likely will appear on screen.

The reality is that most work will finish in 2K so it's not so bad, but this is a real issue for a 4K finish. The only way to judge if something will be in focus at 4K is to view it in 4K.

This will likely be academic once the 1920x1080 output starts functioning, right? Double 1920 gives you 3840, which Stuart mentioned would be a center-extraction "Super HD" viewing mode anyway. But you better have a real 1920x1080 monitor of a decent size with you to judge that focus pixel for pixel.

Mitch Gross
Abel Cine Tech

[BASSAM MSSALATIE]

. So if you are doing 4K finish with your material you must set the flange depth while viewing in 4K resolution. The only way to judge if something will be in focus at 4K is to view it in 4K.



VERY important point .this is the first time i hear that ..thanks
but my question is that
befor long time they were adjusting flange sometimes on set .depending viewing through video assist ??is that time was correct or it is a matter of
optical viewfinder..:calm: :huh:

[Mitch Gross]

VERY important point .this is the first time i hear that ..thanks
but my question is that
befor long time they were adjusting flange sometimes on set .depending viewing through video assist ??is that time was correct or it is a matter of
optical viewfinder..:calm: :huh:

No one would ever set flange depth in this manner. For HD, it is set using an HD viewfinder with peaking or an HD monitor doing the same. For film it is done with a collimator that can physically measure the actual distance from the mount to the film plane.

Mitch Gross
Abel Cine Tech

[jbeale]

As far as I know- correct me if I am wrong- there is no way right now to view 4k live from the camera. To achieve that, you'd need a real-time debayer, and the camera does not do debayer on any output. Right now, all 4k output is generated outside the camera using software (eg. Redcine) running on various different hardware. Are any of those doing true real-time output?

For the purposes of focusing, of course you might only need to view a small fraction of the full frame at any one time. For a small enough window size, I'm sure debayer could be done in-camera, I mean isn't that what the little Canon HV20 does to generate 1080p out via HDMI? I'm thinking about a viewfinder digital zoom mode- it just is not currently available on Red.

[Mitch Gross]

Note that for accurate setting of the flange you do not just want to be viewing the center of the frame magnified. You also need to see the corners of the image. This is to be sure that the flange is resting flat and true, and that corner-to-corner focus is optimized. As lenses are sharpest at their center with drop off to the corners. There can be a range where the center is within maximum tolerance yet only at one point on this range the corners are at their best. To judge this one needs to see the entire image area fromocrner to corner. And yes, in real time.

Mitch Gross
Abel Cine Tech