guess what i found..

**evinsky** · 12-12-2006 10:57 AM

Erik, I think it's fine to respond to a genuine interest in an undeveloped product. As long as it doesn't escalate to a direct sales pitch. I'm sure many here would be interested in having an electronic solution to focus and exposure, and many would be willing to pay $1500 for it. Getting a commitment for 20 units may be tough though, at least on an as yet non-existent product. But Red has managed to gather over 1000 faith based reservations (Including myself) so you only have to capture 2% of the existing market to make your goal.

I for one would be happy to let you know how I think this product should perform.

A knob based, friction dampened focus control with an accurate and repeatable focus draw. A smooth, smaller knob for Iris control, a digital read out is fine for this since aperture changes on some zooms. The control box should be compact enough to fit on the rails like a normal followfocus. But should also have the ability to be remotely operated.

If IS/VR is available great, but it's better to get the focus/iris working smooth and seemlessly.

**widding** · 12-13-2006 05:12 PM

Originally Posted by evinsky

I for one would be happy to let you know how I think this product should perform.

A knob based, friction dampened focus control with an accurate and repeatable focus draw. A smooth, smaller knob for Iris control, a digital read out is fine for this since aperture changes on some zooms. The control box should be compact enough to fit on the rails like a normal followfocus. But should also have the ability to be remotely operated.

If IS/VR is available great, but it's better to get the focus/iris working smooth and seemlessly.

Evinsky, thank you for weighing in on this. IS/VR is actually already working, and has been tested on a 4K camera based on the Panavision/SVI sensor.

We have been tossing around follow focus as a product extension the past two days. I am not promising anything, but it has escalated on our priority list. I see this task as being broken into three pieces:

(1) Optimization of our existing controller for dynamic applications, such as follow focus. We have already committed to existing customers to finish this, understand what is needed here, and this will be completed regardless of whether we actually produce a hand controller. It is important to note, all lenses are not created equally. Those Canon lenses with Ring-type USM motors work best, most "L" lenses have this. All irises are pretty much the same.

(2) Mechanical interface to the camera for the existing lens mount. This can be left until the end. Depending on the camera may be best suited to be done by that manufacturer, us, the owner, or a third party.

(3) A hand controller, that is rail mounted on the camera or used remotely via cable or wireless.

I am interested in having a public conversation (and this forum's audience and comments to date makes it seem the proper location) as to what (3) should entail. I can not make any commitments when or if we will choose to release this product. But I can tell you, this product will only be interesting if it works the way a DP wants it to, rather than how this engineer thinks a DP thinks it should.

The feedback that I have received from a number of DPs has been good enough to give us a concept of what we want to build. The hand controller will have the following:
big knob for focus
small kob for iris
bluetooth radio
display
other buttons/switches (very few!)
battery
mounting features for rails on camera
easily handheld off camera
durable
cost reduced without sacrificing utility/durability

We have taken this to mean a couple of things:

The housing is going to be polycarbonate, rather than metal. Mounting to the rails should be with thumb screws rather than tools. The mounting hardware should be removable if it interferes with the ergonomic handheld use, but all hardware should be captive so it does not get lost.

If two hand controllers are used in a system, the one mounted on the camera can be a slave (mainly providing a radio receiver, and power to the lens), and the second one can talk to it wirelessly as the master. Both displays would be syncronised. The bluetooth interface would allow for: control from a computer/PDA (SPP profile) and/or integration into a larger motion control system; upgrading the software; customizing/calibrating/etc. Basically the desire to keep the unit inexpensive means it needs to have a simple user interface.

Anything beyond the actual use of the lens, and the most basic changing of modes, should be handled from the external PDA/computer talking over bluetooth. We can recognize a unique lens model (but not a unique lens) over the interface, so it would be easy for the follow focus unit to remember any calibrations, i.e. lens step to distance, knob/lens rotation mapping (possibly non-linear, user customized shape), and automatically adjust when a lens is replaced. The unique lens/model is not a problem in single camera setups.

Buttons/knobs/display beyond the minimum is going to both add cost, and clutter that will make this less user friendly. But in a menu mode, the two knobs could be available for navigation, so the interface might not have to be too limited.

We can make the unit smaller than it needs to be, so I would ask a few ergonomics questions: How big (assuming substantially a rectangular solid) is a good size? How big (diameter and thickness) of each of the knobs to be really usable? Does tension of these knobs need to be adjustable? or is something just tighter than "they won't turn on their own" good enough? How far apart should these knobs be?

The lenses do not have absolute encoders, but with some precaution in the implementation of a system, are extremely repeatable (assuming ring USM motors). A homing operation will always be required when the unit is powered up, or a lens is replaced. This is just one of the limitations of using SLR lenses.

For the knobs, incremental encoders are the cheapest/easiest to integrate - these do not have any type of home reference. Stepping up from there are encoders that can have absolute orientation known by rotating them once through after powering up. Is absolute and deterministic orientation of the knobs from powerup to powerup a value add or not? Or is a homing of the knobs, along with a homing of the lens good enough?

Should the range of rotation of the knob relative to the entire range of focus always be the same, or should the mapping of rotation be user configurable, and potentially non-linear. Should the rotation of the knob be linear in distance or linear in angular rotation of the lens mechanism? Is there a need for multiple rotations of the knob to map to the range of focus (i.e. very fine focus mode)?

How much data should be on the display? Is min/max/current f-number, focus count, and distance of focus (if available/calibrated) enough? Is a 12 character by 3 line display, cell phone sized characters, good enough? If not, how many lines/characters?

What I am asking for is help to get from a rough idea, to a product specification. We have been procrastinating entering this market, in no small part because we do not know exactly what this follow focus control should look like.

I am impressed by the excitement that you all have for Red. It is your camera, and it will meet your needs, because you have been involved in its development. If a good specification for a follow focus controller can be developed, that will meet the needs of DPs - I will allow faith, rather than guaranteed orders, be my guide in prioritizing.

Gauntlet has been laid down: Help me navigate to a meaningful product specification, and we may build it.

[Moderator, should this conversation be moved somewhere else?]

Regards,
Erik Widding
Birger Engineering

**evinsky** · 12-13-2006 11:18 PM