In a rash exhibition of the foolishness I am prone to indulge in and having enough collected bits and pieces handy, I tried the Letus Extreme EX1 production bundle on the SI2K.

The SI2K will mount to the Letus rails - just. The rear platform has to be lowered to just high enough for the small lever handle on the wedgeplate to clear the wide camera base.

The front pillar under the Letus body needs to be almost at its full height.

I tried a Nikon 35mm f1.4 for relay however this is too wide and brings in the edges of the prism path as a vignette.

Previous 2/3" tests I did with the old Letus for XL model hinted that a 50mm prime as in the original Minolta direct relay would have a field-of-view of the groundglass which would not be too wide.

The nearest to 50mm on set today was a 58mm Noct-Nikkor which is good for sharpness down to full-open iris.

With the SI2K on the rails and the Extreme perched precariously on its single pillar, with the 35mm f1.4 on front and the Noct-Nikkor 58mm f1.2 as relay, the "A" and "F" blocks on a Lemac chart appear to resolve, or at least are seen by the camera imager as there is a sharp fine moire displayed on the screen within the blocks.

If the Le brothers' direct relay for 2/3" goes into release, is of equal or better sharpness than the Noct-Nikkor and if there is a B4 mount for the SI2K, then 35mm film emulation of acceptable resolution could be a practical option for the SI2K near to immediately by this method until proper and better accessories are available from the respective vendors.

I shot the Lemac chart and a few faces in ambient interior lighting today. I have yet to extract the footage from the camera.

I have not seen RED footage at its highest practical resolution, so I cannot offer any comparison as to the look.

I am of an age where I am not normally given to being particularly animated about things. However, those around me seemed to think I had been enlivened to a little more than my normal demeanour by this morning's experiments.

If you are prompted to experiment, take exceptional care to make sure the Letus rods are well and truly secure and the tripod mount is likewise. The SI2K is a very heavy camera and the Letus hardware is a little bit light for this task. If the tripod head rolls forwards or backwards to its limit, the rods may release the camera unless the clamp screws are well firmed up. The camera needs to be as far forward as possible yet allowing space for the Allen wrench to fit to tighten the front pillar of the wedgeplate mount.

The SI2K/Letus Extreme combination got its first outing today. My alignments were a bit off because the 50mm relay picked up an edge so went to the 58mm for relay.

We shot a little shadow footage of an actual project with and without the Extreme on front and we going to break the camera down and put it away. The director took a look and suggested we leave it there, shoot some more and that he might use "some of that".

He also commented on apparent greater latitude even through a groundglass. How it holds up for resolution may be another story.

One thing readlily apparent right away is that focus is not something to be hunted for or guessed at, best practices required and no less.

Furthur report.


Using a Nikon prime for relay reprises most of the disadvantages of the old Minolta based direct relay except for field-of-view of the groundglass which is close to the EX1 view of the groundglass.

I used a 58mm Noct-Nikkor for relay due to picking up an edge as a result of the hasty assembly of the rig under pressure in the day. With more meticulous setting up, the 50mm would probably be okay.

The relay lens and the bronze adaptor that I turned up to bridge it to the back of the Extreme seems to be temperature sensitive and the relay focus went off after about 90 minutes under production lighting after coming in from outside 8 degree C temperatures.

I'm not sure it is was metal expansion of softening of internal lens lube which might have allowed a slight relay focal shift. A proper relay lens and hard mount design in place of my crude hack would deal with that nonsense.

Unfortunately, I cannot publish this footage as it is subject to copyright. Hopefully soon I can publish some test footage shot elsewhere.

This footage would not be representative of the system as it was scaled to 720P to match the main production camera origination.

The file based recording system, like the EX1 and DR-100 for the JVC is a dreamland compared with the ordeal of the capture process. The bonus is that the origination is already in Cineform and the application of the look file to the raw footage playback in Premiere with Prospect 2K settings, once you work out how to do it is instantanous.

A bit of a footnote to above.

In some of the wide shots with a 14mm, I have observed a faint artifact which presents as very broad bands moving slowly vertically upscreen. They are very softly defined with graded edges and ony apparent over uniformly lit and coloured bright even areas.

My theory is that the motor vibration on a wide shot may be enough to only just move the image on the imager by a portion of a pixel. With the rolling shutter of a CMOS imager in progressive, what would present as an interlace artifact shows up differently.

I think it may be that some pixels "see" less light when the vibration movement is against the direction of the rolling shutter scan and more light when the vibration movement follows the direction of the rolling shutter scan.

I am prompted to this theory by the improvised arrangement being a little loose as the screwhole on the front pillar did not line up on the day and I instead strapped the front of the Extreme to the pillar with a velcro tomato tie.

(Velcro tomato ties which come in a continuous roll are absolutely the handiest things for quick cable restraint on boom poles or anywhere else.)

The SI2K camera body is built like an engine block. The only time it will twist will be if something like an engine block is dropped on it.

It seems that the Extreme needs to be firmly attached to its support rods when used with a CMOS camera and the lens mounts themselves need to be firm. The lenses themselves need to have tight, not sloppy focus movements.

hello, is there anyway you could post a clip of this footage so we can see what it looks like? I always had a thought about using a letus with the SI 2K or the DII from ikonoskop.

Guiseppe.


Sorry about the confusion. Links are in this post :-

http://www.dvinfo.net/conf/let-us-discuss-letus/129125-roiding-extreme-3.html

A rough instruction set can be found here :-

http://www.dvinfo.net/conf/silicon-imaging-si-2k/127880-letus-extreme-si2k.html#post983328