2/3" B4 Mount for Scarlet/Epic? Anyone interested in this?

[Jose Lomeña]

Jeff I tested two years ago with my redone and b4 lens without optical adapter and the only problems was coverage, i dind't saw any chromatc aberration or other problems, but 2k was not good enough for me, at least from the old m sensor... I did the test for a pal 4/3 show and the results was good, but not for 16/9 or hd.

Maybe a metal adaptor + optical multiplier can be a good solution. But only a metal adaptor will not. Maybe i can share some images in the next days with my redone + b4 lens without optical adaptor only metal.

[Pete Lutz]

Here is a video of a lens adaptor with no optics. B4 lens on an AF100.
http://vimeo.com/20729392

It's not fantastic looking, but perfect for some jobs.

[Brian F Kobylarz]

The biggest thing with doing a B4 mount for any single-chip camera is the optical correction. You *WILL NOT* have proper RGB convergence without the optical correction. The physical mount pieces or metal parts are not a big deal, it's the optics you need to make it right.

Pete,

Jeff is absolutely right about using a B4 lens in a single imager camera. You need optical correction if you plan on delivering acceptable results - even at 2K.
Sure, you can try to get away with less if your market or clients will only use Vimeo or small displays - or if they don't seem to care. But what happens someday when surprise - they project it on a large screen? That is your reputation that is at risk. Unfortunately, RED also suffers when someone watching remembers that is what it was shot with.

There is plenty of high quality 2/3" HD glass out there - primes, zooms - I own my own share. The form factor has advantages in some situations - I know exactly why you - and many others - pine for the same functionality in corporate, industrial and broadcast markets. We may yet see some manufacturers resond to this market need.

So why do adapters need optical correction?
The introduction of the B4 interface was a milestone - it was the first time in history that a common lens mount and specifications were adopted for all 2/3" HD cameras. (this was the Broadcast Technology Association global standard BTA S-1005-A adopted in 1994, followed by the EBU Tech Spec in 2002). Before this standard, every camera manufacturer had their own mounts and specifications. I could not use a Panasonic lens on a Sony, or an Ikegami. Sometimes you ordered a new camera based not on its features, but on the glass you owned.

In establishing the B4 standard, a specific 48mm flange back dimension was dictated - this is where the green image "in air" plane is that you see reference marked on the side of the camera - and what you "back focused" to.
If you are only looking at one part of the light spectrum, that would be easy - and the end of the story. Chances are that you are not shooting black & white - where the green channel would be enough for luminance. Color adds a degree of complexity.

There is a brilliant engineer I know from Sennheiser who likes to say that "Everything in life is a wave" (Thank you Volker!!!).
He is right - when you are talking about capturing images, those beams of light travel through the lens at different wavelengths - therefore they focus at different planes. This is what is called "longitudinal chromatic aberration". There is a secondary issue - when they focus at different distances, their size (magnification) is different - this is referred to as lateral chromatic abberation.

In the case of 2/3" HD cameras, the light is split by a prism to land on 3 sensors. The prism assemblies are designed to mitigate both types of abberations by offsetting the sensors - Red is offset by 10 micrometers and Blue is offset by 4 micrometers from the Green image plane. This is a compromise - but take away those modifications and you end up with real problems ( aka inferior results).

The B4 lenses are designed - as a worldwide standard - to pass the light waves into a camera with the above factored in.
That means that without optical correction, any B4 adapter would pass a flawed image onto a single sensor.
There is no work around once you capture it wrong.



I'll divert to a parallel subject:

Let's consider motion picture cameras using color film stock. Since there is no prism, does that mean that there is only a single plane (the film stock) for everything to focus on?
If that was your first thought, sorry!
The film stock has different layers of emulsion - each sensitive to a different part of the color spectrum. Even though the difference is very small, it does exist. Since there is no way to modify the light once it leaves the back of the lens and reaches the film, another approach is used. The solution is in the design of the lens - if you ever have a chance to hear Larry Thorpe talk about lens design - all of the different glass formulas and lens coatings that go into making every wave of light arrive where it is supposed to be - you gain an understanding that it is a true science. Manufacturing high quality lenses is truly both technology and art.


That brings me to the RED Digital Cinema cameras - single sensor, one plane for all waves of light to arrive at the same time.

There is a reason that RED Pro Primes achieve the level of detail they do - exceeding the performance of many lenses designed for motion picture work (that cost many times more). The Pro Primes are designed for a digital cinema sensor, not film stock.

In the interest of fairness, other manufacturers are now producing glass designed for single sensor cameras - we now have a number of exceptional choices. The future looks good (waiting for those anamorphics ... hint, hint).

Choosing the right lens for the project you are working on is, I would suggest, one of the most important decisions you will make. No matter how good the camera, you cannot correct for a flawed optical path. Make the right decision. You may just want to go back to those shots someday.

[Peter Versey]

Very nicely explained Brian .... although I've seen some fabulous images on single sensor cameras with Movie Film Lenses

P.S. was your engineer from sennheiser who said "Everything in life is a wave" by any chance a surfer as well? :-)

[Pete Lutz]

Wow Brian. Thank you for that extremely thoughtful and well written post. I really appreciate the full explanation.

[Jeff Kilgroe]

Very nicely explained Brian .... although I've seen some fabulous images on single sensor cameras with Movie Film Lenses

Fabulous is in the eye of the beholder. The Vimeo clip above where they're making cookies doesn't show us much. It's overly compressed and doesn't seem to hold a lot of detail. I have yet to see 1:1 pixel shots of B4 mount "HD" glass on a single-imager camera where there is no chromatic offset or convergence blur. This doesn't appear as typical CA. What you get is the red and blue channels offset in distance from the sensor. It's not a two-dimentional or axial convergence issue like aligning RGB lamps on a projector or something like that. It's a spatial or depth convergence. It's like having your FFD or backfocus off for the red and blue color channels. Typically the blue offset is nearly unnoticeable at 4 microns, however the red offset at 10 is noticeable and you can often see that red objects, or those with a high concentration of red within their RGB make-up, will always display soft edges, funny halos or even just be slightly out of focus. And, like I said, the blue will not be as noticeable, but still exhibit the same sort of properties. When I say it's not as noticeable, it usually won't be as proper FFD adjustment is usually within 5 microns. So, if your back focus or FFD is dialed in just right, the blue should still work out OK or at least show minimal results and you'll be fine if you're not trying to squeeze full resolution 2K/1080p out of it. Full 1080p is a tall order for these lenses, even if the sensor has the resolution or you use optics to enlarge and cover 3K/4K. Most all 2/3" "HD" lenses lack the resolving power to do much better than about 1080p resolution. Factor in OLPF or prism systems on 2/3" sensor blocks and even on full-raster 1080p cameras, it's rare that you get more than 850~900 lines of nominally measured resolution. Well-processed R3D 2K is right there with that using good 16mm or 35mm glass.

[Brian F Kobylarz]

Jeff,

I believe that Peter was referring to "fabulous images on single sensor cameras with Movie Film Lenses " ...(NOT B4). This was a response to my statement on motion picture film compared to a single sensor.
I agree with him - while an older Angenieux Optimo or Zeiss super speed may have been designed for a film camera, they can deliver exceptional results on a RDC camera. As can many motion picture lenses. All I was trying to point out was that the design of a lens is a complex undertaking that is driven by many factors.

And Peter, the next time I see Volker, I'll throw the surfing question his way! He is back in Germany but makes his way here on occasion.

[Patrick Stringer]

What kind of lens do you have for your HPX500?

[Jeff Kilgroe]

Jeff,

I believe that Peter was referring to "fabulous images on single sensor cameras with Movie Film Lenses " ...(NOT B4). This was a response to my statement on motion picture film compared to a single sensor.

Oh... Yes, I was looking at the example posted above by someone else and then quoted him by mistake. Sorry, Peter! And Peter, you are correct, of course. That's what most of us use on our RED Ones, EPICs, and now Scarlets -- cinema lenses. Most all current and new cinema lenses are designed to work great with film. A few don't -- like the Optimo DP series. And it's not that they can't work with film, but they took advantage of lack of a mirror shutter on cameras like the RED One, to alter the design and the rear element sits close to the image plane. It provides an improved telecentric design with much more even filed illumination. FWIW, the RED Pro Primes look amazing shot on film too.

[Mitch Gross]

Just to note, our HDx35 B4/PL Optical Adapter is specifically designed to get all the colors of the spectrum to focus on a single plane.