ANDREWS MP TEST

[Andrew M.]

Right!

OK, Should I cancel my order for Zeiss Masterprimes? :waaa:

Pawel
www.achtel.com

I just shot the chart with Zeiss Master Primes 65mm (my favorite) and 16mm
Distance to the chart as per specs, so all labelling on the chart is one to one.

It is 4K chart! So no guessing with 1200 line or 1800 or whatever.
What you see is what you get.

Just loading the data in REDCINE.

The moment of truth is near......................

 

[Andrew M.]

Test Setup

Test Setup

Test setup pictures.
Chart is comming ..........................

 

[Andrew M.]

Zeiss 65

Zeiss 65

Just converted TIFF to jpg, no correction
Next 1:1 crops

 

[Ketch Rossi]

Okay, now I'm Ipnotized!

Are we the only ones that like the MP's Andrew?

Ciao

 

[Jarred Land]

The MP's are great lenses, good work man. Its getting a little off track though, We should probably move your test to a new thread.

 

[Andrew M.]

The MP's are great lenses, good work man. Its getting a little off track though, We should probably move your test to a new thread.

Yes, I was thinking where to put it.
I already tested this MP here on the projector.

http://www.reduser.net/forum/showthread.php?t=1475&highlight=zeiss+65mm

This test I did it to compare RED 85 with MP 65.

 

[Andrew M.]

65mm Zeiss Master Prime on RED #906

65mm Zeiss Master Prime on RED #906

Copy from RED 85mm test, thread...................

It is 1:1 crop of right lower corner.
No CA noticeable. Pretty strong vignetting.

Firmwere V 13
Setting:
Aperture T1.3
Shutter 1/48
Distance to chart 3315mm (3.315M)
25% sharpening, applied

So this gives us 1200 horizontal lines pairs resolution per whole 4096 pixel frame length, in the corner, not in the middle.
It mean that Mysterium with OLPF filter applied, and as a whole camera system and lenses plus REDCINE conversion to TIFF gives us 2400 easy recognizable lines (points) on the horizontal length of 4096 pixels. Not bad indeed. I will sharpen this TIFF bit more and see what more we can get out of it.

 

[Andrew M.]

I received answer from ARRI about vignetting problem while using Master Primes on RED.

They suspect that Mysterium is using micro-lenses above each photo sensor.

Well, I will not speculate on this but doing the test of other lenses with the white plain board, bit underexposed will be good to do.

Also I retested 65mm MP today and got bit better results in terms of number of lines per mm of resolution than the picture above.

I contribute this to the fact that I used DPX instead of TIFF and looks like Adobe PS is bit better in DPX. Also REDCINE on the screen looks bit better than in the TIFF file being viewed in PS.
I used ver 15 instead ver 13 of firmware (if it helps anything)

At the end I am getting solid 50 lines pairs per millimetre at 50% MTF for the whole chain MP lenses - sensor-camera-REDCINE-DPX.

Lenses are tested and shows 200lp/mm
Mysterium is tested and shows 73lp/mm (Graeme 78% of 4500 pixels / 24mm)

The whole system from lenses to the final look of the DPX file on my screen
56lp/mm (or 112 lines/unique points per mm)
Below 1:1 crop of 4K frame shoot at 1.3 1/48 24fps shot from the distance of 3315 mm (3.315 meters)

Andrew

 

[Andrew M.]

Anybody knows why I am getting bit better results when rendering in REDCINE to the screen as oppose to the TIFF file?
I tried snapshot of the screen and it is also bit better.

For instance the square C on the screen looks almost like square D when I use REDCINE on the screen, in terms of contrast and clarity.
REDRAW once converted to TIFF shows the square C at very low MTF (modulation)

Converting 1:1 cropped TIFF to 100% jpg doesn’t change anything.
TIFF file looks the same on my screen as the one above when I use the Internet browser.