Home of Professional Cinematography since 1996

class="Paragraph"  

Accepted Methods To Test Lenses

Published : 26th January 2004


CML list

I have read with great interest during the past few weeks the many threads concerning lens quality. (ie: older Angenieux zooms).

From what I can research, there seems to be no universally accepted method to test and evaluate the relative quality of a given lens.

My question to the CML thread is:

1) What methods do you use to test and evaluate a lens?

Do you use QUANTATIVE methods (ie : test charts, lens projectors, etc).

If so, what do you actually measure (lpm resolution, contrast, color, etc) and in what units of measure are those values specified in?

OR

Do you use QUALITATIVE methods (ie. subjective evaluation of a print or CRT monitor image)

Thanks to everyone for their time in advance.

Jim Palmer
UGS/PLM Solutions



From what I can research, there seems to be no universally accepted method to test and evaluate the relative quality of a given lens.

Actually there is. All of the manufacturers, and many of the top rental houses test using modulation transfer function (MTF). MTF is a quantitative number that is expressed as a percent at the measuring frequency. A perfect lens would have an MTF percentage of 100% at the measuring frequency. This measures the contrast of a lens at a given resolution. When measuring lenses it is important for the manufacturers to have a method that is quantitative, repeatable and not subject to how much someone partied yesterday.

Dan Irvin
Former Opto-Mech Engineer



Yes, but MTF is only measuring one part of a lens' performance. There's flare resistance, color accuracy and fidelity, barrel distortion, chromatic aberration, centre to corner sharpness, rectilinearity and more. It's how important one aspect is than another that drives preferences in lenses. I don't think anyone can honestly say that any top notch modern lens system such as Cooke S4, Zeiss Ultra or Panavision Primo is absolutely better than another, only that each has its benefits over the others.

Mitch Gross
NYC DP



The problem with MTF, as we saw on my recent lens tests, is that we see a lot of things in images that 'scientific' tests don’t show.

Cheers

Geoff Boyle FBKS
Director of Photography
EU Based



Congratulations to Wade Ramsey!!

Our esteemed Roy Wagner ASC had a VERY nice comment in the recent AC magazine about Wade! Calling him "one of the greatest teachers of cinematography!"

Way to go Wade!

Cheers,
Jeff Barklage, s.o.c.
US based DP
www.barklage.com



Geoff Boyle wrote :

>The problem with MTV, as we saw on my recent lens tests, is that we see >a lot of things in images that 'scientific' tests don't show.

Absolutely!

We can only evaluate lens performance in workshop conditions. I personally prefer a lens projector and a trained eye to spot optical problems, but this is done using white light ( as is MTF )and we look for resolution, contrast and other criteria in conditions not usually found in the field. These are problems we can fix.

The optical design determines the lenses own character and feel, this is where a lens becomes a tool for you guys....

P.S : I cut my lens teeth on the Ang 25-250 & 35-350mm

Andy Taylor
Camera Engineer
Arri Media
3 Highbridge
Oxford Road
Uxbridge
Middlesex UB8 1LX
UK
www.arrimedia.com
www.arri.com



Geoff Boyle wrote :

>The problem with MTF, as we saw on my recent lens tests, is that we see >a lot of things in images that 'scientific' tests don’t show.

I have designed some film resolution charts (Target) for determining line pair per mm results. In my literature, to support this method, I mention:

In his book “Hands-on Manual for Cinematographers”, David W. Samuelson, a veteran British cinematographer, points out: “Although lens testing by MTF is the most efficient means of measuring lens performance, and the auto-collimator is a useful tool for checking variations in lens and ground glass settings, it is only photographic testing which can satisfactorily evaluate the entire cinematographic chain.” "it" refers to test shooting with the lens on a camera and evaluating the results on the film itself.

Richard Stringer CSC
Toronto



I just shot with Canon K-35 primes(haven't yet seen the results).

They seem to be fairly unique-has anyone else here worked with them?

Best regards,
John Babl
Miami



I have a couple of Canon K-35s (18mm & 24mm) that I use with my Ultracam. I think they're nice and they intercut with the Zeiss Super Speeds well. Quite sharp.

Roderick
Az. D.P. (off to Disneyland
www.restevens.com
12On / 12Off