Amateur Photographer - UK (2019-06-29)

subscribe 0330 333 1113 I http://www.amateurphotographer.co.uk I 29 June 2019 53

Subscribe online at

magazinesdirect.com/18AG

0330 333 1113

Quote code: 18AG

Lines open Monday – Saturday, 8am-6pm (UK time)

*Pay just £24.99 by Direct Debit every 3 months,

with the price guaranteed for the first 12 months

and we will notify you in advance of any price

changes. Offer closes 30th September 2019.

Terms and conditions apply. For full details please

visit http://www.magazinesdirect.com/terms

• Never miss an issue
  with delivery
  direct to your door


• Peace of mind with
  our money back
  guarantee if you
  change your mind


• Access to exclusive
  offers, giveaways
  and prizes with
  subscriber Rewards

Subscribe to

Amateur

Photographer

for just

£24.99

saving 35%

Passionate about photography since 1884

Saturday 6 January 2018

The perfect small camera bag Lovely new Billingham ticks the boxes

Leica CLWhy this gorgeous rangefi nder-style

camera is Leica’s best mirrorless yet

One face

50 ways

TESTED

Fresh approaches for creative portraits

APOY ’s topportraits

Which did well, and why?people shots

Top-notch Tokina

Superb fast 20mm prime tested

Collodioncreativity

A modern twist on a classic technique

Passionateabout photographysince 1884

BUYI NG 12-page

vsulixcEGUIDE& lensescameras 597

e^ money-saving off

PORTRAIT SPECIAL rse

Tech Talk

Professor Newman onÉ

Bob Newman is currently Professor of Computer Science at the University of Wolverhampton. He has been working with the design and development of
high-technology equipment for 35 years and two of his products have won innovation awards. Bob is also a camera nut and a keen amateur photographer

The trials of

product test i n g

P

ity the product reviewer.
Whatever they do,
there will be a group of
people who will insist
that they have been unfair to their
favoured product. One wonders
why, if someone already knows
how good a product is, they would
be consulting a review, but it is a
fact of modern life that many see
the purpose of product reviews
more as validating their own
choices than the informing of a
new acquisition. If the reviewer
gives an honest opinion of a
product, that view will be attacked
as not being objective, which is
likely true, since every human,
however scrupulous, carries a
lifetime of experiences, knowledge
and approaches that undoubtedly
colour what they write.
Only with a hard-nosed scientifi c
method can a review be truly
objective, but even then there
are pratfalls along the way. Having
done an amount of experiment
design in my career, I know just
how hard it is. The fi rst problem is
deciding exactly what the question
is that you are aiming to answer. In
general, this is far from simple.
The two categories of consumer
review mentioned above are
asking two questions. The one
wanting to validate their own
choice want to be reassured that

their choice is best. This is

unanswerable since ‘best’ is

a comparative term and all

comparative terms have to

be related in the end to some

numerical metric. Those wanting

to inform a buying decision want

to understand how well the

characteristics of a product will

meet their needs. The best the

reviewer can do in this case is

to try to hypothesise what some

broad categories of photographers

might be looking for, and relate

their testing to that.

Autofocus dilemma

I was put in mind of this dilemma

in a discussion about which

autofocus system is ‘best’. Modern

autofocus systems have many

operational characteristics, and

which is ‘best’ depends on one’s

own requirements. As an example,

I’ll use a series of photographs

taken during my recent trip to

photograph the motorcycle racing

at the North West 200 meeting in

Northern Ireland.

When taking these pictures, I

usually aim for a degree of speed

blur, which requires panning the

shot. At quite close range, it is not

usually a failure to pan that causes

a low hit rate, rather the ability of

the autofocus system to capture

and hold the subject, and to adjust

focus as the subject-camera

distance changes. This is much

less of an issue at long ranges,

since the relative focus distance

changes less from shot to shot.

In taking such a sequence, the

fi rst requirement is to activate the

AF system when the focus point is

over the required in-focus part of

the subject. The selection of this

point will also determine whether

the autofocus can track effectively,

since the tracking algorithm works

in part by comparing the image

patch of the focus point and

looking for similarities in the next

reading from the AF system (or

more precisely, from the metering

or image sensor). It also tries to

calculate the trajectory of the

subject and use that to predict the

likely location in the next frame.

In this respect motorcycles are

easier than people because their

motion is more predictable. If the

AF system can lock-in and track, it

also has to be able to adjust focus

fast enough to maintain sharpness.

In this sequence, the AF managed

to successfully lock and track, but

in the fi nal frames could not follow

the movement quickly enough.

So, if a reviewer develops an

objective test of autofocus, which

particular aspect should be tested

and which priority should these

different parameters be given?

There are any number of factors to take into

consideration when reviewing products

The autofocus has successfully locked on and tracked until the point of focus is moving too fast for the lens to follow