Using QUEST for orientation discrimination task

1
Hi

for an orientation discrimination task, which of the following
option is better:

1. linearic:
beta=3.5;delta=0.05;gamma=0.5;
tGuess = 10; tGuessSd = 30;
grain = 0.01; range = 60;
QuestCreate(...);
tTest=QuestMean(q);
tTest = max(min(tTest,30),1);
tTest = round(tTest);
... % run a trial with target_angle = tTest. get isCorr
q=QuestUpdate(q,tTest,isCorr)

2. logaritmic
beta=3.5;delta=0.05;gamma=0.5;
tGuess = log10(10); tGuessSd = log10(30);
grain = 0.01; range = log10(60);
QuestCreate(...);
tTest=QuestMean(q);
tTest = max(min(tTest,log10(30)),log10(1));
tTest = round(10^tTest);
... % run a trial with target_angle = tTest. get isCorr
q=QuestUpdate(q,log10(tTest),isCorr)

The reason I ask is that in contrast to experiments manipulating
contrasts, in which the intensity axis of the psychometric function
(X axis) is logaritmic, in this experiment I manipulate angle, and I
expect the X axis of the psychometric function to be linear (see for
example figure 4 in Parkes, Lund, Angelucci, Solomon & Morgan, 2001.
link:
http://www.nature.com/neuro/journal/v4/n7/full/nn0701_739.html).

Thanks,
Oren Flekser