E-PRIME的时间精度问题(操作)

红

X

代表没有达到； 黄

色？代

表边缘达到。

通常第一次测试结果较

差

这是由于启动测试程

序操

作系统不稳定

Test number in lhe series. Letters indicate multiples of 10.

The number of extra trials necessary

in

order to compensate for the tinning errors that

occurred from a less than perfect machine.

Results

from

the

dock testing

rtir^ as

shown

in

Figure 2

r

indicating whether a greater

than billisecond

error

occurred.

A green

indicates

that the

results suggest

the

machine

can deliver millisecond accuracy (as defined

in

Chapter 3-

Cntical Timiiig).

A

red

X'

indicates that

the

test

failed

to reach

the

specific criterio

rt

(%

Missed

Ticks

>1% or

Maximum Missed Tick

>

10ms)

;

therefore,

the

computer could produce serious timing

or display problems.

A

yellow

L

?

JI

indicates timing concerns

(1%>

% Missed Ticks

>

0

1%

or

10ms >

Maximum

Missed

Tick

>

5ms).

The status

bar

moves across

the

screen as the test proceeds. Ideally,

the

result will

be

a

III green

Typically,

the

first test is poor because

the

operating system

stabilizes itself

after the launch

and incremental loading

of

the test program.

Clock Test

% Extra Trials

Required: Timing Bias

Mi ssTick Percent

Detected Ticks

Missed Ticks

Maximum Missed Tick

Maximum duraton of missed ticks. This value shows

:

in the worst

匚

ase

:

what the

timing error would be during a WOOOms span. This should be small (e g.

h

less than or

equal to 5ms). Numbers over 10ms should be viewed as a serious timing problem.

Timing Variance; Squared

Observed timing variance and the measurement enor variance (difference between

Error

expected and actual duration variance). If the measurement error variance is below 1 it

is negligible.

Percentage of extra trials that would need to be

run

in order to compensate for the

timing Errors.

Percentage of missed millisecond dock licks. This should be below 0.1%.

Total number of detected and missed millisecond clock ticks

Total number of missed millisecond clock ticks

Refresh Cycle Test

% Refresh Missed

Percentage of refresh cycles that were missed (i.e,. percentage of time that the onset

of

the

vertical blank signal was not detected).

This

value should

be

below

or

_

equal to

0.1% (one miss

测到的垂直空格信号

的启动的

in

1000 refreshes).

If

the

vertical

blank

simulation code

is

;

then this value

次数。如果太太，

则重新设置

should be below or equal to 50%.

A

miss is defined as

a vertical

blanking period that is not

屏幕参数（分

辨率，颜色深

between

5 and 20ms at 50Hz to

2

00Hz (i e.

r

a miss is

considered to

have

occurred).

If the

%

Refresh

Missed

is greater than 0.1% then the Refresh Rate underestimates the refresh rate.

If

度，刷新周

期），或更换显

the %

Refresh

Mussed is

above

0.1 %

and

the very top portion

of the screen

must

be

updated

卡。

on

single refreshes, adjust

the screen

parameters (i.e.

r

screen

resol

utia

n’

color depth, refresh

rate), or switch to a different video card.

Refresh Missed Count

The actual number of times the onset of the refresh cycle was missed.

缺失的刷新周期，比如没

有检

Refresh Rate

Refresh Duration

Refresh Duration SD

Refresh Duration Max

Refresh frequency in Hertz (cycles per second).

Mean time

(in

milliseconds) bet wee n refresh cycles.

S

伽

dard deviation (in milliseconds) of the refresh duration,

Maximum duration observed of a single relresh cycle This should be very close in value to the

Refresh Duration. It is often a multiple of the refresh cycle due to missing 1, 2 etc. vertical

blank signals.

Vertical Blank Duration

Mean time (in mi Illi seconds) in which the vertical blank hardware bit was set.

Vertical Blank Min Duration

Time (in milliseconds) in which the

vertical

blank hardware bit was set.

E-Pirime

monitors

this

bit

in

order to determine

when

the

refresh cycle begins in order

to

synchronize stimulus

onsets

且

nd to

keep the

monitor from flashing when updates

are done in

the middle of

the

refresh E-Prime

runs well on Pentium 120 class

ma

chi nes and

above

when

the

vertical blank

period

is greater than

0.05ms.

需要分析的时间数据：

Shows the missed tick percentage by test run. This value should be below 0.1% . You may see a larger

number of misses on the initial test runs as the operating system finishes loading and initializing the test software. If this is the only

problem, it can usually be overcome by running a short block of practice at the begi nning of the experime nt. If you have large spikes

during subsequent runs, it indicates that the operating system, other applications, drivers, or hardware are tak ing executio n cycles from

the experime nt. You should attempt to review the state of your system to ide ntify other programs or hardware that may be causing

delays in the experiment.

用来显示

missed tick

的百分比。如果在几次测试之后还是有比较

大的峰，这表明有其他的应用程序，或其他原因造成

在实验时不能达到毫 秒级精度。

Shows the maximum missed tick durati on

by test run. This provides a good measure of the

worst-case time delays caused by the operating system, other applications, drivers, or hardware. It should be < 10ms.

BinCoun

Shows the distributi on of missed ticks across all test runs. Each bin holds the count of observed delays of a specified duration. Bin #1 is

the count of the detected ticks (1ms). Bins 2-20 are the counts of greater delays (respectively). Bin 21 is the count of all delays greater

than 20ms. Bin 22 is the count of any delay that was negative in duration (should not happe n). You can use the bin pro determ ine if

other programs or hardware are consuming time.

RefreshMissedPerce

Shows the perce ntage of missed refresh cycles (i.e., on set of vertical bla nk even ts). This value

should be less than 0.10% . If it is higher, the vertical blank simulation feature of E-Prime should be used (this feature is on by default in

E-Prime). If the rate is > 50% , you should re-run the test using different display parameters (e.g., resolution, color depth, refresh rate) to

see if the results can be improved. If the rate cannot be reduced below 50%, you should con sider changing the video display adapter or

possibly not using the computer for experiments requiring short duration, precise visual

stimulus presentations.