one second at a time

7/25/2019 One Second at a Time

1/19

Getting Faster, One Second at a Time

iSpeed Lap Analyzer

written byTroy Schulz
http://www.nessoft.com/ispeedhttp://www.nessoft.com/ispeed


2/19

Overview

The goal of this document is to help people that may not be familiar with the

iSpeed Lap Analyzer discover how it can help them become a better driver. I'veused this tool predominantly over the last year to help me learn the techniques

required to go from midpac! in the Star "azda Series to a frontrunner.

#hy the title '$ne Second at a Time'% #ell& it's simple. or me& I started ( to )

seconds slower than the times of the top qualifiers in the series. I have ateammate that is one of these top qualifiers and it gave me the opportunity to

witness his techniques very closely. Trying to compare my driving to his wasnearly impossible. The techniques involved are so different that it can seem as if

they are alien. So& rather than compare to the leaders in the series& it's best to

start loo!ing at people that are roughly * second faster. These people havelearned certain techniques that have given them the boost to be closer to the top&

but not yet reached the top. requently& when you loo! at what they are doing&you will find that it's not completely foreign. This allows you to learn the

techniques required to be an alien in small increments.

#hile this guide focuses on the Star "azda and my e+periences with it& it should

not preclude it from being valuable to you in whichever series you call home. Theonly true requirement I believe is that you need to be focused on a single car.

,ot to say you can't run other series while learning your primary car& but you willpic! up the tools you need to be fast quic!er if you spend the ma-ority of your

time in a single car. I !now for sure& that the time I've spent in other seriesdelayed my progress in the Star "azda to some degree.

#hy is any of this relevant in a guide on how to use the Lap Analyzer% Shouldn't

it be as simple as loo!ing at the wheel& bra!e& and throttle inputs and ma!ing the

proper ad-ustments% I sure hope you all laugh at the thought of that. If it werethat simple& everyone would be able to do it. #e all react to inputs differently.

Sometimes these differences can be compensated for using a different setup& butother times we have to alter our reactions. I !now for me& altering my reactions

too! a long time. ut once it happens& there's no going bac!. The only way to

get there is through hard wor! and perseverance.

So& this is where the iSpeed Lap Analyzer comes in. Is it magically going to ma!eyou faster% I wish. The Lap Analyzer is simply a tool. A tool that helps identify

the areas where you're slower relative to others& and it shows you how the inputsdiffer that result in you being slower. /nowing the difference only allows you to

wor! to resolve that difference.

So& lets get started...


3/19

The Application

Lets start by tal!ing a bit about how iSpeed and the Lap Analyzer wor!s& thenwe'll ta!e a quic! loo! at the tools to go over the basics.

There are two applications in the iSpeed suite0 the 1ashboard and the Lap

Analyzer.

oth of these can be found at http022www.nessoft.com2ispeed

Dashboard

The 1ashboard serves two purposes. irst& it can be used to display all the

telemetry in formation in a consistent way for any car you drive. Second& it canstore this telemetry information to the datamart for later use with the Lap

Analyzer.

NOTE: The 1ashboard will record telemetry information to your datamart even if

it is not visible3

The above ,$T4 is very important. 5ou don't have to be using the 1ashboard asa display in order for it to function as a data recorder. It wor!s -ust as well

minimized as it does when it's visible. It's the data recording abilities that thema-ority of this document is about.

I don't have a separate monitor to run the 1ashboard& it -ust runs in the

bac!ground collecting data until I'm ready to do some analysis.

There's a couple options in the 1ashboard's menus that I want to call your

attention to though. They are both in the Save / Exportconte+t menu.

I prefer to have every lap I run saved locally& not -ust my fastest laps. Thereason for this is that sometimes& I run my fastest laps the first day of testing.

ut a couple days later& I can't get even close to those initial times. If I onlyrecorded my best laps& the 1ashboard would not save my most recent laps.

6aving these saved allows me to loo! at what I'm currently doing differently than

my fastest lap so I can understand how to get bac! to the earlier speeds. So& Ialways chec!Auto-save all valid laps (vs best laps only).

Also& theAuto-publish best laps to Datamartshould also be chec!ed. It's

enabled by default& so you shouldn't have to change anything here. 4ven if youfeel you are slow& it is still valuable for you to publish your laps to the datamart.
http://www.nessoft.com/ispeedhttp://www.nessoft.com/ispeed


4/19

irst& you may well be one second faster than someone& and they may find yourlap valuable. Second& if you're struggling to analyze your lap& having it on the

datamart would allow another iSpeed user to analyze the lap for you.

#hen laps are published to the datamart& it saves your best lap of a given type0

race& qualifying& hotlap. It will save a different set every season.

Lap Analyzer

The second application provided in the iSpeed suite is the Lap Analyzer. The Lap

Analyzer is the perfect tool for comparing your telemetry against others.6opefully by the end of this document& you'll understand the Lap Analyzer well

enough to both increase your speed as well as your en-oyment of i7acing.

irst& we'll -ust ta!e a quic! loo! at the different parts of the Lap Analyzer& -ust soyou get a feel for the application. The following page has a screenshot with a

following description of the different parts. 1on't worry too much about all the

details& as we'll cover more in the following sections.


5/19

Illustration 1: Lap Analyzer Application


6/19

Parts

1 Lap Info

The value of the Lap Analyzer comes when you compare two laps. The Lap Infopanel shows you the details of the laps that have been selected. It also shows

you the color each lap is represented by. In all of my samples& the lighter blue

color is the faster lap and the dar!er blue color is the slower lap.

2 Track Map

The Trac! "ap shows you a few interesting things at a glance. It shows you the

line that you too! around the trac! as well as the trend of the throttle and bra!e.5ou can use the scrollwheel on your mouse to zoom in and out. The zoom point

is the mouse cursor.

3 Stats

The Stats 8anel has two separate behaviors. $ne when a single point on thetrac! is selected& the other when a segment of the lap is selected. #hen you

select a given point& the specific value for each telemetry value is shown. #henyou select a segment of the lap& only certain values are relevant so the irrelevant

values are removed and replaced with other relevant information.

Splits

4very trac! has a set number of splits based on how i7acing defines them. These

splits will show up in the Split 8anel and shows the time for each split. It's alsopossible to remove the default splits& and define your own.

! Speed

The following panels all behave the same way. The Start2inish line is the far left

and the far right of the graph& the rest of the lap falls in between. If you select a

segment of the lap& then rightclic! in a panel and select '9oom In'& the graphdata will change to only show the selected part of the lap. To go bac! to the

default view& rightclic! in the panel and select '9oom $ut'.

The Speed 8anel shows your car's speed at a given point on the trac!. A largepart of analyzing a lap will involve this panel.

" Ti#e $elta

The one thing everyone wants more than anything else is to go faster. The Time

1elta panel is where all of this becomes blatantly obvious. The centerlinerepresents the first lap& the line represents the difference in time versus the


7/19

second lap. Above the centerline means slower& below the centerline meansfaster.

Since time is what we all want to gain& this panel is where most of the analysisbegins. :enerally spea!ing& you want to focus on the steepest parts of this graph

first& since that's where you have the most to gain.

$nce you find the wea!est spot using this panel& you'll focus on the followingpanels to understand what inputs are causing you to lose that time.

% &rake 'orce (

The ra!e orce panel shows you bra!e input& from ;< = *;;sing this

panel& you can discover where you are bra!ing too early or late& and too light orhard. This is probably pretty obvious& but what may be less obvious is that this

panel can also show you hardware issues. 8erhaps your bra!e calibration wentoff& and you have *< bra!e on at all times. 8erhaps your 8$T is going out and

you get spi!es in bra!e input while going down the straights. This panel ma!esthese issues stand out very clearly.

) Throttle (

Li!e the ra!e orce panel& the Throttle panel shows your throttle input& from ;ntil then& -ust driveconsistent. As the saying goes& CTo finish first& first you must finish.D

1on't be frustrated if your consistency isn't perfect. 5ou may miss a bra!ing pointhere or an ape+ there. This is normal. #ith more practice& this will becomeeasier. #hat you don't want is these mista!es being catastrophic. 5ou want your

bra!ing points& ape+es& and line to be conservative enough that you can ma!e

mista!es and still recover.

/eep in mind& this isn't the goal& it's the beginning. 5ou want to learn what yourcurrent level is. 5ou're only trying to discover your baseline. $nce you have

that& you can then wor! on raising that baseline to meet your goals.

Time trial's in i7acing are a great way to wor! on your consistency.

$nce you feel you !now a trac! well enough to drive consistent laps& go aheadand start the 1ashboard and go bac! out and put down some laps. 7emember&

the 1ashboard doesn't have to be visible to be able to record laps. $nce that's

done& you can move on to analysis.

3 4 Identify and 5esolve

,ow that we !now the parts of the Lap Analyzer and we have some good data to

wor! with& it's time to get our hands dirty and actually start loo!ing at some laps.

It can be overwhelming loo!ing at telemetry information for an entire lap. $ne ormore seconds seems to disappear very quic!ly. The !ey when analyzing

telemetry is to focus on one corner at a time. 7un the rest of the trac! -ust li!eyou always do& but try to change your approach for that one corner. There will

generally be one corner where you're losing more time than all the others. Start

with that corner& since you have the most to gain there.

As you are learning to approach that corner differently& your consistency is goingto suffer. #ith enough repetition& you'll soon be able to change your technique

for that corner and regain your consistency along with additional speed. $ncethat happens& load your new fast lap and repeat with the ne+t problem corner.

As you repeat this process& you'll gradually gain that second you're loo!ing for.

Then you'll be loo!ing for another.

Loadin- Laps

So& let's start with loading two laps. :o to the 'ileELoad Lap *' menu item.This will bring up a view into the datamart. It will include all of your local laps

along with all the laps from the online datamart. ,avigate to the trac! and car


10/19

that you're wanting to analyze and select your lap& then -clickanother lapand clic! the oad Selected apsbutton. This will load both laps at once and

you'll be ready to do some analysis. As stated before& don't automatically select

the fastest lap available as the lap to compare against. Select a lap that isroughly one second faster.

There are a lot of laps in the online datamart& from drivers with a wide range ofs!illsets. Selecting a lap that's too much faster will ma!e learning too difficult asit will li!ely show that you're slower in every corner an equal amount. It -ust

ma!es it harder to !now where to start& plus it li!ely involved more advanced

techniques that you're not li!ely to be able to use consistently.

NOTE:,ot all laps are created equal. "y point here is that people can go thee+act same lap time in much different ways. If the lap you selected doesn't have

one really clear place where you need to wor! on& you may want to try a lapeither -ust faster or -ust slower than the one you selected.

6ick a 0orner

$nce you load the two laps for analysis& you have to pic! the one corner to focus

on. It should be pretty easy to find. Always start by loo!ing at the !ime Delta

panel.

Let's ta!e a close loo! at the !ime Deltapanel to understand what it's telling us.

The dashed line in the middle represents the reference lap.

The plotted line represents the time differential of the second lap to thereference lap.

If the plotted line is above the dashed line& it implies the second lap isbehind the reference lap.

?onversely& if the plotted line is below the dashed line& it means the secondlap is faster than the reference lap.

The steeper the plotted line& the faster the delta is growing or shrin!ing.

This time delta is from a couple laps I put in with the 7iley 18 at the 6omestead

7oad course. I'm busy preparing for the 1aytona (.@ and since I rarely drive the

TimeDelta 1: Sample

?A


11/19

car& I figured a little practice would be good. This chart is comparing myFualifying time with one of my early practice times. 6owever& it has a couple

interesting things to point out. Also& I went through all the same analysis

techniques to get from the slower time to the faster time.

In this graph& the delta grows pretty quic!ly in three places. Those three places

are obvious places to focus time resolving the differences. 6owever& we need tofocus on one to start& then possibly the others after that. I always focus on thesteepest curve first& when that steep curve isn't followed by a steep drop.

A= A is interesting because its a decent amount of time loss over a long period

and none of that time is regained immediately following the climb. It's a definite

candidate& but not the steepest so we'll pass on this one for now.

B= has a pretty steep climb& but it is followed by an equal amount of drop sothe net result is ;. ,ot really very interesting for analysis.

C= ? is clearly the steepest climb and is followed by no gain in time. This is the

one I'd always start with. ,o need to loo! further.

hat to Look 'or

So now that you've identified the corner to wor! on& you have to figure out whatwent wrong. ortunately& there a fairly small set of reasons why you lose time.

*. $ver1riving GLate ra!ingH

(. $verra!ing

). #rong :ear

@. Too "uch 1ownforce

. Too fast bac! to throttle

B. Too slow bac! to throttle

. #rong Line

Over$rivin- .Late &rakin-/

This is one of the most common mista!es people ma!e. ortunately it's alsoreally easy to identify and also to fi+.


12/19

In !imeDelta "& we get the same time delta graph that we used in our sample.

In fact& we also see the same section referenced as the corner that we'd focus ourattention on. #hy% ecause this is a clear e+ample of $ver1riving.

5ou will always see a slight dip followed by a -ump. The dip means you went in

slightly faster than the reference driver. The -ump means he got bac! in thethrottle sooner and e+ited the corner with more speed. #hat you also see is that

even though the steepness of the curve flattens off& it is always going uphill all

the way to the ne+t bra!ing point.

Loo!ing at !rack#ap $& we see very clearly that the reference lap was able tostay tight to the ape+ of the corner& where the slower lap pushed past the corner

and had to turn more to get pointed in the right direction.

7esolving this issue is as simple as bra!ing -ust a bit earlier and ma!ing sure youcan hit your ape+ and get on the throttle sooner.

A teammate of mine documented this tric! once to me and it became something

that I rely upon to this day. #hen trying to determine your optimal bra!ing point&

first determine your optimal acceleration point. 5ou want to find the spot in thecorner where if you get in the throttle you don't have to lift until the ne+t bra!ing

point. $nce you determine where your optimal acceleration point is& then wor!on timing your bra!ing such that you can start accelerating at that point. It

needs to allow a split second of time to transition between bra!ing and

accelerating. Then& you !now that if you bra!e any later than that spot& there's

TimeDelta 2: Over-Drivin

A

Trac!"ap 1: Over-Drivin


13/19

no way you can start accelerating at your optimal acceleration point and you'regoing to lose time.

I guarantee& you'll lose less time by bra!ing too early than you will by bra!ing toolate. So& it's best to start on the safe side and bra!e a few feet too early than to

try and push too hard.

Over&rakin-

In the process of learning to drive faster& I've went through a few CAh 6a3D

moments. $ne of these was in regards to overbra!ing. I found that if I !ept mysame bra!ing point& but forced myself to release the bra!e a fraction of a second

faster& I could still control the car through the corner and get a perfect ape+&

while maintaining a higher minimum speed for the corner. I too! a long time toprogram my brain into trusting the higher speed& along with a setup that would

oversteer. $f course& it also too! greater car control to handle a looser setup.It's not something that can happen over night& but here's how to spot over

bra!ing.

At point 'A' in "ulti:raph *& you can see that the reference lap is able to !eep a

higher minimum speed for that corner. If you loo! directly below in the ra!eorce < graph& you'll also see that he holds the bra!e pedal -ust a hair longer

than the reference lap. This tiny e+tra bit of bra!e is all it ta!es to cause thefairly significant -ump in the Time 1elta.

The solution& is to force yourself to release the bra!e -ust a tiny bit before you

normally would. This can be incredibly difficult& since your brain is essentially

telling you CIf you release the bra!e now& we're going to wrec!.D

"ulti#raph 1: Over-$ra!in

A


14/19

>nfortunately& your brain may actually be right. If your setup is too tight& youmay not be able to carry that little e+tra speed through the corner. 5ou'll want to

try a number of times and -udge if you feel you may be successful if you get it

'-ust right'. If it doesn't feel li!ely& you need to start pushing yourself to handlelooser setups. :enerally& if you're only loo!ing at a lap one second faster& the

amount you need to loosen the car is not large& so it shouldn't be a huge hurdle.

I've also found that when this technique is the most significant problem for thelap& I had learned enough other techniques that a looser setup wasn't has hard to

handle as it had been when I tried one without the necessary s!ills relating to

bra!ing and acceleration.

There's a reason they say CLoose is fast.D So& eventually you'll have to learn tohandle a loose race car. It's best to wait until you have the other fundamental

s!ills mastered before trying. It will give you a solid foundation to be able tobuild on.

Li!e I said before& you'll be ready to ma!e that -ump when your biggest problemareas for a lap resemble the above graph.

ron- +ear

Sometimes in our testing& we get into certain habits that become really hard tobrea!. or me& one of those habits are my shift points. In the sample below& you

can see two laps that have different shift points at three points on the graph.#hen you see a lap that is faster using different shift points& you may want to

ad-ust your shifting and see if simply changing gears in time with the faster lap

can gain you a portion of the time you're loo!ing for.

5ou can see in %ear%raph $and !imeDelta && that at the point the down shiftis made& there is a -ump in the time delta. This ma!es it pretty clear that

downshifting at that point is slowing him down.

#ear#raph 1: Star "az%a &'i(e% #earin)

TimeDelta *: Star "az%a &'i(e% #earin)

A

A


15/19

elow shows a graph from the 1allara& which allows ad-usting each gearseparately. #hen you see a graph li!e this& you may want to reconsider your

gearing choices. The faster lap Glight blueH is clearly ma!ing more complete use

of the gears available. y !eeping the car higher in it's 78" range& it's li!elygaining a speed advantage.

Although& this is one type of difference where you may want to chec! a coupledifferent laps that are near the times of the reference lap. It may -ust be this oneperson that has odd gearing choices& or it could be a pattern and suggest that a

change in your gears would be required.

Too Mch $ownforce

$ne of the big questions we all face when building a setup is how much downforce do I need to ensure the optimal lap times. Ideally& a higher downforce

setup will gain you more time in the corners than it loses you on the straights. In

the below sample& the faster lap is roughly a second faster overall. If you loo! atthe acceleration curve after each bra!ing point& you can clearly see that the faster

lap Glight blueH gains speed faster than the slower lap Gdar! blueH. This indicatesthat the faster car is using less downforce than the slower car.

This one can be a little deceptive. In this case& since every acceleration zone hasthe same behavior& it's pretty clear. It could also be that the car was draftingsomeone at the time they made the lap. There's no way to !now for sure. $ne

thing to do it to chec! a couple laps around the reference lap to see if it's a trendof the faster drivers to use less downforce.

Too 'ast on Throttle

6ave you ever gotten on the throttle too fast and have to lift to !eep from driving

Spee%#raph 1: Too "uch Down+orce

#ear#raph 2: Dallara &In%ivi%ual #earin)


16/19

off course% #ell& #ulti%raph "shows an e+ample of the telemetry that occurswhen this happens. At pointA& you can see that the speed graph shows that it

grows& right up to the point where I had to lift to !eep the car on trac!. Then the

speed drops below the reference lap and never catches bac! up until the bra!ingpoint to the ne+t turn.

It's pretty easy to see why this causes such a great time penalty and it is alsoeasy to !now when you do this. If this happens a lot for a corner& you may needto move your acceleration point further ahead. 5ou may be trying to be too

aggressive.

Too Slow on Throttle

Just as getting on the throttle too fast can cost time& so too does getting on the

throttle too late. $f course& it's probably a little more obvious why. If youaccelerate more slowly& you're not going as fast and thus loosing time.#ulti%raph &shows two instances of not getting on the throttle fast enough. InA& the result is lest noticeable& but in 'it's pretty clear. There's a -ump in theTime 1elta and it lasts all the way to the ne+t bra!ing zone.

It may seem obvious enough that all you need to do is put on more throttle

sooner. 6owever& it's not always so easy. ?ertain setups can be harder to 'getthe power down' than others. The stiffer the rear suspension or higher the anti

roll bar& the less e+it grip you have. The less e+it grip you have& the harder it is

to get on the throttle quic!ly. The stiffer the rear suspension& the looser the carwill be. I can hear you thin!ing C5ou said 'loose is fast'& so what gives%D #ell& its

a balance& li!e all things. 5ou want it loose enough to get through the entry and

A

"ulti#raph 2: Too 'ast on Throttle

A


17/19

midcorner quic!ly& but not too loose that you can't get bac! in the throttlequic!ly and aggresively.

As you learn the s!ills to get faster& you'll also learn to push your limits in'looseness' further and further.

I#proper 5acin- Line

Sometimes analyzing laps is a bit more complicated. "ost often& you'll loo! atthe Time 1elta :raph and see some lost time& but it's not always as clear what's

happening to cause the issue.

In these cases& it can be necessary to loo! at multiple graphs and numbers and toa little detective wor! to really get to the bottom of the problem.

In this ne+t scenario& we're going to loo! first at !imeDelta . In this graph it

loo!s identical to the $ver1riving e+ample above. The slower car drives in with

too much speed and it costs precious time on the e+it. The interesting part ofthis& is where this occurs. This section of the trac! is the ritton's chicane Gor T

KH at $ulton 8ar!. Let's loo! a little deeper...

,e+t& we have the Speed :raph of the same section of the trac!. #e !now from

the Time 1elta graph that we have more speed coming into the corner& but less

TimeDelta ,: ron Line

"ulti#raph *: Too Slow on Throttle

A


18/19

speed e+iting. 6owever& loo!ing at the speed graph& it's not a clear early bra!ingsituation. The speed of both cars isn't terribly different on deceleration.

6owever& the faster car is able to maintain a higher minimum speed than the

slower car. Loo!ing closely at the Speed :raph will also show that the slower caris more abrupt slowing and accelerating& where the faster car has a more gradual

speed curve.

#e still don't have all the information we need to isolate the problem. So& let's go

to the trac! map to see if indeed the slower car is missing the ape+ or e+actlywhat is happening. Loo!ing at the below segment of the trac!& it should be pretty

obvious the problem. The faster car is the line on the bottom and the slower caris the line on top. The slower car has a different entry into the chicane than the

faster driver. The consequence of this is that the slower driver has to turn

sharper than the faster driver& thus is forced to slow more in order to get the carto rotate properly before he can accelerate. This also forces the slower driver into

a slightly later ape+. Since the slower car had to slow more& he also doesn't use

the entire trac! on e+it since his speed doesn't force him to drift all the way to theedge of the trac!.

Spee%#raph 2: ron Line


19/19

0onclsion

6opefully by this point& you've gone through a few iterations of analyzing yourlaps and have discovered a few tenths already. $r if nothing else& you have a

good idea where you need to focus your time.

As you go through the process& it's important that you try not to get ahead of

yourself or give up -ust because you don't see immediate results& or you don't seeas much as you want.

I started i7acing as a lower midpac! driver in the Star "azda. I've used iSpeed

since the very beginning. I didn't magically -ump ) seconds -ust by analyzing a

few laps and doing everything in this guide. 4very trac! is made up of a numberof corners. 4ach of these corners has something to teach us as a driver. Some

we master more easily than others. Some we'll spend our entire lifetime trying tofigure out.

4very wee! you spend analyzing your laps and attempting to get faster is a wee!

well spent. 4ven if you feel li!e you didn't get anywhere that wee!. Sometechniques ta!e a long time to get right. After a full season of analyzing data and

trying to master the techniques to go faster& you'll be very surprised when you go

bac! to the same trac!s for a second time. 5ou'll load up the same setup youused the last time around and you'll start laughing. 5ou'll thin! to yourself& CI

actually drove this setup%D Then you'll twea! the setup to fit your current s!illset& then by the end of the wee!& you'll be at least three tenths faster than the

previous season. I had some wee!s where I was a full second faster from the

previous season's fast lap.

5ou don't even need iSpeed to see gains li!e this. 5ou need e+perience with thecar and trac! and a desire to push your boundaries. The value of iSpeed is that it

focuses you right in on your wea! points so you can get to wor! and resolvethose issues quic!ly.

Trac!"ap 2: ron Line

one second at a time

Documents