When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
So you've gone to the track and you've recorded some sessions with the PDR. Now you're looking to get faster and wondering what to do with that data you've collected. I wanted to share with the forum a little bit about how I use data to improve my lap times.
I'm going to do this post using the PDR and Cosworth toolbox since most C7 track rats have that in their car. However, the Cosworth toolbox is pretty limiting as far as data analysis goes even though it collects a ton of great data. If you're serious about using data then I'd also recommend getting an Aim Solo DL and running that in parallel on track as it makes it much easier than the PDR. I keep hoping that GM or Cosworth will make some enhancements to the toolbox to make it a more full-featured data analysis product. I'd even be willing to pay for it.
So for starters, make sure you have the MP4 files you want to analyze downloaded to your computer from the SD card and also make sure you install the Cosworth toolbox available here: http://www.cosworth.com/toolbox/
There's also a video posted by Cosworth here that goes over some of the more basic functions in case you need help:
Now my general approach to data analysis is pretty simple. I'm not perfectly consistent lap to lap and I suspect most of you aren't as well. Even when my lap times are very close I've noticed that I'll do some things better or worse on two laps with very close times. My goal is to find those areas and then learn from them to understand what makes my fastest ideal lap. I do this by comparing my fastest lap with my 2nd, 3rd, or 4th fastest laps to identify those differences. So let's start with a real example.
Open up the Cosworth toolbox and select your session with the fastest lap as the "Reference Recording" and the session with the lap you want to compare as "Recording". If you're just starting out it's best to choose laps from the same day so that you know that track conditions (grip, temp, humidity) and vehicle conditions (tires, brakes, suspension) are largely the same.
Now that you have the files open let's go to the view that I will primarily use in the toolbox. At the top select "Corners" and "Chart". Off to the right the Channels you should select "Driver Input". Make sure the laps you want compare are selected at the bottom and finally make sure you select "Distance" not "Time". That compares the two laps based on distance travelled rather than time elapsed so in theory you are always comparing the same two places on the track.
Now, most of my effort will be focused on the time delta graph and the "Corners" view is the only one that will show you that data. You might notice that the time delta suddenly jumps up at the start by ~.2 seconds and then back down again at the end of the lap. I have no idea why the tool is showing that but in the end the result is the same so it won't affect our analysis.
If you look at the time delta at ~1600 ft in to the lap it shows about a 0.2s delta and by ~2900 ft it's down to 0.03s so there is nearly two tenths I gained. By clicking in the chart in the middle of that area (~2200 ft mark) and hitting the 'Z' key you can zoom in on just that section. You can also use the +/- to zoom in on the map - in this case to where I started gaining ground.
I noticed a couple of things about this. First, my line here is closer to the apex curb for longer. That in turn lets me get back to full throttle sooner but I also have to breath off of it. In my reference lap I get to full throttle later but once I'm there I stay in it. In theory, the latter technique is correct but is it?
If I toggle the "Channels" tag over to "Forces" it allows me to see the speed graph. Now the speed graph itself is pretty worthless because it does not autoscale like the other graphs and is always 0-200 MPH. That's pretty worthless when you're trying to see a 2-3 MPH difference.
However, you can see the actual MPH off to the right and use the arrow keys to move along the graph and see how it changes. In this case, you can see that while my speeds at apex were the same that little shot of WOT allowed me to pick up ~ 3 MPH and I carried that speed delta all the way until I started to brake for T3. That's where the time gain came from but unfortunately I gave it all back by over slowing for T3 (min speed 5 MPH slower than my reference).
I think this is a good place to stop for now and I'll add another post later. However, I discovered one other interesting point while looking at this data. I am way overslowing for T1 and getting back to throttle too early. Look at this capture.
As you can see I am back on the throttle and accelerating before I'm even half way to the apex. I should be carrying speed through there and gradually slowing the car until the end of the inside curb, then getting back to the throttle. In this case I don't have a lap to compare to know how much it would gain me but it's something that I'm going to focus on next week when I'm back at VIR.
Finally, here are some Cosworth keyboard shortcuts I've found:
Chart
Z - zoom in
Backspace - zoom out
M - max speed jump to
N - min speed jump to
R - rate of change and can highlight section
T - tile
Map
+/- zoom in and out
Arrow keys move around the map
I noticed a couple of things about this. First, my line here is closer to the apex curb for longer. That in turn lets me get back to full throttle sooner but I also have to breath off of it. In my reference lap I get to full throttle later but once I'm there I stay in it. In theory, the latter technique is correct but is it?
If I toggle the "Channels" tag over to "Forces" it allows me to see the speed graph. Now the speed graph itself is pretty worthless because it does not autoscale like the other graphs and is always 0-200 MPH. That's pretty worthless when you're trying to see a 2-3 MPH difference.
However, you can see the actual MPH off to the right and use the arrow keys to move along the graph and see how it changes. In this case, you can see that while my speeds at apex were the same that little shot of WOT allowed me to pick up ~ 3 MPH and I carried that speed delta all the way until I started to brake for T3.
Sean,
You are pointing out something that has been highlighted in many classroom sessions. Each straight is a drag strip and it has long been established that the car leaving sooner has an advantage that keeps growing as time goes by V=1/2 times Acceleration times time squared. Given cars with equal acceleration and starting from 0 mph that is impossible to overcome. Give the car that leaves first a 2 mph advantage (rolling start) as it breaks the timing lights makes it almost impossible to overcome even for a car that has a lot more HP. When you look at the realities of what happens in a corner acceleration out of the corner between two cars or two different times through the corner for the same car more than likely will not be equal since positioning in the turn can affect how much throttle can be applied and when it can be applied. As your graph shows being closer to the apex gave you several slight advantages when it came to getting that initial jump and rolling start.
A few years ago I did a track walk at the Glen with Peter Argetsinger (his father was the guy that started the road racing in the village and then expanded it to the current track). He has been a pro driver and professional instructor for a few decades. When we stopped in Turn 1 he walked over to the rumble strips near the apex and put his foot down and said when you come through this turn you want your right front tire HERE not here or not here and he indicated those spots by moving his foot by a few inches. Then he said if you do that you will gain 1 mph of exit speed over a similar car that doesn't place the wheel there. You can carry that 1 mph speed advantage all the way to the end of the back straight which takes about 30 seconds give or take based on the car we are talking about. Multiply 30 seconds by 1.5 ft per second which is ~1 mph and you have gained 45 ft over the other car. As we progressed around the track he kept pointing out all of the little differences where small increments of speed could be gained. Put those altogether in a single lap and you have the picture of how the guy in a low powered car can easily run away and hide from a guy driving a super car that has a 20 mph or more higher top speed. It also highlights how important it is to hit your marks consistently on each lap. Besides Cajones that is a prime factor in what separates the pros from the rest of us.
Thanks for posting! I'm brand new to the brand and platform but have been tracking for about 7 - 8 years and love being on the race track. (I traded in a 2013 Audi RS5 for the C7 Z06 w Z07).
The PDR (and accompanying Cosworth tool) is something that I'm pretty excited about (ignoring the 720p video resolution lol). I previously used Harry's Lap Timer on my phone.
So far, I've only played with a bit of recordings and love the amount of detail available. I wish there were more data points available (like boost and IATs).
I can't wait to hit the race tracks again. Just had the DSC Sport alignment done and am still getting used to how the car handles before stretching it's legs on the track.
Thanks Bill, and yes we've all been taught this in class. The thing I like about data and comparing your own laps is that it is proof that even on your fastest laps you likely aren't hitting everything right and that you and your car are capable of doing it.
Frankly, I really wish I had more time with students that have a C7 to actually go over the data with them. I think that a lot of the lessons would make more sense if we could show concrete examples using their own data. I'm no Peter Krause but I think most intermediate students at a minimum would get a lot out of this.
Thanks for posting! I'm brand new to the brand and platform but have been tracking for about 7 - 8 years and love being on the race track. (I traded in a 2013 Audi RS5 for the C7 Z06 w Z07).
The PDR (and accompanying Cosworth tool) is something that I'm pretty excited about (ignoring the 720p video resolution lol). I previously used Harry's Lap Timer on my phone.
So far, I've only played with a bit of recordings and love the amount of detail available. I wish there were more data points available (like boost and IATs).
I can't wait to hit the race tracks again. Just had the DSC Sport alignment done and am still getting used to how the car handles before stretching it's legs on the track.
I really like the PDR. However, from reviewing my videos I doubt the PDR video is actually achieving 720P resolution. On a 32 inch display it is almost impossible to tell the difference between 720 and 1080 and when I review the videos on the in car display they look like the camera is struggling to provide 480P definition. Watching them on my Surface Pro display or 60 inch TV isn't any better.
The biggest problem with the PDR is the inability of the camera to adjust sufficiently to changes in light level as the car turns or encounters shadows going around the track. I have some videos taken on bright sunny days when the car drives into a shadow the only way you know there are cars in front is when the brake lights show as two or three red spots in front. Higher resolution doesn't do anything for you if the screen is black.
Thank you so much for this great thread. I have been tracking much more seriously for just the past couple of years. I am CLUELESS when it comes to data analysis so your post is MOST helpful! Thank you for taking the time to do it. Basically I mainly watched the in car videos and looked at lap times.
As you know, the corvette forum can be a great resource, which is why I like it. On the other hand, post after post regarding "which color is best" and the juvenile and pointless posts are more frequent than ever.
Your post reminded me of why I have remained with this forum for so long, so thank you again. Any further tips are appreciated!!! I ran a 2:21 which was a new personal best for me at the NCM last week. My buddy and pro driver Jim Painter ran it in 2:08 in his 2016 Z06 (non zo7 model, manual). I marvel at just how far away I am from that. I was running in intermediate and was never passed and yet I am still an astonishing 13 seconds off the pace. I keep thinking where on God's green earth am I going to find 13 more seconds and survive lol. So any tips are appreciated. I do have a Z07 and run cup 2's (Jim had hoosiers I think) and I run stage 3 aero but more and more people (including Jim) recommend I go to stage 2 aero, what stage aero do you run and do you have a z07.
Any go fast tips or more examples of how to use the PDR data are appreciated.
Originally Posted by Poor-sha
So you've gone to the track and you've recorded some sessions with the PDR. Now you're looking to get faster and wondering what to do with that data you've collected. I wanted to share with the forum a little bit about how I use data to improve my lap times.
I'm going to do this post using the PDR and Cosworth toolbox since most C7 track rats have that in their car. However, the Cosworth toolbox is pretty limiting as far as data analysis goes even though it collects a ton of great data. If you're serious about using data then I'd also recommend getting an Aim Solo DL and running that in parallel on track as it makes it much easier than the PDR. I keep hoping that GM or Cosworth will make some enhancements to the toolbox to make it a more full-featured data analysis product. I'd even be willing to pay for it.
So for starters, make sure you have the MP4 files you want to analyze downloaded to your computer from the SD card and also make sure you install the Cosworth toolbox available here: http://www.cosworth.com/toolbox/
Now my general approach to data analysis is pretty simple. I'm not perfectly consistent lap to lap and I suspect most of you aren't as well. Even when my lap times are very close I've noticed that I'll do some things better or worse on two laps with very close times. My goal is to find those areas and then learn from them to understand what makes my fastest ideal lap. I do this by comparing my fastest lap with my 2nd, 3rd, or 4th fastest laps to identify those differences. So let's start with a real example.
Open up the Cosworth toolbox and select your session with the fastest lap as the "Reference Recording" and the session with the lap you want to compare as "Recording". If you're just starting out it's best to choose laps from the same day so that you know that track conditions (grip, temp, humidity) and vehicle conditions (tires, brakes, suspension) are largely the same.
Now that you have the files open let's go to the view that I will primarily use in the toolbox. At the top select "Corners" and "Chart". Off to the right the Channels you should select "Driver Input". Make sure the laps you want compare are selected at the bottom and finally make sure you select "Distance" not "Time". That compares the two laps based on distance travelled rather than time elapsed so in theory you are always comparing the same two places on the track.
Now, most of my effort will be focused on the time delta graph and the "Corners" view is the only one that will show you that data. You might notice that the time delta suddenly jumps up at the start by ~.2 seconds and then back down again at the end of the lap. I have no idea why the tool is showing that but in the end the result is the same so it won't affect our analysis.
If you look at the time delta at ~1600 ft in to the lap it shows about a 0.2s delta and by ~2900 ft it's down to 0.03s so there is nearly two tenths I gained. By clicking in the chart in the middle of that area (~2200 ft mark) and hitting the 'Z' key you can zoom in on just that section. You can also use the +/- to zoom in on the map - in this case to where I started gaining ground.
I noticed a couple of things about this. First, my line here is closer to the apex curb for longer. That in turn lets me get back to full throttle sooner but I also have to breath off of it. In my reference lap I get to full throttle later but once I'm there I stay in it. In theory, the latter technique is correct but is it?
If I toggle the "Channels" tag over to "Forces" it allows me to see the speed graph. Now the speed graph itself is pretty worthless because it does not autoscale like the other graphs and is always 0-200 MPH. That's pretty worthless when you're trying to see a 2-3 MPH difference.
However, you can see the actual MPH off to the right and use the arrow keys to move along the graph and see how it changes. In this case, you can see that while my speeds at apex were the same that little shot of WOT allowed me to pick up ~ 3 MPH and I carried that speed delta all the way until I started to brake for T3. That's where the time gain came from but unfortunately I gave it all back by over slowing for T3 (min speed 5 MPH slower than my reference).
I think this is a good place to stop for now and I'll add another post later. However, I discovered one other interesting point while looking at this data. I am way overslowing for T1 and getting back to throttle too early. Look at this capture.
As you can see I am back on the throttle and accelerating before I'm even half way to the apex. I should be carrying speed through there and gradually slowing the car until the end of the inside curb, then getting back to the throttle. In this case I don't have a lap to compare to know how much it would gain me but it's something that I'm going to focus on next week when I'm back at VIR.
Finally, here are some Cosworth keyboard shortcuts I've found:
Chart
Z - zoom in
Backspace - zoom out
M - max speed jump to
N - min speed jump to
R - rate of change and can highlight section
T - tile
Map
+/- zoom in and out
Arrow keys move around the map
I cannot get the reference lap chart to lineup or lay over the subject lap. There is a gap between the two and it is hard to analyze with the delayed gap. It is very frustrating. I have used both the time and distance references and it is the same with both. In fact it does the same thing with video too. The subject lap is slightly off from the reference lap.
I cannot get the reference lap chart to lineup or lay over the subject lap. There is a gap between the two and it is hard to analyze with the delayed gap. It is very frustrating. I have used both the time and distance references and it is the same with both. In fact it does the same thing with video too. The subject lap is slightly off from the reference lap.
Have you seen this poor-sha?
First, make sure that the reference lap is a faster lap than the lap you are comparing. It gets a little wonky otherwise. Then make sure you are using distance rather than time. As you saw in my example there was also a misalignment of the time delta but the other segments lined up.
Also, are these both from the same session or event? I'm wondering if you possibly redefined the start/finish line between the two sessions you are reviewing. It's nearly impossible to define them in the same place so I can imagine that if you redefined it at a slightly different place you would have an offset in the data.
Thanks AVETTE for the kind words and I'm glad you found it useful. It seemed like a number of folks were asking for info like this and a rainy day at the beach finally gave me an opportunity to pull something together. I'll look for more examples to add to this thread.
Jim Painter was also my instructor when I got started in this hobby and has far more experience and talent than I ever will. Without data it's pretty much impossible to tell where he's faster then you but it most likely comes down to not slowing as much going to the corners, carrying more speed through the corners, and getting back to the gas sooner. I'm sure the Hoosiers help but the MPSC2 tires when new are darn near as sticky - they just don't hold up as well.
As far as the aero is concerned, I see no reason to NOT run the stage 3 aero. The testing I've done shows no difference in top speed with and without on the road course and it does help with cornering grip and overall confidence in the car in high speed sweepers. The car I've run for the last two years was a 2015 C7 Z06 (non-Z07) with the DSC Sport controller, Tractive shocks, Pirelli scrubs, and AP Racing brakes. I've recently bought a 2017 Z07 which already has the AP Racing brake and DSC Sport controller as well as the stage 3 aeri. I'll be taking it to VIR for the first time next weekend.
Let me know if you have a specific example you'd like me to look at and use. Unfortunately though the size of the PDR embedded video files makes them tough to share and move around.
Thanks poor-sha for the quick response. The reference and subject laps are from the same session and use the same start finish line. I am almost positive that the reference lap is the quick lap but I will check that. I can set it to both distance and time and the problem is the same. It seems to me that this is a relatively new phenomenon so I have obviously done something wrong.
It is on my office computer so I can take a photo of it on Tuesday and will post up my problem.
This is a cool subject. I obsess over the data after a session and I have leaned a lot by reading a book on how to analyze the data. While other set ups may be better I still believe that this is an extremely cool feature that Chevy put on the car. Mich better than the Porsche guys have at the track.
...Also, are these both from the same session or event? I'm wondering if you possibly redefined the start/finish line between the two sessions you are reviewing. It's nearly impossible to define them in the same place so I can imagine that if you redefined it at a slightly different place you would have an offset in the data.
I agree!!
I don't know what Bullet Tooth's problem is that won't allow the laps to line up the same.
I asked a buddy who ran the same event at Daytona that I did to send me a file of his laps so I could compare them to mine. He sent me a file and I watched it and immediately saw that his S/F line was defined after coming out of the pit-out lane and right as he got onto the main track - which is in the infield. I had defined the finish line on my laps to be the official S/F line on the front straight of the tri-oval used in the NASCAR, IMSA, and all other races at the track.
So, when I tried to compare our laps they were waaayyy off!! I could see his line, but his corners were all labeled differently than mine so the data was very hard to compare and interpret.
There's a ".map.gpx" file created on your SD card when you define the finish line on a track. It's a small file in the range of 150KB, and it contains a whole bunch of lat/long points (one every 2/10ths of a second) all the way around the track until you get back to where you defined the finish line.
I can open the .map.gpx files for the S/F lines I've created at various tracks by just double-clicking on the file, and it opens in Notepad on my Windows machine, or I opened Word and selected the file and opens as a text document. Here's what the first few lines of the file look like, and you can see a couple things I have boldfaced in colors - the red data is the "mark" point where you define the finish line; the blue is the exact same lat/long coordinates but is named "point 1"; and the green is 2/10ths of a second later and is named "point 2".
At the end of the red color data is numpoints="824", because there are a total of 824 points in the file - 824 is crossing the S/F line after one complete lap. This was an out-lap where I wasn't going real fast so I could define the finish line pretty precisely - the time from point 1 to point 824 was about 2 minutes 43 seconds instead of my normal 2 minute laps with no traffic.
Bottom line - this .map.gpx file has the finish line defined right on the official S/F line of the track. If everybody defined their finish line right on the same spot we could easily analyze our laps against a buddy's laps - but like Poor-sha said "It's nearly impossible to define them in the same place..."
What I'd like to see is if we could develop a sticky thread that could be a repository of .map.gpx files for tracks all over the country/world. It's always a PITA to be able to define the finish line when at a new track - especially to get it right at the official S/F line. Actually this should be done by Cosworth and posted on their site!!!
If we had a bunch of .map.gpx files, when you get ready to go to a new track you could just download the file for that track onto your SD card and you'd have exactly the same S/F line as everybody else. That would make it easy when you get to the track - just start recording and never have to worry about defining the finish line. And if you compare files with anybody else and they used the same .map.gpx file, you should have absolutely no difference in distances, etc.
Yes, I have entered my info before but the page doesn't change, doesn't present the opportunity to download.
However, your link does the trick. Thank you!
I just took a photo of my computer screen. I hope you can see what my problem is. I have chosen distance and it does not matter if I select absolute or delta. There is a .2 difference at the beginning that prevents the two graphs from laying over each other.
@BEZ06 are you saying there is a single file for each PDR recording or one for the current start finish line that changes when I go to a new track? If the former, what happens if you save the lat/lon the same for both? Will it update the start point of the data, but leave the video unchanged?
Agree this is a major downfall to the DIY start finish line setting.
When you hit the screen to "define finish line" when out on the track, you need to make a complete lap back to the point you defined, and every 2/10ths of a second the PDR will record the lat/long, course, elevation, time, etc., for each point on the track until you get back to the defined point.
Once you have created the map file (.map.gpx file on your SD card), you can do lap-after-lap, session-after-session using that SD card, and you won't need to define the finish line again for that track (well, that particular configuration!).
Take a look at this screenshot of a folder on my laptop where I've got a bunch of files from different track events. After an event I copy these files from the SD card to a folder on my computer and delete them from the SD card to open up space for my next track event. You can see there is only ONE .map.gpx file, and there are all the .mp4 files with the video and data, and .png files along with each .mp4 file. The .png file is just a small thumbnail picture.
You can see that I put my cursor on the .map.gpx file and it shows a date of 12/9/2016. I "defined the finish line" in Dec 2016, but you can see that the folder contains all these PDR videos from Flat Out event in April 2017. After the Dec 2016 event I copied those files from my SD card to a folder specifically for that event.
Before the April event when I knew I'd be running Daytona again (Rolex24 course, exact same configuration as in Dec), I stuck an SD card into my computer, copied the .map.gpx file from the Dec event to it, put it in the slot in the glove compartment, and DID NOT define any finish line in April - I just touched "Start Recording" before each session.
If you have a buddy who has defined a finish line on a track, you could copy his .map.gpx file onto your SD card, and when you go to that track you won't need to define the finish line - just go out and run the track and your files will give you lap times and lap data.
If you do that, YOU MUST RUN THE EXACT SAME CONFIGURATION OF THE TRACK that your buddy ran, or you won't get the data you want to record.
Take a look at Bill Dearborn's video where he's running the North Course at VIR. He must have had a track map (.map.gpx file) that had points for the Full Course. Scroll out to right at 3 minutes into the video and you can see that he cuts across the center of his Full Course track map:
Bottom line - it would be great if Cosworth (or us here on the forum) could develop a library of .map.gpx files for a bunch of tracks and configurations. Before we went to a new track we could simply download the file onto our SD card and go record our laps. The S/F line would be EXACTLY the same for everybody for easy comparison of laps with a buddy's video.
Before going to VIR in June I got a .map.gpx file (I think from Bill Dearborn). You can see on the map below where the S/F line is located - and it's not at the actual/official S/F line of the track. I opened the .map.gpx file as a text document, and that lat/long is the point that was marked and named as point 1 in the file:
If I'm comparing my own laps - no problem! But if I'm comparing someone else's laps to mine, and they defined a different S/F line, it might be a little harder to correlate the data properly.
Next time I go to VIR I'm going to try to define the finish line a little more precisely on the actual/official S/F line of the track.
I was helping a friend set up his driving modes, and while I was going thru the screens I noticed that his was set ay 0-70, not 60. Didn’t see any obvious way to fix this, and pretty sure he didn’t change it.
HOW TO: Basic Data Analysis with PDR
