How to correctly read a dyno graph (F10 M5 graph shown))

[Dan@BPC]

Hello everyone, my name is Dan Connor. Happy to be a new member of this forum group. I have been seeing a lot of misleading post's of dyno graphs with inflated numbers. In case some of you do not know that it is possible to do this I am going to share how it is done. The purpose of this post is so we can all be more accurate as a community when comparing results.

Before I begin I am aware every dyno reads different. The main area we are looking at is how the numbers are displayed in the scale selected for the graph.

The are 2 main scalars which are SAE and STD. The second part is smoothing factor which is normally 0-5.

SAE (Standard of Automotive Engineer's) is the closest to manufacture spec. It will generally be a lower number. This calculation takes into consideration outside air temp, humidity and altitude just to name few. All these play a key factory in air density which is key in making horsepower/tq.

STD (Standard) numbers are the least accurate and most inflated. This is the graph that I have been seeing the most. This calculation also takes into consideration outside air temp, humidity and altitude. However it uses a higher level of air density will which nets a higher and less accurate number.

To compare our results accurately we need to show SAE numbers. The people running the dyno know about this. The reason they show STD numbers are to make you feel better about the money being spent. They know the car is most likely not going to be used in a competition environment and the number will not be questioned.

Lastly the is the smoothing factor. This is what what it sounds like. It smooths out the peaks and ripples in the graph. When this is set low i.e. 1 or 0 you are being shown a false number off of a spike in the graph.

So to sum it up for all of us to accurately compare dyno numbers we should all ask for and post our graphs in SAE and smoothing factor of 5. Below is a dyno pull on a 2015 Comp package F10 M5 with a flash tune, catless downpipes, stock exhaust, meth kit and 93 octane. The 3 graphs show below are all from the same dyno pull scaled 3 different ways.

The 1st graph is shown in SAE (most accurate).

The second graph is shown in STD. Horsepower/TQ

The 3rd graph is shown in STD with smoothing of 0. This shows how the ripple/spike in the graph can raise the peak number.

Hope this helps everyone and thanks for reading.

Dan

[SECOND2NONE]

I'm glad this was posted.. I was always looking at all the glory dyno runs being posted. wanting to sell my e46 m3.. After talking to my tuner Mike@ maximum psi. I realized that on paper, my car was making the same or more power as the cars I was comparing to..

if you don't mind I will like to post this on other forums if you haven't already

[Thor2j]

This is actually wrong information Dan. Majority of the time BOTH SAE and STD are Corrected values. SAE is corrected to lower atmospheric conditions then STD.

STD:
Air Temperature: 60F
Absolute Pressure: 29.92 inches Hg
Relative Humidity: 0%

SAE:
Air Temperature: 77F
Absolute Pressure: 29.23 inches Hg
Relative Humidity: 0%

Either one of these methods can also be Uncorrected, meaning on this given day at these current conditions this is what u got. To go from corrected SAE to STD is about a 4% loss.

[dadysev]

Not hijacking the thread but this is so similar to the false power rating of audio amplifiers.
Meaning the RMS (route mean square) value is much lower and the real power which is not as glorious as everyone wants. For example 100 watts RMS is really like 250 watts peak to peak which means nothing but to the consumer buying it, they love those high numbers.
Rule of thumb, amplifiers are 60-70% efficient and cannot make power than they consume. So the next time you see an amplifier with a higher rating on the output than the power consumption just remember this post and its false advertising.
To determine the power consumption in, multiply the supply voltage times the maximum fuse rating on the amp, unless its written on it, and thats your power consumption in Watts.

[Dan@BPC]

Originally Posted by SECOND2NONE

I'm glad this was posted.. I was always looking at all the glory dyno runs being posted. wanting to sell my e46 m3.. After talking to my tuner Mike@ maximum psi. I realized that on paper, my car was making the same or more power as the cars I was comparing to..

if you don't mind I will like to post this on other forums if you haven't already

Thank you, feel free to share

Originally Posted by Thor2j

This is actually wrong information Dan. Majority of the time BOTH SAE and STD are Corrected values. SAE is corrected to lower atmospheric conditions then STD.

STD:
Air Temperature: 60F
Absolute Pressure: 29.92 inches Hg
Relative Humidity: 0%

SAE:
Air Temperature: 77F
Absolute Pressure: 29.23 inches Hg
Relative Humidity: 0%

Either one of these methods can also be Uncorrected, meaning on this given day at these current conditions this is what u got. To go from corrected SAE to STD is about a 4% loss.

Thank you for that info. Look at the temperature numbers above, STD uses a lower number, thus its saying that the air density is higher than SAE. So higher air density = higher hp numbers. The reason I posted this information is to show that STD numbers always display higher numbers than SAE. I just wanted to bring this up since SAE is the industry standard way of horsepower and torque correction, helping someone to get a more real world numbers than what most post on the internet. Just so everyone can compare, I've posted the uncorrected numbers from the same run along with the condition information so someone could take the calculate listed above.

[Thor2j]

Originally Posted by Dan@BPC

Thank you, feel free to share



Thank you for that info. Look at the temperature numbers above, STD uses a lower number, thus its saying that the air density is higher than SAE. So higher air density = higher hp numbers. The reason I posted this information is to show that STD numbers always display higher numbers than SAE. I just wanted to bring this up since SAE is the industry standard way of horsepower and torque correction, helping someone to get a more real world numbers than what most post on the internet. Just so everyone can compare, I've posted the uncorrected numbers from the same run along with the condition information so someone could take the calculate listed above.

Yes, STD corrected will always be about 4% higher the SAE corrected numbers. Your post refered to STD not taking into account atmospheric conditions which is not true. Usually both these numbers are corrected unless stated Uncorrected. The problem with Uncorrected numbers is you can't compare 2 different runs.

[Dan@BPC]

Originally Posted by Thor2j

Yes, STD corrected will always be about 4% higher the SAE corrected numbers. Your post refered to STD not taking into account atmospheric conditions which is not true. Usually both these numbers are corrected unless stated Uncorrected. The problem with Uncorrected numbers is you can't compare 2 different runs.

Thank you for clearing that up, I missed that. Let me correct it above. I want to make sure this information is accurate for our community. Anyone else that has information that they would like to add please chime in. It is important we are all on the same page.

[CanAutM3]

I do not fully agree with your post. Since modern turbo engines adapt to atmospheric conditions, using correction factors will skew the result. So it is not correct to say STD or SAE are the more accurate number for these engines. IMO, the uncorrected data is probably the most accurate one. I have read that the SAE and EU regulation even specifiy that for turbo engines that have compensation mechanisms, uncorrected numbers are to be used for the official power ratings.

The fundamental purpose of correction factors is to baseline the numbers so they can be compared to each other. In essence, for the same engine tested in different atmospheric conditions to show the same results. If correction factors are applied to the result of an engine that compensates atmospheric conditions, it will show varying results. Which defeats the purpose.

[Dan@BPC]

Originally Posted by CanAutM3

I do not fully agree with your post. Since modern turbo engines adapt to atmospheric conditions, using correction factors will skew the result. So it is not correct to say STD or SAE are the more accurate number for these engines. IMO, the uncorrected data is probably the most accurate one. I have read that the EU regulation even specifiy that for turbo engines that have compensation mechanisms, uncorrected numbers are to be used for the official power ratings.

The fundamental purpose of correction factors is to baseline the numbers so they can be compared to each other. In essence, for the same engine tested in different atmospheric conditions to show the same results. If correction factors are applied to the result of an engine that compensates atmospheric conditions, it will show varying results. Which defeats the purpose.

I corrected the statement above, its not that the numbers are more accurate, but SAE is the closest to manufacturer's numbers. Yes, the EU has there regulations, but here in the states they have to get certified by SAE. Quote from SAE Wiki:

A few manufacturers such as Honda and Toyota switched to the new ratings immediately, with multi-directional results; the rated output of Cadillac's supercharged Northstar V8 jumped from 440 to 469 hp (328 to 350 kW) under the new tests, while the rating for Toyota's Camry 3.0 L 1MZ-FE V6 fell from 210 to 190 hp (160 to 140 kW). The company's Lexus ES 330 and Camry SE V6 were previously rated at 225 hp (168 kW) but the ES 330 dropped to 218 hp (163 kW) while the Camry declined to 210 hp (160 kW). The first engine certified under the new program was the 7.0 L LS7 used in the 2006 Chevrolet Corvette Z06. Certified power rose slightly from 500 to 505 hp (373 to 377 kW).

As Thor2j stated above, it enable someone to be able to repeat numbers than have the most accurate numbers of that moment in time. And having the ability to have access to SAE numbers from the factory and seeing the SAE numbers at the wheel would be the most "accurate" way to judge the horse power gains.

[CanAutM3]

Originally Posted by Dan@BPC

I corrected the statement above, its not that the numbers are more accurate, but SAE is the closest to manufacturer's numbers.

As Thor2j stated above, it enable someone to be able to repeat numbers than have the most accurate numbers of that moment in time. And having the ability to have access to SAE numbers from the factory and seeing the SAE numbers at the wheel would be the most "accurate" way to judge the horse power gains.

This is also incorrect. Chassis dynos, especially roller inertia dynos, are inherently inaccurate at producing absolute numbers. You will simply not get reliable numbers that approach manufacturer ratings, regardless of the correction factor used. There are simply too many variable parameters that are difficult to control.

Chassis dynos are good for comparative measurement though. Test before and after mods in similar conditions, and you will get a decent appreciation of the relative gains provided by the mods. Use uncorrected, STD or SAE, it does not matter, as long as you use the same correction for the before and after runs and that the runs were done in similar conditions.

Originally Posted by Dan@BPC

Yes, the EU has there regulations, but here in the states they have to get certified by SAE. Quote from SAE Wiki:

It is important not to confuse SAE testing standards with SAE correction factors.

BTW, even the SAE standards specify that for engines that compensate for atmospheric conditions, no correction factor should be applied. See paragraph 5.5.2 below (thanks to dmnc02 for the reference). So anyone applying correction factors to engines that compensate for atmospheric conditions is NOT respecting SAE standards .

[MCarsFan]

We get your points.. Butt Dyno for me is best though.
Where possible I try to be consistent in using the same Dyno for all my testing and for comparison purposes.

[Dan@BPC]

If we refer back to the 3 graphs that were in the first post. They show how with the click of a scaling button you can turn a 676hp/625tq car into a 700hp/655tq car.

We spend a considerable amount of money on exhaust, intakes, flash tunes, piggy backs, turbos ect.. Period bottom line, us as a community need to post SAE numbers with the smoothing factor turned up at or near 5. Without doing this we are not accurately comparing gains. There is no argument or debate. SAE will always be the most accurate. The only reason to post standard numbers with the smoothing factor turned down is to make someone's product show the highest number possible so that you will buy it over the other guys. BMW's are amazing cars! We know this and that is why we drive them. If we want to give ourselves the rep of being the best most educated/knowledgeable group of car enthusiasts out there we all need to get on the same page.

[CanAutM3]

Originally Posted by Dan@BPC

If we refer back to the 3 graphs that were in the first post. They show how with the click of a scaling button you can turn a 676hp/625tq car into a 700hp/655tq car.

We spend a considerable amount of money on exhaust, intakes, flash tunes, piggy backs, turbos ect.. Period bottom line, us as a community need to post SAE numbers with the smoothing factor turned up at or near 5. Without doing this we are not accurately comparing gains. There is no argument or debate. SAE will always be the most accurate. The only reason to post standard numbers with the smoothing factor turned down is to make someone's product show the highest number possible so that you will buy it over the other guys. BMW's are amazing cars! We know this and that is why we drive them. If we want to give ourselves the rep of being the best most educated/knowledgeable group of car enthusiasts out there we all need to get on the same page.

You are completely missing the point. Don't hunt for absolute numbers, it is utterly useless (my M5 makes 600whp with a tune). Look for relative gains (my tune gave me a 10% gain).

From that perspective, it does not really matter what correction factor is being used, SAE is not any more accurate. What is important is to use the same correction method for the before and after runs. I agree about the smoothing though, spikes need to be removed. But the values for the entire curve are far more telling than the single peak value.

[M5Rlz]

Might you care to also enlighten some of our Dynojet friends that their are much lower reading dynos like mustang ones, and that having the higher numbers isn't always a good thing

[allmotor_2000]

Dyno's are a relative measure to tune the car - take it to the track or do some VBOX runs. There were cars making similar (or more) power running 10mph less in the 1/2 mile. Go figure!

[MCarsFan]

Originally Posted by allmotor_2000

Dyno's are a relative measure to tune the car - take it to the track or do some VBOX runs. There were cars making similar (or more) power running 10mph less in the 1/2 mile. Go figure!

That's my point.

[allmotor_2000]

Also, I think we should start posting the drf file or all the relevant dyno parameters including correction factors etc. It is easy to manipulate otherwise!

[Boss330]

Originally Posted by CanAutM3

You are completely missing the point. Don't hunt for absolute numbers, it is utterly useless (my M5 makes 600whp with a tune). Look for relative gains (my tune gave me a 10% gain).

From that perspective, it does not really matter what correction factor is being used, SAE is not any more accurate. What is important is to use the same correction method for the before and after runs. I agree about the smoothing though, spikes need to be removed. But the values for the entire curve are far more telling than the single peak value.

And just to add one thing to what you are saying here which is completely correct BTW:

If using SAE or STD corrections when comparing before and after on a modern turbo engine you MUST have the exact same atmospheric conditions! As you stated (and which was debated in detail on the F8x forums a year or two ago) these modern engines self correct for atmospheric conditions. Adding dyno correction on top of that constitutes a "double dipping" of correction factors. So if the atmospheric conditions have changed on the "after" run (often taken later in the day) the ECU/DME will auto correct boost, timing and fuel for this. But if you use dyno correction for SAE and STD the correction factors will be different between the "before" and "after" runs and thereby skewing the dyno graph between the two runs. On a NA engine the correction factors is there just because of this, to make the end graph comparable even though atmospheric conditions have changed. On a modern turbocharged engine, using these correction factors makes the end dyno graphs void!

European magazines like Sport Auto is now publishing their dyno numbers on turbo engines without correction factors just because of this.

So DO NOT use any correction factors when dynoing modern turbo engines!!! It will just make the dyno graphs inconsistent and NOT comparable!

And then we could also open the debate on dyno inconsistencies, both between different brands of dynos and the strange situation of "low reading" and "high reading" dynojets... Comparison between different dynos is allmost impossible to do with any credibility. Dynos like these are best used to compare a delta gain on before and after runs on the same dyno.

[Msizzle]

I have a modded M6 and I love the car, but I have never seen a platform make so much claimed power and go so "slow" My old v10 M6 went 11.81@ 121 with 508whp and those cars made no tq. These new cars I have seen with literally 200whp more and 300wtq more and not get out of the 11's and not go 130!

[Msizzle]

OP thankyou for making this post, its great to see more technical info on this forum

[Boss330]

Originally Posted by Msizzle

OP thankyou for making this post, its great to see more technical info on this forum

Agreed about tech articles/posts. We did quite a few deep tech articles on both dyno, dyno consistency, dyno measurement problems, correction factors etc over on the F8x forums a year or two ago.

Problem was that many of them went bad after a few pages and the mods closed a few of them after a certain tuner name was talked about (that also happened to be a forum sponsor at the time...)

The OP is a good starting point, but unfortunately has some wrong conclusions...

[M5Rlz]

Originally Posted by Msizzle

I have a modded M6 and I love the car, but I have never seen a platform make so much claimed power and go so "slow" My old v10 M6 went 11.81@ 121 with 508whp and those cars made no tq. These new cars I have seen with literally 200whp more and 300wtq more and not get out of the 11's and not go 130!

Yeah and even your m6 v10 was among some of the best times lol


Personally someone could post m6 made 7909090whp tomorrow. It'll mean nothing to me till it sets some records.