Copying RTT results back to tune file?

[JakeRobb]

I found a four-year-old YouTube video where the guy just highlighted the cells in the RTT table, hit Copy, then moved over to VCM Editor and pasted the values in.

Has anything changed in the years since, or is that still the best way to do this? I had expected that there would be some more automatic way to do it -- something like an "apply RTT tables to tune file" button where you just click it, specify your tune file, click Apply, and it copies all of your RAM tables to the file.

Also: I assume anything you had in RAM tables is lost once the car is powered off. Is that right?

If I use VCM Scanner to switch to RAM tables and then disconnect VCM Scanner, does it continue using the RAM tables, or does disconnecting cause it to switch back to Flash tables?

[Alvin]

RTT really hasn't worked right since HPT editor v3?

You used to be able to copy the histogram fuel trims and paste them into you RTT table. Then go back and clear the histogram table so that you can get new data for the new RTT table. Before shutting down you would copy the RTT table to your editor.

[JakeRobb]

That sounds useful, but I'm asking about a different thing. You're talking about copying from histograms into the RTT tables. I'm talking about copying from RTT tables back into the tune file (e.g. the VE table).

[Alvin]

Before shutting down you would copy the RTT table to your editor.

But you miss my point in why its no longer useful.

If you average a -5 fuel trim and make a RTT correction for it and with the RTT correction it over swung to +5% than the histogram is going to average 0 because you can't clear the -5 out of it. Does that make sense? Even if the RTT made the appropiate change and the fuel trim is now 0.. the average would be -2.5%. Without being able to clear the histogram like in Pre hpt 4.0? days RTT is pretty useless.

[horsepowerguru427]

But you miss my point in why its no longer useful.

If you average a -5 fuel trim and make a RTT correction for it and with the RTT correction it over swung to +5% than the histogram is going to average 0 because you can't clear the -5 out of it. Does that make sense? Even if the RTT made the appropiate change and the fuel trim is now 0.. the average would be -2.5%. Without being able to clear the histogram like in Pre hpt 4.0? days RTT is pretty useless.

Dream crusher lol

[JakeRobb]

But you miss my point in why its no longer useful.

If you average a -5 fuel trim and make a RTT correction for it and with the RTT correction it over swung to +5% than the histogram is going to average 0 because you can't clear the -5 out of it. Does that make sense? Even if the RTT made the appropiate change and the fuel trim is now 0.. the average would be -2.5%. Without being able to clear the histogram like in Pre hpt 4.0? days RTT is pretty useless.

Okay, first let's get some terminology straight. A histogram is a bar graph. In VCM Suite, AFAIK these exist only as a visualization of log data in Scanner; there are no histograms in Editor. As such, it doesn't make any sense to ?clear a histogram.? Want the histogram gone? Close it, or close the log file.

Scanner can, however, clear the PCM's record of long term fuel trims, and that's the thing you should be doing in between successive rounds of tuning. I?m guessing you probably know these steps better than I fo, but for the sake of thoroughness:

1. Run the engine in whatever range of conditions you are currently trying to tune.
2. View LTFTs as a table (which Scanner incorrectly calls a graph).
3. Highlight and copy the LTFT values you wish to use to drive tuning changes.
4. Navigate to the RTT VE table.
5. Highlight the cells that correspond to what you copied in step 3.
6. Paste Special and select "% - half".
7. Clear LTFTs (while datalogging in Scanner, click the green toolbar icon labeled Controls and Special Functions, then click Reset Fuel Trims)

Repeat until LTFTs are all consistently less than +/- 5%.

[Alvin]

7. Clear LTFTs (while datalogging in Scanner, click the green toolbar icon labeled Controls and Special Functions, then click Reset Fuel Trims)

This doesn't clear the current graph or histogram or whatever you want to call it.. So you still have the old data being shown even though you made RTT changes.

Do you see the problem? The problem isn't that you cleared your fuel trims.. its that the histogram/graph still has all the old data loaded up. No way to clear it or reset that histogram.


I was thinking it was HPT4 that made all these changes but it was really pre HPT V3. Before V3 they called graphs histograms.

[JakeRobb]

I think I might understand. Let's focus on this bit.

This doesn't clear the current graph or histogram or whatever you want to call it.

1. The thing that isn't getting cleared is a grid of numbers, right? That's a table. The "Graphs" section supports two kinds of things -- tables (which are not graphs) and histograms (which are graphs).
2. I'm assuming the grid of numbers is showing LTFT values, with a columns for RPM and rows for MAP. Right?

If I have that right, then I think your problem is that the table is showing you data that was recorded in the current datalogging session. It is not showing you the current stored LTFTs; it's showing you the observed LTFTs from your logging session. To clear it, you need a new logging session. Stop logging, save (if you want to) and close the log file, then start logging again.

You could also put the table in the Zoom Data Range mode (rather than All Data), then zoom/scroll to a different range on your timeline. There's no way to scroll to a "future" range such that it's clear until you do some more driving, so this solution doesn't really work if your goal is to clear out the table so you can do some more driving and repopulate it.

[SiriusC1024]

LTFT's are stored by the PCM. They have to be explicitly cleared with Scanner>Vehicle Controls & Special Functions if you want them reset.

A graph is a visual representation of a table. It's an abstraction. The raw data is neither. Whether you see it as a histogram, graph, or a table is only a matter of the way you personally perceive the data. I can look at those tables and see the graph in my head automatically. Really, making a point to distinguish between the two has more to do with where you lie on the spectrum than analytics and actual tuning.

You're going to wear out your welcome if you keep on like this. Why is the guy who couldn't figure out a throttle cable trying to tell a nationally-regarded tuner with 2+ decades experience what's what?

[Alvin]

You could also put the table in the Zoom Data Range mode (rather than All Data), then zoom/scroll to a different range on your timeline. There's no way to scroll to a "future" range such that it's clear until you do some more driving, so this solution doesn't really work if your goal is to clear out the table so you can do some more driving and repopulate it.

You cannot do this while logging.. Try it

RTT is not useful in its current state.



Give up on he semantics on graph and histogram. I've been telling you this entire time the scanner hasn't been consistent throughout it's lifespan. Us old guys who've been using HPT since the serial cable days know them as histograms cause that's what the damn scanner called them. And that is what the scanner called them until recently.

[JakeRobb]

LTFT's are stored by the PCM. They have to be explicitly cleared with Scanner>Vehicle Controls & Special Functions if you want them reset.

Yep, we covered that already; Alvin is having a different problem.

A graph is a visual representation of a table. It's an abstraction. The raw data is neither. Whether you see it as a histogram, graph, or a table is only a matter of the way you personally perceive the data. I can look at those tables and see the graph in my head automatically. Really, making a point to distinguish between the two has more to do with where you lie on the spectrum than analytics and actual tuning.

I can see your point, but consider the perspective of someone new to tuning, who has never used anything but VCM 5. Me, for example. If you tell me to use a histogram, but you actually mean something that is not called a histogram in the application (nor does it fit the literal definition of the term), I’m going to be confused, and I’m going to ask you for clarification. I know old habits die hard, but when asked overtly whether you actually meant “table” when you said “histogram,” you really should make an effort to be consistent with the current naming scheme. As it stands, if I want to know anything about actual histograms (bar graphs), I’m kinda screwed, because all of my searches turn up stuff about working with tables. Without knowledge of the naming history in the application (which I didn’t know about until this thread — and why would I?), it’s been extremely difficult for me to learn how these functions work at all. Searching these forums for “histogram” yields basically nothing at all about using the so-named bar graph feature. Googling “hptuners histogram” isn’t any better — in fact, that led me to the HPTuners YouTube channel; a video explaining how to use tables and never once mentioning histograms — but it matched my search because “histogram” was in the video title. Not a single mention of the name change in the video.

Frankly, VCM Suite is some of the most unintuitive software I’ve ever seen in 37 years of building software. Poor naming isn’t helping. Let’s maybe not make it worse by refusing to adapt when the interface changes terminology?

I don’t know if I’m on the spectrum. Maybe; a lot of software engineers are. But the neuropsychological exam I did a few years back didn’t reveal anything in that area….

You're going to wear out your welcome if you keep on like this.

I’ll try to keep myself in check. For the benefit of everyone else who comes here in the future, I’d encourage you (and everyone) to accept and embrace the new names.

Why is the guy who couldn't figure out a throttle cable trying to tell a nationally-regarded tuner with 2+ decades experience what's what?

As long as we’re engaging in Appeals to Authority (logical fallacies are fun!), I have 3+ decades experience working with tabular data and visual representations thereof, including histograms. I guess that means I should be a throttle cable adjustment expert!

I certainly deserved the throttle cable dig, but FWIW I did figure it out after less than an hour of hands-on time. If only I had more time with the car and less time with the computer….

You cannot do this while logging.. Try it

Okay, then how about the suggestion I gave first?

Give up on he semantics on graph and histogram.

Please see my reply to Sirius, above. It’s relevant to you too.


My favorite thing about this thread is that nobody has even attempted to answer my original question. Is copy+paste from Scanner to Editor still the best way?

[SiriusC1024]

All I'm saying is keep your cool and you'll get all the help you need. Don't get argumentative. Should be "If I'm understanding this correctly" and not "What you mean is" when needing clarification.

I answered the question about the scanner in my brief tutorial. Paste percent-half. The more you work with the tune and see the results, the more you'll be able to intuit whether it's percent-half, percent, multiply a zone by a whole percentage based on an average, magnitude of STFT in respect to LTFT, all that. For now, percent-half will work no matter what, so stick to doing that.

[Alvin]

Good luck!

[JakeRobb]

I answered the question about the scanner in my brief tutorial. Paste percent-half.

I assumed I was pasting percent-half into the RTT table in Scanner, and I assumed I'd be doing that multiple times in a tuning session. At the end of the tuning session, I'll want to capture that information into the tune file. At that point, is the best thing to do to just copy the whole RTT table from Scanner over to the equivalent table in Editor?

Of course, I could paste percent-half into the VE table in Editor, then re-write the calibration to the vehicle each time, but that kinda undermines the point of RTT, doesn't it?

As long as I'm confirming assumptions, I'd like to confirm my understanding of the behavior of "Paste percent-half." If I understand correctly: you start by copying some percent values, e.g. fuel trims. When you paste percent-half, the application will take the target values and reduce them by half of the percentage you copied. It's basically a stepwise way to cut down the recorded errors by half, getting things progressively closer and closer to the right values in the base table.

For an example, let's say I have this in my VE table:

10	20	30
40	50	60
70	80	90

And this in my LTFTs:

10%	5%	3%
-5%	-10%	-8%
1%	0%	-15%

If I selected the entire LTFT table, copied it, then selected the entire VE table and did paste percent-half, The result should be:

10.5	20.5	30.45
39	47.5	57.6
70.35	80	83.25

Do I have that right? (I don't expect you to check all of my math.)

[gtstorey]

"RTT is not useful in its current state."

[JakeRobb]

The more you work with the tune and see the results, the more you'll be able to intuit whether it's percent-half, percent, multiply a zone by a whole percentage based on an average, magnitude of STFT in respect to LTFT, all that.

I kinda wonder if it wouldn't make sense to do a first pass at percent rather than percent-half. Barring erroneous values (shouldn't be many, I'd imagine), it seems like that would greatly reduce the number of tuning iterations -- especially if you had done a lot of driving and worked hard to cover the whole table, such that the LTFTs are in really good shape. A round or three of percent-half after that should should help to resolve whatever errors pop up.

It might be nice if there was an option to paste percent-3/4 or something, to increase the rate at which you approach the targets while still leaving room for erroneous trim values. Of course, you could do this with a transform function applied to the LTFTs before you copy them....

Is that a reasonable thought? Any pitfalls?

I really appreciate all the advice you've given me so far. I'm excited to get past the theory bits of this process and do some real, hands-on tuning! Between my TPS sensor issue, my schedule, and the weather, I don't know when that's going to be....

[JakeRobb]

"RTT is not useful in its current state."

If I've understood Alvin's issue correctly, it's not as useful as it should be -- but that doesn't mean it's not useful at all. You can still make changes on the fly, which is the point. It's just difficult to do multiple rounds of changes without stopping periodically to clear the trims from the log file.

BTW, Alvin, have you posted about this issue before? Or submitted a support ticket? Had any response from the HPTuners folks?

[Alvin]

It's not Alvin's issue. I clearly explained several times what the problem with RTT is in its current state. I can do no more to help you understand that.

And yes.. If you check my post history I've made them aware of this several times when the version came out that effectively killed RTT usefulness.

[JakeRobb]

It's not Alvin's issue.

Sorry, I only meant that as a shorthand to refer to the issue you described.

And yes.. If you check my post history I've made them aware of this several times when the version came out that effectively killed RTT usefulness.

I only found one: https://forum.hptuners.com/showthrea...plying-changes

The response is not great, but they suggested the same thing I did: stop and start logging again. Your response to that was "how do we know the RTT modified tables are still being used?" -- to which you never got an answer, but that seems like something you can check easily enough. In the past seven years, have you taken the time to check?

In my search, I also found this thread: https://forum.hptuners.com/showthrea...s-scanner-data

But the YouTube video is gone. Is that something you've since decided not to share? I'd love to see what you had.

[SiriusC1024]

I assumed I was pasting percent-half into the RTT table in Scanner, and I assumed I'd be doing that multiple times in a tuning session. At the end of the tuning session, I'll want to capture that information into the tune file. At that point, is the best thing to do to just copy the whole RTT table from Scanner over to the equivalent table in Editor?

Of course, I could paste percent-half into the VE table in Editor, then re-write the calibration to the vehicle each time, but that kinda undermines the point of RTT, doesn't it?

As long as I'm confirming assumptions, I'd like to confirm my understanding of the behavior of "Paste percent-half." If I understand correctly: you start by copying some percent values, e.g. fuel trims. When you paste percent-half, the application will take the target values and reduce them by half of the percentage you copied. It's basically a stepwise way to cut down the recorded errors by half, getting things progressively closer and closer to the right values in the base table.

For an example, let's say I have this in my VE table:

10	20	30
40	50	60
70	80	90

And this in my LTFTs:

10%	5%	3%
-5%	-10%	-8%
1%	0%	-15%

If I selected the entire LTFT table, copied it, then selected the entire VE table and did paste percent-half, The result should be:

10.5	20.5	30.45
39	47.5	57.6
70.35	80	83.25

Do I have that right? (I don't expect you to check all of my math.)

You have that right. Except the RTT confusion. I don't use it. Logging, stopping, tuning, gives me a chance to think about my changes and check for other problems. If you RTT you might end up tuning around a problem or into a problem. RTT should only be used to make minor adjustments on a known-good configuration. Any perceived time-savings is offset by the risk imo.