Rounding rules for TP200 and TP360

General discussions of measurement techniques and the results of testing of techniques and equipment.

Moderators: edfrank, dbhguru

#1)  Rounding rules for TP200 and TP360

Postby dbhguru » Tue Oct 16, 2012 12:51 pm

NTS,

  Here is the text of an email from Steve Colburn, Director of Sales for North America for LTI, sent to a number of Ents and affiliates.


Hello Bob, et.al.!

Thanks as always for putting our instruments to the test, Bob. There is nothing like an impartial keen eye to keep us on our toes! My delayed response to this message is due to the fact that I wanted to verify something with our (very busy) Engineering dept. It was to determine the formula used to round the numbers seen in the display of the TruPulse, after a precision distance measurement is completed. Here is the answer:

When units are set to Feet with the increments at 0.5', the unit rounds as follows:

10 = 9.75 to 10.24
10.5 = 10.25 to 10.74
11 = 10.75 to 11.24

When you confirmed with me, Bob, that you positioned yourself right at the changeover point with the higher number displayed, this would put you at the low side of the value, ~ 0.25' less than the display. If you factor this in to the data set attached, the numbers get even better with an average error of ~ 0.08' or about an inch!

This demonstrates that careful use of the instrument with proper procedure can yield very accurate results. We need to keep in mind, however, that most users won't practice quite the level of care as Bob does when measuring and they also will be shooting to more "uncertain" targets like leaves, twigs, candles, etc. At any rate, it sure is nice to see confirmation in the data that the process is working as planned!

Many thanks also for organizing the workshop at MTSF last Friday Bob, you did a great job as usual. Walks through the woods measuring trees with you is quickly becoming one of my favorite activities!


Warm regards,
Steve Colburn
Director of Sales, North America
Laser Technology, Inc.


   By taking the time, it is apparent that we can measure distances extremely accurately with the TruPulse line. From a spot of know distance, the angle can be measured independently to get our best determinations of height to date. Note that Steve says

This demonstrates that careful use of the instrument with proper procedure can yield very accurate results.


   By proper procedure, Steve means the sine method where that method needs to be used. He has to be cautious about what he says for obvious reasons, but he understands what works and what doesn't. We are making rapid strides in getting the message across.

Bob
Robert T. Leverett
Co-founder and Executive Director
Native Native Tree Society
Co-founder and President
Friends of Mohawk Trail State Forest
User avatar
dbhguru
 
Posts: 3921
Joined: Mon Mar 08, 2010 10:34 pm
Location: Florence, Massachusetts
Has Liked: 3 times
Has Been Liked: 1035 times
Print view this post

#2)  Re: Rounding rules for TP200 and TP360

Postby Larry Tucei » Thu Oct 18, 2012 11:55 am

Bob,  Congrats on all you do for North America  LTI. Your tenacity with measurement accuracy has set a high standard for all of us at NTS.  I can't wait to get up to England in the future to learn from the best in the buisness.   Larry
User avatar
Larry Tucei
 
Posts: 1751
Joined: Tue Mar 09, 2010 11:44 am
Location: Southern Mississippi
Has Liked: 632 times
Has Been Liked: 564 times
Print view this post

#3)  Re: Rounding rules for TP200 and TP360

Postby dbhguru » Fri Oct 19, 2012 6:14 pm

Larry,

    Thanks. I've begun testing the tilt sensor on the TP200. That's a more difficult for of testing. So far the average deviation is 0.204 degrees. About half the readings have a deviation of less than 0.2 degrees. However, an occasional readings can be off by as much as +/- 0.5 degrees. The stated accuracy is within +/- 0.25 degrees. The tilt sensor needs fairly frequent calibrations. All in a day's work.

Bob
Robert T. Leverett
Co-founder and Executive Director
Native Native Tree Society
Co-founder and President
Friends of Mohawk Trail State Forest
User avatar
dbhguru
 
Posts: 3921
Joined: Mon Mar 08, 2010 10:34 pm
Location: Florence, Massachusetts
Has Liked: 3 times
Has Been Liked: 1035 times
Print view this post

#4)  Re: Rounding rules for TP200 and TP360

Postby Matt Markworth » Thu Mar 03, 2016 8:54 pm

Bob, All,

When measuring in feet (displayed in increments of .5 feet) with my Trupulse 200, I generally calculate the vertical distance manually rather than relying on the displayed number for VD. When measuring in the smaller increment of .1 yards, I generally just go with the displayed VD number.

For example, I just did a quick test aiming at the ceiling. Here are the results:

TP 200 Displayed VD: 4.5'
TP 200 Displayed SD: 7.0'
TP 200 Displayed Angle: 38.1º

Calculated VD : 4.3'

In this example, I don't see any benefit in letting the instrument round 4.3' up to 4.5'. The distance that was directly measured (the slope distance) of 7.0' already represents a rounded number (unless I just happen to be at exactly 7.0' from the target), so I don't see any benefit to allow the instrument to also round the sine calculation. Does anyone see any faulty logic in this?

Matt
User avatar
Matt Markworth
 
Posts: 1270
Joined: Fri Nov 23, 2012 9:41 pm
Location: Cincinnati, OH
Has Liked: 1058 times
Has Been Liked: 807 times
Print view this post

#5)  Re: Rounding rules for TP200 and TP360

Postby dbhguru » Fri Mar 04, 2016 10:01 am

Matt,

  According to LTI engineers, the rounding point where the display goes from x.0 to x.5 is x.27. I presume that the point where x.5 goes to x+1 is x.77. Not sure why, but that is my understanding. As you recognize, when shooting distance and angle separately, distance can be set to yards and the display reads to the 0.1 of a yard, which of course is 0.3 feet. I think brother Will prefers this unit of measure.

   When employing both the laser and tilt sensor, such as is invoked with the HD and VD returns, the limits of accuracy  of both sensores are brought into play. The interaction is not always predictable, i.e no fixed rule you can go by. Then if the ML routine is used on the 360, the limits of accuracy of the compass enter the equation. Interestingly, the SD, HD, and VD returns of the ML routine are all displayed to 0.1 feet. This is misleading in terms of accuracy, but at least it does not invoke arbitrary rounding to the nearest 0.5 feet.

    More on this topic later.

Bob
Robert T. Leverett
Co-founder and Executive Director
Native Native Tree Society
Co-founder and President
Friends of Mohawk Trail State Forest
User avatar
dbhguru
 
Posts: 3921
Joined: Mon Mar 08, 2010 10:34 pm
Location: Florence, Massachusetts
Has Liked: 3 times
Has Been Liked: 1035 times
Print view this post

#6)  Re: Rounding rules for TP200 and TP360

Postby Don » Fri Mar 04, 2016 9:49 pm

Larry-
I know that anywhere out of the South seems foreign to you...; ~ } but I think you meant New England? I thought, Bob's not going to England, is he?
-Don
PS:I know, picky picky picky...:~)

Larry Tucei wrote:.  I can't wait to get up to England in the future to learn from the best in the buisness.   Larry
Don Bertolette - President/Moderator, WNTS BBS
Restoration Forester (Retired)
Science Center
Grand Canyon National Park

BJCP Apprentice Beer Judge

View my Alaska Big Tree List Webpage at:
http://www.akbigtreelist.org
User avatar
Don
 
Posts: 1402
Joined: Tue Mar 09, 2010 1:42 am
Location: Anchorage, Alaska
Has Liked: 64 times
Has Been Liked: 217 times
Print view this post

#7)  Re: Rounding rules for TP200 and TP360

Postby Matt Markworth » Sun Mar 06, 2016 8:41 am

Bob,

Interesting, I wasn't aware that the 360 displayed increments of .1 ft for missing line. If they ever make changes to their software, it might make sense to show VD and HD in increments of .1 ft as well (or least make that an option that can be turned on/off). Perhaps they wouldn't want to because it may leave the impression that the instrument has a higher degree of accuracy than it actually does, although I believe that it would get us closer to the actual VD or HD distance without the unnecessary rounding.

I wasn't sure if the displayed VD result utilized the already rounded displayed SD measurement to perform the sine calculation, or if the displayed VD result utilized the internal, pre-rounded SD measurement. Based on some testing as shown below, I was able to confirm that the displayed VD result utilizes the internal, pre-rounded SD measurement.

I used arcsine to find a couple angles that would be near to the click-over point for rounding, and found a couple angles that would work with my TruPulse 200 on a tripod positioned 5 ft away from a wall. I moved the Trupulse closer to the wall by a couple inches and positioned it at the inner limit of where the unit would still read a SD of 5', and took a bunch of VD measurements with an angle of 2.8 degrees and 2.9 degrees. I then moved the Trupulse away from the wall and found the outer limit of where the SD would still read as 5'. I then took a bunch more VD measurements at angles of 2.8 degrees and 2.9 degrees. With each press of the firing button I toggled through the results to ensure that SD was showing as 5' and that the angle was one of two that I was recording: 2.8 degrees or 2.9 degrees.

Here are the results:

SD reading of 5', Angle of 2.8 degrees
Inner limit where SD read as 5': VD = 0', 0', 0', 0', 0', 0'
Outer limit where SD read as 5': VD = .5', .5', .5', .5', .5', .5'

SD reading of 5', Angle of 2.9 degrees
Inner limit where SD read as 5': VD = 0', 0', 0', 0', 0', 0'
Outer limit where SD read as 5': VD = .5', .5', .5', .5', .5', .5'

If the unit could show VD in increments of .1 ft, then the displayed VD measurements would have been .2' and .3', instead of 0' and .5'. Based on the finding that the VD measurement is using an unseen, internal SD measurement, it would probably just require a software change to begin showing VD in increments of .1 ft (and maybe a note in the manual explaining accuracy limitations and rounding methodology).

Matt

For this message the author Matt Markworth has received Likes :
Will Blozan
User avatar
Matt Markworth
 
Posts: 1270
Joined: Fri Nov 23, 2012 9:41 pm
Location: Cincinnati, OH
Has Liked: 1058 times
Has Been Liked: 807 times
Print view this post

#8)  Re: Rounding rules for TP200 and TP360

Postby Will Blozan » Sun Mar 06, 2016 9:57 am

Matt,

Brilliant test! Now for another one: do the same but run the ML routine...

Great work!

Will
User avatar
Will Blozan
 
Posts: 1144
Joined: Fri Mar 12, 2010 8:13 pm
Location: North Carolina
Has Liked: 1521 times
Has Been Liked: 431 times
Print view this post

#9)  Re: Rounding rules for TP200 and TP360

Postby dbhguru » Sun Mar 06, 2016 11:52 am

Matt,

  Ditto to what Will said. Excellent work. In my conversations with LTI officials, they have always been concerned about overstating the accuracy of their instruments, but the choice of reflecting SD, VD, and HD values on the display of the TruPulse line to the nearest 0.5 feet is partly a marketing decision. They've always known that the TruPulse surpasses the stated accuracy of +/- 1 foot. However, until we did lots of field testing and fed them the results, I don't think they appreciated just how much better the accuracy is. Marketing decisions don't necessarily jive with the underlying engineering. LTI's use/inclusion of the HT Routine is another case in point. LTI chose to go with the standard timber industry Tangent Method as the way to measure tree height because they were asked to by their biggest customers.

  Yesterday I posted, as you likely have seen, the results of a 66-trial experiment on the accuracy of my TP 360. Thirty-seven trails were with the unit on a tripod and 29 were hand held. The gold standard for distance was my Bosch GLM 80 accurate to +/- 1.5 millimeters to distinct targets. All trials were indoors. The first 37 were in my basement. The last 29 were in the Earle Recital Hall at Smith College. There, distances varied from 5.5 to 63.5 feet. The distances for the first 37 trials (tripod group) varied from 14.0 to 24.0 feet.

  The average SD distance error for the tripod trials was 1.98 inches. The maximum error was 4.8 inches. The average error for the hand held trials was 2.2 inches - surprisingly close, but the maximum error was 7 inches. Not surprising.

  I hope to do a series of tests on the ML routine in the next few days since I am using it more and more. If we all work together, I think we can map out the performance of the TP 360, TP 200, and TP 200X so that users will be able to understand all the common measurement patterns of these instruments. We should never take any thing for granted. Instrument performance is influenced by target color, shape, orientation, texture, reflectivity, and surrounding environmental conditions to include background lighting, whether the atmosphere is clear or not, etc. In many cases, moderate differences in these variables don't seem to affect instrument performance, otherwise they wouldn't be worth much. But there are always the sweet and sour spots and they vary with the brand, and more often, the particular instrument.

  As we all would agree, an important feature is repeatability of a measurement. LTI instruments excel here. Nikon products are often weak. I have no idea why the Nikon Prostaff 440 turned out to be such a good performer. Nikon engineers made compromises in subsequent models, and none of those I own has ever matched the performance of the 440, although I long ago established that it shoots long typically an average of a foot - occasionally more.

Bob
Robert T. Leverett
Co-founder and Executive Director
Native Native Tree Society
Co-founder and President
Friends of Mohawk Trail State Forest
User avatar
dbhguru
 
Posts: 3921
Joined: Mon Mar 08, 2010 10:34 pm
Location: Florence, Massachusetts
Has Liked: 3 times
Has Been Liked: 1035 times
Print view this post


Return to Measurement and Dendromorphometry

Who is online

Users browsing this forum: No registered users and 1 guest

cron