Spiderlight Chart Error?

[simonh]

Sanjay & Tim,

First, thanks for putting together another great article. However, I am sure there are some errors in the Spiderlight charts.

In figure 13 at 9" we can see that the Spiderlight peaks at the dark blue colour (around 1500-1600 PAR) same as on the LumenArc peak. However, on Figure 14 beneath it seems to be cut off at 700-800 PAR.

Likewise, the chart in Figure 14 for the 12" distance seems to have a problem with the Y axis scaling as it runs 0-550 PAR yet the yellow peak is 1000-1100 PAR. This error also seems to have carried forward into Figure 25 where at 12" the Spiderlight shows a small area of coverage (what appears to be actually the 1000-1100 area) above 500 PAR whereas looking at Figure 13 it is quite a large area.

I do not know if this problem has crept into any of the coverage area calculations but can see that the maximum PAR does not match with that in Figure 13 so suspect so from a visual eyeballing of the top down charts?

-Simon

 

[sanjay]

In figure 13 at 9" we can see that the Spiderlight peaks at the dark blue colour (around 1500-1600 PAR) same as on the LumenArc peak. However, on Figure 14 beneath it seems to be cut off at 700-800 PAR.

Likewise, the chart in Figure 14 for the 12" distance seems to have a problem with the Y axis scaling as it runs 0-550 PAR yet the yellow peak is 1000-1100 PAR. This error also seems to have carried forward into Figure 25 where at 12" the Spiderlight shows a small area of coverage (what appears to be actually the 1000-1100 area) above 500 PAR whereas looking at Figure 13 it is quite a large area.

I do not know if this problem has crept into any of the coverage area calculations but can see that the maximum PAR does not match with that in Figure 13 so suspect so from a visual eyeballing of the top down charts?

-Simon

Simon,

Thanks for catching the error in the plot. After working with so many plots and many versions in trying to present the plots in a meaningful manner, a mistake slipped through.

After going over the raw data, and the plots I think I have figured out the problem with figure 13 and 14, and the source of your confusion. The plots in figure 13 are all at intervals of 50, but the legend is showing intervals of 100. At one time we were playing around with plots to see what resolution would be more readable, and tried 2 different scales. We felt that intervals of 50 would be too cluttered, and decieded to go with the interval of 100. Hence you are seeing values that are 2X in that plot. If you take the values and 1/2 them, then you will have the correct plot.

In figure 14, the 6" and 9" data is correct, but the 12" data again needs to be halved.

As soon as Tim gets back from spring break, I will get this figures fixed.

I know I was dissapointed and surpirsed at the performance of the spider light, and in fact re-did the readings at 9" because I thought there was something not right with the data when I first saw it. So I know the values are correct.. its just that the plots that are messed up due to the scale.

The peak values measured were as follows:

6" - 2003 (points around this fell to around 1500)
9" - 795
12" - 545

Once again, thanks for pointing this out.

sanjay.

 

[simonh]

Thanks Sanjay. I knew something didn't quite look right.

A couple of other catches. Figure 25 has the PFO Parallel label used twice, whereas I presume the second one is the PFO Perpendicular. Also, Figure 15 is titled incorrectly I beleive.

Do you have any idea why the data for the Spiderlight didn't show the same distribution pattern as in Richard Harker's data a few years back?

 

[sanjay]

Simon:

I should use you as my proof reader

here are the corrected plots and figures.

Figure 13 - corrected

Figure 14 - corrected

Figure 15- corrected

Figure 25 - corrected

Thanks,

sanjay.

 

[simonh]

Thanks Sanjay.

 

[MrSandman]

U mentioned in your article that you needed someone w/ software who can calculate the areas in figure 25. I have some software at work that can probably do this for you. We use it to measure the areas of the ink imprints that projectiles leave behind. What we do is scan the image, determine the scale, and then click on the image and it spits out an area. (generally speaking). Let me know if you still need someone to do this for you. I may be able to help you out.

Ramon

 

[sanjay]

Let me know if you still need someone to do this for you. I may be able to help you out.

Ramon

I have sent you email, but I did not hear back from you. Just wondering if my emails are getting to you.

sanjay.