Kedok goes CAM

10 years ago, I already did some testing with a camera, but everything was far too slow, and a good camera cost around $2,000 at the time. Because of the high cost, I had to abandon the idea.

Now, 10 years later, I started looking again and found an affordable camera module.

The first tests are very promising. I haven’t really looked into the accuracy in detail yet. My first impression is that it is accurate enough for a pistol target.

Advantages that are becoming apparent:

• Very easy to find the target
• No modifications to the shooting range are required. No LED or halogen lamps are needed, so you 
  can simply use it on any shooting lane
• The target card no longer needs to be mounted with great precision.
• Very simple settings, which are currently made on the computer.
• Small in size.

 Disadvantages:

• An additional computer is required, either a laptop or a specially built computer. 
• There is a significant chance of aiming at the neighbouring shooter’s target. 
• At the moment, mains power is still required, or a laptop with enough battery life for an entire
  shooting session. 
• The accuracy has not yet been proven.

 Stay tuned!

  Benjamin doing the first sound checks

Never in my wildest dreams did I think that one day we could achieve such scores with acoustic sights.

Never in my wildest dreams did I think that one day we could achieve such scores with acoustic sights.

When Jan and I created Kedok, our goal was to develop an acoustic aiming device for pistol target cards. Initially, we were happy when the shooters occasionally hit a 10, thinking that this was the highest possible score with an acoustic aiming device. However, over time, our shooters improved significantly. I suggested using rifle target cards, which are much smaller.
Two weeks ago, Milan, one of our shooters, achieved a score of 608.6 on a rifle target with 60 shots, averaging 10.14 (WOW!). Benjamin also scored very well with 589.0, averaging 9.81.
These results exceed any expectations we had at the beginning. I never imagined that shooting and aiming with sound could be possible.
Unfortunately, Jan can no longer experience all of this. Well, maybe he is helping us from above :)

LED to better results?

While Kedok use a halogen spotlight for aiming at the target many other aiming systems use an
LED at the target.

When Kedok project started I used an LED too, but when I discussed it, many disapproved.
Arguments were:
We are not used to it and we would have to change the shooting lane we already have.
They wanted a system just like Swarovski, with the only change that it can be used with ISSF
approved target cards.
And of course available for a low budget.

Anyway, this was four years ago and in the meantime you see systems like Eco-AIMS and
VIASS using a LED at the target for aiming.
Than there came questions from Kedok users, why don’t we have this?
There is a reasons not to use it: Chanting of the rifle gives a error in aiming.
I discussed it with Daniel Walø and Michael Whapples. With Daniel I did some theoretical tests.
Daniel calculated it, I made a 3D model on my computer and we came to the same results.
There is a significant error if you chant the rifle.
Would it affect the results much if a trained shooter used an LED for aiming? I doubt it, looking
at the scores they shoot.

Halogen spotlights have some disadvantages too:
The quality of the paper target card must be good, shooters hear a noticeable difference in pitch
if the card colour slightly changes.
At our club the SH3 shooters always get a card set from the same factory batch.
Also the card can bend and a shooter will hear this too.
One of the shooters loves the pitch change if the pellet hits the target, Wheee, whee whee :)
This convinced me to make Kedok also suitable for an LED as target.
The question now is, how do we mount the LED if paper card targets are being used?
First attempt was to make target card frame with an LED.
Problems that occurred were that every target card runner had a different mounting system and
therefore it is hard to find a right way to mount the card frame.
It all became a bit heavy and I skipped this idea.
Now we are testing with an LED mounted on the tripod we already have for the spotlight.

The LED setup has only been tested by two shooters so far but soon we will have more data
available.

Nobody is exempt from the surprises of coding!

As a reader of Michael Whapple's blog, where he  mentions the benefits to have a tone jump in the sound when you enter the inner circles. As Kedok has this option in the software, specially made for John. When he moves towards the inner circles above the eight it makes an extra tone jump of 300 Hz; I took up the idea to make it adjustable, so you can move the tone jump from the rim to the centre of the target card. I discussed it with a few Kedok users and I was surprised that some already used the fixed tone jump option in Kedok. Next update of the software will have an option to move the tone jump from 1 to the 9 circle. A few users will test it and I hope they can improve the results with this new feature. Have to dig in other things too:  for example, is there such a thing as  having tone blindness.

Stay tuned.

All credits goes to the users.

Working along with many users of Kedok is great.
Two active Dutch users give me all the input I need to improve many things. While developing, I always make a todo list of improvements the users want, which you also can find in the software source. After a phone call this evening with Nanne we scored three things on the todo list that would improve Kedok. For example: an option to have an advanced shooters settings option in the menu. The settings in advanced mode for the up/down buttons will be in smaller steps and Kedok will stop nagging the shooter by voice with detailed help and only give short beeps. Also, more setting will be adjusted once you who know how Kedok works. I even get much help from the other side of the world from Geoff Hunter. You should check their Facebook page Exsight Tandems Illawarra. The scores they make are seriously good and I enjoy the results.
Our holiday break at the club is over; looking forward to what the future will bring.

Thank you guys.

Statistics are like a bikini.

During the holiday break at the club I started to put some statistics tests on the Kedok hardware. I coded some statistics math on the Kedok to test the analog to digital converter. By putting a reference signal to the input and storing 1000 samples over 10 sec., I could read the standard deviation, standard error, mean and coefficient of variation. All Kedok units I build in the past have a standard deviation of 1.6 in the readings. So there was some noise on the signal. After some research it seems the source was the battery power converter, after putting some filtering on the power lines, it dropped to zero.

What it reveals is suggestive, but does it conceal something vital like a bikini does? We soon know after some tests by Benjamin.

The six, shaped the future for Kedok

The new Kedok six was received very well by Benjamin.
He tested the unit and without any help he could adjust all settings.
If you build things, you don't always see the problems that the user might face.
So I past that test. The sound was ok too.

At the range benjamin asked me, why the display?
You VI/B shooters never think about those who are keen sighted ;)

Next on the list is to improve averaging on the sensor readings for a more smooth aiming.
At this moment we have a simple averaging (SMA) filter and we are going to test with an exponential moving average (EMA) filter to see if this makes the aiming more smooth for the shooter.

For more info about data filtering:
https://en.wikipedia.org/wiki/Moving_average

Best of both worlds

We have the Kedok with an LCD display and one with voice only. I really loved the Kedok voice but it had one problem.
Using it at the range you have no idea what the shooter settings are and you have to ask to know them, which is very disturbing to the others.
Months ago Jan said to me, why not make it with both? I saw too many objections, like how to keep the costs low and is the Arduino CPU powerful enough to handle both. This weekend I started to build a Kedok with voice and a display. It does work very well and the price is almost the same. Think in a couple of months it is finished and tested. Love to see the new born?

Well, here it is:

Small step in frequency, one giant leap for score.

This Friday we did a test with John, a B/VI shooter at our range.
John scores went down to very bad while the year was going on.
He lost all his faith in shooting and wanted to quit the sport.
Ben a shooting trainer looked at what he was doing while shooting and said to me, John is doing well physically, very steady and I have no idea why he shoots so random. He is a good shooter; maybe it is his rifle?
We tested the rifle and and hit bulls eye after bulls eye.
So there was something wrong with John. 
I took all data from Kedok and there was something odd going on.
Seems John did not aim at the highest score.
Shot after shot he did aim at a 5, 8, 2, 10, 4, 9 and 3.
And he was telling me he always did shoot at the highest pitch, while he was not doing it according the data.
He had a doctor's test for his hearing and all seems to be okay.
I have no idea how the human brain works, but there was something going wrong with John's.
Why does he thinks he aims well and not do it right??  A puzzle for me.
I added some extra code to Kedok software for John and it makes a big frequency step if he is above the 8.
Like there is a helper that say to him you are about right now.
He tested it and his scores went back to good.  He came back from the range with a huge smile.
His wife did reply to me with:  finally, you cracked the nut. I will have a nice weekend too :)
Benjamin tested the software first and had no better score with it, so there is no scoring benefit to a shooter using this extra feature.

Great day at the ranges

Yesterday we had a shooting marathon at the range.
Sighted shooters are doing 5 rounds of 50 shots.
Audio shooters 5 rounds of 25 shots.
It's a mix of all shooters, all classes are in the same match.
25 club members came to this match and showed their best skills during this match.
So our B/VI shooters had to compete with an Olympic shooter, young and older shooters using a rifle rest stand.
Scores are by round numbers maximum 100 for a serie of 10 shots.
First round at 10:00, second 11:00, third 12:00 After this a 30 min rest for quick lunch.
Next rounds were at 12:30, 13:30 and 15:00.
It was a great match to see B/VI, elder and young people compete with each other.
The three B/VI shooters were on 4th, 14th and 22th place.
Next year the B/VI shooters also have to shoot 50 shots in a serie.
Benjamin a B/VI shooter got a special reward for being a big asset to the club.
Think our B/VI shooters we are fully accepted as first class shooters, thanks guys.