Notices
Marine Electronics Forum

sonar theory vs. MFD display question

Old 11-16-2016, 08:41 AM
  #1  
Admirals Club Admiral's Club Member
Thread Starter
 
Join Date: Apr 2016
Posts: 210
Received 30 Likes on 16 Posts
Default sonar theory vs. MFD display question

Quick question on sonar theory and how it is displayed on our MFDs:

For example, lets take a B175HW mated to a 7612xsv. The B175HW has a constant beamwidth of 25 degrees. At 100 ft deep, the cone width on the seabed is ~44.34 ft, and the coverage area on the seabed is ~1544 sq ft.

Out of that 1544 sq ft, what part of that circle exactly is the MFD displaying on the screen, which is reflected as a scrolling line?

Hope that makes sense.
Old 11-16-2016, 09:04 AM
  #2  
Senior Member
 
Join Date: Feb 2005
Location: Norway
Posts: 16,432
Likes: 0
Received 1,128 Likes on 918 Posts
Default

Everything within the cone, it's drawing all returns within the cone in one vertical line of pixels on the screen.
Old 11-16-2016, 09:25 AM
  #3  
Senior Member
 
Join Date: May 2014
Posts: 2,078
Received 23 Likes on 16 Posts
Default

This is a question I have been trying to answer for a while but I havent found a good - simple - explanation.

The MFD has to decide which of the many bottom 'returns' represents the actual depth under the boat.

Assuming a flat bottom, the outer edges of the cone will be further away from the transducer and so will look deeper. But how does it handle a sloped bottom or a hump or a dip or a large rock or several large rocks or a series of old pier columns or.....? How does it tell the difference between a fish suspended close to the bottom and a rock? How do things change when on a slope vrs flat bottom?


The only details I have been able to find are complex mathematical formula and algorithms that are way over my head. Its not as simple as averaging the returns or just taking the shortest time and calling that the "bottom" because neither of those will work on an irregular or sloping bottom.

I suspect each MFG handles these things at least somewhat differently.
Old 11-16-2016, 09:38 AM
  #4  
Admirals Club Admiral's Club Member
Thread Starter
 
Join Date: Apr 2016
Posts: 210
Received 30 Likes on 16 Posts
Default

Originally Posted by abbor View Post
Everything within the cone, it's drawing all returns within the cone in one vertical line of pixels on the screen.
Hmm, ok. So that's helpful for fish finding, but kind of not so helpful when searching for bottom structure, correct?

For instance, at 25 degree beamwidth and a depth of lets say 500 feet, the cone width on the seabed is 221 feet with a coverage area of 38,601 sq. ft. If you see an obstruction on the screen, you have to assume it's SOMEWHERE in that 111 ft radius and 38,601 sq. ft. It's easy to get excited and assume the obstruction is right under your boat, when in fact it really could be over a 100 ft away.
Old 11-16-2016, 09:52 AM
  #5  
Senior Member
 
Join Date: Feb 2005
Location: Norway
Posts: 16,432
Likes: 0
Received 1,128 Likes on 918 Posts
Default

StructureScan has 1-2 degrees signal angle in the longitudinal direction, so DownScan is what you should use for structure.
Old 11-16-2016, 10:00 AM
  #6  
Admirals Club Admiral's Club Member
Thread Starter
 
Join Date: Apr 2016
Posts: 210
Received 30 Likes on 16 Posts
Default

Originally Posted by abbor View Post
StructureScan has 1-2 degrees signal angle in the longitudinal direction, so DownScan is what you should use for structure.
Yep so it's more of a "slice" than a cone but even then again, unless the MFD image is 3D, that slice is still being represented as a 2D line and all the data in that "slice" is being compressed into the line.
Old 11-16-2016, 10:03 AM
  #7  
Senior Member
 
Join Date: Feb 2005
Location: Norway
Posts: 16,432
Likes: 0
Received 1,128 Likes on 918 Posts
Default

Yes, but then you use SideScan for determining the sideways distance. StructureScan 3D with three sidescan elements for both sides is excellent for this.
Old 11-16-2016, 10:31 AM
  #8  
Senior Member
 
Join Date: May 2014
Posts: 2,078
Received 23 Likes on 16 Posts
Default

Sidescan and Di are good for details, but the question still remains - how does 2D determine depth? What about in a situation like this in the pic?

What digital "depth" is shown on the screen and how is the bottom drawn? Does it vary depending on which direction you are going?
Attached Images  
Old 11-16-2016, 10:43 AM
  #9  
Senior Member
 
Join Date: May 2014
Posts: 2,078
Received 23 Likes on 16 Posts
Default

Here are three screen shots taken in rapid sequence showing the same relatively flat bottom in SI, DI at 455 khz and 2D at a fixed frequency of 135 khz. Its shallow water but still a relatively wide beam angle.

Its obvious to me that the 2D isnt "averaging" the bottom returns from the entire cone. But how is it determining the depth and differentiating between the logs sticking up and the bottom?

Also notice that where the "bottom" is drawn on the screen does not quite match up tot he digital readout of the depth.

I thought I had a similar series with a more irregular bottom, but cant find them just yet...
Attached Images    
Old 11-16-2016, 10:57 AM
  #10  
Senior Member
 
Join Date: Feb 2005
Location: Norway
Posts: 16,432
Likes: 0
Received 1,128 Likes on 918 Posts
Default

The 2D isn't averaging anything it's drawing everything within the cone. It's collapsing everything within the cone to a line.
Old 11-16-2016, 10:59 AM
  #11  
Senior Member
 
Join Date: May 2014
Posts: 2,078
Received 23 Likes on 16 Posts
Default

Here are some more in deeper water with a much more irregular and sloping bottom with some very large "structure" - pieces of the old Tacoma Narrows bridge collapse.

Note that in the 2D picture, the bottom depth readout is very different from where its drawn on the screen, but it matches up pretty close on the DI screen. There was a significant depth change within a short distance plus the large chunks of debris on the bottom that stuck up a good ways.

Im actually pretty impressed with how well the 2D handles the detail, but still unclear of how it calculates the bottom depth and where its taking that info from exactly.
Attached Images    
Old 11-16-2016, 11:01 AM
  #12  
Senior Member
 
Join Date: May 2014
Posts: 2,078
Received 23 Likes on 16 Posts
Default

Originally Posted by abbor View Post
The 2D isn't averaging anything it's drawing everything within the cone. It's collapsing everything within the cone to a line.
Im not sure what you mean by collapsing. Look at my image above. How is it collapsing that and how is collapsing different from averaging?
Old 11-16-2016, 11:09 AM
  #13  
Senior Member
 
Join Date: May 2014
Posts: 2,078
Received 23 Likes on 16 Posts
Default

Lets take a specific example - as in my pics above. The cone on the bottom covers a large area.

At the leading edge the depth is 66 ft.
At the trailing edge it is 49 ft.
On the port side its 54 ft
On the starboard its 57 ft.
Directly under the boat its 39 ft to the top of a very large piece of sunken bridge that is solid concrete and steel.

What does that collapse down to on the screen and what "depth" is shown digitaly and what "depth is drawn on the screen at that exact moment?
Old 11-16-2016, 11:36 AM
  #14  
Senior Member
 
Join Date: Feb 2005
Location: Norway
Posts: 16,432
Likes: 0
Received 1,128 Likes on 918 Posts
Default

Which depths are shown digitally depends on the the digital depth algorithm, but typically it's the smallest depth.

Sometimes a fish can be drawn below the drawn bottom, this may happen with a wide cone and steep bottom.
Old 11-16-2016, 11:49 AM
  #15  
Senior Member
 
Join Date: Apr 2013
Location: Norway
Posts: 651
Received 33 Likes on 24 Posts
Default

Your sonar takes the most shallow point within your coneangle that it can define as bottom, and draws that as the bottom line in the first row of pixles.

This means that if the bottom is anything but flat as a pancake, the bottom line on your sonar might conceal fish that sits below that line. You can test this yourself quite easily. Place your boat over a steep incline/ decline and drop a lure right under your ducer so you can follow its decent on your sonar. Take note of when the sonar is showing that the lure has hit bottom, and when your line stops going out. You will find that your line continues to go out after the sonar has shown your lure hitting bottom.

Here is another example. This screenshot shows a quite narrow conangle to the left (11 degrees) and a quite wide conenagle to the right (45 degrees). The ducer used is an Airmar P66. That spike you see on the left is actually a wreck.



With the wide coneancgle the wreck is concealed by the bottom line, you can only see a faint hint of it by looking at the bottom-composition. With the narrow conenagle it stands out.

Here is the wreck with scanning sonar. Downscan:

Sidescan:


And again using a bit of sidescan and some PC-software so you can see the depth-contours over the wreck:

Last edited by Team Colibri; 11-16-2016 at 11:52 AM. Reason: typos
Old 11-16-2016, 06:59 PM
  #16  
Admirals Club Admiral's Club Member
 
Join Date: Jan 2015
Location: St. Petersburg FL
Posts: 689
Likes: 0
Received 27 Likes on 26 Posts
Default

Wow. Great screen shots. Rag's
Old 11-16-2016, 07:42 PM
  #17  
Senior Member
 
Join Date: Apr 2014
Posts: 892
Received 71 Likes on 45 Posts
Default

Originally Posted by Yrral3215 View Post
Lets take a specific example - as in my pics above. The cone on the bottom covers a large area.

At the leading edge the depth is 66 ft.
At the trailing edge it is 49 ft.
On the port side its 54 ft
On the starboard its 57 ft.
Directly under the boat its 39 ft to the top of a very large piece of sunken bridge that is solid concrete and steel.

What does that collapse down to on the screen and what "depth" is shown digitaly and what "depth is drawn on the screen at that exact moment?
It puts it all there, in that single column of pixels. It will "color" a pixel at 39, 49, 54, and 57 feet. The strength of the return at each of those depths will effect the color/brightness of the pixel.

Realize it's not just getting those 4 discrete returns. It's also hearing echos from everything else in the "patch" of coverage on the bottom and coloring pixels at corresponding distances from the transducer.

Build that column of pixels quickly, advance it to the left, and build another.

Magically, it usually ends up looking like something we can interpret. Usually.
Old 11-16-2016, 09:00 PM
  #18  
Senior Member
 
Join Date: May 2014
Posts: 2,078
Received 23 Likes on 16 Posts
Default

Originally Posted by abbor View Post
Which depths are shown digitally depends on the the digital depth algorithm, but typically it's the smallest depth.

Sometimes a fish can be drawn below the drawn bottom, this may happen with a wide cone and steep bottom.
It has to be more complex than just using the shallowest return as the bottom or the depth. It seems to me they are also using complex algorithms to draw the bottom.


Originally Posted by Hossharris View Post
It puts it all there, in that single column of pixels. It will "color" a pixel at 39, 49, 54, and 57 feet. The strength of the return at each of those depths will effect the color/brightness of the pixel.

Realize it's not just getting those 4 discrete returns. It's also hearing echos from everything else in the "patch" of coverage on the bottom and coloring pixels at corresponding distances from the transducer.

Build that column of pixels quickly, advance it to the left, and build another.

Magically, it usually ends up looking like something we can interpret. Usually.
Your explanation is basically the same as abbor's and would end up with the bottom getting drawn in at the shallowest depth within the cone.

Clearly that doesnt happen all the time. or even most of the time.

If it was that simple, then some of those pics I showed above would have the bottom drawn at half the actual depth or less because of all the logs sticking up in this picture for example. Plus the depth readout would have been fluctuating wildly as we cruised over this area - which it did not.

In the second pic Im cruising right next to a bridge column, so the transducer is seeing returns as little as 1/4 of the actual water depth, yet the bottom is drawn more or less correctly and the depth readout doesnt fluctuate by a factor of 4 as I pass by.

It seems obvious to me that the mfg's have some serious math going on to handle these situations. Its also clear that they are not 100% fool proof I was hoping for some simpler explanations of exactly how it works in different situations. I doubt that any of the major players are willing to share proprietary techniques though.
Attached Images   
Old 11-16-2016, 10:30 PM
  #19  
Senior Member
 
Join Date: Apr 2013
Location: Norway
Posts: 651
Received 33 Likes on 24 Posts
Default

Turn on A-scope.

Remember that it is only the very first row of pixles that show what is in the cone now, the rest of your screen is history. Now take another look at your own screenshots, one row of pixles at a time.

You will find that your (2D) sonar is indeed drawing the depth at half of what it is, it does so every time you pass over one of those logs. At that time, when the log is in the cone, your sonar "belives" that the top of the log is bottom. Your depth-digits might not change, but that is due to other reasons.

Last edited by Team Colibri; 11-17-2016 at 12:05 AM.
Old 11-17-2016, 08:11 AM
  #20  
Senior Member
 
Join Date: Apr 2014
Posts: 892
Received 71 Likes on 45 Posts
Default

Originally Posted by Yrral3215 View Post
It has to be more complex than just using the shallowest return as the bottom or the depth. It seems to me they are also using complex algorithms to draw the bottom.




Your explanation is basically the same as abbor's and would end up with the bottom getting drawn in at the shallowest depth within the cone.

Clearly that doesnt happen all the time. or even most of the time.

If it was that simple, then some of those pics I showed above would have the bottom drawn at half the actual depth or less because of all the logs sticking up in this picture for example. Plus the depth readout would have been fluctuating wildly as we cruised over this area - which it did not.

In the second pic Im cruising right next to a bridge column, so the transducer is seeing returns as little as 1/4 of the actual water depth, yet the bottom is drawn more or less correctly and the depth readout doesnt fluctuate by a factor of 4 as I pass by.

It seems obvious to me that the mfg's have some serious math going on to handle these situations. Its also clear that they are not 100% fool proof I was hoping for some simpler explanations of exactly how it works in different situations. I doubt that any of the major players are willing to share proprietary techniques though.
It's still drawing all the pixels. Even the pixels that are below the "bottom" But there's another algorithm that determines the depth.

And I have been over schools of bait / structure / etc that have fooled the bottom algorithm. Makes for interesting QuickDraw maps.

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service - Do Not Sell My Personal Information -

Copyright © 2021 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.