Hi all,
I am using RM for the first time to create a side scan mosaic of a seagrass bed for export into GIS for calculation of bed area. The data was collected using a Humminbird solix G3 scanning at 1200kHz.
See below:


I have found it difficult to get the tracks perfectly straight and the overlap, even after adjusting swath range, has made things a bit confusing. The transitions between each track are also clunky. The middle of each track is quite blurred where the depth offset has been applied. Do you have any tips of how I can improve this? Is there a way I can make my unit less sensitive to the boat changing direction slightly?

One thing I have tried is to crop each each track into smaller sections. I was hoping this would let me remove blurred bits and adjust depth offset on a finer resolution. When I crop the track, I find two issue arise, 1) the new cropped section has no side scan data attached to it, 2) the original track still shows the side scan data which I just dropped out. So its kind of like the track dots are removed, but the side scan image remains. Surely I am just doing this incorrectly? Any help would be appreciated.

Also, any tips of how I can improve the resolution of the sea grass bed in the mosaic. I was hoping to be able to assess patchiness and overall area, but this seems hard from this image and noticed variability in the bed shape depending on which tracks I used. The southern edge of the bed is nice, but the rest is vague, how can I improve this?
For those not familiar with the Solix G3, I can collect data in 455, 800 and 1200 kHz. The higher the frequency the more detailed the image is, but at a sacrifice of range.
I was thinking of trying to collect data at 800 kHz instead of 1200, but bump out the range of the swath to much further and conduct less tracks. Do you think this would help with my image quality?

Cheers

You can split the sidescan, but you need to use the "swath control bar" at the bottom of the sidescan window to do this - cropping track-points will have no effect, as you have found out. To crop or split the sidescan swath, first ensure that the segment of interest is selected and then adjust the visible start/end time using the slider at the bottom of the map view. You can then split the segment at the specified end points, which will give you up to three segments instead of one. Each of those segments can now be edited individually for all parameters, including water level offset, brightness, Z-order etc, and any segments you don't want can be deleted.

Go easy on water level adjustment - this is really just to fine-tune areas where bits of water column remain visible in the slant range adjusted image - better to split a segment so that you are operating the offset on as small a section as possible. Aggressively applying water level offset to try and reduce the low resolution centre of the swath will introduce inaccuracies in the slant range correction.

The area directly below the boat is always of a lower resolution - this is simply a result of how slant-range correction works on a sidescan sonar return, and there are fewer data returns available per unit of distance to the side.

If you are running with hardware graphics acceleration enabled, try the "blend closest" option in the mosaic to blend the edges between overlapping swaths. This can often remove visible edges entirely, although you may need to tweak relative brightness/constrast settings of the overlapping swaths before it appears truly seamless. Tip: if you want to adjust image settings for one side of a swath, you can duplicate the segment and hide one side of each identical segment so that you end up with the original swath, but composed of separate segments for port/starboard. You can then adjust image settings per side.

How visible an object is in a sidescan image often depends on the angle the object is being hit by the sonar - objects that are further away, and hit by a shallower angled beam, may throw more prominent shadows that objects that are more directly underneath the boat. Objects may also appear clearer when approached from a certain direction, so try hitting areas of particular interest with multiple passes from different distances and directions until you find what works best.

You also might need to temper your expectations a little bit - what you have already doesn't look too bad to me, and you are not going to achieve a perfect photo-realistic of the bottom, particularly in areas directly underneath the boat.

Hi matt thanks for getting back to me with the tips. I have sorted out the cropping issue now.
When stacking the transects onto the map, can I order the layers to control which sections are visible? At the moment, it seems pretty random which transect is visible when two are over lapping.

Also with the blending options. What do they do? How do I know which method to choose. Can I choose different blends for different parts of the mosaic?

Yes you can adjust Z order of segments by dragging within the list on the right. I suggest having a good read of the sidescan mosaic section of the manual.