Topic: DI View Inaccuracies

[okiedokie]

    Users  of the DI view  may be seeing something different than what they are actually collecting with DI. I recently posted a lengthy example of this under “Improved DI Settings and Observations”. It may have been too long to stay interested in what I think is going on, but I believe it warrants consideration and comment by Humminbird. The following diagrams may more easily illustrate the issue I am trying to point out.

    Assume  a boat is making a SI and DI scan in a very long flat bottomed pool that is 60ft wide and 30 ft deep with some concrete raised symbols and letters on the bottom.

    Figure 1 shows the raw SI and DI images that would be collected.

    Figure 2 shows the raw DI image rotated 90 degrees clockwise to conform with the 2D view.

    Figure 3 shows the raw DI image folded in half to prepare it for display in the 2D format. Notice how the symbols stack on  each other and the J is inverted.

    Figure 4 shows how running 2D and DI in bottom autotrack mode leaves out a lot of DI data that is still actually there as I showed in the posting of “Improved DI Settings and Observations”.

    Figure 5 shows the final 2D/DI/SI display that you would probably see.

    Figure 6 shows a real example of this issue after the autotrack was turned off and the 2D/DI depth was set to 3 times the actual depth.

    Yes this is all speculation but I think it is correct. Buyers of a Humminbird unit with DI capabilities are spending their money for the information it can provide them and trust that it is accurate. I now am not so sure about that with the DI view. I want to believe that Hummbird values its customers and users and would want to resolve this issue in a productive way. If I am wrong, so be it, end of discussion. Currently the only part of the DI image that I believe can be trusted as accurate is the very middle which is shown as the start of the bottom on the display.

[newkid4si]

okiedokie

        I find both of your post fascinating. I have difficulty understanding DI displays. I can't match up the image on the unit with what I know is physically down there.
        Please continue.

             Mike

[okiedokie]

Thanks for the encouragement of my 2 DI posts. I am trying to show why I believe the current DI view is not very user friendly because of the way I suspect the full image has been folded in half over itself at the center to make a 2D compatable view. If a user wants accurate bottom feature DI image views  its my belief that the folded over right side data is inverted and combined with the left side data making a combined mess that I tried to clarify with the symbols on the 2nd post diagrams. Particularly how the O's and X's on opposite sides of the centerline are now seen as 1 combined symbol by the user. To get any progress from Humminbird on this issue I would expect several other users would have to post enough replies to get their attention as it might open up a can of worms they don't want to deal with.
Ken

[newkid4si]

Ken

    The 2D and SI are pretty straight forward. In the DI, I have the benefit of divers physically examining the target to confirm what is really there.
    The images that I was seeing didn't match the target.  I discontinued using that mode. The information was there, but I couldn't understand it.
    With the help of your explanation and drawings, I'm going to review some old recordings again.

                    Mike

[rnvinc]

Here is the "definition" I copied from Greg on another post..(which I think does indeed match what you observe)...

QUOTE...in part

The Di sonar in the HB Si units is a blend of both the left and right Si sonar:
*- take the right Si sonar and rotate it 90 degrees clockwise.
*- take the left Si sonar and flip it horizontally (mirror image) and then rotate it 90 degrees clockwise.
*- overlay the two images and you have a Di sonar image.

UNQUOTE...

Rickie

[okiedokie]

Rickie,

[okiedokie]

Rickie,
    Thanks for the historical confirmation from Greg. It may indicate that this is just the way it is and will be with Humminbird DI view despite the problems of view confusion it causes as I've illustrated.
    If asked what I would like  Humminbird to consider doing about this DI image confusion, I guess I would say add a DI Contour feature to the DI Enhance menu. It wouldn’t have or need the side to side range of the full SI , but WOW! crappie brush piles look out. This would be a full left to right, unfolded, non-inverted DI image that would be accurate and make sense to the user as to what the bottom was really like.
    In regards to the view charts I showed, if there is a question where the F and J and the rest of the T and R went in the final DI view of figure 5, they are still available in the DI image data but not shown because  bottom autotrack depth is active. If the bottom depth was manually set to 60ft for the 2D/DI views, they would be visible as shown in fig 7 below.

[rnvinc]

I will let Greg comment on his observations...but...

I think that this "DI from SI" computer generated DI image is probably just what it is...

If we look at a DI image...we know that the "live" data is at the right vertical edge of the DI image...(everthing left of that is history)...

If we look at a SI image...we know that the "live" data is at the top horizontal edge of the SI image...(everything below that is history)....

If we disect that SI image into separate halves we see...

*Its easy to rotate the right SI half 90° clockwise and the "live" data edge (of the right SI half) fall right into the right edge of the DI image...

*Now if we were to rotate the left SI half 90° clockwise (without flipping it horizontal first)....to keep the "live" left SI data lined up the "live" DI data...now the left SI bottom detail would be at the top of the DI image "upside down"....

The inherent problem in having this "DI from SI" technology displaying a DI image is the lack of a "down pointing" piezo...but instead rendering a DI image from side pointing piezoes...

While the data in the DI image water column (above the bottom detail in the DI view)...may be easily discernible when the 2 SI halves are blended together...the details against the bottom detail get very confusing...as you have seen...

Maybe this was HB's idea with this DI from SI anyway...maybe HB feels DI users are more interested in what's in the water column more than what's against the bottom detail...

Rickie

[okiedokie]

Rickie
Thanks again for the comments. Upon further thought, all of my comments may be factually correct about the DI view confusion but my mistake may be in how I thought the DI feature could be used. Probably the terms Side scan and Down scan led to this error. The term Down scan lead me to expect I could get a good Down scan view of the bottom features. A better term might have been vertical image scan. In this usage the DI shows a vertical image that is more informative than the 2D summed cone image, despite the fact the DI image might combine coincident images  from the right and left that would be hard to interpret. Successful horizontal position GPS marking of bottom targets is not possible with DI. The SI scan might be better thought of as a horizontal image scan and already provides accurate bottom views and GPS marking of horizontal bottom targets.
I still wonder if a DI Enhance Contour view would be more informative to use than the current SI Enhance Contour. If the DI is just a software manipulation of the SI scan which is probably the only scan taken, then maybe there is no real improvement in having a DI Enhance Contour feature except for possibly narrower side to side range settings to increase the available screen resolution.
Never to old or big to be wrong.

[okiedokie]

Thanks Rickie got it

[RGecy]

okiedokie,

When HB first came out with DI we all plotted over the same issues you are bringing up here.  Yes, Humminbird DI that is using a composite of the two SI beams is not truly Down Imaging as if it were a single dedicated beam.  Basically, they take the two beams and combine the sonar returns to create a composite image of left and right. 

I am not sure what you are trying to establish with your post or if you are just trying to make a point.  I personally think HB came out with the Down Imaging just as a way to compete with Lowrance Down Scan and to appease the traditional fishermen who were used to reading 2d sonar in the vertical mode. 

In my opinion, Side imaging offers better detail and if you get used to reading and interpreting it, it offers much more information than Down Imaging.  I think Down Imaging is good for fishermen when using it side by side with 2d to see whats in the water column or the first bit of area on the bottom itself.

There are plenty of older post that discuss all of this in detail here on the forum.

Here is some great reading!   http://forums.sideimagingsoft.com/index.php?topic=678.msg6626#msg6626

Robert

[okiedokie]

Robert,
Thanks for the reply and topic reference. In my last reply I conceded that it was mainly an inaccurate expectation on my part as to what I would see with DI. This might have been corrected by researching the forum topics more thoroughly. DI performs exactly as Humminbird configured it. Without a doubt the SI view is the gold standard of bottom and water column information. The diagrams helped me understand DI better.
Regards