[IGSTK-Developers] Re: Aurora SROM file

[David Gobbi]

Patrick Cheng wrote:
> Hi David,
>
> I think we need to revert to
> igstkAuroraTracker.cxx   V 1.24
> igstkAuroraTracker.h     V 1.16
>
> There are some fixes on style after that, which is minor.

Good, I've looked through the 1.24 code and I agree that
it does exactly what is needed according to the information
that NDI gave you.

> We should also remove the AttachSROMFilesToChannel function and 
> related operations. And we need to be careful when using "Single 
> Sensor Coil Tool".

Agreed.

> It would be nice to add a function to set the valid time period for 
> the transform and set the default valid time period to be around 
> 300-400ms. That should be done in Tracker class.

Agreed.

These are straightforward changes, so I can check them in
today.

 - David


> Thank you,
>
> Patrick
>
> David Gobbi wrote:
>> Hi Patrick,
>>
>> Thanks for running these tests.   The Aurora that I have access to 
>> does not
>> have any 5DOF tools or splitters, so it is impossible for me to run 
>> these
>> tests myself.
>>
>> We will definitely have to revert the code if the changes were made in
>> error.  Let me know exactly which CVS version number you want to
>> revert vtkAuroraTracker.cxx to, and I will verify that aren't 
>> inadvertently
>> throwing away any valid bug fixes.
>>
>> - David
>>
>>
>>
>> Patrick Cheng wrote:
>>> Hi David,
>>>
>>> I did some testing here, here is the result:
>>>
>>> 1. I actually went back the Aurora code before the your fix last 
>>> time. It seems to be working fine, as it suppose to be, according to 
>>> NDI.
>>>
>>> 2. I have attached two 5DOF needles to the same port using splitter, 
>>> and did the following tests:
>>>
>>>   A. Using a "Dual Sensor Coil, 6DOF Tool" SROM file, and attach it 
>>> to port, we will get 1 tool and the result transform is 6DOF
>>>
>>>   B. Using "Dual Sensor Coil, Dual 5DOF Tool", and attach it to 
>>> port, we will get 2 tools with 5DOF transforms
>>>
>>>   C. The problem is with using "Single Sensor Coil Tool",
>>>      C.1. when attached to port, we get 2 tools,
>>>      C.2. when attached to channel 0, we get 1 tool
>>>      C.3. when attached to channel 1, we get 2 tools
>>>      C.4. when attached to both channels, we only get channel 0.
>>>
>>> The only confusion here is scenario C. This behavior is actually 
>>> consistent with what I heard from NDI.
>>>
>>> "You can not use two different SROM files in the same port, if you 
>>> want to use splitter to tracker multiple tools, you have to use 
>>> 'Dual Sensor Coil, Dual 5DOF Tool' SROM file", as in Scenario B, and 
>>> we are not suppose to do Scenario C.4.
>>>
>>> The way Tracker deal with SROM file is:
>>>
>>> 1. Look for SROM files in tools plugged into the tracker, and load 
>>> them in control unit.
>>>
>>> 2. If there is user defined SROM file, it will be send to the 
>>> Virtual SROM unit in the tracker, and this SROM file will overwrite 
>>> previously loaded one from the tool.
>>>
>>> 3. In case of tracking two 5DOF sensor, the tracker will 
>>> automatically add another handle to the port, if this port received 
>>> a "Dual Sensor Coil, Dual 5DOF Tool" SROM file
>>>
>>> 4. Most of the case, the tracker tools all have SROM file burned 
>>> into it, so we can tracker them without attaching any SROM files.
>>>
>>> My suggestion is:
>>> 1. Go back to the previous version of the AuroraTracker code
>>> 2. Eliminate the AttachSROMFileToChannel API, if we want to track 
>>> multiple tools, either do not use any SROM file (will load from 
>>> tools), or use a "Dual Sensor Coil, Dual 5DOF Tool" SROM file.
>>> 3. Modify the way we LoadVirtualSROM files in AuroraTracker class to 
>>> be consistent with NDI.
>>>
>>>
>>> Thank you,
>>>
>>> Patrick
>>>
>>>
>>>
>>
>>
>>
>>
>
>
>