About accuracy: I simply don't know if it will work out, but here are 
the numbers I'm building upon:

- It's actually not really an IMU, but a set of accelerometers and 
gyroscopes.
- The accelerometers have 16-bit converters built in, the gyroscopes 
have 13 bits. This should give me a resolution (not accuracy or 
repeatability) of far less than 0.5°. That's the innermost circle's 
diameter in a telrad.
- Aligning: The sensor data is converted to altitude and azimuth. The 
readout can be offset with known values of a known star at a certain 
time and should be more or less accurate in the region surrounding that 
star.

Example: If I want to find M57 in Lyra, it's probably a good idea to 
align with Vega first, because that's easier to find.

If the IMU idea fails, I can still use (relatively coarse) axis 
encoders, it's not aligning for fotos or GoTo, after all.

The point is not having a lot of information in the display, but 
something like a "spoiler" that's doesn't interfere with the art of 
actually finding an object: It could just be "Yes, you found it" 
indicated by a change in the circles shown, or a marker when I'm closer 
than 1°. The easier variant is showing a directional marker that tells 
me where the desired object is or how far out. There are lots of 
possibilites.

And to prevent misunderstandings: I dont want a separate display in 
addition to the telrad. The OLED display will be built into the real, 
see-through telrad instead. It will display the circles and everything 
else I program into it.

Regards

Christoph

Am 28.08.2013 12:54, schrieb Georg Zotti:
> Sounds like a nice project! But are the IMU sensors accurate enough to
> allow see-through support where what you see on screen centermarks is what
> you see in the main telescope? Such support until now requires aligned
> mounts and axis encoders. Unquestionable, a small dim map screen next to a
> (real, see-through) Telrad that shows the surroundings of where a big
> Dobson is pointing sounds like an interesting project.
>
> There is definite Qt5 support for Raspberry Pi, and OpenGL1 hardware (Atom
> netbooks and such) is supported via ANGLE. So these aspects are all
> catered for. A second application based on Raspberry Pi could be a
> dedicated telescope remote control unit (for GOTO scopes) running
> Stellarium, which will hopefully be even more power-friendly than an Atom
> netbook (but could maybe also be used as travel/email computer).
>
> Both devices don't need 3D foreground rendering, and the "telrad support
> device" can be limited to perspective projection (OpenGL built-in).
>
> Good luck with that!
> G.
>
>
> On Mi, 28.08.2013, 09:53, Christoph wrote:
>> Re VR:
>>
>> I'm currently building a telrad-like device with an OLED display and an
>> IMU that can be attached to a telescope. I'm not sure what I want to
>> display in it, but I'm planning to get the information from stellarium
>> over bluetooth. For that, it would be nice to be able to strip it down
>> in such a way that it can run (just a bare-metal core, no rendering) on
>> a raspberry pi or even a cortex M3. Otherwise I could let it run on my
>> atom laptop - so I'd appreciate a "runs-on-cheap-hardware-path", but not
>> from a 3D point of view.
>>
>> Also, please keep writing plugins as simple as it is now. I'm not
>> familiar with the new features of Qt5, but the current implementation of
>> e.g. StelButton feels a bit clumsy. Qt4 does support widget layouts in a
>> QGraphicsView if the widgets inherit from QGraphicsLayoutItem, but
>> StelButton inherits from QPixmapItem which is a QGraphicsItem. I think
>> much of the code of e.g. the Oculars plugin could be much simpler if the
>> given Qt features were used.
>>
>> Regards
>>
>> Christoph
>>
>> Am 28.08.2013 01:32, schrieb Georg Zotti:
>>> Dear Jörg and Fabien,
>>>
>>> if the old classes return, I hope scenery3D will finally be integrable.
>>> I
>>> did not like the Qt announcement to drop OpenGL1.4, but with ANGLE also
>>> cheap hardware (Atom etc) should be kept alive, so there's a chance.
>>> Unfortunately my students finished their respective tasks just before
>>> that
>>> big merge towards 0.12.0, and left for other business, and my project
>>> time
>>> is up as well. So, when porting to Qt5 is done, I will likely require
>>> somebody to go through and reanimate the scenery3D code. It included
>>> quite
>>> a lot, including cascaded shadow mapping, and allows rendering of and
>>> walking in OBJ models. Of course, Qt5 will require to adapt some OpenGL
>>> hacks which were necessary in Qt4.6.
>>>
>>> How does one move/walk with these glasses, or what would be your aim to
>>> display in eye-separated 3D? In the mid-late 1990s there were
>>> experiments
>>> with VR glasses (with thick cables dangling from the lab ceiling...) fed
>>> by SGIs (remember those?), but the full-cutoff was seen problematic,
>>> leading to AR ideas (with semitransparent glasses). I have since then
>>> imagined semitransparent goggles with explanatory displays augmenting
>>> the
>>> live stars. I think Celestron built such a device some time ago, and now
>>> there are cellphone apps using these sensors, augmenting camera views.
>>> Time flies...
>>>
>>> So, if you wish, (I think we had contact before), feel free to reanimate
>>> this plugin and enhance it with oculus rift, but it should please also
>>> work without.
>>>
>>> G.
>>>
>>> On Di, 27.08.2013, 22:59, Fabien Chéreau wrote:
>>>> Hi Jörg,
>>>> to reply to your email about oculus rift:
>>>> It sounds quite a nice project :)
>>>>    - For 3, you need first to use perspective projection, and I guess
>>>> simply adjust screen size so that the ratio is correct, and FOV so
>>>> that it matches the FOV of the glasses.
>>>>    - For 4, This is not so clear what you want to do here: most of the
>>>> stuff displayed in Stellarium are infinitely far away, so it's
>>>> difficult to add parallax effect. Maybe on the landscape, but we don't
>>>> have 3D models, our landscapes are flat. Maybe you can try to use the
>>>> 3D landscape code done by Georg's team?
>>>>    - For 5, I would suggest you rather implement a special proper
>>>> projection doing just the inverted distortion than the one from the
>>>> glasses.
>>>>
>>>> Fabien
>>>>
>>>> On Mon, Aug 26, 2013 at 5:53 PM, Jörg Müller <nex...@gmail.com> wrote:
>>>>> Hi,
>>>>>
>>>>> and about the Oculus Rift part, which actually was the main part of my
>>>>> mail?
>>>>>
>>>>> Regards,
>>>>> Jörg
>>>>>
>>>>> Am 2013-08-26 17:42, schrieb Fabien Chéreau:
>>>>>> Hi Jörg,
>>>>>> well I had a lot of conflicts with this code indeed. We should try to
>>>>>> re-integrate it properly on the branch, if we decide to go this way.
>>>>>> Fabien
>>>>>>
>>>>>> On Mon, Aug 26, 2013 at 5:33 PM, Jörg Müller <nex...@gmail.com>
>>>>>> wrote:
>>>>>>> Hi Fabien,
>>>>>>>
>>>>>>> interesting news to read. Does this mean my shadow code gets removed
>>>>>>> too? I don't know if GLES2 supports the required float textures.
>>>>>>>
>>>>>>> Apart from that I recently tried to integrate Oculus Rift [1]
>>>>>>> support
>>>>>>> in
>>>>>>> stellarium. The steps needed to be done for integration are:
>>>>>>>
>>>>>>> 1. Headtracking (Moving the view based on the movement of the
>>>>>>> headset)
>>>>>>> 2. Rendering in two viewports each taking half of the screen space
>>>>>>> for
>>>>>>> the left and right eye
>>>>>>> 3. Calculating and adjusting the projection matrix based on aspect
>>>>>>> ratio,field of view and other settings of the hardware and using it
>>>>>>> in
>>>>>>> the rendering pipeline
>>>>>>> 4. Adjusting the view matrix for each eye location
>>>>>>> 5. Correcting the distortion of the hardware's lense using a pixel
>>>>>>> shader
>>>>>>>
>>>>>>> I succeeded in the first two steps, but as for the next two
>>>>>>> stellarium
>>>>>>> is a very special case I failed to properly change the projection
>>>>>>> matrix
>>>>>>> of the rendering pipeline. What do you think about this project and
>>>>>>> would you be willing to help me understand how the transformations
>>>>>>> in
>>>>>>> the rendering pipeline of stellarium work so that I can integrate
>>>>>>> the
>>>>>>> required calculations?
>>>>>>>
>>>>>>> Regards,
>>>>>>> Jörg
>>>>>>>
>>>>>>> [1] https://oculusvr.com/
>>>>>>>
>>>>>>> On 2013-08-26 15:51, Fabien Chéreau wrote:
>>>>>>>> Hi Bogdan,
>>>>>>>>
>>>>>>>> I currently have a bit more free time, and I was actually
>>>>>>>> investigating doing some major refactoring of the project (not a
>>>>>>>> total
>>>>>>>> rewrite but not very far from that :D )
>>>>>>>> Basically what I want is to produce a much lighter version of
>>>>>>>> Stellarium based on Qt5 and OpenGL ES2 features only (it's a
>>>>>>>> requirement of Qt5 itself). This involves:
>>>>>>>>      - revert all changes from GSOC 2012 on the renderer: after
>>>>>>>> having
>>>>>>>> a
>>>>>>>> detailed look I am convinced that all this work is almost useless
>>>>>>>> from
>>>>>>>> the moment on we want to support only GLES2. It also introduced a
>>>>>>>> number of rendering bugs, as well as a large performance penalty.
>>>>>>>> I've
>>>>>>>> done most of this revert in a branch called
>>>>>>>> https://code.launchpad.net/~stellarium/stellarium/simplegles and
>>>>>>>> went
>>>>>>>> from 45 FPS to 62 FPS on my computer.
>>>>>>>>      - suppress completely GL1-related code and focus on
>>>>>>>> shader-based
>>>>>>>> rendering only. This is also almost done in the branch.
>>>>>>>>      - switch to the new Qt rendering system, i.e. abandon rendering
>>>>>>>> based
>>>>>>>> on QGraphicsView / QWidgets and use only classes from the new QtGUI
>>>>>>>> module (the one from Qt5). I am not completely sure but this may
>>>>>>>> involve a complete GUI re-write based on QML. This last task may
>>>>>>>> sound
>>>>>>>> quite ambitious. I'm quite confident that we can reproduce a GUI
>>>>>>>> similar to the current one based on QML, but it may cause some
>>>>>>>> troubles for interactions with plugin etc..
>>>>>>>>
>>>>>>>> Fabien
>>>>>>>>
>>>>>>>
>>>>>>> ------------------------------------------------------------------------------
>>>>>>> Introducing Performance Central, a new site from SourceForge and
>>>>>>> AppDynamics. Performance Central is your source for news, insights,
>>>>>>> analysis and resources for efficient Application Performance
>>>>>>> Management.
>>>>>>> Visit us today!
>>>>>>> http://pubads.g.doubleclick.net/gampad/clk?id=48897511&iu=/4140/ostg.clktrk
>>>>>>> _______________________________________________
>>>>>>> Stellarium-pubdevel mailing list
>>>>>>> Stellarium-pubdevel@lists.sourceforge.net
>>>>>>> https://lists.sourceforge.net/lists/listinfo/stellarium-pubdevel
>>>>>> ------------------------------------------------------------------------------
>>>>>> Introducing Performance Central, a new site from SourceForge and
>>>>>> AppDynamics. Performance Central is your source for news, insights,
>>>>>> analysis and resources for efficient Application Performance
>>>>>> Management.
>>>>>> Visit us today!
>>>>>> http://pubads.g.doubleclick.net/gampad/clk?id=48897511&iu=/4140/ostg.clktrk
>>>>>> _______________________________________________
>>>>>> Stellarium-pubdevel mailing list
>>>>>> Stellarium-pubdevel@lists.sourceforge.net
>>>>>> https://lists.sourceforge.net/lists/listinfo/stellarium-pubdevel
>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------------
>>>>> Introducing Performance Central, a new site from SourceForge and
>>>>> AppDynamics. Performance Central is your source for news, insights,
>>>>> analysis and resources for efficient Application Performance
>>>>> Management.
>>>>> Visit us today!
>>>>> http://pubads.g.doubleclick.net/gampad/clk?id=48897511&iu=/4140/ostg.clktrk
>>>>> _______________________________________________
>>>>> Stellarium-pubdevel mailing list
>>>>> Stellarium-pubdevel@lists.sourceforge.net
>>>>> https://lists.sourceforge.net/lists/listinfo/stellarium-pubdevel
>>>>
>>>> ------------------------------------------------------------------------------
>>>> Learn the latest--Visual Studio 2012, SharePoint 2013, SQL 2012, more!
>>>> Discover the easy way to master current and previous Microsoft
>>>> technologies
>>>> and advance your career. Get an incredible 1,500+ hours of step-by-step
>>>> tutorial videos with LearnDevNow. Subscribe today and save!
>>>> http://pubads.g.doubleclick.net/gampad/clk?id=58040911&iu=/4140/ostg.clktrk
>>>> _______________________________________________
>>>> Stellarium-pubdevel mailing list
>>>> Stellarium-pubdevel@lists.sourceforge.net
>>>> https://lists.sourceforge.net/lists/listinfo/stellarium-pubdevel
>>>>
>>>
>>>
>>
>> ------------------------------------------------------------------------------
>> Learn the latest--Visual Studio 2012, SharePoint 2013, SQL 2012, more!
>> Discover the easy way to master current and previous Microsoft
>> technologies
>> and advance your career. Get an incredible 1,500+ hours of step-by-step
>> tutorial videos with LearnDevNow. Subscribe today and save!
>> http://pubads.g.doubleclick.net/gampad/clk?id=58040911&iu=/4140/ostg.clktrk
>> _______________________________________________
>> Stellarium-pubdevel mailing list
>> Stellarium-pubdevel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/stellarium-pubdevel
>>
>
>

------------------------------------------------------------------------------
Learn the latest--Visual Studio 2012, SharePoint 2013, SQL 2012, more!
Discover the easy way to master current and previous Microsoft technologies
and advance your career. Get an incredible 1,500+ hours of step-by-step
tutorial videos with LearnDevNow. Subscribe today and save!
http://pubads.g.doubleclick.net/gampad/clk?id=58040911&iu=/4140/ostg.clktrk
_______________________________________________
Stellarium-pubdevel mailing list
Stellarium-pubdevel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/stellarium-pubdevel

Reply via email to