Beta 2.5.0.18 - QHY10 Image Format


#1

Hi

I use a QHY10 and understand that the ASCOM driver downloads and presents images ‘diagonally flipped’ from the normal landscape view. ‘Diagonally flipped’ means the image is horizontally flipped and rotated by 90 degrees. Rotating the image is only half the job. Nebulosity 3/4 does the full rotate and flip. I think the SGP change is a very good initiative - it is much better to view images in landscape - but I also recommend that images should be both rotated (I think clockwise) AND then horizontally flipped - i.e. all or nothing.

Cheers
DC


#2

While I appreciate the notion, I am unsure of how to apply this in a universal way. We do not have any code in SGPro specific to the QHY 10 (the way it should be). As such, we also have no way to tell (without plate solving) if an image is flipped. Rotation is easy and, as such, can be universally corrected without issue.


#3

Hi Ken

OK, by ‘horizontally flipped’ I meant ‘horizontally mirrored’, I believe an image format and camera driver matter, not related to telescope orientation or optics etc. Apologies if that caused any confusion. I still think a 90 degree rotation, followed by a horizontal ‘mirror’ in all cases would do the trick. I believe this is what Nebulosity does, and since moving to SGP, I have been performing a ‘rotate and mirror’ as a part of my routine workflow when processing images taken with my QHY10. Fortunately, I do not have to do that with other cameras. Anyway, over to you, SGP is working well for me.

Cheers
DC


#4

With all due respect…

Shouldn’t it be QHY who needs to be addressing this issue at the driver level ???

If a captured image from their camera comes out in a strange orientation surely they should modify their drivers applicable to the QHY10 only which will sort this out at source. I use the QHY8L and don’t have to flip or mirror anything, you shouldn’t have to either…OR…Am I about to be educated here ?

:open_mouth:

Paul


#5

Yes and no. Technically there is nothing broke with their ASCOM driver. It’s just not standard and it operates differently for different cameras (QHY10 is flipped, other QHYs are not) Their driver meets the ASCOM standards but for whatever reason they handle cameras differently. So yes, if someone is to flip the “flipped” issue it should be QHY. But at the same time it’s not technically broken either.

Jared


#6

I’m assuming this is when using the native QHY driver in Neb? Just out of curiosity what does the image look like when using the QHY ASCOM driver?

Jared


#7

Hi Jared

The linked review of the QHY10 was done by Craig Stark in 21012 (http://www.stark-labs.com/craig/resources/Articles-&-Reviews/QHY10-On-the-Bench.pdf) . It shows how the images are presented. I believe Nebulosity was updated some time after the review to perform the rotate and mirror (or ‘mirror about the diagonal flip’ as Craig Stark called it).

BTW, yes, I would have preferred this to be a QHY driver update, but that does not look like happening anytime soon.

Cheers

DC


#8

As expected it sounds like he found the same weirdness and addressed it with his implementation of the native driver (which we don’t implement for qhy).

So at this point I think the rotation fix is all we’re going to do. This is obviously non-standard but it conforms with every other qhy. And is coded in such a way that it’s not actually qhy specific. However the mirror logic would only apply to qhy cameras and even more specifically just the qhy10. That is not something were willing to pull in.

If it’s that big of an issue I would recommend contacting Qhy, after all you’ve (referring to everyone that is having this issue, not JUST you) given them substantially more money than you have us and ultimately we don’t want to bring bad behavior into our application as its very camera specific and that goes against some of the tenants of ascom.

Thanks
Jared


#9

Thanks Jared.

From memory, I think Nebulosity only made the change (rotate and mirror) for the native Windows driver, not ASCOM. So to that extent, you have gone a bit further with the ASCOM driver, ie the 90 degree rotation. No issues for me - landscape is much better for viewing on the screen - I used to do a rotate and mirror, now its just a horizontal mirror, plus making sure calibration images match of course.

Cheers
DC


#10

I have been using Nebulosity for quite a while with different cameras before using SGP and my current camera is QSI 660wsg. Both Nebulosity and SGP use native QSI drivers. I capture lights with SGP and bias, darks and flats with Nebulosity and I had no issues with calibrating SGP’s lights with Nebulosity’s bias, darks and flats. If Nebulosity rotated images differently than SGP, then I would definitely notice bad calibration especially from flats but they calibrated perfectly.

So I think Nebulosity and SGP orient images the same way with QSI camera using QSI native driver. I am not sure with SP’s latest version so I have not yet test it but as per Ken’s suggestion from another thread, I will take bias, darks and flats with SGP from now on to be safe.

Peter


#11

I wonder about the image mirror you describe. I just did some tests with Platesolve 2 and it can’t solve an image that is mirrored in one direction. I have had no problem with PlaseSolve 2 and a QHY10 which indicates the image is not mirrored. (An image mirrored in both directions is the same as rotating the image 180)


#12

This is really not a big issue for me, the system has been working fine, and I just ensure calibration images are of a consistent orientation to the lights. I was just suggesting an all (rotate and mirror) or nothing approach for reasons of consistency.

Note the ‘issue’ relates specifically to the QHY10 camera. I also use a QSI camera, and it is not affected.

I was also concerned whether plate solving would work, but as you say, it does thankfully. See this link if you are not convinced about the mirroring: http://www.stark-labs.com/craig/resources/Articles-&-Reviews/QHY10-On-the-Bench.pdf. The ‘issue’ is completely separate to the normal 180 degree rotation intrinsic to refractors and SCTs etc. Take a shot at something distinctive such as the Horsehead and you will see it pointing the wrong way.

I can easily live with the approach being taken.

Regards to all

DC


www.mainsequencesoftware.com