Combine RGB - No Header data

Yes, this is because they aren't, there will be some differences while integrating each channel due to alignment and such. You need to first register these integrations together. Load them into APP as lights, go to the register tab and all the way down click "save registered frames". These can then be combined.

@vincent-mod

Just a quick answer inherent this topic...

I usually click "normalize", instead of just register.

What is the difference between using normalized files instead of just registered ones?

Thanks

@ippiu Normalization corrects for differences in between the frames for illumination and such, registered frames not. I'm not 100% sure what would be better in this case actually, I normally use registered frames and mix those accordingly. Have to test the difference. 🙂

@vincent-modThank you Vincent.  Will give this a go 🙂 

Ok, to clarify a bit. If you save the normalized frames, Ha will be more similar in signal to the other frames, the others will have more noise then though. So I think that's why I tend to play around with both sometimes. It's been a while so I wasn't sure. Thanks to @wvreeven for clarification. 😉

@vincent-modThank you Vincent.  The software has so much to offer...I will test the suggestion and see how this works out.  I am making good progress with the software of how it works 🙂 Pretty happy with it so far. 

To test this, I have tried to do this on another set of Data. Image shot with Nikon Z6 and Redcat 51.  RGB and Ha.

I have created two integrated files - 1 for RGB and 1 for Ha.  Stars in RGB 10000 and Ha about 4000. 

When I load the two files in "light" APP cannot register the files when I try to register the files.  I have got through all suggestions, but unfortunately no luck. 

Any suggestions, tips of what this could. be?

Hendrik

Best way to combine data with narrowband is to seperate the RGB channels first, in tab 2 all the way down "split channels". These can then be loaded in again as lights and added to their correct channels. Which you can then register and normalize together. If that also doesn't work, I suspect it's either a data issue or the issue we have in APP 1.083 beta 2. For the latter please try 1.082.

I have splitter the RGB data in separate channel. When I register the 3 files with HA, the registration fails. As Ha also was shot with the NIKON Z6 I have split the Ha channels as well.  When I register the 3 RGB files with the 3 Ha files on the Ha Green channel is registered.  Red and Blue fail.  With this in mind I added the 3 RGB Channels and Ha Green channel and started the Registration.  It failed.    Any suggestions? 

I am on APP 1.082 

Mmm, I think it's a data issue then. Can you upload the raw files, including calibration data (masters) to our server for me to check out? Please allow for up to a few days.

Go to https://upload.astropixelprocessor.com and for the username and password, use: upload

Create a directory named “nordiclightphoto-registerIssue” and upload in there. Thank you!

OK will do this. Vincent.  I did one additional test.  I just registered the 3 RGB files.  Registering Correctly.  Clear the buffer... loaded the 3 HA files.  Registering correctly.  When I clear the buffer and load first HA(3)  then RGB (3) - File 4 and 6 fail to register (In this scenario RGB R & RGB B).  When I clear the buffer and load RGB (3) first and HA (3) - File. 4 and 6 fail register (In this scenario HA R and HA B .  

Everything uploaded. 

Ok looking at your data, thanks for sharing.

For the H-alpha I use the H-alpha algorithm in tab 0 (I noticed you tried to split the channels, this algorithm is much better for this and the data you then load in is immediately H-alpha, no need to split. The RGB is better to split as you can then nicely combine the H-alpha with the other split R, G and B mono channels). I see quite a bit of undersampling (causing more square stars). Likely still fine for registration, but just a thing I'm noticing. ISO 3200 is also likely not the optimal setting for your camera, not sure though but there are references online for certain sensors to choose a nice one.

Looking at the master files, the ha-master-flat seems to have a bit on an issue when using the Ha part. I'm not seeing a proper vignetting profile. I think the exposure was too high for that channel, which is always a bit tricky with a DSLR and narrow-band filter, looks like the flat was kind of exposed to get a median value of about 50% exposure for all channels, but the red channel is way brighter, maybe even clipping on the high side. So I'm not going to use it in data integrations, using calibration data that has an issue, will cause more problems down the road otherwise. Something at least to have a look at. 😉

This is the h-alpha integration I got, no issues:

The RGB data; clearing the data and switching to the regular Adaptive Airy Disk algorithm in tab 0. Master-flat here seems ok, so using that.

This is the RGB integration;

Now, using this RGB integration above I'm going to split those channels to get mono R, G and B. I load in the integration, go to tab 2, all the way down select "split channels" and press "save light frames".

Then I clear yet again and load in each mono channel into the correct filter, R for the R, G for the G, B for the B and Ha for Ha.

Then I go to the normalize tab, click on "normalize" and then save the normalized (and registered) frames, with the button all the way down in that tab. And here I stumbled upon your issue. So I tweaked some of the parameters and what worked was to disable same camera/optics, increase scale stop to 8 and choosing triangles for pattern recognition. Bit odd, but maybe caused by the undersampling/less optimal registration in beta 2.

Combined the channels, used the HaRGB formula, added bit of saturation, bit of contrast;

Then I used the light pollution tool to get this;

Followed by color correction, which worked pretty nicely;

Thank you Vincent for the instant response.  On the data...wanted to see how far I can push the SW SA in terms of exposure times...SO I am not too worried about a few things.  ISO...Yes, can be high, but it is in the middle of the inferred SNR for the sensor... 3200 gives me 2 times more light than ISO 800. And as you know...Ha...needs light!  Last night I looked at TAB 0 and saw the single HA process... Now I have figured out how it works...it will become easier. 

But ok...I will do some other test :.). I understand everything what you have written...except.

Then I clear yet again and load in each mono channel into the correct filter, R for the R, G for the G, B for the B and Ha for Ha.  Where do I load and map the the channels TAB 1?  

Further...once registered I go in TAB 9 and select Combine RGB? Correct?  Hope you can point in the right direction on the above question.

Cheers

hendrik 

@vincent-mod Hi Vincent, I figured out how to map the RGB Ha channels to their corresponding filter in TAB1.  Registration went fine...no issues and did not need to tweak any parameter.  Also the Combine RGB worked fine. 

Thanks