IDAS LPS D2 Filter. Is colour balance easy?

[Ross1204]

Hi, 

I have been looking into buying an IDAS light pollution filter as they seem to be the preferred choice most people recommend when similar questions are asked. I currently own a Astronomik CLS CCD clip in filter which has really helped in gaining longer exposure times in my light polluted garden but as others have suggested it can be a real struggle at times to correctly adjust colours (especially stars).

I have also read multiple posts on here and different forums where people have stated that the IDAS LPS (P2 & D1) filters really do help in keeping colours balanced correctly whilst also retaining good star colour when compared to others on the market. 

So my question is I have not really read a great deal of reviews on the newer LED reducing IDAS D2 filter and what that is like compared to the others in the IDAS LPS range.

Has anyone had a prolonged use of the newer D2 filter? and how does it compare to the others when it comes to colour balance and star colours?

Thanks

Ross

[vlaiv]

Here is a good web page on performance of those filters:

https://www.sciencecenter.net/hutech/idas/lps/plots/index.php

Just move cursor over table on the right and place it over respective field to view graph showing particular spectra.

Look at the filter spectrum response and see how "well spread" it is over spectrum. If it covers roughly equal parts of Red, Green and Blue parts of spectrum then you will not have much trouble doing color balance.

You can (very) roughly divide 400-700nm range into three bands - 400-500 Blue, 500-600 Green, 600-700 Red.

P2 has the best native color balance, followed by D1, D2 and V4 at the last place (best suited for emission nebulae).

On the other hand P2 and D1 will suffer from LEDs quite a bit because ~460nm peak is included in both of those filters, as well as big chunk of 530-630nm continuum.

With any of those filters you can try to get white balance "manually" in post processing, but I would advise you to try automatic - photometric color balance as a first step prior to any manual tweaks.

This process will look at the stars and plate solve your image to obtain spectral class for each star (well for set of stars on image, not necessarily all of them). There is transformation from spectral class to R, G and B ratio of each star, and from this information and black point automatic white balance can be derived to best match star colors regardless of filter used (and spectral sensitivity of sensor used).

You can even do this manually using a single star - find "white" star in your image (use planetarium program and identify star that has color close to white - F star for "pure white" or G star - for sunlight white) and scale R, G and B channels according to measured value of the star (select star and measure total signal value over selection in each channel).

[Ross1204]

Hi vlaiv,

Firstly wanna say thank you for taking the time to help with this question, much appreciated. 

So from looking at the graphs for all  four filters it's kinda made me more unsure on which would provide me the best reduction in light pollution, I mean this is a good way though as its thrown up questions to help better inform the right choice.

Over the past few months when the moon has been good I've shot completely unfiltered through a Baader Modified Canon 700d and it's these shots that lead me in to wanting better star colours (I could actually see them). The Astronomik CLS CCD I have is good with increasing contrast on emission nebula but completely whacks out the colours more often than not which makes me always think now the better option at times could be to just remove gradients in post processing.

So I guess if I am that concerned with getting the correct colours the P2 would be the best choice providing I don't suffer from any LED light pollution. I did email the local council and ask about the lighting in my area and was told the budget couldn't yet stretch to provide this area with LED lights (I was obviously made up). The caveat being the "budget couldn't stretch yet" hence why my attention was towards the new IDAS D2 for its LED blocking capabilities. The spectra graphs you linked to have slightly changed the thinking a little cos the P2 would obviously provide the best colour balance and I'm not yet shooting through LED light pollution.

I currently do not own Pixinsight but have read about the photometric color balance option it has which sounds great from all I've read, I'm just not sure I fully understand the manual way you described to give that a go in photoshop. 

Ross 

[vlaiv]

I also use P2 and it works quite well. No LEDs here where I am, but there are quite a bit of regular incandescent light bulbs, that have even worse spectrum signature than LEDs.

Remember it is not only gradients you need to contend with - additional LP signal brings in associated noise that is the real issue.

As for manual method of color balance, I'm not sure that you can do it in photoshop alone. I use one star manual calibration with ImageJ - open source software for scientific image processing. It has bunch of features - so it allows for sort of manual "photometry" to be performed - it will calculate necessary statistics for selected region - so it is the matter of marking a star with circular selection and hitting Analyze / Measure menu option.

Afterwards, there are basic math operation that can be performed on images (like add / subtract / multiply ....), basically everything that you need to get one star calibration to work.

Multi star calibration is a bit more complicated as you need to perform least squares approximation to find matrix transform of a set of vectors (each star R, G and B can be considered as a component to a 3-vector, and usual channel mixer in graphics software is actually matrix transform of such a vector, so given set of vectors - or RGB values of multiple stars, you can find matrix that does closest transform to expected RGB values of actual stars based on their spectral class).

Anyway, if you want to give ImageJ and one star calibration a go, look into ImageJ (download it - no need to install it works in "portable" mode), and if you wish, I'll be more than happy to provide you a brief "tutorial" or "how to" for that.

[Ross1204]

Yeah I must admit the P2 filter is definitely the more appealing as it lets the most light through but all the IDAS literature on multiple astro websites state that the D1 is the better of the two for modified DSLR's. I shall keep looking online for more user testimonies but I know doubt have a better understanding after your advice so thanks again.  

With regards to the tutorial that's entirely up to you, I would happily to look through it  myself and I'm sure others would benefit  in the future if its not been shown before but don't feel like there is any pressure if its too much hassle.

Ross