Quote:
Originally Posted by gregbradley
Mosaics are tricky though. It'd be nice to find some sort of a writeup on tips on how to do them. One guy suggested running Gradient Xterminator on each panel before combining which makes sense. Also perhaps a gradient mask to help wth the blend. Rob Gendler mentioned once he matches the brightness of the background and does not necessarily go for equal exposures on each panel.
Do you have any tips? I have done 2 mosaics now and the difficulty was the blend line. One decision I made was about mosaics would be to take the panels over a few days as you need ideally to image in the same angle of the sky for each panel within a range. If you try to do 5 hours worth the last few will be low in the sky and get too much red. Difficult if you have a time constraint like I do travelling to my dark site and only there 3 or 4 nights. Although at this time of year its not uncommon to get 4 nights in a row that are perfect. 3 is likely, 2 would be a bummer and 1 only would be rare.
Rob Gendler's Vela SNR and Tarantula Nebula mosaics are totally amazing.
I'd like to do something like that at some point.
Greg.
|
Hi Greg,
I look forward to seeing your results with the new gear. Thanks for the info on the filters too. I see Don Goldman has gone down the extreme bandpass with some 3nm filters now. I haven't had to opportunity to use these, but feel certain they'll have similar transmission ratings as the Custom Scientific set.
Re: Mosaics, there are a few options. I've emailed Rob a few times to discuss his techniques. All very helpful. You're right, he doesn't capture each panel with the same length of exposures. I believe the determining factor is frame brightness. As with all mosaics, planning is critical. Rob goes to great lengths in the stage of imaging. By matching a wide field image he then determines how to optimally acquire the data with the most efficient panel overlap. This could mean rotating the camera to different angles to achieve the goal. The wide field image also services another purpose during the registration process - it helps to counter act rotation. Not matter, how good you acquire and plan, rotation is simply a fact you need to deal with. Usually, its quite subtle between the panels, but I've experienced a situation where it was rather severe. I've found TheSky to be rather efficient. Rob has a novel way of using TheSky, Registar and photoshop -
http://www.robgendlerastropics.com/Compositions.html I've used this to try some compositions in which I have not followed up in acquiring data (yet).
I've also asked what he feels is the best acquisition technique. For example, a rotational loop acquisition method i.e take a 900s luminance of panel a,b,c,d, then again a,b,c,d. Or perhaps, all blue channel data on the one night (if possible). It would appear there is no "best practice" approach on the acquisition. Regardless of how you acquire the data, you'll always face the challenge of matching the panels. The "brightness/contrast" tool in PS can greatly assist, along with "select colour" and normal curves on the RGB. If you've taken an LRGB composite, you can process each panel as LRGB or match the luminance, then overlay the RGB. The former is sometimes easier.
There are a few resources on the web that can assist, here's one:
http://www.astromatt.com/Articles/MosaicsPI1.html
Don't bite off more than you can chew...start with straight RGB until you get comfortable at adding luminance (less headaches) or do RGB frames for the bright regions and just add luminance to bring out the fainter regions.
Also, I did include the link in the last post. You need the click the "here" hyperlink. Anyway, Hope this helps.
. Surely those exposure times underneath are wrong?? 
.....ah, he's such a Jase 
