How to Drizzle in PixInsight! Improve your images!

[@thepuffybird]

Thanks! I wasn't sure how to do drizzle in Pixinsight. This helps.

[@jimwaters304]

I have been ‘Drizzling’ for several years. When I do I walk away because it takes forever. I go have dinner, watch a movie …etc :). The results are worth it...! I have found that it also reduces background noise. It does increase the integrated sub file size.

[@omgmoreheals]

Thank you for your efforts educating the community! It is a great charity to us all.

[@MikeTettenborn]

Thanks Shawn. Have been drizzling for the past few months ever since you walked me through this. Glad to have the reference video. Clear skies!

[@_TriffiD_]

5:19 - It actually does show that the drizzle-data is added by displying the "" in front of the image name.

[@seanmolony-redstickastro238]

Saved a big mosaic. Thank you

[@peterkudzinowski1274]

Certainly made a big difference , Thanks for the tip Shawn..

[@paulwilson8367]

I stopped drizzling when NSG didn't support it. Now that it does (for a fee), I needed refresher, so thank you. One thing is, no one seems to know a good way to tell if your equipment is "undersampled". Everyone says "undersampled" and then when I ask, "how can I tell", no one really can.

[@VisibledarkAstro]

Simplest way to know is visually. If you're stars are blocky looking when you zoom in then you are. If they are smooth around then you most likely are not undersampled. I've found anything over 2 arcsec/px tends to be undersampled. Astronomy Tools website, the CCD resolution calculator can help determine your setups arcsec/px. Cheers.

[@VisibledarkAstro]

You may also want to consider using the local normalization now built into WBPP. Not much difference between local norm and NSG results, IMO. And with local norm drizzle is supported. Basically latest ver of WBPP can do it all for you. Save some steps and time.

[@paulwilson8367]

@@VisibledarkAstro yes I saw that. I haven’t tried the WBPP normalization check box yet. I would be interested in the comparison.

[@paulwilson8367]

@@VisibledarkAstro did PixInSight add that because the NSG guy demands a fee?

[@paulwilson8367]

@@VisibledarkAstro why can’t PI make drizzling more automatic? In APP, when you check the drizzle box, it does it for you.

[@JonnyBravo0311]

One thing I didn't hear mentioned (I could have missed it) is that when you drizzle, the resulting file is 4 times as big as the original (assuming a 2x drizzle like you used). I have the 294MM Pro, which has two modes. The default mode produces images that are about 11.6MP with 4.63 micron pixels. The unlocked mode produces images that are about 47MP with pixels that are 2.315 microns. I regularly use the default mode, which is under-sampled with my optics because of the smaller file size. I then drizzle. Makes things a bit easier on my CPU/SDD during pre-processing. I have yet to try doing a 2x drizzle on data collected using the unlocked mode. Mostly because that data really isn't under sampled to begin with, but also because I'm pretty sure my poor MacBook Pro would curl into the fetal position and cry if I asked it to try it 😂

[@paulhumberstone5784]

Superb content, Shawn. Keep up the good work! Following your tutorials plus purchasing your excellent Pixinsight for beginners has taken my images to the next level. Thank you.

[@VisibledarkAstro]

Awesome! Thanks Paul.

[@the_astro_garden]

Thanks for another great tutorial Shawn. Your tutorials always cover off the right details, no steps missed and easy to follow, cheers.

[@VisibledarkAstro]

Glad to help!

[@lucasmaguire7625]

thanks Shawn I was wondering about this

[@MikeJohnson-zh9ue]

I wanted to thank you for this video. I'm new to PI and I just tried this and WoW what a difference!!!!!!!!

[@mschmalenbach]

Thanks - just tried it for first time ever - it took hours to process, but easy to set up following your video - fantastic results! I did it with 4 hours of Ha data on Horshead & Flame nebulae through Apertura 60EDR f/6, ASI1600mm at -20C, 60S guided/not dithered, and due to my gear & seeing conditions I was slightly under sampled. Stars didn't look great. now they do! Subscribed!!

[@mariospenard5125]

Clear instructions! Thank you.

[@olly7248]

Just done this with my images of NGC7822… Game changer for my under sampling Redcat 51 with an ASI 1600mm Pro… thank you soooo much for doing this, relatively new to AP so all appreciated 👍🏻

[@VisibledarkAstro]

Awesome!

[@SharpStarAstro]

Awesome tutorial!

[@grahamwhite75a]

Very clear explanation. Thanks.

[@VisibledarkAstro]

Thanks for watching and the comment! Cheers.

[@robb7342]

Simple and sweet presentation Shawn. Greatly appreciated. I'm just on the edge with W.O.Z73iii & ASI2600MC, so I will have to give this a go and see if there much gained.

[@ScottDavies]

Awesome Shawn 👌

[@davidf9494]

Another great video Shawn ! I'm a regular viewer of your channel but missed this one. Quick question: Does WBPP do drizzle integration automatically when generating the Master Lights? If not, I presume I just have to stop at the pre-integration step and then use Drizzle integration? Thanks and Clear Skies!

[@VisibledarkAstro]

There is a setting in wbpp for applying drizzle. If you contact me via email I can help better. Can send a screenshot showing where it's located.

[@soilwork113]

That made it so easy. thanks a bunch

[@SkyShedPODTeam]

Wow! You're the best Man!

[@malek2900]

thanks, it was really helpfull.

[@stephenlast1897]

Thanks for the clear video good as usual . I shall be trying this very soon !

[@VisibledarkAstro]

Please do! Thanks for commenting.

[@neudan7788]

Thank you Shawn!! I usually image NB under light pollution Bortle 6 skies which is suited to my setup. So when I go to a bortle 3 skys imaging broadband I have under sampeling. Will give it a try next time I go over data. Thanks for the video!!!

[@SteveKennedy2902]

Hi Shawn, I am a subscriber and like many on here, I bought your beginner tutorial. I have used Drizzle Integration twice before and found it to be effective. However, yesterday I had an issue ... I drizzle-integrated my 61 images of the Cocoon Nebula and noticed that the final product had many fat or egg-shaped stars. This was peculiar because some of the stars (the larger ones) were round, but others were not. This had not happened previously. I looked at my original master light frames and saw no issues. I also re-checked each individual RAW files one-by-one using Camera Raw in case I missed something during the blink review and concluded there are no problematic images. In the Drizzle integration, it looks as if some of the smaller stars that are very close together have been merged into one oblong shaped star. After watching your video, it looks as if my error was in failing to check the Winsorized Sigma Clipping pixel rejection algorithm during the Image Integration/Drizzle Integration step. I used the WSC rejection algorithm in the past, but this time, I am nearly certain I failed to click the button. I re-processed with the WSC rejection algorithm checked and the problem is solved. I still find this to be peculiar, however, because I saw no issues with my light frames using Blink or Camera Raw. My best guess is that the rejection algorithm is removing data causing the issue. In any event, your video helped me identify the failure in selecting the WSC rejection algorithm and for that I am thankful. Wishing you clear skies from the Great State of Texas.

[@frl8031]

really awesome and concise tutorial. What happens for normally sampled images (around 1-2"/px)? Do the smaller stars benefit?

[@lizards821]

Hey Shawn I really appreciate you walking through this process and explaining it so well! This is easily the best tutorial I have found so far on how to Drizzle. One question though. I noticed my drizzle integration seemed to ignore my flats, and that makes me wonder if it ignored my other calibration frames as well. My standard imageIntegration file looks like it did take into account my flats, etc. What in tarnation do you think I botched in this process haha?

[@VisibledarkAstro]

Drizzle doesn't do any calibration. You would need to do calibration either manually or use WBPP script in PI to calibrate. The script can be instructed to create the drizzle files (which will be calibrated already) and you then just integrate those files using the drizzle integration tool. Other than this, something may not have gone right with your calibration. Does this help?

[@lizards821]

@@VisibledarkAstro yeah it is really funky. I did calibrate manually but according to some folks on discord they are saying my masterflat is either too bright or too dark,. I am also oversampled with my scope so maybe that has something to do with it. Either way, this is def one of those things in astrophotography that is currently a hair frustrating but also enjoyable in the sense of learning. I’ll keep toying around and see where it goes, I appreciate the response!

[@alanrockowitz379]

I always wondered what the heck to do with the Drizzle files after they get created! Thanks for another great video. One question though--where is the new image that gets created? Is it saved somewhere?

[@VisibledarkAstro]

PixInsight puts it in your workspace after the drizzle integration is complete. You then would need to save it. Thanks for tuning! Cheers!

[@LogansAstro]

Nice tutorial Shawn

[@Spaced_Out_Bill]

I'm trying it for the first time! Thanks for making a video about Drizzling. How does one know if their image is under sampled? Thanks!

[@scottfairbairn6305]

How often do you need to dither your images? Every 5 images? every image?? thank you and great video.

[@Demandred20484]

Hi Shawn, thank you very much for this tutorial. It’s been really helpful. Can you please do a quick vid with a workflow starting with wbpps (only registering) and then the new normalize scale gradient and then drizzle? I’m lost as to which files should I use in the drizzle process that will take the NSG script in effect. Thank you in advance and keep doing the great work your doing so far.

[@VisibledarkAstro]

I honestly haven't been following the NSG script development lately. I did try it a few times but haven't recently. Last I was aware, it did not work for drizzled files. Maybe that has changed. I don't know. Will have to revisit the NSG script progress.

[@Ben_Stewart]

I'm confused. Why did we add the drizzle files if image integration will just produce an under sampled image?

[@seanmolony-redstickastro238]

So the second choice using star alignment is performed after calibrating them? Last question.... does using the drizzle tool cost a portion of resolution on the final image?

[@VisibledarkAstro]

Hey Sean - yes do the calibration first. And no it doesn't cost any resolution. It actually preserves resolution.

[@seanmolony-redstickastro238]

VisibleDark thanks Buddy!

[@louboom5711]

Hi thanks for the video! I’m new to PixInsight/drizzle so forgive my ignorance. Where do all the calibration frames come into play here? I’m wondering how you get the final image? I don’t use WBPP I’m using Deep Sky Stacker thanks.

[@gordonahooper]

Does drizzle apply with the new NSG script process?

[@paulwhitton957]

I’m struggling to get drizzle working with the WBPP followed by NSG ?

[@rohitnair2360]

Thank you for this! Can you tell us how you can identify whether your images are under-sampled? And does this also have any effect on noise?

[@VisibledarkAstro]

Noise should not be an issue. Generally I've found if your setup is over 2arcsec/px resolution than you start getting into being under sampled. To find your resolution I'd need to know the telescope focal length and the camera sensor pixel size (microns). You can do the calculation as well using: ( Pixel Size / Telescope Focal Length ) X 206.265

[@malcolqwe2]

very informative, thanks! In WBPP2 there is now a post processing tab and you can enable drizzle there for each filter, with different settings available. Having these choices is powerful, and intimidating at the same time. The scaling factor in particular seems tricky. What would you recommend as far as these settings are concerned? cheers!

[@weshandrow2011]

Is there anything I should be doing if my setup is over-sampled?

[@VisibledarkAstro]

No not really. What is your telescope focal length and camera pixel size?

[@weshandrow2011]

@@VisibledarkAstro 2500mm f/10, 3.76 pixels ASI 2600mc pro.

[@astromansid]

How well does drizzling work on well sampled images? I have heard that if you drizzle the file becomes bigger, has less noise and overall a better image to work with, that can be scaled back to the original size with a clearer image. I am not too sure how that would be better than working with the original data. I have been drizzling in DSS with some good results but when doing mosaics in pixinsight it would fail with the Mosaic by coordinates, I think it is, it wont platesolve. Whereas it will go through with normal data (I might be missing something though) I plan on trying this way of aligning my data in Pixinsight though. Thanks for the video, really helpful and well presented I think. I like these short tutorials.

[@sarahjanereilly9335]

Do you need to dither between every image? I have data that dithers every second frame. Thanks!

[@VisibledarkAstro]

Always dither! 😉

[@sarahjanereilly9335]

@@VisibledarkAstro does that mean my data is useless or how can I adapt the process so I can use it?

[@craigskinner8489]

Thanks Shawn. If my camera / scope is properly sampled, is there a benefit to doing this?

[@terohiekkalinna]

If I drizzle my data I cannot use NormalizeScaleGradient script, because DrizzleIntegration does not recognise files anymore after NSG. Anyone else has the same experience?

[@invatorke]

Why do you need to add the drizzle files to the initial integration? Does the drizzle integration process makes use of this? Thanks!

[@invatorke]

Never mind, you do need to create the original integration, or you'll get errors during the drizzle integration.

[@leemaisel9794]

Can the Actual Drizzle Integration NOT be done in WBPP? I'd rather my master image be output from WBPP instead of an extra step that loses astrometric info. Am I missing something?

[@VisibledarkAstro]

It's a good question. I don't believe so. The drizzle files need to be created and wbpp will do all that. Leaving drizzle off means no drizzle files will be created by wbpp so you would not have them to do a manual drizzle integration.

[@fgandola76]

Hi, thx for the video. I did not manage to Drizzle while using the Scale gradient normalisation. Any tips or may be a suggestion for a video ? ;-)

[@VisibledarkAstro]

Drizzle and Normalize Scale Gradient (NSG) are not compatible. You can't do both. There is talk the developer of the NSG script may add drizzle capability in the future. We just have to wait and see! Thanks for watching. Cheers.

[@fgandola76]

@@VisibledarkAstro not the answer i was hoping for but the answer i expected ;-) thx for answering and for the content you provide.

[@JoseLausuch]

Can you drizzle to non undersampled data? Any side effect?

[@VisibledarkAstro]

Wouldn't really help if your data is not undersampled.

[@JoseLausuch]

@@VisibledarkAstro ok, I thought you could get some extra detail. Thanks!

[@paulwilson8367]

About drizzling, when I attempt drizzle integration with my strong PC (i9, water cooled, 32Gigs RAM) PI crashes to desktop. I tried 7 times, increased virtual memory, restarted, closed everything un-necessary. Loaded drizzle and L-Norm files from new WBPP. Regular image integration worked and I wisely saved the INT image before Drizzle Integration would wipe out PixInsight. Have you heard about this problem?

[@VisibledarkAstro]

Haven't encountered this problem.

[@paulwilson8367]

@@VisibledarkAstro thanks. On the PI forum I see it being reported, especially for large groups of subs. Your channel is a good one.

[@Unavidadevideos]

great video as always! i tried drizzle integration before the master integration, but i ended with a miss aligned integration (i did a blink and the frames are all aligned)... i used the same .xisf calibrated lights for the non drizzle lights and the drizzle... i use a color dslr, can you help me to figure where it can be wrong?

[@mupangamwanakatwe1300]

Hi Shaun. Which software can one use to determine, in advance, if a particular telescope/camera combination will result in under sampled images. Thanks and regards, Mupanga

[@paulwhitton957]

Thank you for all your videos. I have been using WBPP with NSG after following your videos but can’t get drizzle working after telling WBPP to use drizzle, no integration. Then NSG telling it drizzle which creates all the NSG and drizzle files and opens up image integration - this is where I get stuck ! Any help would be appreciated. Regards Paul

[@Hebert6615]

If the data is not under sampled, is there any benefit to drizzling?

[@VisibledarkAstro]

No there really wouldn't be any benefit.

[@alexgti9345]

Hi Shawn! Thanks again for your very useful tutorials, I was just wondering: how to use Drizzle Integration when you want to use NSG for Image Integration after WBPP for calibrating/registering (not integrating obviously) ?? Have you found out the way to go for this kind of processing please? Thanks a lot and cheers from France! :-)

[@VisibledarkAstro]

As far as I'm aware, you can't do drizzle and use NSG. Not yet anyhow. Its been discussed and may come in future updates.

[@alexgti9345]

@@VisibledarkAstro Thanks for your feedback Shawn! Despite many tries, that was also my temporary conclusion, but wanted to ask you If I was missing something.. Let's just hope it will be part of a future update, because NSG is pretty interesting to fight LP or moonlight! Cheers!

[@Dennis-tf2cs]

Are there any benefits to using Drizzle on data that is not under sampled?

[@VisibledarkAstro]

No there isn't. Cheers.

[@garybarr7116]

What exactly do you mean, “use if your images are under sampled?” I have an SBIG STF8300 and shoot 1x1 bins. Does this mean drizzling will not help me?

[@VisibledarkAstro]

Hi Gary - depends on the focal length of your telescope and pixel size of your telescope. I believe the STF8300 is 5.4 microns. I don't know your telescope focal length. But I've found anything over 2arcsec/px resolution tends to be under sampled.

[@garybarr7116]

@@VisibledarkAstro Hi. I really have learned a lot from your videos. Thank you. Just wanted to say that. My telescope is 1230mm and also, I have one that’s 480mm. Will drizzling help me for both scopes?

[@paulwhitton957]

I’ve now read that the creator of NSG needs to update NSG to work with drizzle.

[@CaptRescue2]

Why does Snoop Dog carry an umbrella....for Drizzle!

[@paulwilson8367]

I see that you are totally aware that, using NSG, we cannot drizzle. In your informed opinion, to you gain more by drizzle plus regular normalization or by NSG?

[@VisibledarkAstro]

Hi Paul - that's kind of an apples to oranges comparison. Drizzle will reveal more detail and improve stars. Yes it can't be used with NSG currently. You would gain more with drizzle and regular normalization. NSG just deals with the gradient only.

[@seanmolony-redstickastro238]

Hey Thanks for doing this video. I've been trying to do this to my mosaic that I under sampled with the QSI 683 on my GT71, that we chatted about, but this really helps! Sean. redstickastro PS I dithered all the data.