Jupiter reprocess from 09/09/23

[Kon]

It feels ages since we had that amazing night in UK. I have not done any Jupiter imaging since that session! I reprocessed the image. After careful analysis it seems that derotation did not help, even between two continuous captures; lost quite a bit of details. This is the best single capture. First image is the new version and the second is the original I posted on 9th September. Let me know your thoughts.

 

 

[Maurizio83]

The new image clearly appears better to me, particularly sharper and more defined.

[Mark2022]

The new one is definitely improved. So what did you do differently, Kon? I see a touch more detail in the Southern Polar hood and a little more saturation overall.

[Kon]

11 hours ago, Maurizio83 said:

The new image clearly appears better to me, particularly sharper and more defined.

Thanks Maurizio. I suppose cloudy nights are good for revisiting nice data.

[Kon]

9 hours ago, Mark2022 said:

The new one is definitely improved. So what did you do differently, Kon? I see a touch more detail in the Southern Polar hood and a little more saturation overall.

Thanks Mark. The main difference is that this image is a stack of 120s rather than derotation of 5x120s. Even derotating the best 2x120s was giving a softer result; still better than the original but the finer details are better on the single stack. 

[geoflewis]

The revision is definitely sharper than the original. How did you derotate, e.g. did you deroate the videos, or the resulting TIFFs from each video? It would be interesting to see the 5 TIFFs side by side, maybe one or more them was poor enough to corrupt the end result....🤔

[Mark2022]

48 minutes ago, Kon said:

Thanks Mark. The main difference is that this image is a stack of 120s rather than derotation of 5x120s. Even derotating the best 2x120s was giving a softer result; still better than the original but the finer details are better on the single stack. 

Sorry Kon, I'm not following. A stack of 120 seconds or 120 subs?  If seconds, a single stack from a single 120 second video? If subs (i.e. frames from the original video) how many frames in total were taken? i.e. What percentage  is 120? So derotation isn't all it's cracked up to be? 

What was the gain and exposure you used and how high was Jupiter at the time? Regain and exposure: I guess you use firecapture or sharpcap whereas I'm using AstroDMx for the moment so my guess is that gain, at least, may not be measured precisely the same in each software.

I've been reading up on pixels and well depth, QE etc as well as gain and exposure and how they interact so I'm beginning to wonder if I'm just not 'filling the well' enough or filling too much with my gain/exp settings.

I noted a very calm, clear night last night at about 2am. If I'd known earlier it was clear,I'd have been out but, at 2am (and the moon and Jupiter were about 60 degrees I estimate) I was flailing having been up at about 7am. Also, within the hour, I'd lose Jupiter behind my house wall/roof.

[Space Cowboy]

Seeing varies so much it's pretty rare to have 10 mins of similar quality. I often come to this conclusion unless the data is noisy and can't support a single stack.

I'm currently working on some new software that will disperse cloud and promote excellent seeing.🧐

[Kon]

4 hours ago, geoflewis said:

The revision is definitely sharper than the original. How did you derotate, e.g. did you deroate the videos, or the resulting TIFFs from each video? It would be interesting to see the 5 TIFFs side by side, maybe one or more them was poor enough to corrupt the end result....🤔

The first 3 of the stack where 'softer' compared to the last two after careful inspection, but somehow the ladt two where very similar but derotation did not help (they were sharper pre-winjupos). I derotate the tiffs. I have never derotated a video. Much of an advantage?

[Kon]

3 hours ago, Mark2022 said:

Sorry Kon, I'm not following. A stack of 120 seconds or 120 subs?  If seconds, a single stack from a single 120 second video? If subs (i.e. frames from the original video) how many frames in total were taken? i.e. What percentage  is 120? So derotation isn't all it's cracked up to be? 

What was the gain and exposure you used and how high was Jupiter at the time? Regain and exposure: I guess you use firecapture or sharpcap whereas I'm using AstroDMx for the moment so my guess is that gain, at least, may not be measured precisely the same in each software.

I've been reading up on pixels and well depth, QE etc as well as gain and exposure and how they interact so I'm beginning to wonder if I'm just not 'filling the well' enough or filling too much with my gain/exp settings.

I noted a very calm, clear night last night at about 2am. If I'd known earlier it was clear,I'd have been out but, at 2am (and the moon and Jupiter were about 60 degrees I estimate) I was flailing having been up at about 7am. Also, within the hour, I'd lose Jupiter behind my house wall/roof.

Sorry, I should have said I stacked a 120s video capture; the capture has around 25000 frames. In this instance I used 50%, 12500 frames to make the tiff above. 

My exposure was 5ms. Do not worry about the gain as we have different cameras; what you are aiming is to have a histogram of 50-70%. Once you have the exposure set, then play with the gain to bring it to that level.

Derotation can help and it certainly helped when I was doing the Venus captures back in June. The other images where a bit softer that did not help with derotation. Pointed by Geof and Stuart above too.

[Kon]

3 hours ago, Space Cowboy said:

Seeing varies so much it's pretty rare to have 10 mins of similar quality. I often come to this conclusion unless the data is noisy and can't support a single stack.

I'm currently working on some new software that will disperse cloud and promote excellent seeing.🧐

Spot on and that's what I noticed after a careful analysis of all the tiffs.

looking forward to the software but I though it already exists and it is called ticket to Barbados!🤣

Edited October 3, 2023 by Kon

[geoflewis]

4 hours ago, Kon said:

The first 3 of the stack where 'softer' compared to the last two after careful inspection, but somehow the ladt two where very similar but derotation did not help (they were sharper pre-winjupos). I derotate the tiffs. I have never derotated a video. Much of an advantage?

In my experience its rare to see worse results after derotation of the TIFFs created from separate videos, unless I included a really bad one, so it's certainly surprising that the deotation of two similar TIFFs resulted in a worse result than ether of the source files. Would you be prapared to share the TIFFs so that we can have a play please?

I don't usually de-rotate SERs, but did experiment with it last year during the Mars apparition. The overheads in file size were huge, but it did allow me to shoot longer SERs, so maybe there could be some benefit sometimes, but my preference remains to shoot a series of shorter videos, process those through AS3! and Registax, then take the resulting TIFFs into WinJupos for derotation. I then take the derotated stack TIFF back into Registax for a tad more wavelets and sometimes a bit of unsharp mask in PS or Affinity Photo. Did you try any further processing after derotation?

[Kon]

1 hour ago, geoflewis said:

In my experience its rare to see worse results after derotation of the TIFFs created from separate videos, unless I included a really bad one, so it's certainly surprising that the deotation of two similar TIFFs resulted in a worse result than ether of the source files. Would you be prapared to share the TIFFs so that we can have a play please?

I don't usually de-rotate SERs, but did experiment with it last year during the Mars apparition. The overheads in file size were huge, but it did allow me to shoot longer SERs, so maybe there could be some benefit sometimes, but my preference remains to shoot a series of shorter videos, process those through AS3! and Registax, then take the resulting TIFFs into WinJupos for derotation. I then take the derotated stack TIFF back into Registax for a tad more wavelets and sometimes a bit of unsharp mask in PS or Affinity Photo. Did you try any further processing after derotation?

Thanks Geof. Yes I tried a bit more processing; I tried two ways. 1. tiny wavelets in Registax and then in winjupos and then back to Registax/Astrosurface. 2. Nearly done in registax but I found that i could not push more after winjupos. 

I am happy to share the images. These are the two best ones straight after as!3. The 358 capture is the one reprocessed, although the 400 should be as good but winjupos made a softer output.

2023-09-09-0358_4-Jupiter_pipp_lapl4_ap288.tif2023-09-09-0400_8-Jupiter_pipp_lapl5_ap171.tif

[geoflewis]

1 hour ago, Kon said:

Thanks Geof. Yes I tried a bit more processing; I tried two ways. 1. tiny wavelets in Registax and then in winjupos and then back to Registax/Astrosurface. 2. Nearly done in registax but I found that i could not push more after winjupos. 

I am happy to share the images. These are the two best ones straight after as!3. The 358 capture is the one reprocessed, although the 400 should be as good but winjupos made a softer output.

2023-09-09-0358_4-Jupiter_pipp_lapl4_ap288.tif 1.15 MB · 0 downloads 2023-09-09-0400_8-Jupiter_pipp_lapl5_ap171.tif 1.15 MB · 0 downloads

Thanks Kostas, I've downloaded them and had an initial play, but ran into a couple of issues with the orientation of the images. I'm so used to having mine horizontal that I really struggled visually applying the wavelets in Registax, then even more so performing measurement in WinJupos as the image is flipped. I got past that and performed the derotation, which did enhance the image, but I think I've been too aggressive in Registax, so will take another look, probably tomorrow.

I must say that the quality of your indivial revised 0358 image is superb and so far I haven't got close to that, but will continue to try. I don't use Astrosurface, so I'm interested to know whether you think it's better than Registax. I think I downloaded a couple of years ago and had a play, but gave up on it.

Edited October 3, 2023 by geoflewis

[Kon]

4 minutes ago, geoflewis said:

Thanks Kon, I've downloaded them and had an initial play, but ran into a couple of issues with the orientation of the images. I'm so used to having mine horizontal that I really struggled visually applying the wavelets in Registax, then even more so performing measurement in WinJupos as the image is flipped. I got past that and performed the derotation, which did enhance the image, but I think I've been too aggressive in Registax, so will take another look, probably tomorrow.

I must say that the quality of your indivial revised 0358 image is superb and so far I haven't got cose to that, but will continue to try. I don't use Astrosurface, so I'm interested to know whether you think it's better than Registax. I think I downloaded a couple of years ago and had a play, but gave up on it.

Thanks Geof. Sorry, I flip them before Winjupos for orientation, I should have done it before uploading here. I let Winjupos measure and adjust accordingly. You can use Io as a guide at the bottom left too.

Regarding Astrosurface, I am finding it a lot better these days. On the 358 image, I can't reproduce these details even with hard pushing in registax. When I am back on my laptop I will message you a screenshot of my initial parameters at least for this image. I also did a mild sharpening too in Astrosurface. I like the denoise on Astrosurface too but somehow colour denoise is a bit aggressive and I use imageanalyzer. I tend to throw a bit of the best features of each software to get to the final image.

I also tried x1.5 drizzle in as!3 but it didn't improve things too much, similar to your experience.

[BGazing]

Indeed a very nice capture, I played around with it a bit, result attached.

I'd say 5ms is really short in good seeing. 7.5ms would get a much better S/N for you. Also, although readout noise is not that important, when stacking many frames it makes sense to up the gain over HGC threshold, check it out.

Also, try metaguide for collimation. Your images have that little bit of softness mine had before collimation tweak.

 

 

Edited October 3, 2023 by BGazing

[geoflewis]

47 minutes ago, Kon said:

Thanks Geof. Sorry, I flip them before Winjupos for orientation, I should have done it before uploading here. I let Winjupos measure and adjust accordingly. You can use Io as a guide at the bottom left too.

Regarding Astrosurface, I am finding it a lot better these days. On the 358 image, I can't reproduce these details even with hard pushing in registax. When I am back on my laptop I will message you a screenshot of my initial parameters at least for this image. I also did a mild sharpening too in Astrosurface. I like the denoise on Astrosurface too but somehow colour denoise is a bit aggressive and I use imageanalyzer. I tend to throw a bit of the best features of each software to get to the final image.

I also tried x1.5 drizzle in as!3 but it didn't improve things too much, similar to your experience.

Thanks Kostas, I'd be very interested to see your settings for Astrosurface. I also don't know imageanalyzer . Clouds rolled in here after being clear earlier, so no imaging tonight, hence I decided to have another play with yours. This is after derotation, but still looks noisy to me, where your single image is very clean...

Edited October 3, 2023 by geoflewis

[geoflewis]

40 minutes ago, BGazing said:

Also, try metaguide for collimation.

I'm not familiar with this, please tell me more?

[Kon]

42 minutes ago, BGazing said:

Indeed a very nice capture, I played around with it a bit, result attached.

I'd say 5ms is really short in good seeing. 7.5ms would get a much better S/N for you. Also, although readout noise is not that important, when stacking many frames it makes sense to up the gain over HGC threshold, check it out.

Also, try metaguide for collimation. Your images have that little bit of softness mine had before collimation tweak.

 

 

 

Thanks for the processing. Looks nice. Good tips on capture but with the 462 I haven't been too worried about noise. Even close to maxing the gain on methane or UV captures I can manage the noise ok. I may have been a bit conservative here.

Regarding metaguide I have read about it but I am not sure how easy it will be with my manual dob. I will give it a try. I always check collimation by star test before capture and I think it is more of focusing in this case. I have just upgraded to a dual focuser.

[Kon]

10 minutes ago, geoflewis said:

Thanks Kostas, I'd be very interest to see your settings for Astrosurface. I also don't know imageanalyzer . Clouds rolled in here after being clear earlier, so no imaging tonight, hence I decided to have another play with yours. This is after derotation, but still looks noisy to me, where your so=ingle image is very clean...

 

Sad face for the clouds. I am eager to get out but Jetstream is looking awful. The weekend is looking promising.

Thanks for the image. It looks nice and better than my derotated attempt last night.

I was introduced to imageanalyzer by Stuart. Worth a play. It has several nice features.

[BGazing]

14 hours ago, Kon said:

Thanks for the processing. Looks nice. Good tips on capture but with the 462 I haven't been too worried about noise. Even close to maxing the gain on methane or UV captures I can manage the noise ok. I may have been a bit conservative here.

Regarding metaguide I have read about it but I am not sure how easy it will be with my manual dob. I will give it a try. I always check collimation by star test before capture and I think it is more of focusing in this case. I have just upgraded to a dual focuser.

I have forgotten you are doing this on a manual dob...you have done a fantastic job, indeed.

I guess Paracorr would help if you are capturing it while it is drifting accross the field?

[Kon]

1 hour ago, BGazing said:

I have forgotten you are doing this on a manual dob...you have done a fantastic job, indeed.

I guess Paracorr would help if you are capturing it while it is drifting accross the field?

I would have though the paracorr to be more beneficial for DSO as the planets drift fairly quickly from the sides (I am sure as!3 sorts these if there is an issue).

An ADC would be better but again hard on manual (looking to diy a platform when I have some time).

[Maurizio83]

Very interesting conversation. I also tried experimenting with derotation once last year, but at the time, I wasn't sure if the benefit justified the additional time required. Furthermore, even though the noise was reduced, I wasn't sure that the fine details had actually improved. Maybe the problem from my location  is that good seeing images alternate with images with worse seeing, which, when used, reduce the overall quality. Probably I made some mistakes, and I need to learn how to use derotation better. Here you can find the two images for comparison ( 9 images derotated vs the best single ) :

https://flic.kr/p/2nXxJhN

https://flic.kr/p/2nXqdJs

Anyway, I am amazed by the use of a manual Dobsonian. I struggle a lot to find the perfect focus point with a C8 (first on EQ5, now on a fork mount), with sidereal tracking. The idea of doing that while manually tracking a planet is incredible. 😅

Edited October 4, 2023 by Maurizio83

[geoflewis]

9 hours ago, Maurizio83 said:

Here you can find the two images for comparison ( 9 images derotated vs the best single )

Both ae excellent images, but I think that derotation has yielded a little more detail.

Edited October 5, 2023 by geoflewis

[Kon]

10 hours ago, Maurizio83 said:

Very interesting conversation. I also tried experimenting with derotation once last year, but at the time, I wasn't sure if the benefit justified the additional time required. Furthermore, even though the noise was reduced, I wasn't sure that the fine details had actually improved. Maybe the problem from my location  is that good seeing images alternate with images with worse seeing, which, when used, reduce the overall quality. Probably I made some mistakes, and I need to learn how to use derotation better. Here you can find the two images for comparison ( 9 images derotated vs the best single ) :

https://flic.kr/p/2nXxJhN

https://flic.kr/p/2nXqdJs

Anyway, I am amazed by the use of a manual Dobsonian. I struggle a lot to find the perfect focus point with a C8 (first on EQ5, now on a fork mount), with sidereal tracking. The idea of doing that while manually tracking a planet is incredible. 😅

Thanks Maurizio. Yes the manual has its challenges but it shows that it is possible to get decent images even with 8".

I agree with Geof that your derotation has better details. Unfortunately UK skies prevents me from experimenting too much as usually 1 or 2 captures are decent to do further. I   might try the video derotation on the image above and see how it looks.