Jupiter 8th Oct still trying.......

[yelsac]

250p scope, QHY5L-IIc, 2mins avi at around 65 fps, Pipp 9k frames down to 6k, best 40% from Autostakkert, probably over did it with the wavelets in registax then a fiddle in Image Analyzer. Still trying to get the balance between detail/noise! I find the image is extremely noisy but didn't want to loose the detail I had 🙄.

Its also 1.5 Drizzle & upside down 😂

Any thoughts advice welcomed & appreciated

[neil phillips]

2 hours ago, yelsac said:

250p scope, QHY5L-IIc, 2mins avi at around 65 fps, Pipp 9k frames down to 6k, best 40% from Autostakkert, probably over did it with the wavelets in registax then a fiddle in Image Analyzer. Still trying to get the balance between detail/noise! I find the image is extremely noisy but didn't want to loose the detail I had 🙄.

Its also 1.5 Drizzle & upside down 😂

Any thoughts advice welcomed & appreciated

Starting to get some nice fine detail coming through

[yelsac]

On 14/10/2022 at 18:44, neil phillips said:

Starting to get some nice fine detail coming through

Yes there's a little, still really frustrating though.

Can't tell if the focus isn't quite there or if it's just ME! with my processing 🤔😁.  Also I'm not sure if the scope is correctly aligned. I've bought a laser collimator which I've tested to make sure it itself is collimated but after using it & when I do a star test it looks slightly out!

I think I could do with upgrading to get a more sensitive camera to.

I know this is a real cheek but...... Could I ask Neil if you''ve got a moment or two would you mind having a quick go at processing the same image to see what you think? I've uploaded the original & the drizzled version.

original.tif

Drizzle15.tif

 

[neil phillips]

23 minutes ago, yelsac said:

Yes there's a little, still really frustrating though.

Can't tell if the focus isn't quite there or if it's just ME! with my processing 🤔😁.  Also I'm not sure if the scope is correctly aligned. I've bought a laser collimator which I've tested to make sure it itself is collimated but after using it & when I do a star test it looks slightly out!

I think I could do with upgrading to get a more sensitive camera to.

I know this is a real cheek but...... Could I ask Neil if you''ve got a moment or two would you mind having a quick go at processing the same image to see what you think? I've uploaded the original & the drizzled version.

original.tif 1.15 MB · 0 downloads

Drizzle15.tif 2.59 MB · 0 downloads

 

Just tried a few things but can't get a result any better than what you're doing. If anything, mine looked far worse. Wouldn't be happy putting it up here. Probably used to my own data I am not sure.

Focus doesn't look too bad.  Though As you say its noisy

[yelsac]

4 hours ago, neil phillips said:

Just tried a few things but can't get a result any better than what you're doing. If anything, mine looked far worse. Wouldn't be happy putting it up here. Probably used to my own data I am not sure.

Focus doesn't look too bad.  Though As you say its noisy

Thanks for having a go Neil really appreciate it 👍

I know it sounds obvious but I'm wondering if getting a more sensitive camera with a much higher frame rate is what I need.

With my QHY5L-llc I take roughly 9k frames per avi at around 67fps at 640x480. Then cut that down with quality estimation to around 6K in pipp. Depending on the quality graph in AS3 I tend to take the percentage about 50% which normally is around 30-40% leaving me with around 1500 to 2000 frames that actually get stacked.

Do you mind me asking with Jupiter what Video size & fps you get with the 462C? Also what percentage roughly (I know it depends on seeing etc..) you stack in AS3?

[neil phillips]

10 minutes ago, yelsac said:

Thanks for having a go Neil really appreciate it 👍

I know it sounds obvious but I'm wondering if getting a more sensitive camera with a much higher frame rate is what I need.

With my QHY5L-llc I take roughly 9k frames per avi at around 67fps at 640x480. Then cut that down with quality estimation to around 6K in pipp. Depending on the quality graph in AS3 I tend to take the percentage about 50% which normally is around 30-40% leaving me with around 1500 to 2000 frames that actually get stacked.

Do you mind me asking with Jupiter what Video size & fps you get with the 462C? Also what percentage roughly (I know it depends on seeing etc..) you stack in AS3?

I am capturing at 200. to 250 frames per second.  capturing 6 minuets and doing winjupos video de rotation. capturing around 70 to 90.000 frames stacking anything from 15.000 to 40.000 frames depending on quality. So of course a lot more

[yelsac]

23 minutes ago, neil phillips said:

I am capturing at 200. to 250 frames per second.  capturing 6 minuets and doing winjupos video de rotation. capturing around 70 to 90.000 frames stacking anything from 15.000 to 40.000 frames depending on quality. So of course a lot more

Woooow a lot more! Do you use this IR890 CH4 filter that you can get with the 462C?

[symmetal]

What's your exposure yelzac? It needs to be around 5mS to help freeze the seeing conditions and to allow a high framerate. 5mS should give a framerate of 200fps at your ROI, ( 1 / 0.005) assuming your camera and capture computer can cope with that. Increase the camera gain to achieve around 60-70% on the histogram. The gain may end up being quite high, such that the preview is very noisy but that doesn't matter much as stacking a few thousand frames will remove that noise.

Use 8-bit capture resolution, and if your camera has a high speed mode then select that as it uses a lower bit conversion rate like 10 bits which is fine if you're capturing in 8 bit. Don't enable gamma correction or debayer during capture to maximize fps. Most capture programs by default, do disable debayering during capture, although the screen preview can still be seen debayered.

A two minute video at 200fps will give you 24,000 frames and stacking the best 20 to 30% will give around 5000 frames or more which should be enough to allow a fair amount of processing before it starts looking noisy.

Alan

[neil phillips]

39 minutes ago, symmetal said:

What's your exposure yelzac? It needs to be around 5mS to help freeze the seeing conditions and to allow a high framerate. 5mS should give a framerate of 200fps at your ROI, ( 1 / 0.005) assuming your camera and capture computer can cope with that. Increase the camera gain to achieve around 60-70% on the histogram. The gain may end up being quite high, such that the preview is very noisy but that doesn't matter much as stacking a few thousand frames will remove that noise.

Use 8-bit capture resolution, and if your camera has a high speed mode then select that as it uses a lower bit conversion rate like 10 bits which is fine if you're capturing in 8 bit. Don't enable gamma correction or debayer during capture to maximize fps. Most capture programs by default, do disable debayering during capture, although the screen preview can still be seen debayered.

A two minute video at 200fps will give you 24,000 frames and stacking the best 20 to 30% will give around 5000 frames or more which should be enough to allow a fair amount of processing before it starts looking noisy.

Alan

6 minute video de rotation is way better than a two min video. Also 5000 or more frames isn't always enough. Especially when a lot of post processing sharpening is used. Hence de rotation

 

Edited October 16, 2022 by neil phillips

[neil phillips]

32 minutes ago, yelsac said:

Woooow a lot more! Do you use this IR890 CH4 filter that you can get with the 462C?

I Have IR850. But its overkill. It lowers resolution too much. Not something I've found particularly useful 

[yelsac]

52 minutes ago, symmetal said:

What's your exposure yelzac? It needs to be around 5mS to help freeze the seeing conditions and to allow a high framerate. 5mS should give a framerate of 200fps at your ROI, ( 1 / 0.005) assuming your camera and capture computer can cope with that. Increase the camera gain to achieve around 60-70% on the histogram. The gain may end up being quite high, such that the preview is very noisy but that doesn't matter much as stacking a few thousand frames will remove that noise.

Use 8-bit capture resolution, and if your camera has a high speed mode then select that as it uses a lower bit conversion rate like 10 bits which is fine if you're capturing in 8 bit. Don't enable gamma correction or debayer during capture to maximize fps. Most capture programs by default, do disable debayering during capture, although the screen preview can still be seen debayered.

A two minute video at 200fps will give you 24,000 frames and stacking the best 20 to 30% will give around 5000 frames or more which should be enough to allow a fair amount of processing before it starts looking noisy.

Alan

Thanks for your reply Alan

Yes my exposure is at 5ms but the maximum FPS is around 68, I can get it up to 120fps if the video size is 320x240 but that's too small for me to capture.

Appreciate your other thoughts 👍

[yelsac]

33 minutes ago, neil phillips said:

6 minute video de rotation is way better than a two min video. Also 5000 or more frames isn't always enough. Especially when a lot of post processing sharpening is used. Hence de rotation

 

Must have a go at winjupos, next session I'll try a 6min vid 👍

[neil phillips]

7 minutes ago, yelsac said:

Must have a go at winjupos, next session I'll try a 6min vid 👍

Its not just about noise either. As 70.000 frames can be run through pipp quality sorting. Which is cherry picking. Improving the overall quality of a capture by reducing large frame counts into lower frame counts. higher quality assortment.

Image de rotation also works. Even using images captured from video de rotation. Or just straight image de rotation

[yelsac]

5 minutes ago, neil phillips said:

Its not just about noise either. As 70.000 frames can be run through pipp quality sorting. Which is cherry picking. Improving the overall quality of a capture by reducing large frame counts into lower frame counts. higher quality assortment.

Image de rotation also works. Even using images captured from video de rotation. Or just straight image de rotation

Interesting so which is better de-rotating an image or a video?

[neil phillips]

5 hours ago, yelsac said:

Interesting so which is better de-rotating an image or a video?

Thats Debateable. Both works well, image de rotation might be the ultimate. Craigs technique of multiple 90 second captures work well.  But i would suggest so is what i am doing. I just prefer to not muck about so much. I find one image measurement and waiting for a new de rotated vid. Is pretty easy. Plenty of tutes out there if you get stuck just put a post up, 

[geoflewis]

On 16/10/2022 at 22:29, yelsac said:

Interesting so which is better de-rotating an image or a video?

I've been doing a bit of both. Using PIPP to create a reduced quality graded SER, which I de-rotate in WJ, then I take a few (typically 3) finished TIFFs from each SER back into WJ to de-rotate and stack those, with a final light wavelets in Registax. I'm still refining my method with this new colour camera, but I'm liking the combination of longer (6 min) SERS, quality trimmed through PIPP and de-rotated in WJ, rather than lots of 1 min mono RGB SERs. Probably Mono RGB with multi short SERS each stacked then deprotated will give the best results, but it's a lot of work for small gains and IMHO not worth it unless the seeing is spectacular and long lasting... Yeah right, this is the UK....🙄

Edited October 18, 2022 by geoflewis

[Eris]

Hi,

I thought I would have a crack at your original image and see if I could denoise it a little.  I put it through Registaks again and applied some gentle wavelets and then another with a little denoise on channel 1 of the wavelets.  Put them both in to Photoshop and blended a little. 

  

[yelsac]

14 hours ago, Eris said:

Hi,

I thought I would have a crack at your original image and see if I could denoise it a little.  I put it through Registaks again and applied some gentle wavelets and then another with a little denoise on channel 1 of the wavelets.  Put them both in to Photoshop and blended a little. 

  

Hi Eris

Appreciate you having a go, it does look a little less noisy than mine 👍

[yelsac]

On 17/10/2022 at 04:27, neil phillips said:

Thats Debateable. Both works well, image de rotation might be the ultimate. Craigs technique of multiple 90 second captures work well.  But i would suggest so is what i am doing. I just prefer to not muck about so much. I find one image measurement and waiting for a new de rotated vid. Is pretty easy. Plenty of tutes out there if you get stuck just put a post up, 

Quick question Neil

I've got a 6min Avi which I'm trying to put through winjupos. I've put it through pipp & quality which has reduced it a bit. Then in WJ I've selected the de-rotate of video, entered the start & end time but no idea how to get the image measurement of the Avi, do you have to stack it first to get an image?

[geoflewis]

5 minutes ago, yelsac said:

Quick question Neil

I've got a 6min Avi which I'm trying to put through winjupos. I've put it through pipp & quality which has reduced it a bit. Then in WJ I've selected the de-rotate of video, entered the start & end time but no idea how to get the image measurement of the Avi, do you have to stack it first to get an image?

I'm sure that Neil will chime in, but yes, you need to produce an image through AS3! (or similar) which is then used for image measurement against which WinJupos will de-rotate the SER.

[yelsac]

Great

So am I right in saying, when I open the stacked image & enter the date/time & my long/lat then save, is that it? Or do I have to do more?

[geoflewis]

2 minutes ago, yelsac said:

Great

So am I right in saying, when I open the stacked image & enter the date/time & my long/lat then save, is that it? Or do I have to do more?

Yes, set your location, which I have saved as a site in WinJupos that I can pick from the drop down list. I produce WinJupos naming convention ready SERs and TIFFs (you can set this at capture in FireCapture, or in PIPP), so the date and time is embedded and automatically read by WinJupos. I use the TIFF for image measurement, then load that image measurement file into the appropriate field in the derotation of video frames tab, with the matching SER to be de-rotated.

[neil phillips]

3 hours ago, yelsac said:

Quick question Neil

I've got a 6min Avi which I'm trying to put through winjupos. I've put it through pipp & quality which has reduced it a bit. Then in WJ I've selected the de-rotate of video, entered the start & end time but no idea how to get the image measurement of the Avi, do you have to stack it first to get an image?

Sorry just saw this. You're in good hands with Geof btw. Hes both a quick learner. And very good at refining any methods discussed. So might as well join in just to add perspective. Yes, what I do is stack the 6 min vid, as per normal. It doesn't hurt, just to do 50% stack size. I don't believe the features are used for de rotating. More outline orientation and time and date are used so blur on the image measurement isn't a problem. So once the stack is done, open in image measurement. In the de rotation vid page, all your times should pop up automatically (after pipp). And it will then give you the reference time which will be the exact centre of the film. So take that date and centre time. and put it on the image measurement.

here

If you get bad edges, try lowering lD Value try different amounts depending how bad it is. Sometimes you may need PS Gimp ect if it doesn't work well. start with low amounts first

[neil phillips]

To do the measurement go to outline frame 

here, then go back to image and save in the same folder as the avi.  image ( just before ADJ) , 

Edited October 21, 2022 by neil phillips

[Pete Presland]

Very nicely captured. Very interesting to see the WinJupos posts as well.