PIPP

[jambouk]

What is the advantage of using this software on AVIs prior to running them through Registax? I hear people talking about it "derotating" the image, but I don't know what aspect of the image it is derotating; if using an equatorial mount there won't be much field rotation, and i don't see how it can resolve the rotation of the planet around its axis.

Thanks for any replies.

James

[knobby]

Only advantage I know if is that it cuts off the unused background so Registax or Auto stakkert can process the frames faster...

[Freddie]

I don't think PIPP derotates (could be wrong) but Winjupos does. Its not derotating due to mount movement, its derotating the planet rotation expected during the exposure. Very important for fine detail on Jupiter. Less so on Saturn but still very useful particularly with large image scale or for RGB imaging.

[Ewan]

As Freddie explained, PIPP as far as i know is really used by most just to select the best frames & order them from best > worst, if you load an AVI file that has 5000 you can tell PIPP to only compile the best 1000 frames ONLY into a new avi file, then when you drop this into AS!2 or Reg6 you can either select all frames to stack OR you can cull frames even further.

As i find AS!2 better than Registax i use that, i have not really seen any improvement in finished output using PIPP then AS!2, to be honest it may make your prefered stacking software quicker in the end when stacking BUT you end up using two programs, i find it quicker overall to only use AS!2 now.

PIPP can do a lot more other than only frame selection but it's all i ever tried, debayering, resize & splitting RGB frames to name but a few, it is very good software for what it can do but trial & error, someone with more insight may not agree but each to their own.

[JamesF]

PIPP can re-order your frames, be they from an AVI file or individual images, by quality. It can also crop the frame down and leave the image centred. Almost as a side-effect it can convert from many DSLR raw file formats to TIFF. The first is helpful when you have, say, a set of solar or lunar disc images that have some cloud in. Those will appear at the end of perhaps 100 to 200 frames and can easily be left out of the stacking process. The second is more useful when using Registax, particularly v6, which gets a bit distressed when images move about in the frame too much. AS!2 seems to cope with that better. The conversion of raw to TIFF is very useful for lunar and solar disc images because neither Registax nor AS!2 can cope with raw DSLR format images.

It does plenty of other things too, but I'd guess those are what people use it for most.

James

[jambouk]

Thanks Ewan, that is just what I thought it was doing. Registax is a bit slow at times, but as you say, it also allows me to decide to only stack the best 1000 frames of 5000 frames, so I'm not convinced PIPP is needed by me at the moment.

Freddie, how does it derotate the spin of the planet? Surely the data captured is the data captured. If one images Jupiter for 10 minutes, there will be a load of rotation, and the surface detail will be blurred if one stacks all 10 minutes of footage. How can PIPP prevent this, other than selecting one continuous stretch of footage which is say 2 minutes long from that overall 10 minute footage, and then inserting that 2 minute stretch into the stacking software?

Thanks all for the replies. Really useful stuff.

James

[JamesF]

Ah, should have added that cropping the image down also makes subsequent processing faster because there's less work to do. For instance, cropping a 640x480 image down to 200x200 (quite possible with an image of Saturn, say) reduces the number of pixels in the image by a factor of just under 8. That will speed up processing a fair bit.

James

[jambouk]

That's true James. Next time I plan to try and image Saturn with a smaller frame than 640x480, as that should also allow me to increase the frame rate a bit.

James

[Freddie]

James, there are a couple of things here to think about. For an AVI it is not the duration that particularly causes the blurring (though it will contribute) as a 10 min AVI will be made up of many thousands of individual images each with an exposure of a fraction of a second. Each individual frame is therefore sharp. The stacking prog will then look for common features on each frame and stack them together to align the features. There is therefore some element of a derotation effect in any stacking prog. Chris (CGary) will be able to explain this much better than I ever could as he wrote PIPP.

For RGB imaging, you then end up with 3 nice sharp images from 3 seperate AVIs but the features are not perfectly aligned as the 3 AVIs may be taken over a 15-20 min period. You therefore need to be able to align these images. This is where WinJupos comes in. You tell it the time each AVI was taken and it will calculate the rotation of the planet over that period and align the 3 AVI files for you. Winjupos and also work on colour AVIs but I don't use that feature as all my images are RGB.

Hope that helps.

[Ewan]

I may end up eating my words from my above post, here is a straight comparison of a Saturn Red channel processed. You have to appreciate the times included selecting the programs, loading etc & settings needed in each case, no post processing was done to any image by me.

These results, can be said, do not prove anything it was just a simple test & i was proved wrong that AS!2 would be quicker, this was really raw & not what i would do to process my images normally, at the end of the day to get the most from your images you would use all the time it needed if you wanted the best possible result.

All frames percentages were based on 20% of 7737 frames.

2 Minutes 38 seconds

2 minutes 45 seconds

3 minutes 3 seconds

4 minutes 11 seconds

Freddie is quite correct about the Derotation, say for argument sake you did shoot a 10 min avi, your sharpest frame was the 1st & last frame, these being 10 mins apart, PIPP would select these (as i expect AS!2 would also) & add them to the stack so yes i guess Derotation would in effect happen but i would really use WinJupos for derotating any images as thats what it is designed for.

[jambouk]

Ewan, I'm not sure what the times represent, or what the experiment is testing / showing. Sorry, I'm being thick as usual.

Freddie, I'm also still not getting the derotation. Again, I do appologise for being thick. If, say, the great red spot (GRP) starts on the left of the planet at the start of the AVI and ends up on the right side (I'm massively exaggerating the situation to try and simplify it for me), how can the programme derotate this? It can't unrotate the planet to put the GRS back on the left as it would need to make up data about what then appeared on the right of the planet, unless it just duplicated the images at the start of the AVI, which doesn't add new data, just duplicates the existing data.

Sorry both. Maybe Sunday Lunch will clarify my head!

I do appreciate you taking the time to try and explain this to me.

James

[jambouk]

I use a colour camera as one shot colour. So I only end up with one AVI.

James

[cgarry]

Hi James,

PIPP definitely does not do any kind of "derotation", not for field rotation nor planet rotation. So I will not add to the derotation discussion which is a much more complex issue and handled by WinJUPOS.

PIPP is just as a processing toolbox that works on raw frames before stacking. The idea is simply to give the user access and control of the captured frames whatever the format they happen to be in. Many of the functions PIPP performs can be performed in other, often non-astro related software. PIPP aims to put all these functions into one program that is written as an astro application from the ground up.

This would be a long post if I tried to cover all the possible ways of using PIPP, so I will go over the main ones that I use:

Quality Selection

This can either be used to remove a lot of the lower quality frames and still let the stacking software choose the best quality frames for stacking (the safest approach) or let PIPP do all the quality selection and make the stacking software stack 100% of the frames. There is also a quality weighting function that can make a subtle improvement to the final result in some circumstances.

One thing worth noting is the quality selection works by analysing the contrast differences on the planet surface. Therefore if the planet is very small in the video, just a few pixels across, then the quality detection will not be very accurate.

Histogram Stretching

For monochrome frames this ensures that each frame is of equal brightness and this can help the stacking software with alignment, especially RegiStax in my experience. For imaging with RGB filters it means that each the resulting videos are the same brightness making it easier to get a consistent colour balance after stacking.

For colour frames it has the same advantages in potentially helping the stacking software with alignment but also gives a consistent colour balance to the output AVIs by equalising the RGB channels.

Centring and Cropping

This function is an obvious one and definitely helps RegiStax more than AS!2. As well as centring and cropping the frames, this function also discards frames where the object is not completely in the frame.

Format Conversion

This can be very useful if somebody finds that they have used a video format for their captures that is not supported by the stacking software. PIPP will happily convert it to a format the the stacking software will support.

PIPP will also convert video files to AVI files using the Ut Video lossless codec. This is great for those who like to archive their raw capture files as it reduces the file sizes to between 25% and 50% of the raw files sizes but does not result in any loss of data. These AVI files can also be processed directly with RegiStax is the Ut Video codec is installed.

My thoughts on using PIPP are that it can perform its processing on all captures from a session in batch mode while I am sleeping, leaving just the lean frames to be processed while I am actually sitting in front of the computer.

I think the use of PIPP breaks down into the following groups:

1. Users who always use PIPP as part of their processing workflow as I do.
   
2. Users who only use PIPP with awkward data that they are having trouble getting the stacking software to work with.
   
3. Users who experiment with PIPP to try to squeeze a tiny bit more detail out of their best captures - the reprocess mob.
   
4. Those who do not use PIPP at all.
   

All equally valid approaches in my mine.

Cheers,

Chris

[Ewan]

if you hover your mouse over the pics more information will show.

Appologies if my post was lacking more information, all the examples were showing is that it really is upto the individual as to how long they want to use obtaining their stacked image, as you can see Registax is fast but not very accurate imo, using PIPP & then AS!2 or just AS!2 only produced what i would have expected, a final image pretty much the same.

In AS!2 it can produce an already sharpened preview image, which is saved, you can use this image to refine more with wavelets if you find sharpening from a 'fuzzy' image a little difficult in wavelets, AS!2 also lets you change the amount of RAW data that is merged into a sharpened image if you want.

As for the Derotate question i understand what you mean, it would seem that all information on the left side of the planet would be inhanced & all data on the right, as it spins out of view, would be as existing & no new data can be collected, yes you are correct to a point.

The most i ever used Derotate on was a 6 minute avi & in that time the rotation is not extreme as all, it's very subtle & hardly noticeable, the best example is if you also have one of Jupiter's moon's in shot at the same time, you will see without Derotate the moon will be slightly oval, with Derotate in will be a lot rounder, this indicates that WinJupos has indeed done it's job & produces a cleaner sharper image.

As i have already said though james, the processing time & programs used really is dependant on what sort of image, from your data, you feel you can end up with. Bad data in bad data out.

As for the OSC you can still shoot 2-3 avi's & use Winjupo's if you like, i guess if you shoot 1 x 18 minute avi you have a chance of cloud, planes, power loss etc but shooting 3 x 6 minute avi's cuts down the risk but again it's up to you.

Astronomy is really a lot about trial & error, your location, equipment & skill level are not going to be identical to anyone else's so we can only advise what we have tried ourselves, it is up to you to take the advice or ignore it & persevere.

All the best James & clear skies.

[jambouk]

Right, that is helpful. Straight from the horses mouth; not that I am liking you to a horse!

I've just run an old AVI through PIPP and limited to the best 50% frames, then run that file through Registax6. It was quicker to process, and the resultant image is nice. I'd need to do my own comparison of the various combinations to look at image results. Cutting a few minutes off the process isn't that important to me, but having Registax crashing when I dump a big file on it is. PIPP seems easily ish to use (I need to learn how to make it crop the image).

I still have no idea how any software can 'derotate' the planetary spin, that might have to be something to learn about another day....

Cheers all and thanks for being patient. I'm sure there will be more questions as time goes by.

I do appreciate it.

James

[Freddie]

No problem James. It's all a learning curve and SGL is a good place to learn.

I don't claim to fully understand exactly how winJupos works but it obviously uses the time the pics were captured to work out by how many degrees the planet will have rotated between pics. It then "rotates" each pic by the required amount to ensure all the features are aligned. Over a long timescale you can see some detail missing on the limbs as you rightly say it can't generate data you have not captured. You will always have some data for every part of the final image, its just that you will have less data for the limb that is rotating out of view, the middle will have a full set of data as it is always in view and again less data for the limb that is rotating into view. Hope that makes sense.

Give WinJupos a go next time you capture some colour AVIs and you will probably get a better understanding of what it is doing. I know Ewan and Stuart (spacecowboy) use the colour AVI processing options so have a look at their posts for what they have done. I only use it for RGB processing.

[Ewan]

Not me :-), i only use the DMK""618 + filters, i do have a Toucam Pro II but do not use for anything now.

Agreed that the centre of the image will only have the most detail but it really is worth using Winjupos for longer sets.