Registax 5 Problem

[Radar]

Was shooting Mars the other night, any time I did an avi with 3000 frames the program would crash at the end of the avi. Could only do 2000 frames.

Any ideas what would cause it? Ram?

Cheers

Ray

[timthelder]

Possibly running short on RAM. I think my lappy has 2gig. I had some problems with R5 with a large # of frames, but using the extended mode(see the boxes in the top task bar where it says default or vista and the "use extended mode for" box) I have "always" checked, and for over 3469 frames I specify either 1>gig or 2>gig.

Maybe that'll help.

[Kokatha man]

Was shooting Mars the other night, any time I did an avi with 3000 frames the program would crash at the end of the avi. Could only do 2000 frames.

Any ideas what would cause it? Ram?

Cheers

Ray

Hi Ray - here's a screenshot of how a Mars processing in Reggie5 can look.....I regularly have been processing 12600 frames in him this apparition.

Check top almost centre for what Tim is talking about.....I'll get that tute running asap:smile:

[Radar]

Thanks guys, and sorry, I must have been in lala land when I posted. It is IC capture that won't catch the frames. Maybe there is a setting in there that requires something. I might even try and shut down all other programs in case it is a case of low ram. I will let you know if I find a fix.

Ray.

[Kokatha man]

.....if it's the "IC Cature has encountered a problem....." one Ray, then it is a b#@*$! of a glitch that it throws up periodically!

What happens (if it is this) is that the file headers aren't written properly (these are done at the end of each avi file) and render the avi unplayable when you try to do so in Windows Media etc.....a fella in the UK was going to assist me to rectify the problem, but I haven't got back to him since the Xmas break.....

Another annoying aspect of this particular problem is that you virtually have to use the power button on your pc/lappy to switch off - nothing responds after this error box comes up.....restarting the pc and IC Capture usually fixes the problem - although sometimes it just keeps on reoccurring....!

Having said all that, I hope that it wasn't said.....but at least you'll know what it is if it pulls this stunt on you in the future anytime....!:biggrin:

[Radar]

I think it is this problem Darryl. Because the AVI records perfectly and then when it hits 3000, which is the frame I have set, IC capture comes up with that message. I'll google and see if there is a fix for this yet.

Thanks for the help.

Ray

[Kokatha man]

Hi Ray - sorry, no "fix" per se for that one from what I've seen so far on the net.....as I said, it is the failure to write the header for the file that allows the system to determine what it is and run it: you can find my correspondence on the matter by doing a Google on "replacing faulty avi file headers"

So far, all the other fixes I've seen seem to be for avi's that AREN'T Y800 codec jobs.....:confused:

I'll get back to the chap I mentioned (Adam) who was going to run a sharetime session with me on my pc to step me through his program.....from my own perspective this particular problem has been a real b$#@*! because, as everybody knows from Murphy's Law, it's when you close a really first-rate avi capture that the error occurs....!

I've lost some excellent Mars avi's lately - and it allways seems that it's the one you manage in those fleeting periods of best seeing....!:eek::Oh_No::crazy::crazy::crazy:

[Kokatha man]

ps: you might also try just running an avi manually without presetting the stop time.....just manually clicking the stop button when you want to finish.....it may stop the error coming up so frequently....!:confused:

btw, you can run an avi capture for Mars for up to around 7 minutes or more.....

[Radar]

Cheers mate, I will try that.

Ray

[Radar]

Just an update. I have been succesful in getting some whopper AVI's of Mars. 7 and 8 thousand frames. I did a couple of things to the computer and it seemed to help. First I closed all other programs down. I also didn't connect starry night to the telescope. But what I did do (and I think this is the solution), is turn off all the excess windows during the capture of the AVI. For example histogram. Once it is set, I close it. These changes are making IC Capture work nicely.

Now, next issue, how to process large AVI's? Anyone? Please? :confused::biggrin::eek:

[Kokatha man]

Up the top in Reggie 5 where it says "Use Extended mode for" have the "Always" radio button clicked and on the LHS of these buttons "Extended Mode" on default button clicked Ray.....then just go ahead - I process up to 13000 frames with the DBK that way.....:biggrin:

[Radar]

Cheers mate. When I get to the optimize and stack section, whether I optimize, or optimize and stack or create reference frame the image turns back into black and white. The Use Debayer is still ticked. Any ideas welcome.

Ray

[Kokatha man]

I'm presuming you're capturing in Y800 codec Ray (you should've been!:biggrin:)

It sounds like you are using the older version of Reggie 5 if that is happening.....the newer version (which you can get off his website) sorts that problem out if you check the boxes as per my last post: Cor fixed the loss of colour bug out after extensive consultation with yours truly....!:blink::Chessy_Smile::tongue:

ps: try using a reference frame that is "average" - rather than one of the best you can find when you hand-pick at the start.....my initial trials seem to suggest from the various graphs along the way that alignment and Q difference are noticeably enhanced.....and go for "Centre of Gravity" alignment which dispenses with having to determine an appropriate FFT value.....:smile:

[Radar]

Cheers mate. That fixed it. I'm still processing. Will post soon (hopefully).

Ray

[Radar]

I haven't finished this processing yet but the missus wants to watch a DVD.

This is what I have for now. Thanks for the tips mate.

3263

[Kokatha man]

That's looking promising Ray: being a view of what is popularly known as "The Boring Side" or face of Mars there aren't a lot of features/albedo markings, apart from those around the N & S Polar areas to capture.....but you have essentially caught what is there to see, and importantly a number of cloud formations (those white patches/spots on the disk) that have been on show of late, which (Rob)clayton informs me are from the accelerated "melt" of the NPR.....:thumbsupmate:

I suspect that quite a bit more could be wringed out of this image without making it over-harsh.....has any wavelet panel filtering and/or deconvolution been done to this image.....and presuming it is captured with the OSC did you do a split into RGB channels after final wavelet application in Reggie to work on each channel seperately in any other (eg AstraImage, etc) post-processing....?:confused: This certainly assists maximising each channel's resolution etc, and allows you to bring out the clouds to best when working on the blue channel.....

I'm looking forward to seeing what else comes out of your stack of avi's - get anymore captures of 'em while you can Ray, Mars is fading and shrinking fast now! :frown: (This weekend could well be my last, or second-to-last Mars imaging sessions.....and if the weather is half-decent I'll be getting a view of the more interesting aspect/face of The Red Planet here in Sth. Oz!:smile:)

[Radar]

has any wavelet panel filtering and/or deconvolution been done to this image....

Not yet. Only in Regsitax. I do have the astra image pluggin on photoshop and I'm about to post another version that has had wavelets from astra image, and then noise removal from noise ninja (also a pluggin).

No deconvolution as of yet, not sure what it does, but it is in the astra image pluggin.


and presuming it is captured with the OSC did you do a split into RGB channels after final wavelet application in Reggie to work on each channel seperately in any other (eg AstraImage, etc) post-processing....?

No splitting as of yet mate. Would love to learn this part to as well. When I get some spare time I'm going to crawl over Registax 5 with a fine tooth comb.


get anymore captures of 'em while you can Ray, Mars is fading and shrinking fast now! (This weekend could well be my last, or second-to-last Mars imaging sessions.....and if the weather is half-decent I'll be getting a view of the more interesting aspect/face of The Red Planet here in Sth. Oz!)

This weekend I'll be shooting my final AVI's of the martian surface.

Ray

[Kokatha man]

.....I will get something posted soon Ray, I've been in a bit of limbo-land since the old fella passed away just under 3 weeks ago.....am caught up with lots of paperwork etc but am slowly applying a bit of AA therapy (made a solar film filter for the 5" apo and its finder out of 6" and 2" pvc pipe caps yesty).....also trying to get the mess in my studio together (as well as my head....!:smile:)

I'm not sure about Noise Ninja for planetary images.....I haven't been able to use it myself (I have the full version) - but if you can use it to decrease noise without diminishing the image clarity/quality that's something you'll be able to point me towards!

Deconvolution works wonders on Mars particularly (I'd almost say it is an absolute essential) and the basics of it is that using either Lucy Richardson (more aggressive) or Maximum Entropy (not as "effective" but more gentle) types of deconvolution employs those ubiquitous algorithms to determine/predict/enhance minute subtle image data and (seemingly) create something out of b&##@! all previously visible....!

Depending upon a lot of factors (that I haven't a clue about!!!:eek::biggrin:) the "curve radius" is a value you select that when applied a specific number of times (the "iterations" you select) enhances the image.....selecting 1.5 as a value for the curve radius is a good starting figure, and anywhere between 5 & 10 for the iterations.....you'll quickly appreciate if you've been too savage with your values - if 1.5 doesn't work up the value to around 2.0 - then 2.5 or so if need be.....and if any of a range from say 1.2 to 3.5 (depending upon if you're using L-R or ME) seems too aggressive then drop the iterations to 5 or 6 in value.

Many people do an initial L-R decon job and then resize their image and apply another L-R application.....I prefer the milder ME decon for my 2nd application if I do it that way.....and you will find that for any second application you'll likely need to lower the "curve radius" value significantly.....some signs of over-agressive application (ie, incorrect curve radius or number of reiterations) are images where contrast differences just go "throught the roof" (really harsh, hard-edged definition which is obviously OTT) or coming back the other way, where there begins to get a grainy aspect to the image noticeable in the lighter areas of the disk.....you'll find this region of decon values gives you much more gentle, balanced contrast within the various planetary detail areas and more delicate, finer resolution of detail overall.....

The trick is to try and remove all of the noisy, grainy stuff (I term it "measles-like appearance") by selecting combinations of curve radius/width and iterations values that reveals detail, doesn't bung in too much noise/grain and doesn't overdo contrast to produce a false, hard-edged and over-emphasized dark detail product.....

Good luck with all this.....it is a real art and takes lots of practise and revision to get what you want - I'm still experimenting, it makes Jupiter look like "easy street" in comparison:eek::wink:).....here's a link to a Dutch fella who got a cracker balance in processing on a pik he posted on CN recently.....you'll get the idea from it!

http://www.cloudynights.com/ubbthreads/showflat.php/Cat/0/Number/3620905/page/0/view/collapsed/sb/5/o/all/fpart/1

Of course, a much simpler method to boost the detail resolution/contrast (which you can use in conjunction with deconvolution but needs a lot of care balancing the 2) is when you are in the final wavelet section of Reggie 5 after you have stacked.....click on the "wavelet filter" button amongst those on the upper right of screen listed under "Functions" and when this window opens, try increasing the "initial" and/or the "step" values in their little boxes at the bottom left of this window panel and watch your image develope in front of your eyes....! Again, you have to watch for OTT aplications, but it's much easier than deconvolution first up.....:biggrin:

[Radar]

Didn't get time to have a play with the image today. I'll hopefully try this tomorrow afternoon. I'm also going to attempt a few final AVI's of Mars tomorrow night.

BTW, I'm surprised by the result on my last image. I used an average frame and got a better result. How? I have no idea. :eek:

Cheers

Ray

[Kokatha man]

BTW, I'm surprised by the result on my last image. I used an average frame and got a better result. How? I have no idea. :eek:

Cheers

Ray

I made that suggestion above Ray because someone else had imparted it to me recently.....gave it a try meself and although I haven't run a proper comparative test by doing an exact same processing using a high definition reference frame (my normal practise) I was very impressed with the display graphs you can view in the Reggie stages when working with an "average" frame.....extremely uniform graphs.

The reasoning put to me for this method's efficiency is that Reggie does a lot better aligning the images (and sorting them I'm assuming) when it has a "run of the mill" quality frame in terms of generalised planetary features.....rather than trying to compare each frame to one you've picked that is clearly much better in detail than many/most.....

I want to bounce this and other queries off of Cor Berrevoets sometime to get a bit more of an intimate understanding of such things myself.....:wink:

[Asimov]

Yep, I've had this problem with Reggie 5 & Tim has nailed that particular problem.

As to IC capture; what a piece of crappy capture software it really is..Most times, I'll have to reboot as many as 8 times before it'll start up correctly. One of the many 'secrets' to this crappy program is to (as hinted at by KM) is to 'sneek up on it' - using it manually, but rather than hit start & stop, hit pause...

I could rave on about particular hair pulling problems with IC capture forever! LOL!