DSS processing speed

[Tinker1947]

I use, a Due Core 2.80GHz, 8 Meg of Ram, 64 bit Windows 7, will run DSS quite fast 20 Subs, Darks , Flats and Bias files takes less then 10 minutes, DSS also reports using 2 Processors .

[Gina]

With just a small handful of stars it only selected one frame

It's now stacking. It selected 30 of the 46 frames to stack.

[Gina]

Nearly done

[Tinker1947]

Will there be a picture...

[Gina]

Doesn't look like it!

DSS produced an Autosave.002 file with M42 in it but I'm having a problem getting anything to read it. None of my Linux image viewers will read it. The Mac reads it but it is all black. The single frame I processed at first does show M42 (very poorly) on the Mac. I thought the Canon software, DPP would read TIFF files, but no it just shows a rectangle with a cross in it.

I've just reloaded it into DSS and there is a reasonable image of M42. Now to find out how to get it into something I can post...

[dph1nm]

You need to save the file from DSS as a tiff (see the menu on the left of the DSS window) - the Autosave is an internal DSS format and not meant to be read by anything else.

NigelM

[martin_h]

The Autosave is in 32 bit format. On the left frame of DSS - under Processing, load that frame back in, then under the same heading...Save picture to file......

Select whatever format your software will read...16 bit Tiff or Fits

[Gina]

I'm NOT impressed! The result of stacking 30 images in DSS is far worse than any of the input images!! Here's the result saved as TIFF file and opened in GIMP to scale to fit on here.

[martin_h]

Cant see it!, but it will be dark, hence the need for stretching.

[Gina]

Hmm...it's disappeared

I'll try again.

[Gina]

And cropped and stretched... I didn't remove the colour - DSS did that. The stars have grown considerably in size too. I presume I'm doing something wrong.

This image reminds me of a Muppet Or a PacMan with several eyes.

[Spikey]

That doesn't look right Gina, have you got x2 or x3 drizzle checked? The saturation needs boosting up to about 14-15% to start seeing the colour.

Can you post up one of your subs ?

[dph1nm]

Here's the result saved as TIFF file and opened in GIMP to scale to fit on here.

If you are processing in GIMP you *must* adjust the image in DSS image so it looks good (RGB, Luminance, Saturation)on the screen, then save the tiff with "Apply setting to saved picture" checked.

This is because GIMP scales the 16bit DSS image to 8bits by dividing all the number by 256, so you just end up with a very dark image if you don't do this.

NigelM

[Gina]

If you are processing in GIMP you *must* adjust the image in DSS image so it looks good (RGB, Luminance, Saturation)on the screen, then save the tiff with "Apply setting to saved picture" checked.

This is because GIMP scales the 16bit DSS image to 8bits by dividing all the number by 256, so you just end up with a very dark image if you don't do this.

NigelM

Ah yes I've had another DSS run with a smaller set of images (6) from which it chose 3 to stack. I then adjusted the luminance to brighten it up and saved it as a TIFF file with adjustments applied. I opened this file in ImageJ and it looked very much better than before complete with colour. I then saved this as a JPG to open in GIMP to crop and scale for posting here. GIMP read it as a greyscale image!!

I'll have another look at ImageJ (as this is the first time I've tried it and it's rather strange) and see if it will crop and scale the image and save in a suitable format.

I realise that the camera uses only 14 bits of data and therefore the RAW file data wants multiplying by 4 to fill the 16 bit range of the format.

[malc-c]

Gina,

I think you should be getting better images that that. I've just stacked two subs totaling 6 mins - no darks, no flats etc.

The raw result is shown in image one, then by moving the sliders to place the RGB peaks in the curve of the luminescence graph and clicked apply the image was cleaned up (see image 2). These were raw subs from my 400D.

You can then select the "Save picture to file" and select whatever format you wish

[malc-c]

One thing I noticed in your original screen capture is the lack of data in the peaks in the histogram and that the luminescence line is linear rather than a curve. What exposure and duration are your subs ?

[Gina]

Thanks Malcolm I get the technique now

I took a number of images at 2, 5, 10 and 15 seconds at ISO 6400 plus a few darks. At more than 15 secs the tracking failed. I think the original DSS stack was the 5 sec set of images.

[Gina]

I'm now running DSS on 24 10sec images (ISO 6400). Threshold set to 7% with SS finding 3 - 6 stars in each image. I think this should be good.

[Gina]

It wasn't - it only accepted one frame to stack!!!! OK try again Let's see how 5% works.... Talk about trial and error

5 secs later... NOPE! Only one frame - 4% then... 400-500 stars.

[malc-c]

No expert, but subs that short simply won't have enough data, and at that ISO there will be a lot of noise, and it's this that's causing the poor results. Why is the tracking failing after 15 seconds - your EQ6 should allow you to get at least 60s - 90s before it needs guided exposures. Sounds like your PA is way off if you get trails after 15 seconds..

Those subs I used were guided, 200s @ 800 ISO. I think that once you start gathering data at these sort of settings then DSS will come into its own

[malc-c]

NOPE! Only one frame - .

Frame... are you "capturing" live view type avi and then trying to stack ??

Gina, may I suggest that you iron out the issues you have, such as polar alignment, ATP's control of the shutter, guiding etc and then once everything is hunky dory, then start gathering data with 3 - 5 mins guided subs at 800 Iso. Given those lovely dark skies you must get down your way you should get a lot of good data in 3 mins

[Gina]

I've got to work on my PA - I'm sure it's out. But when I try to use the polar scope I can't see any stars. May try drift alignment. Next week, next month... whenever the stars appear again

So, lower ISO and longer subs then... I have some taken a few days ago at ISO 1600. Like this.

[malc-c]

Gina,

I've managed to replicate a similar result with just my two stacked 200 sec exposures - by shifting the mid tones for the illumination. But my guess is that your problem is simply lack of date to stretch.

[malc-c]

Gina, when you say you can't see any stars in the ploarscope, have you rotated the DEC axis so that the hole in the DEC shaft lines up with the polar scope... (not trying to state the obvious) - I would of thought you could see Polaris from the observatory position ?

[malc-c]

That 1600 sub still looks really grainy to me... has it been zoomed and cropped ?