Atik HORIZON Histogram

[Marky1973]

Finally got back to imaging after doing very little this year while sorting out a new set-up - ED72/Evoguide/Atik HORIZON. - and got set up outside last night for a quick test to see how things were working together. After lots of fun with spacers etc I know have the camera able to focus and set up on the Soul Nebula for some Ha imaging. Guiding was good, 600 second subs were fine so away we went. I am using Atik's Artemis software.

The first preview came in on ARTEMIS and I was one target, the stars were focused (well-enough at this point, I'm just trying to make sure it works) although a bit stretchy in the corners, but I don't have the ED72 flattener yet.  I could definitely see the Soul in there. So I left it running for a while and then went back to check after an hour and to have a quick play with DSS to see what the stack would look like. 

However, I loaded the FITs subs into DSS and they were all just black and they wouldn't stack. Even dropping the star detection threshold to the lowest 2% yielded only one or two stars and even then DSS would only stack one image. If I "stack" that one image it can be stretched in DSS to show up some detail, but this obviously isn't quite right.

I loaded a single sub into GIMP and again it was black, I had to do some pretty aggressive stretching in curves and layers to get some detail out - but it was there.

Anyhoo, went back to Artemis and tried again, this time putting the gain setting on HIGH as recommended for narrowband imaging and ran another 10 minute sub to get the following in Artemis

Again there is detail in the preview but it loads up into DSS and GIMP as a black image - with my old CCD camera  there was always some detail in the individual frames, especially with a 10 minute sub. DSS did manage to detect more stars at the lowest star detection threshold but, again it didn't feel quite right that you woudl have to do this to be able to stack the images. It was 3am at this point and I didn't have time to run a series on subs with the gain on HIGH, so that will be the next experiment.

Looking at Artemis and the histogram, I then started to get confused. Previously I have always aimed for around a third from the left for the peak in the histogram but, with the Horizon, it seems to struggling to get off the starting blocks.

Shouldn't the histogram be further along for a 10 minute sub? Am I missing something obvious in the settings for Artemis or in the way CMOS cameras work? Please say I am as this was really frustrating - my old experience with CCD was very much plug and play, but this seems to be harder than it needs to be. Any thoughts, very welcome!

Cheers

Mark

PS - I wasn't sure if this was better in the Camera forum or the Software forum as it could be the camera or Artemis...or even in the non astro lounge as it could well be a case of a bad workman etc etc ?

 

[david_taurus83]

I'd try using another software like Sharpcap to rule out the camera. Can you post up a single sub and I'll open it in Pixinsight?

[Marky1973]

Hi David

Plan was to try Sharpcap but couldn't get it to connect last night - at 3am my technical skills start to wane! ? Will try it out next time I am out - just get on well with Artemis (or I have in the past) as it is so simple (I like simple)

I've uploaded two files - ignore the random file names I was just fiddling about.

The first is the 10 minute Ha exposure "as is" in Artemis and the second in the 10 minute Ha exposure with GAIN set to HIGH. There is definitely data in the images, I am just used to seeing something when opening a single sub in GIMP/DSS or PI in the past.

 

Capture_043.fit

15.fit

 

Looking at the second image in GIMP this morning (refreshed and with a level head) there is certainly more data in there than I thought and I suspect I could stack a number of those images to get something out...I am just confused by the Histogram - should I just ignore it?

Thank for taking a look - no rush but any thoughts would be appreciated

 

 

[david_taurus83]

Ok, so I had a look in Pixinsight. I'm no expert and I hope someone else who knows what their looking at will pipe in! I was out imaging in Ha myself last night but with a DSLR. I've extracted the red channel so working with with gray like you are. In image statistics my max ADU for my sub is almost the maximum at 65534 but yours is only 4094? Also, if I hover over a star in my my image the K value maxes out at 1.000 but in yours its only 0.0625. I'm guessing the K value is the pixel brightness? So if mine is at 100% then yours is only just over 6%? Sorry if my interpretation of those figures are wrong. Its like your exposure wasnt long enough or the gain was down too low but you'd think a 10 minute sub at low gain would still produce a decent image. Its strange because a STF stretch does reveal neboulosity. I also tried the star detection in DSS and it couldnt pick out any stars. Only thing I could suggest is to try a different capture program. I use APT for the DSLR and Sharpcap for my ASI120MM but I dont know if they will pick up Atik. I'd like to think Sharpcap does.

[Marky1973]

Thanks David. It is a bit odd...the nebulosity is there, but the ADU Level seems really low - I was imaging at around 21k with my old Atik 414EX CCD and was expecting something similar here. When I was using a DSLR a few years back it was again at about 1/3 over from the left. There is no way a 10 minute exposure wouldn't bring something in, so maybe I just need to ignore the histogram? With my levels, I am sure I wouldn't expect to see anything. Will have to experiment a bit more next time out and try Sharpcap/APT etc and see if that makes any difference!

Cheers

Mark

 

[spillage]

Have you tried APT? When going from my 500d to the 1600 (same chip) I too was surprised how low my histogram is in narrowband at unity (think unity is 139 gain) even at 900s.

[Marky1973]

Hi Mark

Sounds like a low histogram is normal then, perhaps? I haven't used APT since I stopped using my DSLR, so might have a look at that as well as SHARCAP to compare the histograms between software - although the rain now seems to have set in for the foreseeable future!

[newbie alert]

Following too as a I'm not getting what I thought I would from artemis with my Atik 16hr

I've just looked at your first image on the top of the thread (can't open files on your links) and I see you have the stretch ticked..on mine if I unclick the pre stretch I can adjust the black and white point on the sliders.. clicking on the pre-stretch puts it back to 0 on both..

[Marky1973]

14 minutes ago, newbie alert said:

Following too as a I'm not getting what I thought I would from artemis eith my Atik 16hr

In terms of the histogram or something else?

[RayD]

What gain are you using?  It may be worth trying a higher gain setting.  If the gain is too low then the stretched capture in Artemis will show, as it is, but the fits will be unstretched in DSS.

Just a thought.

 

Edit:  Sorry just saw you set gain to high.  It looks like a gain issue if the stretched image is showing some data but nothing in DSS (unstretched).

[Marky1973]

32 minutes ago, RayD said:

What gain are you using?  It may be worth trying a higher gain setting.  If the gain is too low then the stretched capture in Artemis will show, as it is, but the fits will be unstretched in DSS.

Just a thought.

 

Edit:  Sorry just saw you set gain to high.  It looks like a gain issue if the stretched image is showing some data but nothing in DSS (unstretched).

Hi Ray, yeah I had it on the high gain preset and it wasn't much better - I did set the gain higher manually but then started to get all sorts of amp glow and rubbish in the image - one extreme to the other..... what annoys me most is the need to experiment and the rubbish weather! ?

 

[RayD]

4 minutes ago, Marky1973 said:

Hi Ray, yeah I had it on the high gain preset and it wasn't much better - I did set the gain higher manually but then started to get all sorts of amp glow and rubbish in the image - one extreme to the other..... what annoys me most is the need to experiment and the rubbish weather! ?

 

That's really frustrating, Mark.  Especially as you are using the native capture software which, whilst pretty basic, is actually very good indeed.

Has this just started or is this a new set up and doing this from the kick off?

[Marky1973]

Just now, RayD said:

That's really frustrating, Mark.  Especially as you are using the native capture software which, whilst pretty basic, is actually very good indeed.

Has this just started or is this a new set up and doing this from the kick off?

Yeah, first night out with the Horizon. Was using an Atik 414EX before this and never had a problem with ARTEMIS - its a great bit of software and so easy to use - which is the main reason I like it! ? I went with the Horizon, even though it is more expensive than the similar ZWO1600 because I like the brand/service and the software - it works with INFINITY as well so want to have a play with that when I am properly up and running...I am sure it is probably an "expectations" thing and I just need to get used to working in a slightly different way.

[RayD]

Yes, as you rightly say, it is probably a learning curve and what settings to adjust etc. 

[newbie alert]

1 hour ago, Marky1973 said:

In terms of the histogram or something else?

No in terms of image quality..nowhere near my DSLR.. thought it be better..must be doing something wrong 

[Marky1973]

16 minutes ago, newbie alert said:

No in terms of image quality..nowhere near my DSLR.. thought it be better..must be doing something wrong 

Stupid question - is that after stacking/processing? The images that appear in ARTEMIS are autostretched to give a rough idea of what you are getting, but the final images after stacking and processing are always much better. 

[newbie alert]

1 hour ago, Marky1973 said:

Stupid question - is that after stacking/processing? The images that appear in ARTEMIS are autostretched to give a rough idea of what you are getting, but the final images after stacking and processing are always much better. 

I've tried aligning in the dawn program and then putting it into dss to convert from fits to tiff..and also tried stacking in dss.. 

[Marky1973]

47 minutes ago, newbie alert said:

I've tried aligning in the dawn program and then putting it into dss to convert from fits to tiff..and also tried stacking in dss.. 

I haven't used DAWN so can't help you there, but there is not always much to be seen in DSS. Try my basic tutorial HERE and it might help let you see what is in there. Once all stacked, you will need to further process in something like Photoshop/GIMP/Pixinsight/Startools etc.  I'm still at a basic level though I am afraid! ?

[newbie alert]

44 minutes ago, Marky1973 said:

I haven't used DAWN so can't help you there, but there is not always much to be seen in DSS. Try my basic tutorial HERE and it might help let you see what is in there. Once all stacked, you will need to further process in something like Photoshop/GIMP/Pixinsight/Startools etc.  I'm still at a basic level though I am afraid! ?

I know how to use dss , and sort of know my way around ps but at my level...i posted some stacked images to a fairly compitant friend who said there wasn't much to see and the image was very dark ( black)..

Sure I'm doing something wrong at capture stage..

You say you don't use dawn,what are you using maxim or nebulosity or something?

[spillage]

Can I just check and it might sound daft, but have you read up on this camera and the similar asi1600. The cmos sensor is supposed to be completely different from a ccd. Just wondering if you are expecting the horizon to behave the same as your old ccd?

Like I said I am not sure what presets you have but try something the same as unity gain on the asi and run off 10 300s subs. See what the stacked version looks like.

I have never used a ccd but from what I have read these cmos camera like lots of short subs.

[Marky1973]

2 hours ago, newbie alert said:

I know how to use dss , and sort of know my way around ps but at my level...i posted some stacked images to a fairly compitant friend who said there wasn't much to see and the image was very dark ( black)..

Sure I'm doing something wrong at capture stage..

You say you don't use dawn,what are you using maxim or nebulosity or something?

Sorry, didn't mean to teach you to suck eggs! ? 

I normally use ARTEMIS to capture, DSS or PIXINSIGHT to stack (although my PI free trial ran out a while ago - just building up the courage to pay for it.....although compared to the price of everything else....) then PI to process and maybe a bit of post-processing in GIMP. I did use to have Photoshop, but got fed up with the monthly costs for not much use, so went back to GIMP.

If you haven't, try STARTOOLS, a lot of people swear by it. I never really got on with it, but you can have a free trial which, I believe, is indefinite - you just can't save the final images. 

Trouble is, if there is nothing apparent in the images, the data might still be there - personally (apart from my latest efforts) I think the capturing part is the easiest once you have a system in place that works for you. The processing is what I find harder...so easy to get it wrong.

[Marky1973]

2 hours ago, spillage said:

Can I just check and it might sound daft, but have you read up on this camera and the similar asi1600. The cmos sensor is supposed to be completely different from a ccd. Just wondering if you are expecting the horizon to behave the same as your old ccd?

Like I said I am not sure what presets you have but try something the same as unity gain on the asi and run off 10 300s subs. See what the stacked version looks like.

I have never used a ccd but from what I have read these cmos camera like lots of short subs.

Yeah, read up on them a fair amount and realise they apparently prefer shorter subs - I just assumed it would cope with a longer exposures (it should at that price! ? ). I need to spend more time experimenting but definitely had the gain up on HIGH as recommended in the ARTEMIS manual - I guess I just need more time to get used to it....and perhaps ignore the Histogram! ?

[newbie alert]

2 hours ago, Marky1973 said:

Sorry, didn't mean to teach you to suck eggs! ? 

I normally use ARTEMIS to capture, DSS or PIXINSIGHT to stack (although my PI free trial ran out a while ago - just building up the courage to pay for it.....although compared to the price of everything else....) then PI to process and maybe a bit of post-processing in GIMP. I did use to have Photoshop, but got fed up with the monthly costs for not much use, so went back to GIMP.

If you haven't, try STARTOOLS, a lot of people swear by it. I never really got on with it, but you can have a free trial which, I believe, is indefinite - you just can't save the final images. 

Trouble is, if there is nothing apparent in the images, the data might still be there - personally (apart from my latest efforts) I think the capturing part is the easiest once you have a system in place that works for you. The processing is what I find harder...so easy to get it wrong.

No offense taken..  I'm ok with PS,I'd like P.I. but  I think I need to get  used to ps first..not really that keen on startools personally 

With a nebula shot taken in HA,I stretched and stretched the data bloating the stars and still no nebula data showed..i thought I'd missed the target but looking back and matching the stars showed I was in the right area..i am undersampling with my current scope thou..

[Oddsocks]

On 11/08/2018 at 09:27, Marky1973 said:

Shouldn't the histogram be further along for a 10 minute sub? Am I missing something obvious in the settings for Artemis or in the way CMOS cameras work? Please say I am as this was really frustrating - my old experience with CCD was very much plug and play, but this seems to be harder than it needs to be. Any thoughts, very welcome!

1. Histogram would be further to the right with your old CCD because a fixed pedestal offset is added automatically (depending on the camera model and driver) plus, CMOS has a much lower read noise than the CCD, much of the 'empty' space to the left of the main histogram with a CCD image is read noise and the offset pedestal.

2. Focus is very poor in both images, looking at the centre of the field, in particular at the fainter stars, you can see the stars are donuts, this means much of the stars energy is spread over too many pixels resulting in lower peak FWHM scores and is most likely the reason for DSS failing to find any stars. For your telescope, the Horizon camera and a Ha filter the fainter stars should be occupying just two or three pixels but your faint stars are spread over twelve or more. Perhaps the seeing was not so good on this session but the poor focus is not helping (centre crops from both images attached).

3. The FITS headers for both images show the change in preset was applied. In image 'Capture_043' gain was 1 and offset 16, with CBLACK 19 and CWHITE 59 while image '_15' had a gain of 30, offset 115, CBLACK 160 and CWHITE 1486. Can't see anything wrong with these values.

Recommendations:

Make sure you are using the latest version of Artemis Capture, there have been several changes since the Horizon camera was released.

Make sure that focus is spot on, this will increase the FHWH scores and DSS should be able to find the stars and stack the images properly.

 

[Marky1973]

1 hour ago, Oddsocks said:

1. Histogram would be further to the right with your old CCD because a fixed pedestal offset is added automatically (depending on the camera model and driver) plus, CMOS has a much lower read noise than the CCD, much of the 'empty' space to the left of the main histogram with a CCD image is read noise and the offset pedestal.

2. Focus is very poor in both images, looking at the centre of the field, in particular at the fainter stars, you can see the stars are donuts, this means much of the stars energy is spread over too many pixels resulting in lower peak FWHM scores and is most likely the reason for DSS failing to find any stars. For your telescope, the Horizon camera and a Ha filter the fainter stars should be occupying just two or three pixels but your faint stars are spread over twelve or more. Perhaps the seeing was not so good on this session but the poor focus is not helping (centre crops from both images attached).

3. The FITS headers for both images show the change in preset was applied. In image 'Capture_043' gain was 1 and offset 16, with CBLACK 19 and CWHITE 59 while image '_15' had a gain of 30, offset 115, CBLACK 160 and CWHITE 1486. Can't see anything wrong with these values.

Recommendations:

Make sure you are using the latest version of Artemis Capture, there have been several changes since the Horizon camera was released.

Make sure that focus is spot on, this will increase the FHWH scores and DSS should be able to find the stars and stack the images properly.

 

Thanks Oddsocks....looks like I am probably worrying too much about the histogram. Frustrating about the focus though as the Bhatinov pattern looked spot on to me - might need to wear my glasses next time! ? Hopefully have the latest software as I only downloaded it a week or so ago, but will double check before next time.

And, as I was typing this I realised, with a massive slap to the forehead, that I changed from my luminance filter to the Ha filter without refocusing....as much as I hate to admit that, it does feel better out in the open! ?

Cheers

Mark