BYEOS, or camera, or data transfer, or DSS...?

[squeaky]

Hi all,

Something is wrong somewhere but I'm at a loss as to where.

I'm having to use an ISO of 12,800 to see anything at all in a 30second sub...

This is M74, honest , and the nebula should pretty much fill the entire frame. The short vertical line of stars on the right hand side are Mag 12 - 13 ish

Equalising it to help show faint background detail basically says there isn't any:-

DSS (3.3.3 beta 47) point blank refuses to accept that my CR2 files are colour and loads each sub as grey. When I finally get it to even see stars at all the histogram has usually been hard left and very low - but now, although it reads them in as grey it's seeing tens of thousands of stars and it doesn't matter whether I use the built in white balance or my camera white balance (set to auto) and the end result seems to be reading the huge amount of noise that imaging at such a high ISO would be expected to produce even when noise reduction is on. My first pass through got me a white screen. This the output of 17 frames after setting the threshold to around 40% instead of the usual two - with noise reduction on and auto white balance:

Grey and completely flat. Equalising it still shows flat. Not even any stars. Beating it up with major lifts on levels and curves fails to show anything at all.

I used BackyardEOS to get the subs and had it set to save the images to my laptop rather than the camera simm.

So is it the Camera? Or BYEOS? Or the data transfer? Or DSS?

Or none of the above and it's me?

[Star Gazer]

Hi Squeaky,

You need to work through a process of elimination.

First off eliminate BYE. Just use bulb mode with a timer shutter release of just take a 30 sec exposure using nothing but the camera.

Second listen to the shutter and time it with your watch (is it 30 secs?).

If the camera seems OK you might want to putting BYE back into the equation

Not sure what camera you're using but just try it with a normal lens, even in the daytime.

Not sure what scope if any you are using?

I'm not a lover of DSS, my results are usually B&W or so low saturated that i've more-or-less stopped using it. If it won't accept your CR2s you need to check they are supported. New Canon cameras are often not supported by third party software until quite sometime after they come on the market.

Also I remember there are some settings on how CR2s are processed in DSS so check out its settings.

..and ninthly...

[EDIT] Oops I missed the equipment in you're signature - I guess you're using a 550? Still my original idea of methodically eliminating each variable is you path to success

[Steve Ward]

Hi Squeaky,

You need to work through a process of elimination.

First off eliminate BYE. Just use bulb mode with a timer shutter release of just take a 30 sec exposure using nothing but the camera.

Second listen to the shutter and time it with your watch (is it 30 secs?).

If the camera seems OK you might want to putting BYE back into the equation

Not sure what camera you're using but just try it with a normal lens, even in the daytime.

Not sure what scope if any you are using?

I'm not a lover of DSS, my results are usually B&W or so low saturated that i've more-or-less stopped using it. If it won't accept your CR2s you need to check they are supported. New Canon cameras are often not supported by third party software until quite sometime after they come on the market.

Also I remember there are some settings on how CR2s are processed in DSS so check out its settings.

..and ninthly...

[EDIT] Oops I missed the equipment in you're signature - I guess you're using a 550? Still my original idea of methodically eliminating each variable is you path to success

You'll find the data and colour is all there in the file DSS chucks you , it just needs dragging out .

Most people are shocked at what sometimes comes out of DSS but adjustments in it are pretty powerful ,

Steve.

[Steve Ward]

Richard ,

I had a couple of occasions when BYEOS trial went all random on me half way through a set.

Main reason I didn't go for full version.

Steve.

[squeaky]

I swapped out the USB cable. I think it's sorted things. I'm currently imaging Mizar and Alcor at ISO 400

Tracking is off (nowt new there) in this example - but Mizar, on the right, really is a double, and the fainter star below centre mid-line is 7.55

I had a quick play with a better sub and got stars down to Mag 10.

The problem I had with M74 is that while the book says it's Mag 9.8 (yeah right) there's nothing in my FOV brighter than Mag 12 - so I have to give the ISO a boot just to be able to see where I am. Then I was taking 10 off 30s subs and then using previews to re-centre because of drift and to rotate the camera a little since I was out there three hours to get 135 frames, of which only 29 were at or near the standard posted above, and DSS chose a whole 17 of them. I really REALLY have to get this tracking sorted out.

As for DSS and CR2's... I have the the beta version that loads them OK - but it still insists on processing them as GREY ... and they're ruddy well not because I'm using a DSLR.

To get ANY colour out of them I have to MAX the saturation lift in DSS.... THEN hit saturation again in PS and we're talking 3-4 iterations at +25%.

[squeaky]

OK... it's my tracking and DSS between them.

I finished up with 84 subs before the cloud beat me. Bear in mind that because of poor tracking and drift I would reposition the scope to get those two stars back into much the same position in the FOV as I started from. As per usual I got a grey looking result from DSS but there WERE two faint smudges on there.

I equalised to see what was there...

DSS wouldn't stack on the first run because it never saw more than six stars in the image. So I turned noise off and it saw around 45 stars. Yeah right.... it saw NOISE and stacked that. The same, I guess, will be true of the M74 subs - only when they got smeared everything was so faint that it all disappeared.

I may be able to beat this for the Alcor Mizar set. Just up the ISO or time a bit to ensure that I can see more than eight stars in the subs.

Whether or not I can do this for M74 is another ball game altogether given how faint everything is.Signal to noise levels when the brightest thing there is Mag 12 and I'm looking through a summer haze lit by LP just plain isn't going to work... I just wish I could show you how bright the clouds looked in my sky last night but my camera is body only so I don't have any lenses that will do a wide field view.

[squeaky]

Handed the jpg's to registax6...

Hmm... so if I were to cull all the dodgy frames by hand and hand it only "MY best" I might get somewhere. TBH - right now I could get a better result by tweaking a single good sub.

Not really worth the effort. I'm shattered. If we get a reasonable viewing night tonight (by the rubbish sky standard I'm getting) I'll have another pop at them. This time though, instead up just going for a full run I'll do ten subs then stop while I hand those to DSS and see what it does. Keep doing that with changes to ISO and duration until DSS can work with them and then try for a full run. After spending three hours night before last getting my 135 subs only to find that 17 of them got stacked - whole nine minute's worth after three hours - I'm darn well not getting caught like that again. I could have cried.

Now to get back to the shop and find out why I've got no answer back from OVL about setting up clutch friction etc. <sigh>

[earth titan]

Are you guiding?

30secs is a short sub. For a subject such as m74, mine are normally 300 at least at an ISO of 800 or 1600. If you're not guiding than 300 secs could give trailing, but at least you would capture something and know if it is you or the camera.

Typed by me, using fumms...

[squeaky]

Gawd no. I'm going to need to - it's obvious my Dobsonian is unlikely to get beyond 30 sec subs and at the moment because of problems with the mount mentioned elsewhere I'm struggling to get 10's.

So it's a case of collecting LOTS of short subs.

To make matters worse it's looking more and more as if DSS doesn't process CR2 files from my Canon 550D properly either. (And I'm using the beta because the latest published version doesn't even load them properly, much less process them.) For the moment I think I'm going to have to grab JPG's only because I suspect I'll get better results from those, despite detail loss, than I will from CR2's.

So... guiding I need, plus save up for an EQ6 type mount, which is going to take a while, plus mounting a Flex on an EQ6 is no easy proposition I'm told.

Then all I need is clear skies

I've not had a fully clear sky here since April. <sigh>

[JamesF]

With lots of short subs are you expecting to increase the signal or decrease the noise?

James

[squeaky]

I don't "expect" anything much - just a slightly better end result, one way or another, from more than "a few" subs.

Noise is pretty much taken care of before the sub hits the drive. The camera is set to take and subtract a bias after every shot, and I've got long exposure noise reduction on (low).

Between that and the fact that every frame is in a different position because of rotation and tracking - I really should get a reasonably clean noise free result. At least, I should at ISO 1600 on a JPG versus the 12,800 I have to use on CR2.

I "think"

Not had a sky with even moderate haze since then, much less a low level of haze. I suspect that to get anything for M74 I'll need a completely haze free sky and, for preference, be at a darker site than I get at home.

[JamesF]

If there's very little noise in the image then I'd not expect lots of subs to give you much more than a few. For more signal (which is how I read what you're after), I'd expect to have to take longer subs. Adding more subs at a shorter length will, as I understand it, reach a point where the returns are negligible relatively quickly, especially where your subs aren't that noisy to start with.

My understanding is also that the CR2 files are pretty much "as the data comes off the sensor" and the JPG versions are based on that data with various filters applied (eg. white balance) and then fed into a lossy compression algorithm, so I'm not sure why you'd need different ISO settings depending on whether you're using the JPG or CR2 image. In theory at least the ISO1600 CR2 should contain at least as much signal (and probably more) than the ISO1600 JPG, though you might need to get aggressive with levels and curves to find it.

James

[squeaky]

The thing was - in order to SEE any stars in the subs as they showed in BYEOS when in CR2 I had to use 12,800 just to see enough stars to know where I was and to reset for framing after every ten subs.

I understand what you're saying, and fully concur - but I don't seem to have a lot of choices. In CR2 I can't get my histogram "off the floor" even at 12,800. I end up with the curve hard left, usually clipped, and the peaks are low - barely 10% of the vertical range. If I save those subs in CR2 and JPG - then the JPG's have the curve showing between mid scale and well right, and the peaks are clipped at the top because they are so high. That's a huge difference between the two.

If I were to drop the ISO for the CR2's to 1600 I'd have absolutely no idea where I was, nor how much to correct for drift because all I'd see would be a flat grey field. How do I beat that?

[JamesF]

I'm not familiar with BYE, but in APT if you set it to grab both the CR2 and JPG (R+L) then I think it displays the JPG on the screen whilst saving both to disk so it's possible to process the CR2 files. I don't see that gets you very far with respect to DSS finding too few stars to stack though. Perhaps M74 just isn't a viable target for the 550D with 30 second subs?

James

[squeaky]

That would make most DSO's not viable. In CR2 anyway.

I am going to have to try JPG's only and see what I get.

[squeaky]

OK - this is getting silly....

Lousy haze this evening but did a quick test shoot of Vega just to see if my secondary's dew heater worked (yes ) and didn't distort my stars. (can't tell)

So I took ten 10sec subs at ISO 800 and saved as RAW + JPG.

DSS found ZERO stars even with noise off for the CR2's

The JPG's produced this...

I have absolutely no idea of what's going on here.

[squeaky]

It's got to be DSS...

Here's a CR2 opened in Paint Shop Pro - untouched except for resizing and saving it as a PNG to post here...

And the same frame but the JPG version originally saved from (RAW+JPG) resized and saved as a PNG to keep things more or less equal...

[JamesF]

No idea on that one. I wonder if waiting until the moon is less bright might help by allowing the camera to pick up more stars in the background sky? Failing that I think i'd wait until the moon isn't so bright and then take a number of widefield shots with the kit lens at various focal lengths to see if DSS can stack those ok just to see if DSS can handle those CR2 files and that it finds enough stars.

But I'm clutching at straws a bit myself here...

James

[squeaky]

No kit lens.

"Equalising" shows around 20 stars in that FOV.

CR2 on the left - the same frame saved as the JPG on the right

The "standard" DSS doesn't open CR2 files from the 550D properly - you only get a sort of "pillar box" of the image. The beta, which I am now using, does open them correctly - I'm just not convinced that DSS is processing them correctly. How can it find ZERO stars in CR2's given the subs like the one I posted above?

[JamesF]

Perhaps it's worth getting in touch with the developers then. Ask if they can explain why it isn't finding any stars in your CR2 file?

James

[squeaky]

Aye!

[squeaky]

At Last!

A quick test through the usual haze and LP - DSS stacked 5 of ten 10 sec subs of CR2 at ISO 6400

A bit noisy but not worth processing more - but I'm back on track

[JamesF]

So do you understand what the problem was, or is it one of those magical "it just started working again" things?

James

[squeaky]

I did a whole bunch of things today.

Reset the Synscan to Factory Defaults.

Put my Lat and Long in PROPERLY. Guess which muppet was at 1.10 East instead of 1 deg 6 mins East?

Hunted down and removed every last scrap of DSS before re-installing 3.2.2 and then adding in the 3.3.3 beta 47 AND moving it from the directory it unpacked into across to the standard x86 folder.

Upped ISO, even for short subs from 1600 to 6400. I can deal with noise IF I get an output from DSS. At 1600 the CR2's don't seem to grab anything.

So... um.... take your pick

More to do yet... because I also put the camera back onto (almost) defaults everywhere, so there must be improvements I can make somewhere.

The resistance heater I made for the secondary works a treat - not a sign of dew on it at all, so that helps; and I finally got my light shroud finished and fitted. (Dew shield yet to do).

All in all it's been a very busy few days.

I'll need to redo my balance weights table after fitting the dew shield. It might be easier this time round because I've got a 55cm magnetic knife rack fitted to the lower half of my OTA so I may only need the one weight and a sliding scale.

Got a whole bunch of test stuff to run through DSS - but not tonight, I'm shattered. Time for bed.