DSS only processing 19 of 32 flat frames selected

[Ouroboros]

DSS is only processing 19 of the 32 flats I created.  As far as I can see there's nothing unusual about 13 of the unused files. I'm selecting all of them when loading the files into DSS. I've counted them. Any idea?  

[happy-kat]

Are you loading then all into the one tab or have you used group tabs?

[tooth_dr]

Have you tried just choosing the 13 it rejects and see what happens? I’ve had this happen to me in the past where it dropped some frames but I never worked out why. 

[Mr niall]

Are they all the same length of time? I had a thing recently where I messed up somewhere and could only get it to stack images that were the same length as the first (eg reference image)

[Davey-T]

Have they got Fits header files ? if so you can check for any differences.

Dave

[Ouroboros]

Lots of ideas there. Thanks. I've turned in for the night now, so will check these out tomorrow. @Davey-T they're canon RAW files so have exif data. 

[Ouroboros]

23 hours ago, tooth_dr said:

Have you tried just choosing the 13 it rejects and see what happens? I’ve had this happen to me in the past where it dropped some frames but I never worked out why. 

OK tried that. It turns out it was the first 13 flats of the 32 taken that DSS didn't use.

Processing using  the 13 ignored flats, DSS selects nine of them to calculate a master flat.

Is there some threshold used somewhere? 

[Stub Mandrel]

Check they are all the same exposure.

[Ouroboros]

21 hours ago, Stub Mandrel said:

Check they are all the same exposure.

My flats were taken using Back Yard EOS using the AV-Flat setting. Always worked OK in the past .... I think. 

All the 19 files were taken with 1/400 shutter. Thirteen other files rejected in the first instance by DSS were slightly shorter of either 1/500 or 1/640.  Doesn't seem anything odd there I'd have thought.  Maybe DSS  doesn't like that for some reason. 

 

[michael8554]

The AV flats I take from a flat panel vary slightly in exposure from day to day, but on the day the exposure time does not alter.

Try taking flats at 1/400th instead of AV ?

Michael 

[dph1nm]

I don't think DSS will stack images (whether flats or lights) of different exposure times. You can get round this by editing the images' properties in DSS and setting them all to the same exposure time (might have to be 1 sec, as I could never get the editor to accept shorter times).

NIgelM

[Ouroboros]

22 minutes ago, dph1nm said:

I don't think DSS will stack images (whether flats or lights) of different exposure times. You can get round this by editing the images' properties in DSS and setting them all to the same exposure time (might have to be 1 sec, as I could never get the editor to accept shorter times).

NIgelM

That would explain it. I guess this must have happened before. I just haven't noticed. As it happens the 19 flats accepted resulted in DSS giving a perfectly satisfactory image. I just don't like a mystery. I'll have a look at Backyard EOS tomorrow to see what other ways are available to record flats. Cheers. 

[ollypenrice]

I would never do all the calibration and lights stacking in one go. Before I use a flat I want to look at it. You get a notion of whether or not it's a believable flat, and sometimes they are are not and you can see that they are not. If it's OK it becomes a master flat and, in an observatory setup, it might last six months.

Olly

[Stub Mandrel]

1 hour ago, ollypenrice said:

I would never do all the calibration and lights stacking in one go. Before I use a flat I want to look at it. You get a notion of whether or not it's a believable flat, and sometimes they are are not and you can see that they are not. If it's OK it becomes a master flat and, in an observatory setup, it might last six months.

Olly

You can do that in DSS if you want ?

It's much more sophisticated than it's reputation... Because it serves up 'recommended' settings for what you put into it with a 'one click' stacking solution don't assume you cna't delve under the hood!

[Ouroboros]

I look through all my lights, flats and bias before callibrating to make sure there are no anomalies. I see there are some nice ways of doing that in PI - i.e. blink routine I think - that lets you look at each stretched sub in turn. Haven't got around to using them yet though.