PNG to TIFF and stacking advice

[Gastraea]

Hi All
Lots of questions but please feel free to ignore stuff

Firstly I've discovered that pixinsight will only open png as 8bit so can anyone suggest a easy way to batch convert 16bit png to tiff?

Secondly although I had wonderful success with my first image using my ASI2600 I seem to have run into an issue with my most recent imaging session. The first time I put the gain at around 375 (out of ignorance mostly) but this time I choose a gain of 100 as that seems to preserve the most dynamic range while still having acceptable read noise.
In sharpcap the auto stretched images looked acceptable but it seems that DSS is butchering the stacking with the output looking heavily quantized basically useless. opening the images in GIMP show that the faint detail is still there and the files are 16bit. I'll try to use pixinsight to get a more acceptable stack but I have to convert to tiff first.

Anyway philosophically do people usually use a bit more gain to give some more separation between the thermal noise and minimum signal and live with the reduced dynamic range or live with the weak signal?

Finally should I be thinking about switching to NINA? I now have and electronic focuser but sharpcap autofocus/focus routines seemed a but clunky


I've attached a 5min sub in Ha that I manually stretched in GIMP. Sadly the DSS output of 24 subs looked like it's a 4bit image.

[mura_gadi]

Quick look, this blog seems to have a few good suggestions for your batch file conversions.

https://blog.media.io/image-converter/png-to-tiff.html

[Startrek]

I’ve had my 2600MC ( same sensor as the 2600MM ) for nearly 2 years now under heavy LP Bortle 7/8 in Sydney and Bortle 3 at my retirement getaway on the south coast NSW and run the camera via Ascom driver in APT. I mostly use Gain 0 ( highest dynamic range) to take advantage of full well and 14 stops of DR. I use use Gain 100 ( Lowest read noise ) sometimes in Sydney to reduce some noise. Check out the performance charts for this camera on the ZWO site , they provide good information. I image both in broad band and narrowband
OSC and use Flats and Bias frames only for Calibration which provides a clean flat even field.
I have no issues with noise in my images at all !!!! Even at Gain 0
I use DSS to stack my Fits and post process in Startools.My DSS stacked fit file is loaded into Startools as purely linear ( no tweaks)
Camera is cooled to -10C more than adequate and I dither every frame
Attached is my procedure for DSS
Maybe give Startools a trial , totally different concept ( signal tracking engine ) of processing but gives Pixinsight a run for its money and you can generally produce a good image in less than an hour
Cheers

[Startrek]

Here’s some a YouTube on Startools presented by the developer Ivo Jager
He compares Startools to Pixinsight and demonstrates the differences in both softwares
It is quite long but worth viewing if your interested , it exposes my myths about how processing should be done to achieve a good outcome and not waist signal along the way
Cheers

https://m.youtube.com/watch?v=8Luu2-BMXzc

[Gastraea]

thanks for the info

[Ryderscope]

An important question John is to check why you are starting with PNG files in the first place. Most image capture software programs will allow for the capture of raw data from the camera in FITS format which is a much better starting point. These can be read directly by Pixinsight and even though I don’t use Startools or DSS I presume that those programs would be able to read FITS files. PNG files should only be used at the final end of the processing activity to save files for distribution/publishing.

[Gastraea]

It's just ignorance really. Sharpcap seemed to have png by default and so I was just using that. I have since found out how to change it so I'll be using FITS in future. I might also change over to NINA, I have yet to experiment with it.
The forecast is clear sky on Saturday night so I might give NINA a bash.

[Ryderscope]

That's good to hear John. I always dive down into the fundamentals when trouble shooting these issues so getting the image capture sorted is a good start.

I used SGP Pro but did have a peek at NINA a few months ago. NINA looks to be a very good package (and its FREE) so should be a good starting point. It is not quite as comprehensive as SGP Pro as yet but I'm sure will be eventually.

[Gastraea]

I think I have discovered that DSS is also opening the png files as 8bit. So it seems last time I got lucky as I wasn't imaging in the lowest few bits. It also means I should probably go back an reprocess the last data I used for the Rosette nebula.