Tiff file size really small!

[Rustang]

I'm now looking to process my first Solar Ha data and Ive done the stack in AutoStakkert and I'm surprised to see that the Tiff has come out small as in 573kb, have I missed a setting somewhere!?

Edited August 27, 2022 by Rustang

[UKDiver]

Is the Tiff compressed too? The format does allow for lossless compression.

[Rustang]

16 minutes ago, UKDiver said:

Is the Tiff compressed too? The format does allow for lossless compression.

I dont think I compressed it, is that a setting in Autostakkert? Is it best to then?

[Elp]

Don't think you can compress in AS3. What percent of images or how many images have you specified to stack? It's defined in the top right set of values in the settings window (they're displayed four in a row on two sets of rows, the left cell is the one you normally alter and leave the rest in the row at zero, the top row defines what percent you want to stack, the bottom row defines how many images, you can only do one or the other).

[symmetal]

What does the image itself look like. AS will only save files in an uncompressed format.

Did you select 'Surface' and not 'Planet' on the Image Stabilization option. 

Did you change the 'Image Size' sliders at the top left of the preview window from what they were when the file was loaded and is the 'remember' box unticked.

When you placed the AP grid was the whole sun surface covered by the alignment squares.

Alan

[Rustang]

32 minutes ago, Elp said:

Don't think you can compress in AS3. What percent of images or how many images have you specified to stack? It's defined in the top right set of values in the settings window (they're displayed four in a row on two sets of rows, the left cell is the one you normally alter and leave the rest in the row at zero, the top row defines what percent you want to stack, the bottom row defines how many images, you can only do one or the other).

I'm still very new to it, I left it on 50% Does the ROI you chose affect the file size? 

Edited August 27, 2022 by Rustang

[Rustang]

26 minutes ago, symmetal said:

What does the image itself look like. AS will only save files in an uncompressed format.

Did you select 'Surface' and not 'Planet' on the Image Stabilization option. 

Did you change the 'Image Size' sliders at the top left of the preview window from what they were when the file was loaded and is the 'remember' box unticked.

When you placed the AP grid was the whole sun surface covered by the alignment squares.

Alan

It looks ok for my first go,  set to surface, no sliders changed and I believe the remember box was un ticket. The whole sun was covered with the alighment squares. 

What should the average solar file be in size then? 

Edited August 27, 2022 by Rustang

[symmetal]

9 minutes ago, Rustang said:

It looks ok,  set to surface, no sliders changed and I believe the remember box was un ticket. The whole sun was cover with the alighment squares. 

What should the average solar file be in size then? 

If you captured the full frame size of your camera then the stacked file should be 1920 x 1200 x 2 bytes = 4.6MB. AS creates a 16 bit stacked image hence the x2

If you used a ROI then it's  ROI width x ROI height x 2 bytes in size.

There will be extra bytes added to the file for the file header but the above figures are close enough. 🙂

Alan

Edited August 27, 2022 by symmetal

[Rustang]

40 minutes ago, symmetal said:

If you captured the full frame size of your camera then the stacked file should be 1920 x 1200 x 2 bytes = 4.6MB. AS creates a 16 bit stacked image hence the x2

If you used a ROI then it's  ROI width x ROI height x 2 bytes in size.

There will be extra bytes added to the file for the file header but the above figures are close enough. 🙂

Alan

That's probably it then I used a smaller ROI,  I didn't realise it would produce a smaller image size,  it's all learning 😊

Edited August 27, 2022 by Rustang

[Elp]

ROI will create a lower resolution, it's the same as with planetary imaging, you're effectively cropping your sensor so if you then zoom into the image it will seem more pixelated then at native resolution. If you check the drizzle option it will increase the no of pixels via computer interpolation making the resolution and file size larger.

The above calculation gives a close approx file size but technically isn't correct as a 16 bit image has 2^16 shades of colour/grey tone depth, it's a massive difference from an 8 bit image which is 2^8, and there are 8 bits in a byte, file size depends largely on the bit depth being used. The more accurate calculation I believe is H resolution X W resolution X bit depth, then divide by 8 to get bytes, divide by 1024 to get kilobytes, then divide by 1024 to get Megabytes. If using jpeg which you should avoid as it compresses it will be around an 8 bit format, better to save as 16 bit or 32 bit fits or tiff though you might be hard pressed to see a massive difference between the lot unless you pixel peek, you'll see a marginal difference between jpeg and tiff/fits.

It is why for DSO imaging in particular that processing is carried out at a higher bit depth (especially in mono) prior to saving as an RGB or jpeg. It depends on the bit capability of the software being utilised.

Edited August 27, 2022 by Elp
Dither word changes to drizzle

[Rustang]

9 minutes ago, Elp said:

ROI will create a lower resolution, it's the same as with planetary imaging, you're effectively cropping your sensor so if you then zoom into the image it will seem more pixelated then at native resolution. If you check the dither option it will increase the no of pixels via computer interpolation making the resolution and file size larger.

The above calculation gives a close approx file size but technically isn't correct as a 16 bit image has 2^16 shades of colour/grey tone depth, it's a massive difference from an 8 bit image which is 2^8, and there are 8 bits in a byte, file size depends largely on the bit depth being used. The more accurate calculation I believe is H resolution X W resolution X bit depth, then divide by 8 to get bytes, divide by 1024 to get kilobytes, then divide by 1024 to get Megabytes. If using jpeg which you should avoid as it compresses it will be around an 8 bit format, better to save as 16 bit or 32 bit fits or tiff though you might be hard pressed to see a massive difference between the lot unless you pixel peek, you'll see a marginal difference between jpeg and tiff/fits.

It is why for DSO imaging in particular that processing is carried out at a higher bit depth (especially in mono) prior to saving as an RGB or jpeg. It depends on the bit capability of the software being utilised.

Thanks for the info, is the dither option you mentioned in firecapture? If so does it dither the same as when I use APT and PHD2!? 

[Elp]

Sorry, drizzle not dither, been a while since I used it.

https://www.autostakkert.com/wp/enhance/