STF ruining my image

[LarryC]

If I process an image through DBE, BN, CC, then a masked MLT and MMT, remove the mask I have a nicely noise-reduced image.  I can turn SFT on or off and it looks fine.  If I save the image as an XISF, then open the image, it continues to look fine.  If I turn SFT off/on with the top tool bar button it looks fine, but if I turn SFT off with the tool bar button, then apply SFT with either the STF process or crtl-A, I can see the image becomes significantly degraded as if MLT and MMT were never performed. At that point the tool bar button, crtl-A and the SFT process all just turn off or on SFT of the degraded image.  It appears to be a problem with the STF process.  I've tried every combination of resetting SFT, saving the file with SFT off, etc. and the result is the same.

Any thoughts would be appreciated.

Larry

[pfile]

all that's happening is that the new image has less noise, so the STF as computed on the new image is more aggressive than the STF computed before all the noise reduction has been done.

it's a common thing... not sure it's a big deal. if you like how the image looks with the old STF, then make it permanent by transferring that STF to the HT tool and doing the stretch.

rob

[jkmorse]

Larry,

Gotta say, as much as I love PI, this may be the most irritating thing I experience and will report it in the Bug Reports section.  As Rob says, it is very common and therefore should have a simple solution rather than having to go in and manipulate the STF settings to get back to normal during linear work.  The problem is for me this often happens after I apply just DBE or ABE and still have plenty to do in the linear stage, such as MLT, Mure Denoise or Deconvolution.

Hoping Juan can find a solution.

Best,

Jim 

[Alejandro Tombolini]

Hi Jim, the easy way is to save the icon of your current STF, just drag the blue triangle on the workspace and apply it to the image after DBE.
Saludos, Alejandro

[pfile]

i can't see how it's a bug - the auto STF has to take some statistical properties of the image in order to come up with something, and i'm sure the total noise plays a part in that. the STF computed would probably be totally appropriate for an image that did not go thru NR but had a similar noise profile...

rob

[LarryC]

I don't know if I would call it a bug, which is why I posted it here, and I have no idea how STF calculates it's settings, but whatever it's doing, it does it very well early in processing but frequently in my hands very unsatisfactorily later in linear processing. 

I appreciate the tip to use an early saved STF icon. 

[msmythers]

I've never considered STF as anything more then a way of seeing my image(before stretching) with a very aggressive stretch, not a final stretch. I can then decide how to proceed. The thing most seem to not realize is that you can change the autoSTF parameters. On a Windows computer it's a control click of the Auto(Atomic Symbol) STF button. That brings up the AutoSTF parameters. You can save a new setting and that will be the AutoSTF from then on. You can always reset back to the original defaults and then save those if desired.


Mike

[LarryC]

Mike,  thanks for the tip.  I didn't know the settings could be adjusted.  I'll play with this tonight.

Larry

[jkmorse]

Alejandro,

Great idea.  I realize this is a touchy subject and maybe I am asking too much.  Your idea is what I was looking for, a quick solution that does not require me to play around with the sliders to get an image I can use effectively.

Best,

Jim