"Grey scale tiffs" v "RGB tiffs" in LRGB processing????

[SlimPaling]

Hi ...

I have noticed that after stacking a series of images in Maxim and producing a "FIT file" ... if I then save it, within MaximDL, as a "TIFF File" and then open it up in Photoshop if comes out as a 16bit RGB tiff.

If I ask "Fits Liberator" to save the same "Fit file" and ask it to convert it to a "TIFF file" it comes out as a 16bit Grey Scale tiff when I open it up in Photoshop.

 

Looking at a particular YouTube video showing somebody processing LRGB files using Photoshop ... he is using "16bit Grey scale tiffs" without any problems.

 

Does it matter if I use "RGB tiffs" or "Grey Scale tiffs" when it comes to processing LRGB files in Photoshop????

 

If there are clear benefits for me to be using "Grey Scale tiffs" in Photoshop .... is it  possible for Maxim to save the "Fits files" as "Grey Scale tiffs" rather that "RGB tiffs" ... I can't seem to find any option in this regards?

 

Cheers Mike

[symmetal]

The RGB file contains three grey scale images which are assigned as being the R, G and B components of the RGB file. A grey scale file saved as an RGB file has the same grey scale image saved three times inside the RGB file which makes the RGB file three times the size of the grey scale file.

Each colour channel in the Photoshop RGB image is a grey scale image so assigning your separate R, G and B images to the RGB channels of the Photoshop RGB file it makes sense for the R, G and B images to be in grey scale format. Trying to assign an RGB image into a colour channel I assume it will convert the image to grey scale (averaging the three identical RGB channels of your RGB image to one grey scale image) before assigning it to the channel. It may say it can't do it, requiring you to choose a channel from your (monochrome) RGB image to assign.

If Maxim only allows you to save grey scale files as RGB (I'm sure it would allow grey scale saves, (cue for Maxim expert to appear)) just convert the RGB file to grey scale in Photoshop before assigning it to a channel to make things easier if Photoshop complains. No data is lost or gained in doing this.

After the RGB image has been created in Photoshop from your R, G, and B images, you create a new layer and copy your L image to that layer. It doesn't matter if the L image is in grey scale or RGB format it will be converted to RGB automatically when it is assigned to the new layer as the layer below (your created RGB image) is already an RGB layer.

Alan

[Stub Mandrel]

A similar thing caught me out. It turns out FITS liberator will offer you three layers for an RGB image, you have to select, stretch and save each layer in turn, 1 as red, 2 as blue etc.

[Oddsocks]

Just checked with some MaxIm DL6 (v 6.18) acquired images, aligned and combined,  they are saved as 16bit greyscale tiffs provided the source images are from a mono camera and the fits header is correctly formatted. Opening these in Photoshop CS6 they are correctly identified as 16bit greyscale.

If I then repeat the process with some DSLR raws they are saved as 16bit RGB tiffs and open as RGB files in Photoshop.

Maybe the fits header for your source images is incorrect, or missing, so that MaxIm doesn't know what type of file it should be and defaults to RGB?

William.

[Stub Mandrel]

I think the best route for the OP is to use the channels tab in PS to extract the three channels.

[symmetal]

4 minutes ago, Stub Mandrel said:

The OP's problem is that FITS Liberator is saving TIFFs as greyscale, not Maxim.

My previous post explains why this is.

I think the OP's problem is that Maxim is saving stacked greyscale images as RGB. Fits Liberator saving them as greyscale is the reasonable option. 

I assumed that the OP was using a mono camera as if he was using a colour camera he wouldn't ask the question. 

Alan

[Oddsocks]

Just out of interest Mike, which Moravian drivers are your using with MaxIm?

I know that the direct plug-in driver they still supply for MaxIm is buggy,  is not recommended and no longer supported by Diffraction Limited. Moravian recommend to use the ASCOM driver with MaxIm.

[Stub Mandrel]

24 minutes ago, symmetal said:

I think the OP's problem is that Maxim is saving stacked greyscale images as RGB. Fits Liberator saving them as greyscale is the reasonable option. 

I assumed that the OP was using a mono camera as if he was using a colour camera he wouldn't ask the question. 

Alan

Yes - I realised that and edited my post - too late it seems!

[symmetal]

1 hour ago, Stub Mandrel said:

Yes - I realised that and edited my post - too late it seems!

Sorry Neil, I was too quick off the mark. 

Alan

[Oddsocks]

Mike.

In addition to the replies to your parallel thread running on the Diffraction Limited forum an extra bit of info that may help you pin down the source of the problem is to use the fits header fuction in MaxIm ( and most other fits readers ) to inspect the NAXIS value.

With a single sub from the camera open on the MaxIm desktop use the fits header function from the top menu bar, look at the list of keywords displayed, a mono image will only have the keywords NAXIS1 and NAXIS2 listed and if you saved this single image as a tiff it should be saved as a greyscale. Do the same with the result of the stack, the same rules apply.

If you see the keyword NAXIS3 this means the image is a colour image comprising three planes, RGB, and would be saved as an RGB tiff.

If NAXIS3 is present in the single subs from a mono camera then the driver, MaxIm’s implementation of the driver or the camera’s firmware is faulty or the wrong version. The type of sensor, mono or colour is determined by the camera’s firmware and there have been a few problems caused when the user, or factory, load the wrong firmware for the camera.

If NAXIS3 is missing in the single subs but present in the stacked image that would point the finger at MaxIm and you should send a selection of single subs in a single folder to their ftp incoming server and ask them to investigate further, unless the loaded subs are grouped by filter and RGB is the selected 'Set Group' method..

HTH.

William.

[SlimPaling]

21 minutes ago, Oddsocks said:

Mike.

In addition to the replies to your parallel thread running on the Diffraction Limited forum an extra bit of info that may help you pin down the source of the problem is to use the fits header fuction in MaxIm ( and most other fits readers ) to inspect the NAXIS value.

With a single sub from the camera open on the MaxIm desktop use the fits header function from the top menu bar, look at the list of keywords displayed, a mono image will only have the keywords NAXIS1 and NAXIS2 listed and if you saved this single image as a tiff it should be saved as a greyscale. Do the same with the result of the stack, the same rules apply.

If you see the keyword NAXIS3 this means the image is a colour image comprising three planes, RGB, and would be saved as an RGB tiff.

If NAXIS3 is present in the single subs from a mono camera then the driver, MaxIm’s implementation of the driver or the camera’s firmware is faulty or the wrong version. The type of sensor, mono or colour is determined by the camera’s firmware and there have been a few problems caused when the user, or factory, load the wrong firmware for the camera.

If NAXIS3 is missing in the single subs but present in the stacked image that would point the finger at MaxIm and you should send a selection of single subs in a single folder to their ftp incoming server and ask them to investigate further.

HTH.

William.

Hi William ... and others who are trying to help me sort this out ?

Yes ... I am using my Moravian G2 Mono camera .. so I am expecting grey scale FITs images.

I have just been checking the fits headers in some of my images .... my original fit images all have "NAXIS / NAXIS1 / NAXIS 2" (but no NAXIS 3) in their headers ..... however when I looked at the images that Maxim has saved they all include " NAXIS / MAXIS1 / NAXIA2 / .... AND NAXIS3"

It seems that Maxim is adding NAXIS 3 to the headers when it is told to convert the FITs image to a TIFF !

Am I doing something wrong here ... ? I think it a good time for me to ask Maxim support for a bit of help here. If I hear anything useful from them I will post it back here.

Cheers Mike

 

[Oddsocks]

3 hours ago, SlimPaling said:

It seems that Maxim is adding NAXIS 3 to the headers when it is told to convert the FITs image to a TIFF !

The NAXIS3 keyword would have been added at the save as .fit stage, the ‘Save As’ TIFF only reads the fits header NAXIS3 keyword to detemine if it should save as greyscale or save as RGB, it wouldn’t change it.

Next time I’m logged into the observatory Windows machine I’ll have a look at some of the image combination options in MaxIm.

You could try carefully checking a mini stack of say ten subs, make sure they are all only NAXIS1/2 and then look to see if the output file fits header has become NAXIS3, that would show for sure a MaxIm incorrect stacking option selection or a bug.

William.

[SlimPaling]

Hi William ...

I have just check every one of my calibrated and aligned subs and they all have only NAXIS1/2 ...... none have NAXIS3.

However NAXIS3 seems to appear from nowhere after I stack the files into a single Fit file .... is this a bug or am I doing something wrong in Maxim?

Cheers Mike

[Oddsocks]

I've just looked at the observatory Windows machine, luckily it's not doing anything tonight.

When you load up the images in the stack make sure that the option 'Auto Color Convert' box is not checked under 'Mark added items as'.

Picture attached.

I'll look for anything else and post again if I find something.

[Oddsocks]

I can't see anything else that could cause it Mike.

Can you place four of five native camera subs in a dropbox or google drive and I'll see if I can produce the same problem.

Are you up-to date with MaxIm? current version is v6.18 which is what I am using.

 

Edited, no need to upload images, see next post.

[Oddsocks]

Okay just a thought, the penny has dropped, perhaps time for a silly question.

When you load up the subs into MaxIm's stack window are they arranged as in my last post image above, a single group, if so, saving the output of such a group will be a mono .fit and a mono .tiff because with only a single filter group the image can only be mono.

Or, when you load the subs are they sub-grouped into separate filters R+G+B as in the image below, if so when you save as .fit NAXIS3 is added and when you save as .tiff it would be a colour RGB because this is what has been created, a three channel RGB image.

Scroll down for more....

 

If this is the case, and I suspect having strayed away from the original opening question, this may be the issue, we can step back to your original starting post.

When you stack a series of grouped filter subs like this in MaxIm the output .fit file will have the NAXIS3 keyword and saving the file as .tiff it will be a RGB colour image.

If you import the .TIFF into Photoshop it will be a composite RGB and you can then use the CIE Lab mode to extract the luminance data and split the RGB file in RGB channels to work on the individual colour channels. I won't try to explain that to you as I don't work with Photoshop enough to give an expert view, there are plenty of other SGL members here who can explain better than me working with RGB files in Photoshop.

If you import your stacked colour .fit file into Fits Liberator you can select each colour pane, or channel, and export each pane as an individual .tiff, these will be greyscale images and each one corresponds to one of the individual RGB channels of the colour .fit that MaxIm created from the RGB filter files. The downside of Fits Liberator is that it does not output a separate luminance channel and you would have to combine your three greyscale tiffs in Photoshop to extract luminance data for use in LRGB processing.

So, back to the original question, if you want to create individual greyscale images for each filter channel in MaxIm, first make sure 'Classify by Object' and 'Classify by Filter' is selected in the stack window, select and load all your subs using the 'Add Files" button, then right-mouse-click the containing 'Group' folder and select 'Set Group' and 'Individual Planes', run the combine stack, each filter group will be output as a single greyscale .fit.

Select each .fit group image open on the desk top in turn and save as 16bit TIFF for export to photoshop and save as .fit for archive if needed. You will end up with a set of greyscale .tiff's, one for each filter and you can use these to work with Photoshop as separate greyscale channel images L, R, G, B etc and follow the processing path appropriate for this type of file.

Which method is best, export as individual greyscale channels or as a single combined RGB I'll leave others to debate...

Hope that is the mystery solved, let us know how you get on.

William.

 

[RichLD]

If you don't get to the bottom of the source problem the easiest route is to change the image mode in PS. Image>Mode>Grayscale.

[SlimPaling]

PROBLEM SOLVED :-))))

After a great deal of detective work I spotted that the problem starts as soon as I dis the stacking process in Maxim. I then realised that the "Auto Color Convert" button was ticked. After un-ticking this button and re-stacking the problem disappears ?

I am not sure if & when this button got clicked as I don't normally click on things unless I know what they do .... anyway all seems well now :-)))

Thanks everybody for helping me to get to the solution.

Cheers Mike