RAW workflow with Panoramic software

[scrollop]

Hello,
  I've been experimenting with PTGui and Autopano and would aprpeciate your help in determining a good workflow when processing panos with RAWs.

I entered a series of RAW files into both, and though they gave good results in terms of stitching, they don't have many colour/exposure options (especially, it seems, autopano).

PTgui gave a more neutral/natural look, whilst autopoano seemed to increase saturation and contrast more. I couldn't find decent RAW controls in either, though, I guess, they're not designed as RAW converters, primarly.

Do you:

enter the RAWS into a RAW processor first (eg. Capture 1)--> output them as TIffs then into PTGUie/Autopano (with minimal exposre/colour processing as that should have been corrected with Capture 1), then output as psd (as the other formats seem to have size contstrictions) into Photoshop, if needed?

Thanks for your help

[Bart_van_der_Wolf]

Do you:

enter the RAWS into a RAW processor first (eg. Capture 1)--> output them as TIffs then into PTGUie/Autopano (with minimal exposre/colour processing as that should have been corrected with Capture 1), then output as psd (as the other formats seem to have size contstrictions) into Photoshop, if needed?

Hi,

I Raw-convert the images to 16-b/ch TIFFs, which allows to remove Chromatic Aberrations and Vignetting and get the best possible Raw conversion. I do not apply geometric lens-correction because PTGUI does that automatically. The Colorspace that I tag the TIFFs with will be preserved by PTGUI (I'm still a bit puzzled about the Color management by AutoPano).

Since I use C1 as my Raw converter, I can also use Diffraction correction, and I might add a very small amount of sharpening and very minor noise reduction (I am a low ISO shooter anyway). The sharp input files allow to very accurately pick the control points.

PTGUI can output to a.o. TIFF (or .PSB for files larger than 4GB). Those results can be output sharpened in Photoshop or a similar editor.

Cheers,
Bart

[scrollop]

Thank you, that sounds very logical.

BTW in terms of colour spaces, you're right about PTGui:

https://www.ptgui.com/support.html

"PTGui is currently not a color managed application, so it ignores any color profile embedded in the images. This means that images are always displayed in the working color profile (usually sRGB). If the images were taken in a different color profile they will be shown with incorrect brightness or color.

However the operations performed by PTGui (stitching and blending) are color neutral; therefore the output panorama will be in the same color space and the final result is not affected. "

http://www.kolor.com/wiki-en/action/view/Autopano_-_Panorama_Editor_-_Color_Correction

"Default color space is sRGB and can't be modified"

[kirkt]

I use PTGui, so take that as the context for this brief workflow description.

I usually convert my images in a dedicated raw conversion application (Capture One, for example).  In this step, I apply basic adjustments, without any lens correction - obviously all adjustments made are applied to all images.  I save a small JPEG (2400 px or so on the long edge, for example) in sRGB and a full size 16bit TIFF in the desired output color space.  With Capture One, these can be output simultaneously with recipes.

I use the small JPEGs in PTGui to perform the stitch and make any adjustments to the stitch (adding control points, etc.).  The use of small 8bit JPEGs keeps the work interactive for large panos and speeds up the process.  I make the final, full-resolution output from the small JPEG stage and inspect it in PS for any blending or alignment errors.  Once that is done and all is correct, I save the project and also save a template for the stitch.

Next, I create a new project and load all of the full-res, 16bit TIFFs.  All that should be required at this stage is to apply the previously saved template and output the result.  If masks are needed, you can create them in PTGui or create them ahead of time in PS and save them as an alpha channel in the TIFF - PTGui will see those masks and apply them.

The output will be in the same color space as the input TIFF files - you may need to Assign this profile to the resulting output when you bring it in to PS or similar program.  Using the small JPEGs as proxies for the stitching process is extremely useful for pano sets with a large amount of images (like a equirectangular set made with a non-fisheye lens) or for HDR sets where you stitch the JPEGs to get the stitch correct and then apply the template to a set of HDR input files for the final.

Kirk