Thanks for the suggestions. Looks like I'll edit the brightness before I photomerge. I might give the warp idea a go as well. ... View more

Thanks for the suggestion. Yes, that's the way to do it, using four layers:   Layer 4… Matte (making the sky transparent, plus any gradients) Layer 3… Mountain Image (with Track Matte Key, matted using Layer 4) Layer 2… Rolling credits (transparent PSD with text and images) Layer 1… Mountain Image.   ... View more

Thanks for all the feedback. I read through it all, ready to try again this morning. Last night before posting, I must have tried a dozen times to keep the Background in place (a true, locked Background) while aligning the top layer (B, one of about 10 layers). I re-arranged layers, unlocked all of them, locked all of them (except B), and various other combinations of trying to align B with the Background. The only thing I didn't try was to quit PS and start again.   Something must have gone haywire with my PS last night, because this morning when I fired up, the alignment worked correctly. But, I did notice that the alignment unlocks and renames the Background layer, and doesn't return it to a true Background. That itself suggests that last night PS was unlocking the Background in preparation for aligning, and then inadvertently moving both because of temporary insanity.   Thanks again. Next time I have a problem with PS, I'll try quitting. ... View more

Thanks for the suggestions. I can't get the "L" and "R" boxes on the right to work on my CS6 version. Very strange. I would have thought that with all the complex routing and assigning available in Audition that it could do this simple job.   But I do have the solution: that good ole reliable audio workhorse that is always in my Dock just in case: Audacity! ... View more

Thanks for the answer. Didn't realize you could just lock the pixels in a layer via the paintbrush icon. ... View more

Goodness, gracious me, Bojan – you've solved it. As simple as dragging the source selection to any-sized image in another tab. Something I'll remember till the end of my Photoshop days.   Thanks heaps. ... View more

Thanks Semaphoric, for the hint.   I have not used Quickmask before, but I gave it a go and couldn't get it to work. My selection source was twice the size of the destination image, and when I entered Quickmask in the destination and pasted, the selection rectangle was twice the size of the image (looking promising), but there was no way I could make the selection smaller so that it fitted into the destination. And when I exited Quickmask, everything disappeared anyway.   If you'd like to see if you can get this working: • start with a document A of any size • duplicate and change the new document, B, to half size • create a selection in A, and see if you can get a reduced size version of that selection into B.   The two workarounds I listed function okay, but I'm interested to find out if there are any other methods. ... View more

Is there a good reason why PS doesn't allow selections to be transferred from one image to another?   Some background This is to do with generating versions of scanned slides that become part of an AV in Premiere. The shows are always 1920 x 1080, and slides are normally cropped to that aspect ratio. Often that is easy to do because of too much sky or foreground, for instance.   But sometimes I come across a selection of slides that are well taken, so really can't be cropped. In that case, I may decide to crop them identically to ~3:2 (even the vertical slides), so that when they cross fade on screen, the audience is not distracted by varying position of the edges.   In this case I'm going further. On screen, I want to retain the rough-edges and the round-corners of the paper-mounted slides. So I choose one slide, generate an outline of the slide image (see below), and use that as a mask in Premiere so that all slides have the same appearance on screen -- as if viewing through a rough-edged, round-cornered window. For various reasons, this time I want to apply the selection to all slides in Photoshop, and not use a mask in Premiere. So, once the selection is saved in the Source image, I want to:   Copy that selection to another image. Move and scale the selection (but keep it's aspect ratio) Apply a Layer Mask > Reveal Selection so that only the area inside the selection is visible Crop to selection and resize to 1080P Save as PSD Import into Premiere, where I now have a rough-edged, round-cornered image.   To answer your "selecting the white part" query, here are my notes on the process:   TO IMPORT A SELECTION Photoshop appears not to allow the user to import a selection into a document that has different dimensions. There are two workarounds:   1. Change dimensions. Make the dimensions of all documents the same. Do this by noting the size of each document, choose the largest vertical dimension (V) and the largest horizontal dimension (H), then increase the canvas size to V x H for all documents. 2. Copy and Paste to a layer Note that this technique only works for hard-edged selections. It does not capture feathering. - Open the relevant selection channel on the source document and make sure the channel is active. - Copy all of the channel and paste into a new layer in the destination document. - Scale the new layer so that it just fits inside the image. If it isn’t made to fit, the whole of the selection cannot be selected (next step). - With the Magic Wand set to a large Range such as 250, select all the white area. Note: Channels use a different colour space from the layers. They use Working Gray, so to accommodate the different whites, the Range may have to be altered to suit. For example, for my special Kodachrome colour space, the range has to be set to 235. Any larger and the whole canvas is selected. - Save the selection. The new selection now has the same aspect ratio as the source, but can be moved and scaled as required. - Delete the pasted layer.         ... View more

Thanks for the detailed response, Rick. However, my post was obviously not clear enough.   The film, shot at 16fps, was wet-gate transferred frame by frame, at 2.5k, ProsRes 4444, on a filmfabriek HDS+. Scanning costs were about $600. I only wanted the best (thus my pickiness with how I'm going to edit this film for an end result on Blu-ray). Previous scans of other old films were done at Videopro in Denmark on a Lasergraphics Scanstation, but for various reasons that option wasn't suitable this time.   Every time I have scans done, the first reel scanned is my test reel. Surprisingly, the filmfabriek (a $50k machine) which a young fellow in Melbourne has recently purchased for his business (and as far as I'm aware after extensive research, is the best scanner available in Australia), gave overall superior results to the $250k Scanstation. So, this time, the film was sent to Melbourne.   To be clear: there's no interlacing, no video transfer involved, no degradation so far of any kind, other than a direct frame-by-frame scan. Now I want to find out under what conditions Warp Stabilizing is possible and/or desirable.   As for the way I've blended frames, to my eye it gives the best result. Before deciding on this method, I compared three techniques in going from 16fps to 24fps:   Frame blending turned off, which simply repeats every second frame. The film will be sharpest, but with jerky motion. Frame blending turned on. Adds frames as above, but blends every frame with the two around it. Every frame is therefore blurred, but the motion should be the smoothest – but blurred. Add frames as per 1, but the added frames are a 50:50 blend of the repeated and the following frame. My idea was that I retain original sharpness for two out of three frames in the final film.   I applied the techniques to the same 15-second clip, converted to Blu-ray for viewing on our 3-metre projection screen, invited my partner and three friends for the test, and away we went.   Test 1 I simply asked them could they see any difference in sharpness of jerkiness between the three clips. No one could, except that my partner (accustomed to my numerous tests, and with a more critical eye than most), said that number 1 looked "jagged". No one could see any difference in sharpness between all three.   Test 2 I explained the difference between the three clips and what to look for. This time they could all see the jerkiness in #1, but no difference in jerkiness between #2 and #3, and no real difference in sharpness between any of them.   My conclusion: for a general audience viewing well-photographed home videos, there was no difference between #2 and #3, projected at a viewing distance of 1.5 screen widths.   Later, I repeated the test myself, really looking for the differences. It was not a blind test, I knew what was on the screen, but in terms of sharpness and fluid movement, to my eye #3 was the winner.   I doubt whether the combination of scan and blend method described above can be bettered. Now I want to experiement with stabilizing. Some of the film was shot on tripod, so it's pretty good for an amateur effort. But other scenes are handheld, or taken on an aircraft or while bushwalking, and it's those that I want to try and improve.   Ques To sum up: what is a suitable technique for ending up with 24 fps video from a 16 fps film, scanned at pretty near the highest quality available, with 300 razor points (cuts, scene changes), frame blended as per the script (every second frame repeated and 50:50 blended with the next), and each scene individually colour and contrast corrected, some of which I may want to stabilize?   As much as anything, this is a technical exercise. I want to find out the best way to do it, and I want to try it.         ... View more

Thanks for the suggestion. I'll try nesting, otherwise it will be a manual re-calculate. ... View more

Thanks for the suggestion of Split Channels. I'll start playing around to see if it's of use, but I'm really aiming for a non-destructive technique and one that I don't have to create more files. ... View more

Thanks for the suggestion. I'll give it a go. ... View more

Another question. I want all my edits to be non-destructive, including Channel Mixer. All these images are edited in PS, saved as Tiff, then saved again as max-quality jeg, 1080P, for use in Premiere.   My workaround for the colour-cast problem in grayscale Kodachrome images is to save the edited image as Tiff (as normal), then convert to Grayscale and flatten (an extra step), then save as jpg for Premiere.   Is there any other technique I can use non-destructively in PS, that changes the image to true grayscale, hopefully for all colour profiles? ... View more

Thanks for the suggestion. I've been having a detailed look at this problem, and if you have the time, maybe you can confirm my findings.   The problem with colour casts when a monochrome Channel Mixer is applied, seems to occur only when the profile is the one that I specially created for my Nikon Coolscan V ED when scanning Kodachrome slides. It's the profile that is embedded in the attached image in my first post. When I use Channel Mixer with an image that has an sRGB-type profile, such as from a digital camera, or a scan of Ektachrome, the grayscale image from Channel Mixer is true grayscale.   I have confirmed this on my system by testing a dozen or so images, and checking the luminance values with the Info tool. In all cases so far, when a monochrome Channel Mixer is applied, a Kodachrome profile results in "a" and "b" values under the Lab panel; but all sRGB-type profiles have "a" and "b" values that are zero.   Is this some sort of bug in Channel Mixer?   ... View more