Photoshop :: Merge To HDR Pro Preview And Final Result Very Different
Oct 14, 2012
I'm using CS5 and beginning to use HDR. When I merge to HDR pro and make adjustments to images the preview window is extremely different from the final image. I'm merging original camera raw images from a Nikon D200. Workflow in ACR is set to 16bit. Output from HDR pro is also 16bit. I adjust using the preview window, but after final processing the resulting image is grossly different, much more saturated and more vibrant among other differences.
In CS3 I use a Dust & Scratches filter on a layer in a high resolution (3200+ppi) 16bpc image, then traditionally change the layer blend mode to Difference, make a duplicate of a channel which has the clearest information of where the filtered changes were made, and use levels to increase the contrast on a copy of that channel. Occasionally I get this result though - the Levels adjustment preview shows spots clearly, but clicking OK results in something very different from the preview. This only seems to happen occasionally.and result looks like this:
As the purpose of the channel is to make a mask, the result isn't as good as I'd like. I assumed that the problem might have been the grey starting colour of the channel, so I pushed the 0 up to 9 in an attempt to get a uniform K value of 100% on the darkest areas, and then ran a second Levels adjustment to make the white spots brighter. The second levels adjustment resulted in pretty much the same result as above, again, not what was shown in the preview. I also assumed that it may be down to numbers and that levels is occasionally too limited to deal with all the shades of grey on a 16 bit channel. Probably wrongly as this normally works exactly as required.
This only happens occasionally, and isn't that big a deal - I can quite happily go and use a Curve instead. I'm just wondering why the preview in the adjustment window is different from the actual result. The fact that it only happens on some of my images is what concerns me - purging all beforehand does not make any difference.how to get accurate results in the preview? Is the preview using a different method to show the results? Unfortunately I won't have a copy of CS6 to try this on until a couple of weeks from now, so not sure the best way for someone to try and replicate it on newer versions.
I've recently tried using the Photoshop HDR feature. I have selected three different exposures in Lightroom and chose to merge them as an HDR image. All went well until the preview. I've found some settings that work really well. I've decided to go for the 16-bit path, not using Lightroom's 32-bit image capability, and let Photoshop do the tone mapping for now.
Here's what my preview image and the entire tool window looks like: [URL]
And this is the result one step later. It's totally over-saturated and has other merge errors. It basically doesn't even look similar to the preview. What's the preview good for if it doesn't show what I'll get? How should I decide what setting works good if I can't see it? [URL]
This is Photoshop CS6 x64 on Windows 7 SP1. The source images are 16-bit TIFFs as they were exported by Lightroom.
I'm trying to create a star icon with lines through it. I used the star shape builder tool and then the line segment tool, but the lines are not being recognised as part of the star shape. For example, when I try to turn the star, the lines remain in the same place. This is pretty much exactly what I am trying to make, but I can't form the segments into one cohesive, editable shape.
I need to make a bunch of barcode tags for a client. I know CorelDraw isn't the best tool for this, but it's what I have and this is a one-time job.
To partially automate the process, I'm trying to use print merge to make the parts of the tags that aren't barcode, but I include the text that belongs in the barcode on the page. The print merge works correctly, I get what I expect. The next step would be to cut the barcode text, choose Insert Barcode, and then paste the barcode text into the barcode dialog.
Except, when trying to paste the text intended for the barcode I discovered that I was actually copying the <var1> field--not the text. I verified this by copying the "text" from the print merge document and then pasting back into the same document--what I got from the paste was not the expected text, but <var1>.
Since the barcode type I was using does not support the "<" character, it gave me an "invalid character" error.
My attempted solution was to publish to a PDF file and then import into a new document. This worked except for some reason CorelDraw makes a new layer named, inexplicably, "Layer 1" (there's a "Layer 1" by default, why make a second layer with the same name?) BUT when the import is done the new layer is NOT chosen in Object Manager, the original "Layer 1" is. Which means, when I go to copy the text and insert the barcode it ends up UNDER the other parts of the barcode tag which are on the NEW "Layer 1".
In Cs6 whenever I touched any of the sliders the image in the preview panel changed quite nicely, now when I start the HDR Pro it comes up with a black image in the preview panel. Thats easy to fix though I switch from 16 bit or 8 bit to 32 bit the preview shows up and I can click on the 16 or 8 bit modes and it shows a preview.
Then when I adjust any of the sliders nothing happens. I can play with the sliders as much as I want nothing happens. I have been using the 32 bit mode in cunjunction with the 'tone in ACR' setting which gives great results, but now I wanted to play with the scott5 preset. If I click ok it creates the image beautifully. So not only won't it preview but my image is completely different to what is in the preview panel.
Why cant I save a cdr job with all the merge fields setup so that next time I open the file I can go straight to 'merge to new document'.
So Ive set up a 10 to view raffle tickets job. Inserted 20 numerical merge fields, positioned, resized, applied font size color alignment etc etc etc.. Takes ages..... So then merge to a new document, works nice and I can save it fine.
There are a hundred reasons why i may have to close the 'master' merge file before the job has been successfully printed. It may be a repeat print job.
Any which way I really dont get why the live merge data fields are lost on closing the file.
I created 61, 9 layer composites that are text only, but I can not figure out how to export it with the alphas for use in keying in Final cut. I exported as Targas, psd's and eps's but nothing give me the alpha and I end up with a flat image that I can only to an add key with in Final Cut...this loses all of my shadows and glows, etc......
I have layers for different possible end picture effects. eg text in different colors some of it on slopes etc. I would like to send the client a few different choices without having to flatten or "merg visible" layers. If I do that then I loose the possibility of changes that the client might want.
I really like how this picture came out and I was wondering how I could get a similar aspect? I know you can play with curves, levels, and saturation but I'm sure its more then that ..
I've been asked to do a poster A2 size 594mm x 420mm 300dpi but when i enter these dimensions my pc is slowing down dramatically. whats the best way to achieve better performance?
If it's possible to attach File Info such as Copyright to a (final) .png image? I could have sworn that I've done it before but just now it didn't work.
I'm making an ad with alot of bright colors. It's gonna be printed to a glossy photo paper for distribution, but I'm having a hard time getting the monitor colors to match the printout colors.
I started as RGB workspace, then once finished I converted to CMYK, but the colors drastically get dull.
I know that it is normal for colors to change during RGB to CMYK transition, but is there any way I can preserve those bright colors?.. (it's mostly bright orange and green.)
A feature notably absent from the Oil Paint filter is any way to temporarily disable the effect (e.g., not preview it).I happen to like using Oil Paint to make refllections in the surface of water smoother and more "liquid" (which involves turning off the "Shine"), and it would really be nice if I could easily gauge the amount of the effect easily by blinking back and forth between original and processed image with a quick "disable preview" functionality.
With most standard filters, one can either uncheck a [ ] Preview box, and/or click the mouse button down on the image to temporarily show what the image is like without the effect.Yes, I know I can turn the Stylization down to 0.1 and see how a very minimal amount of the effect can be applied. And I can look at the original image in Photoshop proper, but of course that can't be zoomed while the Oil Paint filter is active.
Also, the zooming / panning seems kind of stiffly implemented.In summary, the more or less standard and expected filter features I'd like to see added to this (new class of) GPU-accelerated filters are:Preview/No Preview capability. It doesn't have to update Photoshop, just the preview display in the filter itself. I suggest a [ ] Preview checkbox and/or an on-mouse-click temporary effect disable.
Modal click and drag Zooming vs. Panning - i.e., based on a Zoom or Pan function selected (e.g., via an icon along the left) vs. having to hold modifier keys down. I personally would want it set to Zoom by default, not Pan. Since the filter has scroll bars (a Good Thing), I would have thought Zoom would have been a better default as it is. I have implemented these things in my own GPU-accelerated plug-ins, so I know that all the infrastructure Adobe needs to support them is already in there. These changes are controls only and should be trivial to implement.
I have been trying to save images from one file to a final modified one. Two problems:
1) Some images are unedited and save sometimes but not always. Why? 2) Some are a combination of an original plus a transformed close-up from another image. When the finished edited image saves, it is much smaller than the few unedited images that did successfully save. Why?
The images are digital photos of my jewelry. In order to decrease the number of pages on my impending website, I am attempting to transfer, when one is available, a resized close-up to the other shot of the full set (close-ups are what sell the work!). Some of the images show 25% on the top bar but that doesn't seem to make a difference as to whether it will save or not. Others (I notice they seem to be more horizontal images) are 16.7% ? I tried to upload example images to illustrate what I am trying to do but kept getting an error message.
Is there a way to see the end result of a pattern as you are creating it?
That is a way so you can see the pattern tiled all over the document. So say you are working on a 100x100 area to be the pattern in layer 1, but you want to see this replicated over an entire document. Maybe you have an overall image/document that is 800x600 say, and you work on the 100x100 area in a layer. Anyway some way of doing is so that when you make a small change in the 100x100 area of the layer you're using for the patter creation, you immdiately see the replicated result of the 800x600 area.
Scaling on a pattern layer sucks. When you scale a layer up or down (anything other than 100%), the end result is blurry and ugly. Here's an example: URL....
I would love to be able to choose the scaling algorithm used. In this case, the raster layer was scaled down with a bicubic algorithm, which worked fine for it.
I was scanning some old family photos and came across one that isn't an original photo but looks like it was printed on one of those little remembrance cards that are passed out at funerals and it is made up of little dots. The scanned image ends up with a kind of plaid pattern on it because of the dots:
I took a picture with my cell phone and it looks better.I have never tackled anything like this before and was wondering if there is a standard way of improving this kind of picture to look more like a photo.
I cut with the pen tool. The path was not complicated and had many straight edges to follow, so I didn't have to click many points. After I completed and closed the path, I sel the icon, load path as selection and then the path became this bizarre zig zag, almost electric marching ants effect. Then I deselected and there were like hundreds of anchor points at every centimeter or so. The edges of the selected area however were normal and smooth.
Save For Web offers 5 resampling algorithms (the same 5 as in Image Size, ignoring "Bicubic Automatic", of course) for when the output is being resized.
All choices except "Nearest Neighbor" output an identical image (which is different to all results of Image Size). Although the "Bilinear" output contains the same pixels as the output of the 3 bicubics, its file size is slightly different than their shared file size.
I have a big problem in my adobe software (Photoshop and Fireworks by-example). When, for example, I backup my project psd, short layers that I did well at the opening of what I save almost all my layers are disorders .
I'm trying to optimize the rendering time on pre-rendering my Final Gather map.
I've set it to every 5 frames and halved the resolution of my FG map compared to the final render but I'm wondering if some of the other roll-outs such as the Image Precision, Soft Shadows, Glossy Reflections and Refractions make a difference to the FG map. Can I knock these settings down a notch or two, or do I have to keep them the same as the final render?
After I installed design 2011 student version, I can't move the final gather presets or for that matter there are no presets available to change. And the values go erratic often. I switched to scanline and back to MR but that doesn't rectify it.
And also there are no geographic locations available in my daylight system
I am working with a network and it works for standard rendering.
I want to use it to pre-generate the FG maps. I have tried the "generate now" button, but that makes the FG calculate locally on the current machine... even with Net render checked.
When I try doing a render using the frames just as junk, it renders out, but does not save the FG map.
I have copied an pasted a pdf into model space to show in a viewport for the finished sheet. The pdf appears just like I want it in the viewport and when I do a preview it looks fine. However when I plot the final pdf (dwg to pdf) the viewport is blank. But, the pdf shows up in at leaset one other viewport on the sheet.
I recently bought a Nikon D7100. I took some test shots and imported the images to Lightroom 4.4, then generated 1:1 previews for all the images.
Previews zoomed to 1:1 in Library Mode are noticeably soft. The image below is a screen shot of the Library preview:
Moving to Develop View produces a higher-resolution 1:1 preview. The image below is a screen shot of the Develop preview:
Look at the bricks and window screens to see the difference.
This difference occurs immediately after import (with a User Preset applied during import). Once any Develop work is done, the Library preview updates and displays at full resolution.
This problem makes doing initial editing/selection of images time consuming, because I can't determine the sharpness/quality of imported images until I've done some kind of Develop adjustment on each image.
