Photoshop - Flatten before printing?

Discussion in 'Digital Darkroom' started by r_resnick, Nov 6, 2005.

  1. In Photoshop CS, when working with an image with several adjustment layers, is it always necessary to flatten the image before printing?
     
  2. I don't. I try to, but sometimes with many layers, flattening changes the look, so I just print the full file...no problems...maybe occasionally a little slower, but that's been it in my experience.
     
  3. The reason why your image changes color after flattening is because you're using the layers in the wrong order. If done right, flattening will not change the image at all.
     
  4. Danny, what's an example of the "wrong order" in which the colors change? I've not experience this, so I must be doing it right, but I'd still like to know more about what you're describing. Thanks.
     
  5. The main reason to flatten or merge the image is to condense the information so you can make the file smaller and more portable. If you are only going to print, there is no reson to flatten other than to shorten the time it takes to print.
     
  6. >>The reason why your image changes color after flattening is because you're using the layers in the wrong order. If done right, flattening will not change the image at all. Hogwash. There is no such thing as 'correct' if you are trying to be creative. And indeed, even if you are 'doing it right' which I'll take to mean not moving adjustment layers or their relatives around after creating them, at the point you have a lot of layers, you simply can not avoid the change when flattening...but that certainly is not to say what you have done is 'wrong'. This is a stifling approach to creativity. Shawn
     
  7. Even if you don't do it yourself, wouldn't the program have to flatten the image in order to print it?
     
  8. Conrad is exactly right. The print driver will always "flatten" the file as it preps it for the printer. Your work file does not need to be flattened.
     
  9. Flattening should not change the look. Can you upload a counter example?
     
  10. I will. I'll send the only example I think I have right now when I go home (flat/not flat screen shots). It's not something that happens often, but only when I end up with something like this: http://www.photo.net/general-comments/image-attachment?comment_id=3075162
     
  11. Flattening should not change the look unless you have the "wrong" layer selected and are basing your print evaluation on that. I usually flatten and then sharpen before I print. One makes sharpening faster, two makes it print faster, 3, ensures I'm seeing what is going to be printed. Usually I'll make a working copy, saving all the layers in the first copy (always keep the original as is and make edits on a dupe.) You can then keep or toss the printed copy.
     
  12. Shawn, what happens when you make copy 5 the active one? Does the look change?
     
  13. As promised. I have this: it's all layers visible, then flattened, i.e., there's nothing funky to alter the image like unseen (and therefore deleted) layers, or any such thing. The image layer can be seen in the link in my last post... Sorry for the grid but it's an 800meg file and it takes a long while to open (noticed my error after closing)
     
  14. Can you post a cropped and resized version of the PSD file somewhere? I am curious to replicate the problem.
     
  15. >>Shawn, what happens when you make copy 5 the active one? Does the look change? OK you peaked my curiosity. No, it didn't change. nor in choosing another as the active layer. And...very frustratingly...I've never paid attention to this before but... Once I'm done, I copy all layers (copy merged) and paste a 'whole' layer on top. In other words, I give myself a layer to destroy with sharpening, without ruining anything for the future. IT DOES THE SAME THING - screws up the look in exactly the same way as merging/flattening! These may seem little difference but they are very annoying to my eye, espscially when the whole raison d'etre for something like this IS the tonality...grr... Now I'm unhappy. That means it's time to cook some spaghetti and make little slurpy sounds just so my cat goes nuts... Shawn
     
  16. Emre, if this works, maybe upsize it to 240dpi and scale it such that it's just over 800meg...see if it changes the look when you flatten it then? No idea, but it's definitely a size thing, it seems...it flattens file at pee wee sizes... Don't know if this will work - a zipped psd (PLEASE delete this once you're done, Emre, I mean from the server here so it's not wasting space - do what you will with it, though...) If this upload doesn't work...I can zip it and email you if you want, say a 9 meg file? I'll email you if you'd like. (You are a moderator here, no?) If not, moderator please delete my post if it's too pregnant...
     
  17. oq

    oq

    I've ran in to the same problem several times, last time was last week. That was a b&w photo with about 9 curve adjustment layers. It usually happens when I work with photos that need a lot of work and use many adjustment layers all doing major changes in themselves. my best guess is that it's some averaging error in photoshop, but I really have no idea as I'm not by any means a coder.. What I do is quite simply to try to keep the amount of layers down. As I only use curve layers and maybe an occasional hue/saturation layer and I see it changes after mergin I simply add another curve layer to correct the change, as it often is quite uniform or within a 'hard-worked' area.
     
  18. Does this happen when viewing at 100%? It could be caching issues with the screen image being generated from smaller resolution cached version causing round off errors which do not occur in the final image. Have you tried flattening the image in subsets of layers to isolate the cause of the compositing error? It could be a PS bug. some thoughts, Sean
     
  19. I got the file, but it is flattened; I need the layered one. Can you try uploading it to rapidshare.de? My guess is that your computer does not have enough memory for the layered file, and Photoshop is cutting some corners.
     
  20. One other suggestion, just for curiosity sake. Why not try making each layer invisible in turn, one at a time so all the other layers are visible and see if you come accross a setting there that gives you the image you're seeing flattened. I'm not sure what to tell you if you do re-create it that way, but maybe that will give you a hint.
     
  21. More likely is I screwed up LOL. I uploaded it to a friend's server. Lemme know when you've got it so I can delete it. It's very small scale, relatively speaking, and that might be my problem - the originals I'm dealing with are (layered) almost always 1/2 to 1+ Gig.
    Try HERE
     
  22. Last but probably not least, you might just say PH**k it and print it without flattening if you get the result you want. does it really matter at that point? Of course, it will matter if you send it to a service with a Noritsu or light jet as they often just want flattened images. Sorry if none of this helps.
     
  23. That's a good idea Barry. With respect to the post and poster, I'd be interested in figuring out what is going on, but not so selfishly (my file can do what it will...) ...is this happening to anyone else?
     
  24. Shawn, Take your demo image, flatten it at 100% zoom, then do it again at 50% zoom. Your "changes" are in facts display errors caused by accumulated numerical round off errors from the approximations of your image data in the caches at below 100% zoom (viewing actual pixels). You do get some boundary effects in flattening, but expanding the canvas by 20 pixels in each direction before flattening and then cropping it back to size will fix that. But since this only affects an area a single pixel deep at the edges of the image it matters little if you matte your prints. Also, these boundary affects only have a magnitude of 1 or 2 out of 256. In short, you will not see it in a print. You can find more info on potential causes of the boundary errors at: http://www.mathworks.com/access/helpdesk/help/toolbox/wavelet/ch06_a20.html To see this for yourself, flatten the image, select all, copy all, go back two history states and paste the flattened layer as the new topmost layer. Next take the new layer, set the blending mode to difference, and then create a levels adjustment layer above that. Try to find some differences. If the boundary errors really bug you, then expand you canvas by a few pixels in each dimension, flatten, and then crop back to your original data. Now it should all match up perfectly. hope this helps, Sean
     
  25. OK very cool! So, in a sense, it IS a size thing...just not something I'd have noticed merely BECAUSE OF the size. Thank you very much. I'm actually going to bookmark this page so I can use it over the aeons... I'll have to try it on the original Mammoth Mama. Let you know the results. I hope this was helpful for others, too:) Seeing as I stole the thread and made accusations redressing someone for doing it 'the proper way'. See...sometimes logic is better than a haphazard approach masquerading as something 'higher'... Shawn :) Thanks Sean!
     
  26. Late post on this, but I loaded CS2 last week, was using CS and I was working on a file Friday and it had several layers and when I went to flatten, the same loss discussed happened and I tried several things I think it occured when there were several adjustment layers some of them levels. I sort of got around it by creating layer slices when I wanted to work adjust levels on just a small area, such as various highlights. I was basically playing and am not sure how I was creating little "mini" layers, but this seemed to solve the problem and make for a smaller file size as well.
     
  27. Barry, Did this occur when you were viewing at 100% zoom? If not, then caching is likely your demon. The caches in PS are used to speed up display and this can lead to display errors (what you see is not what you have) so when you flatten you get what you have rather what you saw. There is no bug here, just a facet of using caches to create a rapid approximation for screen display. enjoy, Sean (who would like to see an example where this occurs at 100% zoom)
     

Share This Page