Announcement

Collapse
No announcement yet.

Why I hate frequency separation

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Re: Why I hate frequency separation

    I was just wondering, is there anything else like this technique in the field of skin retouching? I mean, this HP technique, in europe come up like for 3-4 years even if most of the professionals knows it for more years.

    What else is coming up to the ground? I understand it's always the same stuff and the traditional approach is the best, I work with Photoshop like 15 years and I can tell that with the basics you can do everything but you see what I mean....HP is like a trending technique...so just wondering, what else will show up?

    A little OT, what do you think about the new Path Blur in CC 2014? I like it but I was hoping to see something more close to Virtual Rig with more control over the path.

    Comment


    • #17
      Re: Why I hate frequency separation

      HP is not a technique. What you do with it can be.

      Just like layers.

      Comment


      • #18
        Re: Why I hate frequency separation

        HP is a technique because its a set things that combined can make something right? So for me it's a technique. Layers are a function/tool of Photoshop. My personal point of view...

        Comment


        • #19
          Re: Why I hate frequency separation

          Just like a wrench is not a technique, but using it may be, so is the layer not a technique but a tool.

          Comment


          • #20
            Re: Why I hate frequency separation

            Originally posted by Isedo View Post
            Hi guys, have you noticed that frequency separation become some kind of "overused"technique in skin retouching?
            Since it's out, and more in these days, everyone is teaching and using FS. It makes me feel like the new HDR when everyone just use it and use it and use it.

            Do you know what i mean? It's an awesome technique but it's "to much" out there. It'a technique used for many years right now and still people use it as a "innovative" way to retouch skin....

            It's quite boring...what do u think about?
            I think it's largely the province of the online retouching community. I don't know any working retouchers who use it, at least not to the degree that numerous online tutorials would leave people to believe. Better results can be achieved with simpler methods, and in my experience, the less complex solution tends to be the most realistic-looking one.

            Comment


            • #21
              Re: Why I hate frequency separation

              Originally posted by byRo View Post
              Not really..
              The 3D bump map serves exactly the same purpose as the FS High Part, It keeps the fine texture separate from the overall shapes and colours.

              If I still remember how to attach images, you should find two examples here. One is a FS High Part and the other a 3D Bump Map. I think you'll see that they are pretty much the same thing.

              In fact I sometimes think that I could apply adapted and resized 3D Bump Map images to faces that have lost all skin texture - thus "inventing" a missing FS high part.

              I don't understand the reasoning here. I wouldn't want to map that to a 3d model. The brows would come out weird. I mean I get what you did. You took it into whatever 3d program and projected on a dummy head of some sort, then brought the resulting bitmap into photoshop. I just have no idea why you set it up that way or why anyone would use that for a 2d image. It's just used to convey extra depth in CG images without the rendering expense of extremely fine displacement mapping that would require excessive tessellation levels at render time.


              Originally posted by Isedo View Post
              It makes me feel like the new HDR when everyone just use it and use it and use it.
              There are interesting uses of the principles of HDR. Have you ever looked at any of the open source raw processors? If you had a completely linearized workflow, you would have a lot of interesting control when it comes to separating out elements in a scene. You see many photos with a more surreal quality linked on here where people have taken distinctly different densities by region without going for a cartoonish effect. It's much much easier to do if working with linearized data, and ICC v4 profiles can actually support negative tristimulus values, so you have less to worry about when making intermediate adjustments.

              It's also used in film to generate reflections on computer generated components. The problem I see here is that you're examining its use by people who just want to toy with their images.

              Comment


              • #22
                Re: Why I hate frequency separation

                Originally posted by klev View Post
                I mean I get what you did.
                Unfortunately, klev, it would seem ... not quite...

                I didn't project anything. The 3D Bump Map I attached is just a "ready-made" image which is used with a common commercial 3D character (Daz3D V4).

                Originally posted by klev View Post
                It's just used to convey extra depth in CG images without the rendering expense of extremely fine displacement mapping that would require excessive tessellation levels at render time.
                Yes, that's all I was trying to say. In 3D, as in FS, the fine detail is treated separately from the overall shape - so, it's no wonder the two different (3D Bump / FS High Freq.) images look similar.

                Back to the subject of FS....
                I've watched some of the online videos that "sell" FS and, as isedo said in the OP, they are just cashing in on this "new" technique. In these videos they separate out the High Freq. part and the just blur away all the rest - throwing out much of the face's structure along with the "grunge".
                To me, a better way would be to separate into three frequency layers:
                1. "High", with pore texture etc
                2. "Mid",mostly grunge
                3. "Low", facial structure

                Yep - deGrunge again, now in "innovative" FS form!!!

                Comment


                • #23
                  Re: Why I hate frequency separation

                  Originally posted by byRo View Post

                  Back to the subject of FS....
                  I've watched some of the online videos that "sell" FS and, as isedo said in the OP, they are just cashing in on this "new" technique. In these videos they separate out the High Freq. part and the just blur away all the rest - throwing out much of the face's structure along with the "grunge".
                  To me, a better way would be to separate into three frequency layers:
                  1. "High", with pore texture etc
                  2. "Mid",mostly grunge
                  3. "Low", facial structure

                  Yep - deGrunge again, now in "innovative" FS form!!!


                  Actually you're not that far off. A fair amount of academic research has been done in the area of both decompositions and convolutions. The uses would be contrast enhancement when normalizing an image, such as an HDR backplate, and edge finding, both without introducing visible ringing/haloing to the image.

                  Originally posted by byRo View Post
                  Unfortunately, klev, it would seem ... not quite...

                  I didn't project anything. The 3D Bump Map I attached is just a "ready-made" image which is used with a common commercial 3D character (Daz3D V4).

                  Whoever made it may started it that way after laying out their UV map.

                  Comment


                  • #24
                    Re: Why I hate frequency separation

                    What's wrong with a time saving measure? Hey, it's not push button. Results can be obvious, or not.

                    Your employers would not be happy that you aren't embracing a quicker way to do things, and you freelancers are going to lose clients if you insist on charging for five to ten more hours to essentially get to the same place.

                    Comment


                    • #25
                      Re: Why I hate frequency separation

                      Absolutely agree with Benny. For an editorial, we have sometimes less than 2 days to retouch 8/10 pictures. It could be different for advertising but 10 hours for one picture in editorial... it's impossible. If your employer is agree....you have a lot of chance!

                      Comment


                      • #26
                        Re: Why I hate frequency separation

                        8/10 images is borderline impossible. Maybe if it's a really clean shoot, or if it's some minor work. Just setting up the conversions and masking out 10 images in detail will take that long.

                        For beauty 1-2 images per day is max.

                        That is if you're the only one doing the actual work.

                        Edit: By masking out I mean setting up color/contrast etc.
                        skoobey
                        Senior Member
                        Last edited by skoobey; 07-15-2014, 08:48 PM.

                        Comment


                        • #27
                          Re: Why I hate frequency separation

                          Yes, we have to be very very fast but one more time, it's for editorial, not beauty campaign.

                          For me, FS is a very good way to do in 5 minutes with exactly the same results (with some dodge and burn after) as taking 30/60min of dodge and burn only.

                          Comment


                          • #28
                            Re: Why I hate frequency separation

                            But fashion is all about making selections. I mean yes, I could make it look more finished in half an hour per image, but without precision, it won't look as good.

                            Half an hour to an hour is for lookbooks only I'd say.

                            Comment


                            • #29
                              Re: Why I hate frequency separation

                              Originally posted by typiac View Post
                              Yes, we have to be very very fast but one more time, it's for editorial, not beauty campaign.

                              For me, FS is a very good way to do in 5 minutes with exactly the same results (with some dodge and burn after) as taking 30/60min of dodge and burn only.
                              I always laugh whenever anyone quotes a full day of work for an editorial image. They don't even mention how revisions or anything, but they would never keep up with their work at that pace. I don't understand how you would cut 30/60 minutes down to 5 with FS. No one has been able to show me that thus far, yet I've read similar claims from others. The past ones simply indicated people who overdid the work, thus making it take longer than necessary.

                              Comment


                              • #30
                                Re: Why I hate frequency separation

                                Originally posted by klev View Post
                                I always laugh whenever anyone quotes a full day of work for an editorial image. They don't even mention how revisions or anything, but they would never keep up with their work at that pace. I don't understand how you would cut 30/60 minutes down to 5 with FS. No one has been able to show me that thus far, yet I've read similar claims from others. The past ones simply indicated people who overdid the work, thus making it take longer than necessary.
                                First at all, I didn't say that I take one day for an editorial image. I said that I have one day to retouch THE ENTIRE editorial!

                                Winning more than 30 minutes could be very easy with FS but once again, I said we always need DB after. FS could be necessary...or not, exactly like DB. It depends of the photographer, the model, the lights, the kind of magazine, the country...

                                Comment

                                Related Topics

                                Collapse

                                Working...
                                X