First Attempt at Greenberg Effect?

[TimeActor]

when you read my posts before than you´ll see what i prefer for my style.(i wrote it down) it is not a problem to integrate those smoothing look.
In some pictures (selfportraits) in the fotocommunity you can see a little bit of them on clothes.

[TimeActor]

yes i mean you can create the "plug in Highpass" (in your words) without highpass. its a combination of other original ps filters.

[secretagents]

Flexmanta, what's wrong with the high pass filter? that's something that would be interesting to know.

With regards to flat images I am believing that there is no entirely flat images (with 100% uniform luma) you are ever going to shoot (unless you shoot in complete darkness with no flash) so that there is always some luma information that makes the picture readable that can be manipulated as to obtain an image that appears less flat from which you can then start for getting the Hill or Greenberg looks.

If I understand your desire to know Photoshop up and down and be able to do the most with it's buit-in tools, and admittedly you know it quite well, for the reasons you say, it strikes me that you categorically reject even the notion you could use third party plugins. I think that is some mental rigidity from yours you should get rid off IMO (I hope I don't offend you saying that as it is not meant to). And of course I don't mean by this to say that you should use portraiture for skin work or Nik Color effects for color adjustments which I am agreing with you and TimeActor are the perfect examples of, shall I say, useless/dangerous plugins.

[TimeActor]

you´re right when you say "useless/dangerous".
any external plugin makes you dependently i think.
i mean your style is nothing when you have not the prefered external plug in.
thats why i´m with you and say useless and dangerous too ;-)

[Der_W]

What's wrong with the highpass filter built into Photoshop is that it calculates the lightness information a little wrong (at least for a frequency separation).
If you want to really find out (and as well find a better approach) go take a look here: http://www.modelmayhem.com/po.php?thread_id=439098 (very great thread and worth reading it all).

[flexmanta]

Here is an explanation to why the high pass filter is inaccurate for frequency separation.
On the attachment you can see 3 columns. The first column corresponds to the un-separated image (original), the second column contains the same image separated with the high pass filter and the third column contains the same image separated by using the apply image command.

On the third row, you can see the differences between the original and the separated image. As you can see, the image separated with the high pass filter is very different to the original. That doesn't happen when using the apply image command, which results in a separated image that looks exactly like the original one and is therefore a correct signal separation (Fourier series http://en.wikipedia.org/wiki/Fourier_series).

[dmeadows]

I've been experimenting with the apply image alternative for the last couple of weeks, and I'm pretty damn impressed with it

[p3r10]

Originally Posted by flexmanta

Am I the only one who uses no plugins?!

No flexmanta, i don't use any plugins now. I find no need for them nor use with what i do. And yet i create better photos then those of plugins...its how you understand the core aspect of photoshop...using the right tools for the right job..

[Freya]

Originally Posted by flexmanta

Am I the only one who uses no plugins?!

No flexmanta, you're definitely not! For starters, I wouldn't have the money to buy all these different 3rd party plugins, and wouldn't spend it on them knowing that I can achieve the same results using the tools already in PS. You just need to get to know the program really well and you will realise that there is no need to buy extra plugins, and that is a plus to your work inside of PS because in that way you really are understanding how it works and how many different ways there are to achieve the results you desire.

[Quantum3]

Originally Posted by flexmanta

Here is an explanation to why the high pass filter is inaccurate for frequency separation.
On the attachment you can see 3 columns. The first column corresponds to the un-separated image (original), the second column contains the same image separated with the high pass filter and the third column contains the same image separated by using the apply image command.

On the third row, you can see the differences between the original and the separated image. As you can see, the image separated with the high pass filter is very different to the original. That doesn't happen when using the apply image command, which results in a separated image that looks exactly like the original one and is therefore a correct signal separation (Fourier series http://en.wikipedia.org/wiki/Fourier_series).

Flex, which are those techniques? Because I just know Degrunge and Splitting...

[flexmanta]

Techniques for skin? Cloning, healing, dodge and burn, inverted high pass at high radius (obtained through the apply image command) and whatever you want for color correction. That's all there is to know. There is nothing else. Splitting is not actually a technique, because a split image looks the same as the non split one. Splitting is preparing the image for the different techniques. For example, why clone all skin when you just want to clone texture?

Unless the client NEEDS you to blur or degrunge or do whatever single pass modification and puts money in your hand, i wouldn't do it... I wouldn't use it for my portfolio either, and I'd also keep it a secret.

The best techniques are to know your photoshop, to be prepared to spend a long time and to be a detail freak.

[Quantum3]

Originally Posted by flexmanta

Splitting is not actually a technique, because a split image looks the same as the non split one. Splitting is preparing the image for the different techniques.

I'm wondering what else can be done through splitting, because I just use it to fix the texture problems and sometimes, luminance problems (in the LF). Would you mind listing, or at least, wirtting a bit what else can be done through splitting?

Thanks,

Mart

[flexmanta]

Ok. 90% of the times i split an image, i delete the low freq layer. I like to keep a layer containing all the pore texture of the image in case another part of the image is lacking texture. In that case, i will cut a chunk of high and place/warp it whenever i need it. When i don't need it, i normally turn that layer off. This is normally a pain in the ass, because you have to make sure you have liquified everything that needed liquifying. If you liquify the merge and have a high texture layer saved there to use as source, then make sure you save the liq mesh so that you can use it on the high res layer. I don't know about other people, but my initial cleaning, no matter what process i used, i like to keep on one single layer. That is, i have an original layer, then a liquified layer, and then a liquified cleaned layer. Even after splitting the latter, i will again, merge all the junk together in order to have one single liquified clean layer.

The only time that i split an image and keep both low and high layers, is normally when i work with very sharp images of high megapixel cameras. I do this particularly with beauty photography, as the skin and hair are very very important. And again, once im done cleaning, I'll merge.

On a fashion photo or any other framing that is not a close up to the face, I almost never find any reasons for splitting the image.

In essence, splitting is for me useful when im going to retouch high resolution skin, so that the cloning and healing tools are more effective, and to have the possibility to retouch the low separately (say by using very subtle median on it).

Another very useful thing about splitting is to extract texture from 3rd party sources. Say you need to add glitter to lips or eye makeup. On a new document you create noise, or maybe get a photo of sanding paper, split it and use the high layer to linearly light the lips/eyes. But again, if the term splitting applies only when you keep both layers, that would not be just high pass instead of splitting.

I'm sure there are many more things that can be done with splitting. I still have to learn them though, or come up with my own.

[Quantum3]

Originally Posted by flexmanta

Ok. 90% of the times i split an image, i delete the low freq layer. I like to keep a layer containing all the pore texture of the image in case another part of the image is lacking texture. In that case, i will cut a chunk of high and place/warp it whenever i need it. When i don't need it, i normally turn that layer off. This is normally a pain in the ass, because you have to make sure you have liquified everything that needed liquifying. If you liquify the merge and have a high texture layer saved there to use as source, then make sure you save the liq mesh so that you can use it on the high res layer. I don't know about other people, but my initial cleaning, no matter what process i used, i like to keep on one single layer. That is, i have an original layer, then a liquified layer, and then a liquified cleaned layer. Even after splitting the latter, i will again, merge all the junk together in order to have one single liquified clean layer.

The only time that i split an image and keep both low and high layers, is normally when i work with very sharp images of high megapixel cameras. I do this particularly with beauty photography, as the skin and hair are very very important. And again, once im done cleaning, I'll merge.

On a fashion photo or any other framing that is not a close up to the face, I almost never find any reasons for splitting the image.

In essence, splitting is for me useful when im going to retouch high resolution skin, so that the cloning and healing tools are more effective, and to have the possibility to retouch the low separately (say by using very subtle median on it).

Another very useful thing about splitting is to extract texture from 3rd party sources. Say you need to add glitter to lips or eye makeup. On a new document you create noise, or maybe get a photo of sanding paper, split it and use the high layer to linearly light the lips/eyes. But again, if the term splitting applies only when you keep both layers, that would not be just high pass instead of splitting.

I'm sure there are many more things that can be done with splitting. I still have to learn them though, or come up with my own.

Oh... I do the same. I have heard that works very good for sharpening by using a gradient map and heard that people does almost everything on these layers... I tryed sharpening there, but found nothing better than doing it by highpassing/darken&lighten.

[Der_W]

Sharpening is usually done via a curves adjustment layer clipped to the high-layer. You simply increase the blacks and whites to give it a sharper effect. The only advantage over the original high pass is that it's a little more accurate in the shadows/highlights (as flexmanta showed before).
Personally I wouldn't do this because it also increases the noise in many cases, so what I'd do (for sharpening that is) is delete the low layer and use surface blur on the high layer to get rid of the noise. Afterwards you can decrease the opacity to lower the effect or even use the curves layer as above to increase the sharpening effect.

Another nice thing that can be done via splitting is removing crumbles from clothing (http://www.modelmayhem.com/po.php?th...8#post11387548).

[Quantum3]

Originally Posted by Der_W

Sharpening is usually done via a curves adjustment layer clipped to the high-layer. You simply increase the blacks and whites to give it a sharper effect. The only advantage over the original high pass is that it's a little more accurate in the shadows/highlights (as flexmanta showed before).
Personally I wouldn't do this because it also increases the noise in many cases, so what I'd do (for sharpening that is) is delete the low layer and use surface blur on the high layer to get rid of the noise. Afterwards you can decrease the opacity to lower the effect or even use the curves layer as above to increase the sharpening effect.

Another nice thing that can be done via splitting is removing crumbles from clothing (http://www.modelmayhem.com/po.php?th...8#post11387548).

By using a gradient map you can set the values more accurately, the result is the same I think. Regarding the highpass technique, you can use the command blend if to get ride of the blwon highlights or the darken shadows

[dvbear]

Originally Posted by dmeadows

I've always agreed with the photographers, although reading this has made me think on it a little. What you do have with Photoshop is the ability to change any pixel in an image to any luminosity value you can think of. Does this mean that the lighting of an image can be completely backwards engineered? Well yeah, it pretty much does. If you had the patience, you could paint a Greenberg image by reference pixel by pixel.

This is not the most efficient way or something retouchers look forward too but it is very inevitable. Just like how we made the jump from using darkdrooms to being able to do it all in photoshop, one day we will also have programs that will help us emulate light. In fact I'd say we're very close. But this really detoriates the creativity and hands on experience of what photography really is.

Originally Posted by TimeActor

go on and dream further!
There where a plugin that can create a hard work style like Dave Hill, Jill Greenberg and so on...hahaha good joke ;-)

Come Back to earth. i mean this not bad!
You can create much in photoshop but nothing at all!

The Look like DH, JG isn´t possible without the right light setting!

To some extent this is very true. Creating a plugin to do what Dave Hill and Jill Greenberg spend the big $$ and time to master is ludicrous. Like saying people can one day fly no?

Now imagine if the Wright Brothers thought like that...

If everyone thought like this we would still be living in caves.

Just my two cents.