R A V E N D R I V E wrote:
Thanks

No problem
Post your results

Beautiful Sundays wrote:
I posted a step-by-step on the original page, and just saw your action plan.

Concerns:

1) Gauss on a standard layer. My question is why knock yourself out to improve upon High-Pass and then compromize it with a fixed Gauss? Your action allows no flexibility except opacity. Suggest Smart Object/Smart Filter w Surface? (thereby infinitely adjustable).

This is by no means THE way to do it. Just an easy to follow step-by-step separation workflow.

But the gausian blur is done on a smart object layer.
Maybe you're still running the first action I uploaded today.

If you'd like a surface blur there, you can easily change the action.

Didn't want to scare people by baking in a surface blur that will take 5 minutes to finish

Beautiful Sundays wrote:
2) HF healing layer....Angela's blank layer plan is fine...she simply forgot to delete the 'eyeball' selectively (current and below samples VISIBLE layers as you know...no eyeball, no sampling)

True, no eyeball, no sampling.
Also, no eyeball, no accurate visible reference.

Because the only layer you want to be sampling from is the HP layer.

This method was to give the same result as the blank layer while maintaining an accurate visible reference of the shot.

Beautiful Sundays wrote:
3) All of the curve adjustment layers seem to me to be 'filler'.....nothing wrong with filler mind you, just would trade them all in for a smart object LF layer w SB.

I can get a better result by doing some good old D&B-ing than by using blurs for everything.

Beautiful Sundays wrote:
Other than that pretty cool....impressed that you took the time

Thought I wouldn't?

Lanenga wrote:

This is by no means THE way to do it. Just a easy to follow step-by-step separation workflow.

But the gausian blur is done on a smart object layer.
Maybe you're still running the first action I uploaded today.

If you'd like a surface blur there, you can easily change the action.

Didn't want to scare people by baking in a surface blur that will take 5 minutes to finish


True, no eyeball, no sampling.
Also, no eyeball, no accurate visible reference.

This method was to give the same result as the blank layer while maintaining an accurate visible reference of the shot.


I can get a better result by doing some good old D&B then by using blurs for everything.


Thought I wouldn't?

1) Smart object now? Much better:) If it were me (it's not) I might mention that Gauss is faster but cruder, and that SB is an alternative...does take a bit longer but so what..     

2) Re visibility...actually is ok on HF alone OR with an original underneath (edit..Richard Vernon suggests using the HF alone @ 100/100)...the concern w the vid was that Ang was sampling the LF layer...yikes...

EDIT--BETTER METHOD ABOVE

3) DB/curves layers are part of every portrait workflow. My point was that they aren't necessarily a part of the FS action plan. If you are posting an FS plan, then they are 'filler'; if you are trying to post a 'complete' skin-portrait action set then maybe...but not exactly complete...see what I mean?

4) Impressed that you took the time? A little...you were basically telling OP yesterday that his repost was redundant, yet you took the time and energy to make a cool action set (although you did indeed create a new thread for yourself).

I'm glad you did...I will tweak it to my tastes (as mentioned), and use it often (with yr permission of course).

Beautiful Sundays wrote:
"It's not just taking, you'll have to give something too."

I assume you refer to yourself....yes indeed posting this action set was a nice 'give'.

I 'gave' by posting a written plan on the original thread.

Angela 'gave' by creating a video which helped launched this whole thing.

OP 'gave' by tempting Angela into making the video.

All-in-all the world seems a better place

I "gave" Carlos and Photons a hard time, which caused them to examine my crude techniques, which, then....never mind.

Seriously, nice work everyone.  I for one appreciate the time taken in the last 12 hrs by Lanenga, Photons, and Beautiful Sundays(what are the other days of the week like?) to further refine and simplify this process. Today is truly a "beautiful sunday."  Off to shoot a couple of fashion models.

Angela Michelle Perez wrote:
Hey I re did the video because I was informed of my mistake in there  so the new video without that is here, Instead of working on a blank layer I worked on the high frequency layer itself as to not confuse anyone by my mistake.  But yeah the previously one was deleted so you might not be able to see it on the link in your OP. The last thing I want is for people to become more confused by my mistake so I tried to show the simplest most basic way to do it in the retake.

New link if you want to switch it
http://www.youtube.com/watch?v=dMyaeZmkZD8

Added the new link!
Thanks

Neil Snape wrote:
Nice collection of tutorials here and thanks for the actions.

I still don't get it. By using this you are making a grey layer with a mask with modulating density from a previous layer by using apply.

So you have your so called low layer and the HP layer.

So if it is true that you are using the HP layer to affect only texture, the low layer if retouched separately will be punching holes in the colour info, and vice versa for the HP.

To me this would be pretty close to using an Lab copy and hacking the L layer, except with this HP technique at least you don't have an green magenta yellow blue layer.

Using just 2 frequency layers, a High and a Low, might look like the same as using LAB. But it's not since LAB completely splits Luminosity and Color which this frequency split will not.

For greater detail on the how and why I would like to point you to the HighPass Sucks (+ solution) thread by SRB Photo.

And if you don't want to read all of it, you could start Here

It's a bit of a read, but it contains the information.

This thread and action are meant as an easy step by step intro to what Frequency separation could do.

Frequency separation will also allow you to split your image in more than just 2 frequency bands.

DerW mentioned a technique here

Neil Snape wrote:
I'd like to see some hi-res samples of what can only be done with this to see where it is useful. So far onscreen res to me looks like plastic zombies. Yet a lot like that look so that is okay too.

I personally don't like to work on the Low layer.
Maybe change a little hue here and there, but for smoothing out the shot I still prefer some good old D&B-ing

But I am a big fan of the control over the HP layer.

rey sison photography wrote:
Short hijack:  Carlos, after this thread, can you start a new one called, "dodge and burn sucks." Hijack over.

Haha no. D&B RULE!

rey sison photography wrote:
Still loving this technique. I taught these steps to another photographer last night. It was amazing that I could recall everything without having to refer to notes or refer back to this thread. It's all so logical once you understand why things are done.

When can we see some applications?

Neil Snape wrote:
So if it is true that you are using the HP layer to affect only texture, the low layer if retouched separately will be punching holes in the colour info, and vice versa for the HP.

Your theory would be correct, were that the premises were true [not your fault, you're seemingly basing on what you watched / read - it's just bad info].  In fact, frequency separation (at least every form which I would advocate for 'normal' retouching) should always have color in both the 'High' and 'Low' frequency layers.  It's just that, whereas you can think of the 'Low' layer as being either additive from black or subtractive from white, the 'High' layer is defined by its difference from middle gray.  And if you look closely (more closely than most do), you'll see that in fact it has a number of subtle color variations within it - created by those colors' components within the spatial frequency range.  That is, the larger the 'High' frequency swath is (say, 20px vs 5), the more color will be contained in it.  In fact, if you had a small image (say, 250px on the long axis), and did your FS at 250px, you would end up finding the Low layer to be generally gray while the High layer would have all the color in it.

I've said before and will reiterate again (my apologies for not being around to do so earlier when we might have prevented some of the new confusion):  except in extremely rare cases, if you're using FS to clone / heal / patch / CAF / etc., it is not a good idea to fully separate color and luminosity (or 'texture') - it will not give a natural result at full resolution.

SRB Photo wrote:
it will not give a natural result at full resolution.

Thank you

I go away for a weekend and 3 threads happen!

IMHO - overuse of blur in the LP looks awful for beauty if not done selectively and I mean REALLY selectively

Texture is not just one radius, when you blur everything but one radius it looks fake.

As someone mentioned before there's ppl who will notice the difference and ppl who won't

x

Lanenga wrote:
To those of you that have downloaded this action:
I just realized I accidentally set the D&B Curve Adjustment layers blending mode to soft light instead of Normal.
So I fixed this in the action.

You could change it yourself or download the action again(Same link)

Thankyou. 

Lanenga wrote:

Haha no. D&B RULE!


When can we see some applications?

Ok, just for you Carlos. Had to go easy on the photoshop because this is mainly for the MUA portfolio(built in excuse). This was a challenging complexion for me. Actually, they all are. Even tried to do a little D&B. Hopefully, it's good enough for government purposes.

Before


After

rey sison photography wrote:
Ok, just for you Carlos. Had to go easy on the photoshop because this is mainly for the MUA portfolio(built in excuse). This was a challenging complexion for me. Actually, they all are. Even tried to do a little D&B. Hopefully, it's good enough for government purposes.

Nice job!
You are getting there

Neil Snape wrote:
The most important question , which I've asked many times, is show me some examples of what you need this particular method for, where it is superior or the only method to use over the way I currently do it, and have done so from the beginning for many many spreads and ads.

It seems the method is not hard to do, perhaps the underpinnings are out of the visual domain and quite technical.

There are so many ways to do things in Photoshop to come up with good results. Sometimes there is only one way way that achieves correct results.

I'm sure that HP is very useful when it is the one thing, the best method. I just want to see where.

Natalia's sample video linked here shows, the application in hair shadows.
Is this a case where HP is the only tool, or the best tool/method?

The other video for skin ( not Natalia's) is not convincing, and although may be a fast hack, isn't for me.

You don't need it for anything. (maybe in forensics as I believe that is how all this got started).

It's just another way of doing things. And if done correctly it could speed up your workflow.

We're not saying this is the way to do it and the only way to do it. We're just showing another technique/tool.

Try it, see if you like it. See if you can find use for it in your process and if not, then put it away again.

You could do entire retouches with just a brush(if you're a really good painter). It is just a matter of preference, time and control.

Lanenga wrote:

You don't need it for anything.
It's just another way of doing things. And if done correctly it could speed up your workflow.

We're not saying this is the way to do it and the only way to do it. We're just showing another technique/tool.

Try it, see if you like it. See if you can find use for it in your process and if not, then put it away again.

You could do entire retouches with just a brush(if you're a really good painter). It is just a matter of preference, time and control.

What can I say that I didn't already say?
I'd like to try it it but would like to know why and where it is useful.

I have not argued in any way what it is for. I can't , I'm ignorant as to what exactly it is for. Hence , show me , I am quick to learn.

The above posted sample is proving that the method can do something.If I say anything it will be an unsolicited critique.

Neil Snape wrote:
What can I say that I didn't already say?
I'd like to try it it but would like to know why and where it is useful.

I would like to suggest to do some more reading on what it is.
then give it a try so you understand what it is and what it does

And when you get it and you like it, you might find a way to use it.

Neil Snape wrote:
I have not argued in any way what it is for. I can't , I'm ignorant as to what exactly it is for. Hence , show me , I am quick to learn.

If you won't try it you will never learn.

If you tried it and you get stuck you could ask new questions so we can help you further

If you tried it and you don't like it. Then you could share with us what you did, why you didn't like it and/or what you would change.

rey sison photography wrote:
Not that it matters now but I inserted the wrong after photo the first time. Just switched it now.

Neil and anyone else who has doubts. Don't use what I just posted to judge the technique. I am not a high end retoucher. Just someone new trying to learn as much as he can.

LOL don't worry, finding techniques that work for you is the most important thing. All of my recent pictures were done to please me, not a client as they had been done in the past. 

Adding new tools and techniques is an ongoing thing for all.

High end retouching , in my terms would be those who have obligations and understandings of the destinations, and client expectations which is in most part not the concern of us on MM.

Neil Snape wrote:
I'm sure that HP is very useful when it is the one thing, the best method. I just want to see where.

MP Retouch wrote:
I love this technique for digitally pressing clothes. During the initial separation, make sure all the detail/pattern on the clothes is on the HF layer. Clone/paint/d&b on the LF layer to remove the larger tonal values of the wrinkles, then clone on the HP layer to remove left over HF wrinkle data.

https://www.modelmayhem.com/po.php?thre … 98&page=18

Neil Snape wrote:

What can I say that I didn't already say?
I'd like to try it it but would like to know why and where it is useful.

I have not argued in any way what it is for. I can't , I'm ignorant as to what exactly it is for. Hence , show me , I am quick to learn.

The above posted sample is proving that the method can do something.If I say anything it will be an unsolicited critique.

Hey Neil:

Is it possible that because of your style(which I think is terrific) this technique would serve no purpose?

respectfully,

Rey

Neil Snape wrote:
Maybe someone else will.

That was my best shot: remove larger scale problems while leaving a fine cloth fabric pattern.

If that doesn't sell you then nothing will.