How much are 2 hours worth?

[kannetth]

Hi there,

I'm here to ask what do you think of this before/after that was done in 2 hours.

In this particular case here's what I did:
- Frequency separation for the skin
- Dodge&burn
- Brightened up and coloured the eyes
- Selective sharpen
- Liquify to re-shape the face
- Quick selective colour to even out the reds
- Overall brightness/contrast
- Color toning

And what I didn't do:
- Nothing was done to the hair
- No in-depth selective color
- No dodge&burn on the arm (this one I just forgot)

So please let me know your thoughts.

EDIT:
Forgot to mention that the photo is not mine, but a free stock image.

[capice]

Re: How much are 2 hours worth?

hard to say since it's a bit small....

[kannetth]

Re: How much are 2 hours worth?

Sorry about that.

Much larger version HERE.

[capice]

Re: How much are 2 hours worth?

no problem,...but hard to judge when so small. The link works perfect to give some comment.
First, you have chosen a very difficult image to work on.
Imho, you have lost too much detail on the skinpart, far too blurry....reshaping is working well.
The hairstrands are sharpened too much or at least stand out too much. Most worrying I think is the loss of skindetail

[kannetth]

Re: How much are 2 hours worth?

I'm also not satisfied by the skin, it looks overdone. I believe that I should have edited it by grades and not all at once like I did. Correcting the big flaws together with the small ones destroyed the texture. I'm going to try again.

Also, I'm not a big fan of the Liquify tool and I use it rarely so it's nice to hear that I did good on this one and didn't get carried away with it.

The image is not the easiest to work with, I know, but I chose it on purpose to challenge myself.

Thank you very much for letting me know your thoughts, it means a lot for me to have my work commented.

[klev]

Re: How much are 2 hours worth?

I'm also not satisfied by the skin, it looks overdone. I believe that I should have edited it by grades and not all at once like I did. Correcting the big flaws together with the small ones destroyed the texture. I'm going to try again.

You'll have to be more specific. If it was me I would start with the things that really bother me. If I see acne, it's gone. I click the changes on and off to check for undesirable side effects. Contrast should be built up partly via shading, not just by cranking levels or curves as high as they'll go. It takes a light touch.

Also, I'm not a big fan of the Liquify tool and I use it rarely so it's nice to hear that I did good on this one and didn't get carried away with it.

Getting carried away is a bad thing, but it's not just that. The lighting isn't the greatest along the cheek, which makes it a slightly difficult decision. I would have probably gone "slightly" straighter on that shadow even if that isn't realistic. It's just not attractive this way. The shadows are just really blocked up, which doesn't help. Regarding liquify specifically, it has a little bit of a learning curve. It's okay to turn on the mesh at times. If you see overlapping grid topology, that is a bad thing no matter how the tool is written.

The image is not the easiest to work with, I know, but I chose it on purpose to challenge myself.

That may be lack of experience. The image isn't that interesting to me, but there's nothing inherently too difficult. If you are going from raw, I wouldn't add so much damn contrast. It blocks up shadows, and that can make things unnecessarily difficult.

[kannetth]

Re: How much are 2 hours worth?

You'll have to be more specific. If it was me I would start with the things that really bother me. If I see acne, it's gone. I click the changes on and off to check for undesirable side effects. Contrast should be built up partly via shading, not just by cranking levels or curves as high as they'll go. It takes a light touch.

As for the detail loss the problem is probably that I used the healing brush on the high frequency layer sampling too zoomed in, while probably it's better to start from afar and only then zoom in building it up.
Contrast was made by dodging and burning on two curves layers, one lighter and one darker, painting white at 5% flow on the masks (so no cranked "S" curves).

Getting carried away is a bad thing, but it's not just that. The lighting isn't the greatest along the cheek, which makes it a slightly difficult decision. I would have probably gone "slightly" straighter on that shadow even if that isn't realistic. It's just not attractive this way. The shadows are just really blocked up, which doesn't help. Regarding liquify specifically, it has a little bit of a learning curve. It's okay to turn on the mesh at times. If you see overlapping grid topology, that is a bad thing no matter how the tool is written.

I'll try to do better on the cheek, unfortunately I'm not yet very comfortable with the "typical" look I should achieve, so I've just tried to keep faithful to the original and not going too far. I couldn't do much for the clipped shadows.

That may be lack of experience. The image isn't that interesting to me, but there's nothing inherently too difficult. If you are going from raw, I wouldn't add so much damn contrast. It blocks up shadows, and that can make things unnecessarily difficult.

It surely is lack of experience, I'm not a professionist and I'm improving day by day. The image is not mine, as I said I've found it on the internet at it was randomly picked with these criteria: a nice but not wonderful woman (everyday-like) with possibly difficult skin and hard light. Hi-res. I started from JPEG and the contrast, as you can see, was already there (I've just changed it locally with D&B)

Thank you very much for your in-depth analisys and you suggestions, I will treasure it.

[klev]

Re: How much are 2 hours worth?

As for the detail loss the problem is probably that I used the healing brush on the high frequency layer sampling too zoomed in, while probably it's better to start from afar and only then zoom in building it up.
Contrast was made by dodging and burning on two curves layers, one lighter and one darker, painting white at 5% flow on the masks (so no cranked "S" curves).

Ah it may have just been the original. Since I haven't seen the original, I can't comment there. I've used a lot of different methods for the basic mechanics. It's generally a matter of what gives me predictable results without homogenizing smaller details unless it's intended. I never use that split frequency layers, as I didn't see any advantage. Personally I would say try not to ever go past 100%. You should be able to rely on a steady hand and comfortable tablet to screen mapping rather than zoom. Consider the topic of paintings. Artists have represented their message for centuries without the ability to view things in a tessellated manner. Going in to 200-400% just makes things take longer, and it generally provides worse results as you lose a sense of what is actually happening in the image. It makes some things jump out at you due to the lack of context, where otherwise the skin might not look so bad once you get rid of a couple bits of acne.

I'll try to do better on the cheek, unfortunately I'm not yet very comfortable with the "typical" look I should achieve, so I've just tried to keep faithful to the original and not going too far. I couldn't do much for the clipped shadows.

You can also pick a new image if that one is too limiting.

It surely is lack of experience, I'm not a professionist and I'm improving day by day. The image is not mine, as I said I've found it on the internet at it was randomly picked with these criteria: a nice but not wonderful woman (everyday-like) with possibly difficult skin and hard light. Hi-res. I started from JPEG and the contrast, as you can see, was already there (I've just changed it locally with D&B)

Thank you very much for your in-depth analisys and you suggestions, I will treasure it.

Yeah that might not be the best starting image, as it's very limiting due to initial adjustments that were applied before you worked on it. That does obviously add some amount of challenge, but it's probably the wrong type of challenge if you are getting started. You want something malleable that can adjusted to taste. You'll find that the basic "cleanup" of skin is ultimately one of the easiest things, even if precision is required. Dealing with emphasis and details can be much more difficult. Dealing with hair can be much more difficult. Dealing with complex anatomical structures around the eyes, parts of the forehead, elbows, shoulders, knees, etc. is much more difficult. At that point you have to think about how it should be represented, which is not something that can be solved via a stamp tool.