Portrait shooting and processing workshop

Thanks to everyone who attended the session last night at Holmes Chapel Camera Club.  Despite the technical problems with tethering a D800 to a Windows 8 laptop (my old D700 always worked flawlessly when tethered) we achieved the main goals of the workshop.  There are a number of articles on this website looking at specific bits of process and techniques which I’ve linked to on this page, however I thought it would be useful to summarise the things we looked at last night.

Many thanks to Clara for modelling for me.  Not easy seeing your face up that close.

Equipment used

  • 2 x SB900  (OK I used a 3rd as a commander, but you could get away without that)
  • One lightweight lightstand
  • One Voice-Activated Lightstand (or “Graham” as he likes to be called – thanks Graham)
  • One Lumiquest LTP fold-up softbox
  • One 24″ pop-up softbox from Meking studios – Like a Lastolite Easybox.

Alternatives

Any old softboxes will do.  Really.  Anything that diffuses light.  You could also use a shoot through umbrella.  I used the Lumiquest as the main light as its smaller and as it would be closer, this would matter less.  Also, I wanted to try it out 🙂   Any old speedlights will do too – 2 Yongnuo YN460-II’s would have done the job just as well.  You just have to walk to the lights to adjust the power, and be careful not to move them.  You should really meter the light if you are going to use manual flashguns, however you can just take shots and adjust until it looks right on the monitor.

Exposure settings

Manual mode, 1/125th of a second at f/5.6 and ISO 100 (the “native” ISO for my sensor)  was enough to eliminate the ambient, and give a light grey background.  To make the background darker, either move the model and lights away from it, or turn the lights up.  In TTL mode, this is achieved by simply stopping down to f/8 or f/11.

iTTL metering

The wonder of Nikon’s Creative Lighting System  (also similar on Canon and Sony). iTTL instructs each remote flash to fire a test flash.  The camera measures the result and sets each remote flash to the correct power to evenly light the subject to achieve a mid-gray exposure and then takes the shot.  You can add compensation to each flash group to create the lighting ratios you need between the lights.  I ended up with the main light 1.3 stops more than the fill light.

The beauty of iTTL flash like this though is that Graham could move the fill light up and down, and the exposure would still be right – still 1.3 stops lower than the main as the camera would adjust the power to compensate.

You can take this shot with manual lights, it just means you either need to meter the light, or keep checking the image on the back of the camera and the histogram until it looks right..

Focussing

Focus on the leading eye.  I used a single focus with the centre cursor.  Focus on the eye, and then compose the image.  Seems the AF on my lens was back focussing a cm or so.  I need to calibrate the AF.  You could also position the cursor over the eye and shoot.  You may not get the framing just as you want it this way, but hey – you can crop it later.  The focus will be more accurate this way, as you are not rotating the camera, and a little slower as the outlying focus cursors tend to be lesser single line jobs rather the cross sensors nearer the middle.

Layout and framing

I like a bit of space for the subject to look into, and as we read images left to right, I like to put the subject on the right, looking left and towards the top of the frame.  So I take most of my portraits in “landscape” mode.  If the subject is looking straight on, then you can get away with the vertical format.

Processing

Set the white and black points, and adjust colour balance in your raw processor.  There is a lot of extra data not visible on the screen in your raw file, which can be pulled in selectively or globally.  Some of this will be discarded as soon as you fit the data into a colour space.  This happens when you render a bitmap to edit in Photoshop. by doing “edit in Photoshop” from Lightroom or Adobe Camera raw.  the develop module in Lightroom *is* ACR by the way.  Even more data is discarded when you finally cram it all into sRGB for displaying on your monitor or printing.  What you see in Photoshop on your screen is an 8 bit sRGB image, no matter what you set the colour space, as this is all your screen can display.  Unless of course you’ve bought a 12 bit Eizo monitor….

Processing in Lightroom

Nice and quick.  After the global adjustments, just get to work with the brushes, on the eye whites, iris, lips, teeth, hair and skin.  Details here:- https://owenlloydphotography.com/?p=483

Processing in Photoshop

Open the image in Photoshop and….

  • Use healing brush to remove skin marks and blemishes,
  • Use Liquify to tidy up bumps, hair shape, alter expression (mouth, eyebrows etc)
  • Create a new layer and smooth the skin
  • Create a new layer with blend mode screen to adjust the eye whites
  • Create a new layer with blend mode Soft Light to adjust the Iris
  • Create a new layer with blend mode Multiply to adjust the lips
  • Create a new layer with blend mode screen to adjust the teeth
  • Create a new layer with blend mode normal for dodging and another for burning.  Dodge and burn at about 9% with a soft brush
  • Select red channel and apply un-sharp mask at 100%, radius of 1 to 1.4 and a threshold of 3.
  • Flatten image and save.

For each of the new layers, you’ll want to apply a mask by alt-clicking the add layer mask icon.  Then use a white brush on the mask to paint back in the new layer.  Adjust opacity until it looks natural   Keep clicking the new layer on and off to see before and after as you adjust.  When you’re happy, merge down.  You don’t have to do this, however i find I will endlessly fiddle with each layer if I keep them around – commit and move on!

The skin smoothing technique

For the skin smoothing layer do this:-

  • Set blend mode to Vivid light
  • Invert the layer
  • Apply High Pass filter at around 10-13%.  Adjust this at 100% zoom, it will be different for each sensor and scale of skin.
  • Apply Gaussian blur at around 6%.  Again, some trial and error will produce the right value.  The aim is to get even hue and light eliminating large blotchy patches whilst retaining pores and micro-fine hair detail on the skin.
  • Apply mask and paint effect back in on the skin only.  Stay away from the edges and remove the effect from any shadow lines by painting in black on the mask.

 

Final touches

  • Split toning – light blue preset colour for the shadows and light yellow preset for the highlights.  This has the effect of further balancing skin hues, taking the edge off the colours and creating a print magazine look to the image.  Play around with the sliders.
  • You may want to de-saturate a little, and play with the ratios between vibrancy and saturation.
  • Try a vignette.  Soft and subtle.
  • Sharpen eyes, lips and nostrils using the adjustment brush in Lightroom

Below are the results of processing the same image in Lightroom only, and in Photoshop as I did in the session.  There’s not a lot in it between Lightroom and Photoshop really – you get more control over healing and can use tools like liquefy in Photoshop (I haven’t used liquefy on these images), however the Lightroom process takes only 5-10 minutes – especially if you save presets for all the different brush values.

 

Further reading

All of those guys and many others have recorded video workshops on Kelby Training.  This is a subscription service, however for someone like me who isn’t doing this day in day out, I found it a great way to learn the basics – so by the time I turn up in a studio with a model and make up, plus assistants, I have a fair idea of what is possible and how things will work.

Owen

 

From Camera with black and white points set

_OHL5715

Processed entirely in Lightroom_OHL5715-2

The full Photoshop banana – finished in Lightroom_OHL5715-Edit-2

 

 

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.