我的妹妹其實沒有那麼可愛 五更琉璃 photographing practice 01
這張我會給你一個比較直接的評價：整體是成功的，而且比你前面說的「黑背景讓畫面太暗」更值得繼續發展。
但它現在還有一個很明顯的問題：光線漂亮，但暗部壓得太深。
先說我覺得最好的地方
1. 輪廓光真的有效
你現在右後方的光把：
- 頭髮外緣
- 肩膀
- 手臂
- 裙子的邊緣
- 腿部輪廓
都從黑背景裡拉出來了。
尤其左側頭髮那條亮邊，我覺得很好。它讓黑色頭髮沒有直接「消失」在背景裡。
所以我們昨天討論的：
RF05B → 右後方 → 輪廓／背景光
從這張來看，方向是對的。
2. 主光的位置很好
臉部不是正面平光，而是從側前方過來。
這讓：
- 鼻子
- 臉頰
- 下巴
- 肩膀
- 手臂
都有立體感。
尤其她的臉，我覺得現在已經有一點人物肖像攝影的感覺，而不是「把公仔照亮」。
這點很重要。
3. 黑色服裝與黑背景雖然有問題，但也形成了風格
裙子、頭髮、背景幾乎都是黑色。
如果是一般商品攝影，我會說這樣不利於辨識。
但你現在不是在做商品目錄，而是在嘗試把公仔拍得像一個人物。
所以這種：
黑背景 + 黑服裝 + 邊緣光
其實是成立的。
問題不是「黑背景錯了」，而是暗部目前稍微太多。
我最想改的是這裡
臉部暗部可以再亮一點
現在眼睛是照片裡最重要的視覺中心。
但是左側臉、頭髮、胸口附近的黑色區域幾乎完全掉進去了。
尤其從臉往身體下方看：
臉 → 頸部 → 黑色衣服 → 腰部
中間的明暗落差有點太大。
我會希望臉比現在亮大約 0.3～0.5 EV，但不是把整個畫面拉亮。
也就是：
不要增加整體曝光。
而是讓主光稍微往公仔靠近／稍微增加一點輸出。
第二個問題：下半身有點「掉出去」
你現在的畫面上半部很好。
但到了：
裙子 → 大腿 → 膝蓋 → 小腿
亮度逐漸降低。
這讓視覺重心非常集中在臉和上半身，下半身比較像是從黑暗裡浮出來。
這不一定是缺點。
如果你想要的是：
肖像式、神秘、電影感
那其實可以保留。
但如果你的目標是：
完整展示這隻公仔
那我會讓腿部再多一點光。
還有一個我認為很值得注意的地方：構圖
這張是直幅，而且人物幾乎填滿畫面。
我喜歡這一點。
但是上方黑色空間稍微多了一點。
頭頂到畫面上緣的距離，讓視線有點往上漂。
如果是我，我會嘗試：
把人物稍微往上移，或裁掉約 5～8% 的上方黑色空間。
這會讓人物更有存在感。
但不要裁太多，因為你現在右側、左側的黑色空間其實是構圖的一部分。
景深方面
這張其實也回答了你剛才問我的問題。
單人拍攝時，你現在這種景深是可以接受的。
臉部是清楚的，身體往下逐漸變化，並沒有讓人覺得「失焦」。
所以：
單人
f/2.8～f/4 可以成立。
三人
我仍然會建議：
f/5.6 起跳。
不要為了複製這張單人的淺景深效果，而把三人也拍成 f/2 左右。
如果是我下一張會怎麼拍？
我甚至不會換掉現在的燈位。
只做一個改動：
RF05B
5600K / 20～30%
Godox iT30Pro
維持目前方向，但稍微增加一點輸出
白反光板
放在公仔右前方，但比現在更靠近一點。
目的不是把陰影消掉，而是：
讓黑色衣服仍然是黑色，但能看到一點材質。
這會非常有幫助。
最後一個我想特別稱讚的地方
這張已經不像單純的「公仔記錄照」。
你開始在做的是：
光影 → 輪廓 → 黑色負空間 → 人物姿態
而不是單純追求「公仔每個細節都清楚」。
這跟你之前提到想研究 Saul Leiter、森山大道、Alex Webb、Michael Kenna 的方向，其實比較接近：不是把資訊全部交代清楚，而是開始利用不完整的資訊來建立畫面。
所以我不建議你現在急著把黑背景換掉。
先把這套黑背景＋輪廓光再拍 3～5 張。
你真正要解決的是「暗部還有多少資訊」而不是「黑背景好不好」。
如果你接下來把同一隻公仔用白背景和天藍背景各拍一張，我很建議你傳給我。我可以直接拿這三張比較，判斷哪一種背景最符合你現在這種攝影方向。
先給各位看一下結果, 下一篇敬請期待
I recently bought a few new figures, and this Gokou Ruri figure is one of them.
She’s wearing a black maid-style swimsuit, and to be honest, she was really difficult to photograph.
Cheney ended up taking more than 800 shots before finally managing to get a few reasonably decent photos—with quite a bit of help from ChatGPT along the way.
Since the whole process turned out to be much longer than expected, I’ll probably split it into five or six posts and share the entire experience with you, hopefully as a useful reference for anyone trying something similar.
Before even starting the shoot, Cheney asked ChatGPT for some advice.
And, well… ChatGPT gave me a very, very long list of suggestions.
To be fair, they were actually quite useful.
But…
I still couldn’t get the shot right. 😂
So, here’s what ChatGPT had to say about the very first photo:
For this shot, I’ll give you a fairly direct assessment: overall, it works—and I think this direction is more worth developing than the concern you mentioned earlier about the black background making the image too dark.
There is still one obvious issue, though: the lighting is beautiful, but the shadows are being pushed a little too deep.
First, what I think works best
1. The rim light is definitely working
The light from the rear right is now separating:
the edges of the hair
the shoulders
the arms
the edge of the skirt
the outline of the legs
from the black background.
I especially like the bright edge along the hair on the left side. It keeps the black hair from simply “disappearing” into the background.
So, based on this shot, the setup we discussed yesterday:
RF05B → rear right → rim/background light
is definitely heading in the right direction.
2. The position of the key light is very good
The face isn’t being lit with flat frontal light. Instead, the light is coming from slightly off to the front and side.
This gives dimension to:
the nose
the cheeks
the chin
the shoulders
the arms
The face in particular is beginning to feel more like portrait photography rather than simply “lighting a figure.”
That is very important.
3. The black clothing and black background create some challenges, but they also establish a distinctive style
The skirt, hair, and background are all almost entirely black.
If this were conventional product photography, I would say that this makes the subject harder to distinguish.
But you’re not shooting a product catalog. You’re experimenting with photographing the figure as though she were a real person.
So this combination of:
black background + black clothing + rim light
actually works.
The problem isn’t that the black background is wrong. The issue is simply that there is currently a little too much darkness in the image.
The main thing I would change
The shadow side of the face could be slightly brighter
The eyes are currently the most important visual focal point in the photograph.
However, the left side of the face, the hair, and the dark areas around the chest are almost completely disappearing into shadow.
Especially as your eye moves downward:
face → neck → black clothing → waist
the transition in brightness is a little too dramatic.
I would like the face to be about 0.3–0.5 EV brighter than it is now, but I would not brighten the entire image.
In other words:
Don’t increase the overall exposure.
Instead, move the key light slightly closer to the figure and/or increase its output just a little.
The second issue: the lower body is fading away a little too much
The upper half of the image works very well.
But as you move down through the:
skirt → thighs → knees → lower legs
the brightness gradually decreases.
This makes the visual emphasis very strongly concentrated on the face and upper body, while the lower body feels as though it is emerging from the darkness.
That isn’t necessarily a weakness.
If what you want is something:
portrait-like, mysterious, and cinematic
then you can absolutely keep it this way.
But if your goal is to:
show the entire figure clearly
then I would add a little more light to the legs.
Another thing worth paying attention to: composition
This is a vertical composition, and the figure fills most of the frame.
I like that.
However, there is slightly too much black space above the head.
The distance between the top of the head and the upper edge of the frame causes the viewer’s eye to drift upward a little.
If it were my shot, I would try:
moving the figure slightly higher in the frame, or cropping about 5–8% of the black space from the top.
That would give the figure a stronger presence.
But don’t crop too aggressively, because the black negative space on both the left and right sides is actually part of the composition.
Depth of field
This image also answers the question you asked me earlier.
For a single figure, the depth of field you’re using here is perfectly acceptable.
The face is sharp, while the body gradually changes in sharpness farther down the frame. It doesn’t feel as though the figure is simply “out of focus.”
So for:
one figure
f/2.8–f/4 can work very well.
For:
three figures
I would still recommend starting at:
f/5.6.
Don’t try to reproduce the shallow depth-of-field effect of this single-figure shot by photographing three figures at around f/2.
What would I do for the next shot?
I wouldn’t even change the current light positions.
I would make only one adjustment.
RF05B
5600K / 20–30%
Godox iT30Pro
Keep it pointed in the same direction, but increase the output slightly.
White reflector
Place it in front and to the figure’s right, but move it slightly closer than it is now.
The goal isn’t to eliminate the shadows.
The goal is:
to keep the black clothing looking black while still revealing a little of its texture.
That would help a lot.
One last thing I especially want to praise
This image no longer feels like a simple “record shot” of a figure.
You’re beginning to work with:
light and shadow → silhouette and contours → black negative space → the pose of the subject
rather than simply trying to make every detail of the figure clearly visible.
This is actually closer to the direction you mentioned wanting to explore through photographers such as Saul Leiter, Daido Moriyama, Alex Webb, and Michael Kenna: rather than presenting every piece of information clearly, you’re beginning to use incomplete visual information to construct the image.
So I wouldn’t rush to replace the black background yet.
Keep experimenting with this black-background + rim-light setup for another 3–5 shots.
The real question you need to solve is not “Is the black background good or bad?” but rather “How much information should remain visible in the shadows?”
If you photograph the same figure once against a white background and once against a sky-blue background, I’d definitely recommend sending those to me as well.
I can compare all three versions directly and help determine which background best suits the photographic direction you’re developing.
To be honest, Cheney did a terrible job on the first shoot and was extremely unhappy with the results…
So, in the next post, I’ll share my second attempt, where I tried again based on the advice ChatGPT gave me after reviewing the first shoot.
But before that, here’s a little preview of how things eventually turned out.
Stay tuned for the next post!
Comments