WOULD YOU PLEASE CLICK THE LIKE BUTTON? 拜托动动您可爱的小手点个赞好吗？

👋 Today I’m excited to bring you a brand-new series: Moody Photography (First Release: v261106) 📸✨

The main reason I started this series is to organize my personal collection of high-quality reference packs and turn them into distinct photography-style LoRAs. My plan is for each version in this series to capture a specific realistic photography style.

🤫 About the Source Material (v261106): As for what this very first edition is based on... sorry, I can't say it out loud here! 😂 If you're really curious and dying to know, feel free to drop me a DM.

🛠️ Dataset & Privacy/Copyright Protection: Long story short, this dataset was carefully curated down to 1,500 top-tier images out of a raw pool of 7,500 photos. Every single image underwent face processing/anonymization beforehand. The goal of this LoRA is strictly to learn outfits, photography aesthetics, camera angles, and poses, keeping it clean of real-person likeness and copyright issues.

⚙️ Recommended Settings:

• Recommended Weight: 1.0 works great right out of the box.

• Recommended Checkpoint: Best paired with Moody MOSS MIX for optimal lighting, texture, and realism.

💸 Pricing & Workload Behind the Scenes: A quick heads-up: this release will be priced slightly higher than usual, though let's be real—it's still dirt cheap! My standard and character LoRAs are usually just $0.88, and for this one, I’m only asking $2.88 USD (CivitAI takes half I take half). Come on, the cloud GPU server rental for training alone cost way more than that, i might break even if i can sell like 30 pieces? (And of course, supporters in my Discord will get a free direct download link, as always!)

The modest price bump comes down to:

1. Dataset Acquisition: The original reference sets were paid purchases out of my own pocket.

2. Massive Processing & Tagging Effort: Filtering 7,500 photos down to 1,500, running face processing and watermark removal on every single image, and writing meticulous captions took an insane amount of time.

Honestly, the workload for putting this LoRA together is right on par with doing a full checkpoint fine-tune—and I mean a real Moody model fine-tune, not those lazy 5-minute checkpoint merges floating around (looking at you, Moody Vivid Mix v2 🤦‍♂️😂).

📈 Future of the Series: I’m putting this first edition out there to test the waters and see how the community responds. If the feedback is great, I’ll definitely keep the series going! If not, well... this might just be the series finale 😂.

I really hope you all enjoy it! Feel free to share your renders and feedback below! ❤️🙏

Want a Krea2 Version? Comment below and I will consider making one.

👋 今天给大家带来的是全新的 Moody Photography 系列（首发版本 v261106）📸✨

其实开这个新系列，主要是想把我自己平时私藏的一些高质量图集，按照不同的分类方式整理出来，做成各种不同摄影风格的 LoRA。我的计划是：每一个版本都代表一个独立系列的写实摄影风格。

🤫 关于第一版（v261106）的素材来源： 这次带来的是……抱歉，这里我还真不能公开告诉你哈哈 😂！如果你真的特别好奇想知道（好奇宝宝们），欢迎群里来问。

🛠️ 训练集与版权说明： 总而言之，这一次的训练集是从 7,500 张图片里精挑细选出 1,500 张制作而成的。为了避免涉及到太多真人肖像与版权问题，所有的训练图片都已经预先做了脸部处理。这个 LoRA 主要是拿图集来学习里面的服装穿搭、摄影光影风格、镜头角度和人物姿势。

⚙️ 推荐参数与搭配建议：

• 推荐权重（Weight）： 保持默认的 1.0 即可，直出效果非常稳定。

• 推荐底模（Base Model）： 强烈建议配合 Moody MOSS MIX 一起使用，光影质感与写实度的契合度最佳。

💸 关于定价与工作量： 这一版的价格会比平时高一点点，但其实也没多贵好不好！平时普通的风格 LoRA 或角色 LoRA 我也就收个 8 毛钱（$0.88），这个我也就收 2.8 美金（$2.88 USD）- 然后还要上供给CivitAI拿一半。拜托，光是租云端服务器跑训练的电费和算力费都远不止这个数了，不卖个30份都难以回本…… (当然，Discord 里的老铁和赞助支持者会有专属的免费下载链接，记得去频道里领！)

之所以比平时贵一点点，主要是因为：

1. 素材本身有成本： 这些私藏图集全都是真金白银付费买来的。

2. 极度耗时的前处理与精标： 从 7,500 张里一张张筛出 1,500 张，逐张做脸部处理，去除水印，再加上详细打标，耗费了巨大的心血。

老实说，做这个 LoRA 的工作量，其实完全不亚于完整做一个底模微调（Finetune）——当然，我说的是真正 Moody 模型那种级别的微调，绝对不是市面上随便拿几个模型合在一起（Merge）就敢叫“微调”的那种，毕竟那种微调大概五分钟就能做完吧，比如 Vivid Mix v2 🤦‍♂️😂。

📈 后续更新计划： 这次首发先放出来看看大家的反响如何。如果大家喜欢、反响好的话，这个系列我就会继续做下去；如果反响一般的话，那这一版大概就是本系列的“绝唱”了 😂。

希望大家喜欢这次的新作品，欢迎大家多多返图交流！❤️🙏

Krea2 版本？ 想要的话下面留言，人多我就做。