Like the work I do and want to say thanks? Buy me a coffee or Support me on Patreon for exclusive early access to my models and more!
NightVisionXL started as a lightly trained, photography focused SDXL model that specialized in photographic output. Over the past few months, NightVision has received dozens of trainings and tunings chasing the dragon that is pleasing aesthetic output. With the release of version 8.X, I feel like I'm much closer to my goal of effortless beautiful output that is aesthetically pleasing without requiring confusing negatives or comma separated nonsense.
NightVisionXL speaks natural language. Prompt it like you would describe a scene to a human being, and you'll be surprised by how well it can understand and comprehend your prompts.
I'm trying something new with V8 of NightVision, and I hope the results show in the output. I continue to scrape beautiful imagery that I think will train well, and this particular update is no exception with hundreds of beautiful license free AI images that I've sourced from MJ, Dalle3, Civit, Playground and my own outputs. On top of that, I'm also utilizing HF datasets for license free photography and the LAION-POP dataset, filtered to remove low quality images prior to training. Captioning is done using my own Spicy Burrito captioning tool (available to my Patreon subscribers on discord!) that creates incredibly verbose and effective captions that are essential for good coherent output.
On a final note, I've had multiple requests to "turbo-ize" this model. I have no plans to mix turbo into NightVision, as the licensing on NightVision is nice and open right now with the SDXL licensing, and I'd like to keep it that way. NightVision isn't fast, but I promise you the longer gen times are worth the wait!
This is NightVision XL, a lightly trained base SDXL model that is then further refined with community LORAs to get it to where it is now. NightVision XL has been refined and biased to produce touched-up photorealistic portrait output that is ready-stylized for Social media posting! NightVision XL has nice coherency and is avoiding some of the weird body issues and biases that are starting to plague some of the other photorealistic models. Further, NightVision XL produces rich deep blacks and great evening/night time scenes. It can also produce ridiculously bright output as well!
NightVision XL is capable of both SFW and NSFW output. As with all of my other models, tools and embeddings, NightVision XL is easy to use, preferring simple prompts and letting the model do the heavy lifting for scene building.
NOTE - This version includes a baked VAE, no need to download or use the "suggested" external VAE.
WARNING - DO NOT USE SDXL REFINER WITH NIGHTVISION XL
The SDXL refiner is incompatible and you will have reduced quality output if you try to use the base model refiner with NightVision XL.
Follow me here by clicking the heart ❤️ and liking the model 👍, and you will be notified of any future versions I release. I also need your help with feedback, please please please post your images and your honest feedback below, I will use your feedback and your output to help guide future revisions!
Patreon coming soon!
Description
NightVision XL Beta 0.7.0.2 Release Notes 08/30/23
This version has been updated with a light fine tuning I trained on several hundred analog photographs with hopes to keep improving eyes, teeth and skin details, continuing to chase the dragon of fine photography style output. Version 0.7.0.2 seems to operate best in the 4 - 6 CFG range, with color pop really hitting around 7.5 or so, but can push into crushed blacks/whites. I'm noticing a nice film grain effect on darker images now, and overall coherence seems to have been improved.
Known Issues
Hands are improving, but still luck of the seed
color temperature has changed a bit since 0.6.8.1, may be a bit more contrasty. I've gotten good feedback from testers on the color, but it is less "neutral" - This is really only noticeable if you're doing after-work in PS/LR and are playing with color temp values, and it's not bad, just different.
male genitalia still non-existent - Every attempt to fix penises so far has lead to coherence problems (and tends to put them everywhere)
Stretchy nude bodies - long torsos, overly long legs, long necks - This is an issue that comes all the way up from base, but I'm going to start tracking it here as I want to try to banish it in future versions.
FAQ
Comments (13)
NightVision XL Beta 0.7.0.2 Release Notes 08/30/23
This version has been updated with a light fine tuning I trained on several hundred analog photographs with hopes to keep improving eyes, teeth and skin details, continuing to chase the dragon of fine photography style output. Version 0.7.0.2 seems to operate best in the 4 - 6 CFG range, with color pop really hitting around 7.5 or so, but can push into crushed blacks/whites. I'm noticing a nice film grain effect on darker images now, and overall coherence seems to have been improved.
Known Issues
- Hands are improving, but still luck of the seed
- color temperature has changed a bit since 0.6.8.1, may be a bit more contrasty. I've gotten good feedback from testers on the color, but it is less "neutral" - This is really only noticeable if you're doing after-work in PS/LR and are playing with color temp values, and it's not bad, just different.
- male genitalia still non-existent - Every attempt to fix penises so far has lead to coherence problems (and tends to put them everywhere)
- Stretchy nude bodies - long torsos, overly long legs, long necks - This is an issue that comes all the way up from base, but I'm going to start tracking it here as I want to try to banish it in future versions.
牙齿！！！ 终于啊等你好久了！！！！牙齿！！！！
I agree, teeth have been troublesome! I'm glad you like it 👍❤️
Personally, I feel 0681 version is easier to get a good picture with CFG=7, especially the pretty face and hair, the teeth are really easy to chipped. 0702 feels more real compared to 0681, but I feel it's not easy to get a good picture with CFG=7. Meanwhile, compared to 0681, the beauty of 0702 version is a little bit less pretty, and the hair is sparse and lacks of color and luster at the same time. The hair is also thinner and lacks color and luster. Maybe I didn't test enough, I feel that 0702's teeth improvement is not too obvious. Personally, I prefer the 0681 version~ I made an x/y/z chart for comparison.
个人感觉0681版本在CFG=7的情况下更容易得到好图，特别是漂亮的脸蛋和头发，牙齿确实容易崩掉。0702与0681相比感觉更真实，但感觉CFG=7的时候不容易出好图，同时相比0681，0702版本的美女，变的稍微不那么漂亮了，头发也变的稀少同时缺少颜色和光泽。也许是我测试的数量不够多，感觉0702的牙齿改进不太明显。个人比较更喜欢0681版本呢~ 我做了一张x/y/z图用来比较。
Thanks for the feedback! There was a shift towards more realism with 0.7.0.2 as I trained more analog photography in. There is another update I have in the pipeline coming that brings colors back more in line with 0.6.8.1, and fixes (or mostly fixes) the long elongated torsos issue from base. If you want to pop onto my discord server, I'd be happy to run some test x/ys with you to see if we can identify if the issue is fixed in the upcoming version, and if not, it can be something I add to my known issues list to work on moving forward.
谢谢您的反馈！在0.7.0.2版本中，我确实增加了更多的模拟摄影元素，以追求更真实的效果。我已经有一个即将发布的更新，它将把颜色调整得更接近0.6.8.1版本，并且（大部分）修复了基础模型中长身体的问题。如果您愿意加入我的Discord服务器，我很乐意与您进行一些x/y轴的测试，以确认即将发布的版本是否解决了这个问题。如果没有，那么这将是我未来需要解决的已知问题列表中的一个项目。
@socalguitarist OK, let me in~ my Discord name is ggyydream
This model is incompatible error when using this in NMKD GUI, can anyone helpfully correct me?
the sweat and water effect in this model looks incredible, just love it!
Weird, I always thought you would win the SDXL contest, but not even on the list. What did I miss?
I spoke with Civit about it - The contest was only for true pure models - it would have been kinda unfair if my train + merges were included against only trained. While I did throw up my "compliance model", NightVision XL was still originally posted as a merge, thus not qualified by the rules. I'm not upset about it and actually agree with the decision 👍 - it's okay though, all the love and support my models have gotten from the community has been awesome and is a great "booby prize" in my eyes! ❤️
thank you for showing a variety of sample images. most posts are just one style. i skip over their 6gb downloads if it's not what i'm looking for. but, i'm downloading your's now. :) thanks again.
I don't know why, but what I see is all red crosses on a white background when I view the images with FastStone Viewer in thumbnail view.
Well, i guess Jesus loves you
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.