[Bug]: PNG info not working

[guzuligo]

Is there an existing issue for this?

• [X] I have searched the existing issues and checked the recent builds/commits

What happened?

When I put any image into PNG info, nothing shows. It either shows me none, or some basic PNG related info.

Steps to reproduce the problem

I'm not sure what caused it. I just add any image to PNG info and nothing shows.

Probably an extension I added, or maybe latest version issue.

What should have happened?

PNG info should show the prompt, seed and other a1111 related info.

Commit where the problem happens

65995a2

What platforms do you use to access the UI ?

Windows

What browsers do you use to access the UI ?

Mozilla Firefox, Brave

Command Line Arguments

--xformers --api

List of extensions

<META HTTP-EQUIV="CONTENT-TYPE" CONTENT="text/html; charset=utf-8">
<TITLE></TITLE>
<META NAME="GENERATOR" CONTENT="OpenOffice 4.1.13  (Win32)">
<STYLE>
	<!-- 
	BODY,DIV,TABLE,THEAD,TBODY,TFOOT,TR,TH,TD,P { font-family:"Arial"; font-size:x-small }
	 -->
</STYLE>

Extension

ABG_extension

DreamArtist-sd-webui-extension

SD-latent-mirroring

a1111-sd-webui-haku-img

a1111-sd-webui-tagcomplete

auto-sd-paint-ext

custom-diffusion-webui

ddetailer

depth-image-io-for-SDWebui

depthmap2mask

embedding-inspector

multi-subject-render

openOutpaint-webUI-extension

openpose-editor

prompt-fusion-extension

sd-webui-additional-networks

sd-webui-controlnet

sd-webui-model-converter

sd-webui-riffusion

sd-webui-supermerger

sd_dreambooth_extension

sd_save_intermediate_images

sd_smartprocess

seed_travel

shift-attention

stable-diffusion-webui-Prompt_Generator

stable-diffusion-webui-aesthetic-gradients

stable-diffusion-webui-anti-burn

stable-diffusion-webui-daam

stable-diffusion-webui-embedding-editor

stable-diffusion-webui-embedding-merge

stable-diffusion-webui-inspiration

stable-diffusion-webui-instruct-pix2pix

stable-diffusion-webui-prompt-travel

stable-diffusion-webui-sonar

stable-diffusion-webui-tokenizer

stable-diffusion-webui-two-shot

training-picker

video_loopback_for_webui

Console logs

Nothing out of the ordinary, except for few of those at the startup:

No Image data blocks found.
Error loading embedding pp1.preivew.png:
Traceback (most recent call last):
  File ".\AUTOMATIC1111\modules\textual_inversion\textual_inversion.py", line 206, in load_from_dir
    self.load_from_file(fullfn, fn)
  File ".\AUTOMATIC1111\modules\textual_inversion\textual_inversion.py", line 155, in load_from_file
    name = data.get('name', name)
AttributeError: 'NoneType' object has no attribute 'get'

Additional information

No response

[CCpt5]

Along the same lines, the ability to have PNG Info change (and/or not change) the model when sending to TXT2IMG, IMG2IMG, etc has been broken/missing for about 3 weeks.

Issues: https://github.com/AUTOMATIC1111/stable-diffusion-webui/issues/7620

Discussion: https://github.com/AUTOMATIC1111/stable-diffusion-webui/discussions/7427

[AlUlkesh]

@CCpt5 Part of what my PR does on that matter, just made it into a recent commit. It should work now (except for sendto inpaint) as overrides. You should see something like this:

This is rather hidden at the bottom of the screen by default. I highly recommend to move it up:

[CCpt5]

Thanks! This is helpful - I hadn't noticed that.

So, basically the model checkpoint listed at the very top will stay as it was prior to the PNG Info send but an override is placed in that tab? I'm glad that works, but I hope in the future there's an option to fully change the model vs. sending an override. It's a little confusing (imho) to have a different model fully listed (name/hash) in the Quicksettings list checkpoint selector typical bar than what's being used (only listed via hash in that new override tab).

Personally I'd love it if the functionality of the old "When reading generation parameters from text into UI (from PNG info or pasted text), do not change the selected model/checkpoint." were restored. I'd added that to my quicksettings top line and it was simple to toggle on and off whenever needed. It's still on my top bar in the latest commit, but the toggle/setting no longer works.

Glad there's progress being made here, though. It's a very important feature in my workflow.

Thanks to everyone working on the project - it's an amazing tool!

[emotionaldreams]

Thanks! This is helpful - I hadn't noticed that.

So, basically the model checkpoint listed at the very top will stay as it was prior to the PNG Info send but an override is placed in that tab? I'm glad that works, but I hope in the future there's an option to fully change the model vs. sending an override. It's a little confusing (imho) to have a different model fully listed (name/hash) in the Quicksettings list checkpoint selector typical bar than what's being used (only listed via hash in that new override tab).

Personally I'd love it if the functionality of the old "When reading generation parameters from text into UI (from PNG info or pasted text), do not change the selected model/checkpoint." were restored. I'd added that to my quicksettings top line and it was simple to toggle on and off whenever needed. It's still on my top bar in the latest commit, but the toggle/setting no longer works.

Glad there's progress being made here, though. It's a very important feature in my workflow.

Thanks to everyone working on the project - it's an amazing tool!

facing same issue. still not working at all for me for weeks. and i don't see that setting in the image posted .. im on ver 1.7