I love captions

[robzonbi]

I'm getting old and forgetful. I recently left social media and constructed a raspberry pi picture frame as per Wolfgang's lovely website. I love the software and it's now how I save and display my memories.

Since downloading all my data from Facebook, the pictures date data have all been set to the date which I downloaded it all. And the location data is gone. And I do miss having a note about the time place and context of the photos

It would be AMAZING if we could update the caption, date and location info via MTQQ. Something like pausing the frame on the picture in question, using a txt panel to update a designated metadata entry...

I'm a tinkerer more than a programmer, so forgive my ignorance if this is foolish.

[paddywwoof]

Hi, this is an interesting idea. It seems like it should be possible to change EXIF info using the modules we already use. Or else just change the info on the sqlite db. However I imagine it's non-trivial to do. As I have previously wondered about doing exactly this same thing I will check what would be involved. I won't promise to do more than that!

[robzonbi]

That's cool! I just wanted to put the idea out there. We don't know if a seed can grow if we don't atleast plant it, right? Cheers!

[estaman42]

Not technical here but why not a text file in the same directory as the photo with the same name. It's fairly future proof and I think the software already adds the background as it goes.

[helgeerbe]

I would use an external tool like https://exiftool.org/gui/ and for geo tagging: https://geosetter.de/en/main-en/

You can do this on your local computer and just copy the saved images again into your pictures folder on picframe. Picframe will notice, that images are changed.

[robzonbi]

This is what I have been doing. But to streamline the experience of using the frame, it would be nice to not have to use an entire suite of different apps. One for control, one for meta data, one for upload, one for editing... And so on. If I were an app developer I'd create one app to it all, simply for the love of the cause. In lieu of that, I just wondered if SOME features could be integrated together.

Just ideas. I really love the software.

[jgodfrey]

Since downloading all my data from Facebook, the pictures date data have all been set to the date which I downloaded it all. And the location data is gone

To be clear (in case it's not), I'd assume that Facebook has removed all of the metadata from your photos (during the original upload, storage, or final download). When a photo doesn't contain a basic EXIF "Created Date" record, the frame software instead displays the creation date of the disk-based file itself - which is what you're seeing.

While I can understand your desire to do "everything" in a single application, every user of picframe would likely define "everything" differently. For you, that includes adding and editing metadata. For someone else, that might include automatic resizing, interactive cropping, cloud-based sharing, or even basic photo editing. Really, these features are only loosely related to the actual purpose of picframe.

From an application complexity perspective, adding such complex features is a slippery slope and can ultimately lead to a bloated, buggy, hard-to-support tool that no longer does anything well.

My preference is to use small, proven, efficient, purpose-built tools for these related tasks. But, that's just me - to each his own.

[estaman42]

Sorry this is so long and rambles. I think this is a very good point. in layman's terms it's that old saying of hurting the thing you love. When I had a bit of a think about it, there've been a lot of tools I've used over the years and loved but the constant add-ins to make it better have in the end destroyed it. I am observing that journey in discord at the moment. When I first started using it was amazing. It just did the job but now there's more and more stuff creeping in and I'm starting to look for an alternative. Maybe as you've mentioned, the best of both worlds is keeping it as a display and having a separate program that allows you to put a lot of that stuff in. Picture frame just displays it. It gets down to use cases in a lot of ways. I've always taken photos and when digital came along I embraced it and now I have a very large picture library and on top of that im reluctant to delete anything out of it. I realised as I think the op is getting at that photos are my way off keeping a digital diary because when I look at a photo even if it's a bad one that I've taken somewhere, I get a whole stack of memories come back associated with that day. That's why I find this software so exciting! I'd call it a digital photo album rather than a picture frame but that is my use case. At the moment I'm having a nice cup of tea watching some photos I haven't really looked at in some years. The software presentation is absolutely beautiful and I've got some nice music playing in the background. so thank you to the developers! You've made something that I've often lamented the lack of. why in this digital age when people have so many photos on phones and so on, does the software for presenting them not mimic the original format of when you stuck photos in a photo album? It's not just about display, it's about presenting your photos in a meaningful way that looks beautiful. Is there a contribution link for this project as I'd like to? So the features I'd like to see would be a page turn animation, the ability to center text and maybe put a heading in. I've already tweaked one of the settings so the frame around each photo is white and a lot thicker like a standard old fashioned photo border. The home assistant integration is excellent as well and as per my use case, I'm trying to get it to Auto generate buttons off the folder names in my photo directory. Not having a lot of success with that at the moment, but I'll keep at it. As I said this is a long ramble but thank you to everyone involved in this project

[robzonbi]

Again, I'm not a high level programmer, so forgive my ignorance. But if a raspberry pi has exiftool installed, is it complex to have an MQTT topic accept a payload to be entered as the IPTC caption for the currently displayed picture?

I honestly don't know.. maybe it is.

Or could the picture frame publish a topic in which the currently displayed pictures path and file name are the payload? So other scripts could be written by users to manipulate said file?

I dunno I'm just spit balling.

[helgeerbe]

that should be quite easy. That‘s what ChatGPT propose

#!/usr/bin/env python3
import os
import json
import subprocess
import paho.mqtt.client as mqtt

# --- CONFIG ---
MQTT_BROKER = "localhost"
MQTT_PORT = 1883
TOPIC_IMAGE = "picframe/image"       # from picframe
TOPIC_CAPTION = "picframe/caption"   # from your sender
PICDIR = "/home/pi/Pictures"         # same as 'picdir' in picframe configuration.yaml

# Store the currently displayed image path
current_image_path = None

def set_iptc_caption(image_path, caption):
    """Set IPTC caption using exiftool."""
    try:
        subprocess.run([
            "exiftool",
            f"-IPTC:Caption-Abstract={caption}",
            "-overwrite_original",
            image_path
        ], check=True)
        print(f"[OK] Caption set for {image_path}")
    except subprocess.CalledProcessError as e:
        print(f"[ERROR] exiftool failed: {e}")

def on_connect(client, userdata, flags, rc):
    print(f"[INFO] Connected to MQTT broker with result code {rc}")
    client.subscribe([(TOPIC_IMAGE, 0), (TOPIC_CAPTION, 0)])

def on_message(client, userdata, msg):
    global current_image_path

    if msg.topic == TOPIC_IMAGE:
        try:
            payload = json.loads(msg.payload.decode("utf-8"))
        except json.JSONDecodeError:
            print("[WARN] Non-JSON payload from picframe/image.")
            return

        filename = payload.get("filename")
        subdir = payload.get("subdirectory")

        if not filename or not subdir:
            print("[WARN] Missing filename/subdirectory in image message.")
            return

        path = os.path.join(PICDIR, subdir, filename)
        if os.path.exists(path):
            current_image_path = path
            print(f"[INFO] Now showing: {current_image_path}")
        else:
            print(f"[WARN] Image file not found: {path}")

    elif msg.topic == TOPIC_CAPTION:
        caption = msg.payload.decode("utf-8").strip()
        if not caption:
            print("[WARN] Empty caption received.")
            return
        if not current_image_path:
            print("[WARN] No current image path set yet.")
            return
        set_iptc_caption(current_image_path, caption)

if __name__ == "__main__":
    client = mqtt.Client()
    client.on_connect = on_connect
    client.on_message = on_message

    client.connect(MQTT_BROKER, MQTT_PORT, 60)
    client.loop_forever()

Haven’t testet it. But idea is
That you create an Text input field in home assistant to show the caption. Write an automation, that sends on change the caption.

Picframe will pickup the new caption on next display.

If this works. Create a service for this script to run it on boot.

[helgeerbe]

By the way. Picframe is under heavy refactoring right now to make it more robust and maintainable. New architecture will contain a FileManager class. This could be a starting point to write a sidecar file with Meta Infos.

[helgeerbe]

By the way. Picframe is under heavy refactoring right now to make it more robust and maintainable. New architecture will contain a FileManager class. This could be a starting point to write a sidecar file with Meta Infos.

[paddywwoof]

Not looked to see how it fits with above but you can put text in the overlay or clock locations in /dev/shm/. The page turn transition might be easy to do if you're happy with a simple straight line between foreground and background image. Picframe has an option to use a different shader in the configuration.yaml file. Wolfgang originally fancied different effects but decided simple fade worked best. I can post some alternatives when I get opportunity next week.

Paddy