Possible defective BeadaPanel 5P (or bad firmware)

BeadaPanel versatile media display
bobg2222
Posts: 21
Joined: Sat Feb 24, 2024 1:29 am

Re: Possible defective BeadaPanel 5P (or bad firmware)

Post by bobg2222 » Mon Mar 18, 2024 8:42 am

Thanks for the diagram. I actually got close to those dimensions using the Pythagorean theorem. It means the digits on the display will be a little smaller I guess. I'll have to think about that some. It would be nice to just try the 6P, but it's pretty expensive--and I'd have no use for it if it doesn't work out.

esky-sh
Posts: 1449
Joined: Sat Dec 20, 2008 4:21 am

Re: Possible defective BeadaPanel 5P (or bad firmware)

Post by esky-sh » Mon Mar 18, 2024 7:55 pm

You can send us a png file in 1280x480 resolution. We can help you test in lab to get a fps parameter with a 6P.

bobg2222
Posts: 21
Joined: Sat Feb 24, 2024 1:29 am

Re: Possible defective BeadaPanel 5P (or bad firmware)

Post by bobg2222 » Mon Mar 18, 2024 8:37 pm

I tried your suggestion of sending the 5P (in continuous PNG protocol) an image and then immediately sending it again, then timing how long that second image takes to send. I'm using the high resolution timer (10^7 tics/second) and get, for example, 29864 tics or about 3 ms. And THAT would be 333 frames/second! Unlikely!

I'll attach a typical image I generated as if it were a 1280x480 display so you can work out how fast it displays on the 6P

Thanks!
Attachments
Image1.png
Image1.png (24.54 KiB) Viewed 124 times

esky-sh
Posts: 1449
Joined: Sat Dec 20, 2008 4:21 am

Re: Possible defective BeadaPanel 5P (or bad firmware)

Post by esky-sh » Tue Mar 19, 2024 4:30 am

I ran a javascript on my host PC today with the png picture you attached. The target is a 6P with latest firmware(715).
And the result is 500 frames/65 seconds, so the calculated fps is about 7.7.

Here is snippet of my test code just for your reference.

Code: Select all

        async function fire(bp, interval=0, loop=false ) {
          let www=0;
          await bp.sendPLHead({'cmdType': 5, 'fmtStr':'image/png'});

          do {
            for (const v of inputArray) {
              for (const t of v.files) {  

                let slices = sliceFile(t);
                for (const n of slices) {
                  if (abort) {
                    throw new Error('User abort!');
                    return;
                  }
                  let x = await readSlice(n);

                    let y = new Uint8Array(x);
                    await bp.sendMedia(y);

                  }
                
                 ++www;
                if (www%20===0)
                    console.log("20 frames");
              }          
            }        
          } while (loop);

bobg2222
Posts: 21
Joined: Sat Feb 24, 2024 1:29 am

Re: Possible defective BeadaPanel 5P (or bad firmware)

Post by bobg2222 » Thu Mar 21, 2024 9:36 pm

Is your javascript using continuous PNG or Start-image-End? I think it's the former.

If I try to do something similar on the 5P with my code, i.e., time how long it takes to send 1000 images then close the display and wait for that to return, I get only about 34 seconds (for 1000 images!) for a framerate of 29.4 frames/second.

But if I use Start-Image-End, 1000 sends take about 154 seconds for a framerate of 6.49 frames/second--similar to your code.

esky-sh
Posts: 1449
Joined: Sat Dec 20, 2008 4:21 am

Re: Possible defective BeadaPanel 5P (or bad firmware)

Post by esky-sh » Fri Mar 22, 2024 1:10 am

It is not surprising you got those figures on a 5P:
1. Yes I use continuous png protocol to send my test stream.
2. For the result difference between your 5P and my 6P(29.4fps vs 7.7fps), there are two major reasons:
  • Resolution difference(800x480vs1280x480). The larger the picture size, the longer the CPU processing time
  • On 6P, native LCD orientation is portrait(480x1280), we use a internal HW rotate module to do the rotation job so user can send landscape (1280x480) images on their host. Of course this extra rotation job will consume extra CPU resources and take extra memory bandwidth.
3. If 7.7 fps on 6P is below under your expectations. You can consider to directly send portrait(480x1280) images to bypass HW rotation and this will help you archive a target fps at 20.

bobg2222
Posts: 21
Joined: Sat Feb 24, 2024 1:29 am

Re: Possible defective BeadaPanel 5P (or bad firmware)

Post by bobg2222 » Sun Mar 24, 2024 4:32 pm

"3. If 7.7 fps on 6P is below under your expectations. You can consider to directly send portrait(480x1280) images to bypass HW rotation and this will help you archive a target fps at 20."

Is this right? If you send a 480x1280 image, it will display as portrait (and not squished up in landscape)?

If you send 480x800 on the 5P, it DOES display squished up in landscape. For the 5P, I rotate the image (on the PC) just before sending to the display, i.e., I "compose" it as 480x800 then rotate it.

esky-sh
Posts: 1449
Joined: Sat Dec 20, 2008 4:21 am

Re: Possible defective BeadaPanel 5P (or bad firmware)

Post by esky-sh » Sun Mar 24, 2024 9:00 pm

You will need to put a special marker in Panel-Link tag to avoid this squished up.

bobg2222
Posts: 21
Joined: Sat Feb 24, 2024 1:29 am

Re: Possible defective BeadaPanel 5P (or bad firmware)

Post by bobg2222 » Wed Mar 27, 2024 10:05 am

"You will need to put a special marker in Panel-Link tag to avoid this squished up"

How do you do that? Is it in the start tag?

esky-sh
Posts: 1449
Joined: Sat Dec 20, 2008 4:21 am

Re: Possible defective BeadaPanel 5P (or bad firmware)

Post by esky-sh » Wed Mar 27, 2024 4:23 pm

We will update documents for the tag format. Stay tuned.

Post Reply

Who is online

Users browsing this forum: No registered users and 9 guests