Onyx Boox as monitor

[mdp]

Originally Posted by 2new

Here is another test with text using a modified Heckbert approach. Much faster, each image is about 2k using png format. My ereader only has 2.4ghz wifi so I can't test reverse adb.

I see a lag just above .5s; in the past and in other contexts I counted screen reaction times just above 0.5s for GU (normal, 16 shades) mode and half of that for A2. You seem to be approaching normal E-Ink's EPD reaction times.

What do you mean with «modified Heckbert approach»? Image warping? Placed where and how exactly?

[2new]

Originally Posted by mdp

I see a lag just above .5s; in the past and in other contexts I counted screen reaction times just above 0.5s for GU (normal, 16 shades) mode and half of that for A2. You seem to be approaching normal E-Ink's EPD reaction times.

What do you mean with «modified Heckbert approach»? Image warping? Placed where and how exactly?

I am taking the raw captures from screen reducing them by using quantization. It's a combination of mediancut and reducing the number of colors to make an indexed image. My latest interation I make the image transparent to increase the ability for contrast. After that's done I compress the image and send it over the stream.

There's definitely room for coding performance gains but it's pretty fast as is. I think network and hardware are beginning to be a bottleneck since it works near flawlessly on a standard tablet.

[fAX]

After the latest updated, `adb reverse` stopped working; probably, due to USB debugging got disabled. Probably, I missed something, but how do you get it back?

Thanks!

[fAX]

Oh, my bad; just in case is somebody struggling, it's under Settings->Application.

[2new]

The default background for an mjpeg stream is black. So I thought might be interesting to compare against pixel qi screen. Wow, pixel qi is so much easier to read when the background is black and fonts are white.

[jamespeter26]

Originally Posted by eldon

Hi,

I am hoping that someone here can help me with the Monitor Mod on the Max 2. (Monitor_Mod3.apk and onyx-monitor-move-pointer-hack-7a.py)

I have it almost working. I can see that using two fingers for right click is registering the x,y coordinates but when the cursor tries to move python gives me the below error. Any advice on how I should proceed?

I am using Windows 7 and Python 2.7.15. I do not know if it matters but I am using a Macbook with Parallels (Windows 7).

C:\Python27>python 7a.py
07:49:40.418000 | 0 | 1751, 814
07:49:40.418000 | 261 | 1751, 814
07:49:41.350000 | 6 | 1724, 877
07:49:42.147000 | 1 | 2347,1059
07:49:45.885000 | 0 | 1904, 957
Traceback (most recent call last):
File "7a.py", line 232, in <module>
main()
File "7a.py", line 137, in main
mouse.press(x, y) #LMB Down
File "C:\Python27\pymouse\windows.py", line 33, in press
self.move(x, y)
File "C:\Python27\pymouse\windows.py", line 71, in move
windll.user32.SetCursorPos(x, y)
ctypes.ArgumentError: argument 1: <type 'exceptions.TypeError'>: Don't know how
to convert parameter 1

Thanks!

Eldon

Dear all,

Congratulations for this great thread which definitely convinced me of ordering the Max2 to use it as a monitor. Overall I'm fully satisfied with the visual quality and with the hacks developed throughout this discussion thread.

I have been experiencing the same error described by Eldon (see quote above). It looks like this comes from using float X and Y variables for mouse events instead of integers. Nothing to do with Parallels or anything. Adding a few int() here and there was enough to fix it, so I'm attaching a proposed update which is now working for me.

I also inverted the scrolling direction by adding a minus sign, I just find it more natural because it’s identical to how I scroll pages on my smartphone. I guess this is more of a personal preference.

In addition, I've also added a couple of lines which will work only for Windows users, to automatically turn High Contrast on whenever the script is launched, and off when the device is disconnected. Easy to remove for those who don't wish to use this functionality.

I'm currently investigating ways to auto-play the script when the device is connected, to make integration even more seamless. So far, AutoPlay doesn't pop up when I plug the Max2 to my computer... if someone has any thoughts about that, I'm interested!

Please note that the MobileRead forum doesn't allow attachments with the .pyw extension, but if you replace the .py with .pyw I also optimized the script to run in the background with no open console window. Just need to launch it when the device is connected and the process will exit once the USB cable is unplugged.

In terms of graphic optimization, I was wondering: don't you think that a dedicated ICC profile could be a good option to obtain a cross-platform standard for pre-processing? The profile could be designed specifically to process the display data flow into 16-tone grayscale and incorporate Cqman's tips:
- Saturation 0% (grayscale)
- Brightness +25%
- Contrast +50%
- Gamma 1.00
- Digital vibrance +50%
(or any better setting according to calibration).

Then it would be very easy to add this profile on any OS (Windows, Mac or Linux) and make it the default profile for the Onyx monitor only. This way, we would no longer need to activate High Contrast or anything...

I'm switching between my regular screen and Onyx several times a day depending on whether I need to read PDFs or draw colorful sketches and PPTs and I find it a bit annoying that I need to activate High Contrast every time. In addition, my system doesn't allow me to use Cqman's settings on the software side.

The only thing is, I don't have the tools nor the skills to develop an ICC profile. If anyone is more conversant with those things, could you please let me know if 1) this sounds like a good idea and 2) how difficult this would be to implement? Otherwise I can try to look into it but cannot promise anything...

[2new]

Originally Posted by jamespeter26

Dear all,

Congratulations for this great thread which definitely convinced me of ordering the Max2 to use it as a monitor. Overall I'm fully satisfied with the visual quality and with the hacks developed throughout this discussion thread.

I have been experiencing the same error described by Eldon (see quote above). It looks like this comes from using float X and Y variables for mouse events instead of integers. Nothing to do with Parallels or anything. Adding a few int() here and there was enough to fix it, so I'm attaching a proposed update which is now working for me.

I also inverted the scrolling direction by adding a minus sign, I just find it more natural because it’s identical to how I scroll pages on my smartphone. I guess this is more of a personal preference.

In addition, I've also added a couple of lines which will work only for Windows users, to automatically turn High Contrast on whenever the script is launched, and off when the device is disconnected. Easy to remove for those who don't wish to use this functionality.

I'm currently investigating ways to auto-play the script when the device is connected, to make integration even more seamless. So far, AutoPlay doesn't pop up when I plug the Max2 to my computer... if someone has any thoughts about that, I'm interested!

Please note that the MobileRead forum doesn't allow attachments with the .pyw extension, but if you replace the .py with .pyw I also optimized the script to run in the background with no open console window. Just need to launch it when the device is connected and the process will exit once the USB cable is unplugged.

In terms of graphic optimization, I was wondering: don't you think that a dedicated ICC profile could be a good option to obtain a cross-platform standard for pre-processing? The profile could be designed specifically to process the display data flow into 16-tone grayscale and incorporate Cqman's tips:
- Saturation 0% (grayscale)
- Brightness +25%
- Contrast +50%
- Gamma 1.00
- Digital vibrance +50%
(or any better setting according to calibration).

Then it would be very easy to add this profile on any OS (Windows, Mac or Linux) and make it the default profile for the Onyx monitor only. This way, we would no longer need to activate High Contrast or anything...

I'm switching between my regular screen and Onyx several times a day depending on whether I need to read PDFs or draw colorful sketches and PPTs and I find it a bit annoying that I need to activate High Contrast every time. In addition, my system doesn't allow me to use Cqman's settings on the software side.

The only thing is, I don't have the tools nor the skills to develop an ICC profile. If anyone is more conversant with those things, could you please let me know if 1) this sounds like a good idea and 2) how difficult this would be to implement? Otherwise I can try to look into it but cannot promise anything...

Can you take a video of your setup? Im interested in the lag delay with the cursor. Does high contrast mode make it smoother?

[jamespeter26]

Originally Posted by 2new

Can you take a video of your setup? Im interested in the lag delay with the cursor. Does high contrast mode make it smoother?

Unfortunately it has no impact on the lag delay of the cursor. High Contrast (Start Menu > Settings > Ease of Access > High Contrast > Turn on High Contrast) will only improve readability of Windows Explorer, Office apps and some websites which are designed to detect High Contrast and adjust their layout. The cursor is still a little bit delayed, but nothing unbearable.

I can unfortunately not post a video but you can very easily try for yourself how High Contrast will modify your display experience, even without using the Monitor Mod3 app or the Python script. The script only combines the two and makes integration more seamless.

My course of action is the following:
- plug the HDMI cable
- launch Monitor Mod3 app on the Boox Max2
- plug the USB cable (USB debugging should be already activated on the device with ADB installed on the computer)
- launch the Python Script manually, it will run in the background (no open console) // NB: I am still trying to find a solution to execute the script automatically whenever the device is connected to my computer
- the Python Script will automatically execute shell command "sethc 251" which triggers High Contrast, i.e. upon launch of the script, a window immediately pops up asking confirmation whether I want to activate High Contrast
- Click "Yes", High Contrast is activated
- I can then use the device as a touchscreen monitor with scroll and right click
- When I want to stop using the device as a monitor, I just need to unplug the USB cable and the same High Contrast confirmation window automatically opens to ask me whether I want to disable High Contrast
- Click "Yes", High Contrast is disabled
- Unplug HDMI, stop the Monitor Mod3 app

[tomat]

Hello.

There might be a fix for the blurriness issue:

Problem: When the Boox Max 2 is connected as a monitor using the HDMI the resulting image is blurry. The reason is that th HDMI connector suggests (through EDID) a resolution of 2104x1560 instead of the correct resolution of 2200x1650.

Solution: Microsoft says that it’s possible to override the wrong EDID: https://docs.microsoft.com/en-us/win...-monitor-edids

I’m considering to buy a Boox Max 2 but I’m hesitating because of the blurriness issue.
Did anyone try Microsoft’s solution succesfully?

[Waiting4somethin]

I don’t have time lately to mull over this subject (or indeed check the recent activity of this thread) but I almost bought the Max2 on a few separate occasions now, only each time I’ve been put off by videos and user comments on YouTube.

I looked at the latest dasung paperlike, marketed as having a turbo mode which appears to update or perhaps respond faster than I’ve ever seen with an ereader.

Ive only just learned about the more reasonably priced pixel Qi, an lcd without a backlight. Downside is that it’s old, 10.1” and 1024x768 resolution. But it runs at normal lcd speeds.

[mdp]

Originally Posted by Waiting4somethin

I looked at the latest dasung paperlike, marketed as having a turbo mode which appears to update or perhaps respond faster than I’ve ever seen with an ereader

I remain nailed skeptical about products that are told to defy the known limits of established hardware (as in: "the ES133TT3 can update its dots 8 times per second but in our product it's much more" - much more information is needed).

Originally Posted by Waiting4somethin

I don’t have time lately to mull over this subject (or indeed check the recent activity of this thread) but I almost bought the Max2 on a few separate occasions now, only each time I’ve been put off by videos and user comments on YouTube

I also have very little time at the moment, but I would like to re-attempt VNC on the Max2 with a few more tricks and see where it brings. I have faith it will not be bad.
Of course, expecting LCD-like results on current EPD is something that only the youngest would do.

Originally Posted by Waiting4somethin

Ive only just learned about the more reasonably priced pixel Qi, an lcd without a backlight. Downside is that it’s old, 10.1” and 1024x768 resolution. But it runs at normal lcd speeds.

I was so intrigued at the time, and the reported results still look somehow interesting. I am still uncertain about the factors that killed it.

AAH! AH! AAA!, emotion. I just found out that Mary Lou Jepsen «has been a professor at MIT». Some things make sensitive people like yours truly tick.

So I mean, the resume makes proud statements about effectiveness,

She built a business to develop new screen architectures ready for mass production with less than a year of development time (display innovation typically takes a decade or more to reach fruition). Pixel Qi technology shipped several million units with >10x lower power consumption and sunlight readability while preserving color, video and image quality.

but still it is not clear why those million units remained a niche instead of opening a way.

Interesting info comes from Michael Kozlowski: after MLJ left, and big contracts (the military) were failed to be signed, the original factory denied availability of production for small amounts.

Pixel Qi Shifting their Business Strategy away from Consumer Electronics
Pixel Qi Receives a Massive Setback with Chunghwa Picture Tubes
Pixel QI is Officially out of Business

[j.p.s]

Originally Posted by mdp

Pixel Qi Shifting their Business Strategy away from Consumer Electronics
Pixel Qi Receives a Massive Setback with Chunghwa Picture Tubes
Pixel QI is Officially out of Business

There are some formatting errors in the above URLs, those below work.

https://goodereader.com/blog/electro...er-electronics

https://goodereader.com/blog/e-paper...-picture-tubes

https://goodereader.com/blog/electro...ut-of-business

[mdp]

Originally Posted by j.p.s

There are some formatting errors in the above URLs, those below work

Corrected. (Missing trailing quote in the address. PhpBB is not very forgiving.)

[whopper]

Time to step up the game from onyx with a new device.

Their old mantra of "is impossible to add light to 13.3" devices" seems to be wrong

https://youtu.be/vnUACe8Bsyg

And with the new competition from another manufacturers launching devices with everything included from the start, their strategy of milking the customer with small improvements every year while abandoning the software of the old ones seems to be finished.

either that, or them as a company.

[mdp]

Originally Posted by whopper

And with the new competition from another manufacturers launching devices with everything included from the start, their strategy of milking the customer with small improvements every year while abandoning the software of the old ones seems to be finished

Which device? I am uninformed. You linked to a Dasung monitor, then write «another manufacturers launching devices with everything included», and a Dasung monitor contains (does) a fraction of what an Onyx tablet does... Where's this competition.