Should Pyra have a Camera?

[Natsu]

This is a official replacement module.


On DigiKey they are better and cheaper.


And yes, even 30 euros is "very very" worth it.

Then show me a better and cheaper one.

Link


Could be place were the headphone port was move to and move the port back.

Does MIPI support camera, I wonder? If it does that would mean we could use a single cable for everything in the lid: LCD, i2C, a GPIO or two, and the camera. If it doesn't then that may end up being the biggest challenge: how to get the signal from the camera to the SoC.

I'm pretty sure it does.

 

[ZetaNeta]

Does MIPI support camera, I wonder? If it does that would mean we could use a single cable for everything in the lid: LCD, i2C, a GPIO or two, and the camera. If it doesn't then that may end up being the biggest challenge: how to get the signal from the camera to the SoC.

Yes. It does.

This is a official replacement module.


On DigiKey they are better and cheaper.


And yes, even 30 euros is "very very" worth it.

Then show me a better and cheaper one. Link


Could be place were the headphone port was move to and move the port back.

Go search for one youself.


There are just too many!

 

[Natsu]

Go search for one youself, lazyass.

There are just too many!

Yeah. Many of them don't have a real picture, no picture and unrelated tech for the Pyra.When you said that, I thought you had a camera in mind. Seems to be easier just to look through replacement parts.

 

[Tenka]

It's silly that you two can argue with each other too.

 

[Natsu]

It's silly that you two can argue with each other too.

Less of an argument and more of a clarification.

 

[Drammurt]

Gonna drop my 2 cents here. If you have followed the $25 Raspberry Pi-camera board come to life, you'd have noticed that creating a camera board is not a matter of plug and play. The following two posts by David Plowman explain pretty well the work that had to be done. Apparently it was not the hardware-part that took so much time, but the translation from raw sensor-data to usable pictures. This means, that if you want one or more camera's on the Pyra, you should have a digital imaging expert in your team.

Creating the camera board

Creating the camera board - Part 2

Also, as far as costs go... these addon boards costs $25 for a 5 MP camera with surprisingly good quality. That is $25 for a camera board developed by a non profit organization and volunteers and produced in large numbers. I wonder what the added costs are for the Pyra. I'd rather not have a camera added because of the added costs, extra work and because I have much better devices with optical image stabilization to take my pictures and video's with.

 

[ZetaNeta]

Gonna drop my 2 cents here. If you have followed the $25 Raspberry Pi-camera board come to life, you'd have noticed that creating a camera board is not a matter of plug and play. The following two posts by David Plowman explain pretty well the work that had to be done. Apparently it was not the hardware-part that took so much time, but the translation from raw sensor-data to usable pictures. This means, that if you want one or more camera's on the Pyra, you should have a digital imaging expert in your team.

Creating the camera board

Creating the camera board - Part 2

Also, as far as costs go... these addon boards costs $25 for a 5 MP camera with surprisingly good quality. That is $25 for a camera board developed by a non profit organization and volunteers and produced in large numbers. I wonder what the added costs are for the Pyra. I'd rather not have a camera added because of the added costs, extra work and because I have much better devices with optical image stabilization to take my pictures and video's with.

Obviously not plug and play, yet, dont think we dont know this.


Yet, we are not on Rasp.Pee.

 

[_wb_]

Gonna drop my 2 cents here. If you have followed the $25 Raspberry Pi-camera board come to life, you'd have noticed that creating a camera board is not a matter of plug and play. The following two posts by David Plowman explain pretty well the work that had to be done. Apparently it was not the hardware-part that took so much time, but the translation from raw sensor-data to usable pictures. This means, that if you want one or more camera's on the Pyra, you should have a digital imaging expert in your team.

Creating the camera board

Creating the camera board - Part 2

Also, as far as costs go... these addon boards costs $25 for a 5 MP camera with surprisingly good quality. That is $25 for a camera board developed by a non profit organization and volunteers and produced in large numbers. I wonder what the added costs are for the Pyra. I'd rather not have a camera added because of the added costs, extra work and because I have much better devices with optical image stabilization to take my pictures and video's with.

There are $5 USB cameras that give usable pictures -- that is the end-user price bought in single units. They have their own case, a USB cable and usually a built-in mic and sometimes a led -- we would only need the sensor and the component that processes the raw sensor data into something useable. I think it should be possible to do it for less than $5.

Also I don't really need a 5MP camera -- something that is good enough for video chat is OK for me (640x480 @ 30fps or so).

 

[ZetaNeta]

I wanna a HD camera in pyra. I really do. I mean to have it as a only device with me most of the time.

 

[klapse]

So far nobody has been able to link to a camera module with open linux drivers.

My ex-GF did software development for Nokia cameras. It is not easy stuff. It appears to be unlikely that we will find someone who can donate the skills necessary to integrate a HD camera. Therefore my "no" vote.

 

[WizardStan]

$10 module 640x480, 30fps.

Don't see any actual drivers, but it's standard 8bit parallel data (which the OMAP5 understands, therefore we just need the OMAP5 camera drivers) and well documented i2c command interface which would make it reasonably straightforward to write our own interface.

edit:

2.5mm cube for $5

That is so unbelievably tiny. It uses something called "Serial Camera Control Bus" (SCCB) which is apparently a standard, may or may not be easier to write a driver for than the i2c controlled one, and no idea about MIPI interface.

 

[_wb_]

It's a bit strange in my opinion that people are voting "No" not because they don't want it, but because they assume it will be too difficult to implement and/or too expensive. When such polls appear about other potential features (e.g. internal storage, memory, processors, e-ink displays, etc), cost and implementation effort seem to be irrelevant, but when it's about a relatively simple, relatively cheap feature like this, suddenly everyone is an expert who knows first hand how tremendously difficult and expensive it is.

A simple 640x480 30fps USB 2.0 webcam costs very little. The OMAP5 has plenty of USB 2.0 inputs. Can't we just use the components from one of those simple webcams and use those? No real hardware effort besides finding a place for it in the case, no software effort besides making sure that the relevant kernel module is included.

 

[ZetaNeta]

 I would pay a maximum of 50 euros extra for a HD camera. But they cost maximum 30.
 

It's a bit strange in my opinion that people are voting "No" not because they don't want it, but because they assume it will be too difficult to implement and/or too expensive. When such polls appear about other potential features (e.g. internal storage, memory, processors, e-ink displays, etc), cost and implementation effort seem to be irrelevant, but when it's about a relatively simple, relatively cheap feature like this, suddenly everyone is an expert who knows first hand how tremendously difficult and expensive it is.
 
A simple 640x480 30fps USB 2.0 webcam costs very little. The OMAP5 has plenty of USB 2.0 inputs. Can't we just use the components from one of those simple webcams and use those? No real hardware effort besides finding a place for it in the case, no software effort besides making sure that the relevant kernel module is included.

Very true. We should include "(dont think bout the price)" in each pool i guess.
 
 

So far nobody has been able to link to a camera module with open linux drivers.
 
My ex-GF did software development for Nokia cameras. It is not easy stuff. It appears to be unlikely that we will find someone who can donate the skills necessary to integrate a HD camera. Therefore my "no" vote.

All of those cameras got a linux driver for sure. Atleast in the sunxi project. All the drivers (as i know) are open.
USB UVC cameras are out of the question immediately, as its a universal format.
There should be much more cameras actually.
 

 

[leathco]

I gotta say yes, especially since I plan on replacing my phone with a Pyra.  Either both a forward and rear facing camera or a single camera that can rotate, whichever would be cheaper to make.  So many uses on Pyra for a cheap camera, such as basic video conferencing.  Also, I don't think it would add much in the price if the camera were a low-res camera, something like 640 x 480 resolution.  Obviously, a 720p camera would be nicer but would cost a lot more.

 

[Kiga]

Some people would not buy a Pyra just because it haven't a camera  while they want to replace their smartphone. Since it cost only $5, I see no reason to avoid it. I'm sure that it would lose a lot of people just because there isn't a $5 camera.

 

[TrashyMG]

Some people would not buy a Pyra just because it haven't a camera  while they want to replace their smartphone. Since it cost only $5, I see no reason to avoid it. I'm sure that it would lose a lot of people just because there isn't a $5 camera.

component may be $5, but there is a ton of other considerations design wise.

 

[leathco]

Some people would not buy a Pyra just because it haven't a camera  while they want to replace their smartphone. Since it cost only $5, I see no reason to avoid it. I'm sure that it would lose a lot of people just because there isn't a $5 camera.

component may be $5, but there is a ton of other considerations design wise.

Won't the only design wise things be where to put the camera and where to run the power and data wires?  The cheapest and easiest way would be to put the camera either beside or above the LCD, have a spot for the camera in the mold, then run the wires through the hinge, perhaps on the same trail the LCD cable runs and then branch it off to it's spot on the board.  To me it would seem the backlighting that's already done would be harder, as you have multiple spots to power and place LEDs, but with the camera it's just one spot.

I DO see design becoming slightly more complicated if it's a dual camera setup, as now the outside lid would also require a camera cutout.  I see significant design issues if the camera is on the hinge and can rotate, as that would require at the minimum notches for the camera to catch and lock, a hollowed channel for cables to pass through, and significant testing to verify that no cables inside could be caught and damaged by said rotation.  But a single camera should be easily doable.

 

[ZetaNeta]

I gotta say yes, especially since I plan on replacing my phone with a Pyra.  Either both a forward and rear facing camera or a single camera that can rotate, whichever would be cheaper to make.  So many uses on Pyra for a cheap camera, such as basic video conferencing.  Also, I don't think it would add much in the price if the camera were a low-res camera, something like 640 x 480 resolution.  Obviously, a 720p camera would be nicer but would cost a lot more.

A "Very Very HD" may cost up around 10 euros. Sure?

Some people would not buy a Pyra just because it haven't a camera  while they want to replace their smartphone. Since it cost only $5, I see no reason to avoid it. I'm sure that it would lose a lot of people just because there isn't a $5 camera.

component may be $5, but there is a ton of other considerations design wise.

I see no trouble

 

[TrashyMG]

I see no trouble

Of course you don't.

 

[comradekingu]

Yes, because more money spent on storage means a faster device. More money spent on this is additional complexity and makes for a mediocre camera/camcorder.