Good news about AF100 firmware

No. 8/10 bit is highly unlikely to be a firmware function.

That's what I was afraid of. But where does the actual downconvert happen from 18bit dsp to 8bit out?

BTW, this is far and away the only feature I'd really like to change other than turning audio ON during VFR mode.
 
SHHHH!

Stop talking about hacking the firmware until more of the cameras shipped to us. I don't want Panasonic to realise and close it down by 'locking' the code...

That would be foolish for Panasonic. Making the firmware accessible so that customers can modify it, makes the camera that much more flexible and popular. Locking out hacks doesn't help Panasonic.
 
That's what I was afraid of. But where does the actual downconvert happen from 18bit dsp to 8bit out?

BTW, this is far and away the only feature I'd really like to change other than turning audio ON during VFR mode.

It depends how much of the design is in ASIC (a custom chip, unchangable, cheap) and how much is in FPGA (a programable chip, but more expensive). If it is all in FPGA, then anything is possible. BUT, you need to know a great deal about the FPGA code and the I/O functions of the FPGA. That's an enormous reverse engineering task.

edit - There may be hope it is FPGA since the volume of this camera is probably low enough that a custom ASIC might not be justified. Anyone want to open their's up and take pictures? :) Then the engineering experts can tell us what the chips are and if there is hope of a superhack.
 
That would be foolish for Panasonic. Making the firmware accessible so that customers can modify it, makes the camera that much more flexible and popular. Locking out hacks doesn't help Panasonic.

Without a doubt. When I bought my WRX the dealer quietly whispered to me that one could modify it, make it into a 400hp, 500hp car. Of course this was never in the official documentation from Subaru.

In the same way Panasonic's official statement (mostly aimed at Vericam owners) would be "we do not approve of this hacking and what you are doing" but behind closed lines they know just the idea will sell a lot of cameras.
 
I retract my statements about 8/10 bit. Indeed, with DSP, just about anything is possible. I don't follow how one could re-write DSP code to do this, unless there is a complete file system where the modules are visible and one can figure it out. Hat's off to the hackers if they can.
 
I dunno. if you are going to hack something panasonic might not want you to Hack, like the AF 100, would this be a great place to have the discussion about the hackability of the camera? It would seem like plotting your bank heist on the steps of the police station, no?
 
Does this mean that 50 Mbit recording may be possible? Or is that not even a possibility?

If you were to combine the vast benefits of AVCCAM with functional 50 Mbit or beyond recording (functional, meaning not six seconds worth of recording) this camera would be very significantly more interesting.
 
I think a movie function should be added. You know, a button you can turn on and off that automatically generates lighting, art direction, etc...
 
A small wish-list gathered from my own nits and others' posts:

- killing headphone output and/or audio meters in VFR
- 1:1 pixel zoom
- ISO being selectable from the rear wheel
- more frame rates, down to, say, 1fps
- relocating/swapping record button(s)

Nice! Thanks for posting my nits which area identical really. Vitaliy, what I am curious of is if they feed a full resolution signal to the LCD display, can we send a control message to the scalar in it to change it to a 1:1 crop on center at least. That's what I was thinking would be the easiest hack, but not to say the hack is easy. In any case, this is really a pretty small list of nits and I REALLY hope Panasonic can address a few of them at least. But, I just wanted to say, even if none of these things are possible from the people hacking the firmware, the info is really cool to learn. I am an electrical engineer who works on cmos chip design so moving up to system level information is neat. By the way, for those asking about DSP chips, they have some AMAZING code in them. I've programmed thousand dollar DSP chips for a research project back in my college days and was amazed at what many of them include. Given these are probably video specific, they probably are not as general purpose DSP as what I used, but nevertheless neat.
 
bumping the record rate to 35mbit or higher would be on the top of my list. i'd much rather we try 35 and proving it reliable before going higher
 
Back
Top