Messages

i uploaded a test video.  7 seconds, 254 MiB....
used a very high f-stop and very high ISO to get a lot of noise.

will shoot trees outside on weekend.

http://upload.g3gg0.de/pub_files/58018fa07888c1c9023edf21ee15eab3/MVI_3014.MOV

DOWNLOADING! Thanks!  by the way, if you could record mounted on a tripod, with and without modified bitrate, id would be of huge help for comparinson!

hehehe I know, but since I dont ave any knowledge that could help in the development, I'm trying to give my grain of salt..

But it all seems to be in vain...   I even wrote to geohot and had a similar reply saying not to worry.. but this developer seems quite
jumpy about this funding issue... so I think this has gotten into a dead spot.  At least for now..

Now that you mention de 7D being shared between 3 people..   Im curious..  I've read about all the roadblocks you've encountered...
But I was wondering, besides receiving inside documentation..(not likely)..  and getting other cameras..   is there any specific/special equipment that could help in the reverse engeneering process?

I've been reading some discussions about Qscale and Bitrate regarding image quality.
I have 2 questions that I dont know if they've been anwered indirectly, so Im asking here instead of creating a new thread.
Specially since developers are stating that increasing bitrate doesnt improve image quality.

1. What encoding factor could improve final results in darker areas and gradients? (banding/blocking) ( www.davidarango.com/H264.PNG )
2. In your experience with the image quality affecting factors.. theoretically speaking, could something be done software wise to improve the downsampling/line-skipping method to produce sharper images?

[SPECULATIVE]

I own a 7D, and I'm fully aware of the current limitations of ML development.

BUT, this post is just merely for SPECULATIVE purposes.

I have NEVER used ML, and I dont know much about the limitations of current builds in regards of the available hardware.

BUT, it has always been spinning in my mind... If ML does what it does whit just a single DIGIC processor...

So, ML developer's... given your experience..
What features could be possible if the dual DIGIC processors are unlocked to work with ML?

Clean HDMI out?
Higher FPS?
H.264 encoding improvement?

As I say, this is just for speculative purposes...

Ideally speaking.. other than inside information and documentation from canon (which is highly unlikely), what tools/resoruces would be needed to successfully reverse engeneer the 7D?

- At least how many 7D bodies to tear appart/brick?
- Special software/equipment?

[So my question is, what do you think about this method, and what would your suggest this developer in terms of funding his project and also keeping peace of mind while doing it.]

Thanks a lot for your reply!

I underestand what you mean about copyrights and reverse engeneering, that answer should comfort "someone1.00".

Now, when it comes to funding the "research" I proposed the following using Kickstarter.

Kickstarter doesnt allow or wont risk funding anything related to violating copyrighted work.. clode included..

So I proposed that the Kickstarter iniciative could be "masked" as a "research" project, with the sole purpose of finding out
what the camera's hardware has capable of beyond its current capabilities.

Obviously with no metion of what other kind of reverse engeneering would be done.

That way he could be free and comfortable publishing any modified software afterwards...


So my question is, what do you think about this method, and what would your suggest this developer in terms of funding his project and also keeping peace of mind while doing it.

Thanks again for your response!

[But, what are the risks for a developer when reverse engeneering a camera's firmware?]

Hello to everyone!
Im starting this thread because I'm actually getting active in a thread at DVXusers.net where someone is developing a custom firmware for the NEX-5N.
http://www.dvxuser.com/V6/showthread.php?259944-Sony-NEX-Axx-Firmware-hacking/page7

There has been some talk about funding and even starting a Kickstarter iniciative under the excuse of research purposes.

Cases like the PS3 with GEOHOT are sacring the developer, specially because the thread is getting pretty popular... but if he wants to go ahead with development he needs funding to buy the cameras.. etc..

So here comes my question.

GEOHOT's lawsuit, as far as I know, was MAINLY because what he was doing allowed people to use pirated game copies..

But, what are the risks for a developer when reverse engeneering a camera's firmware?

Canon has posted their kernel sources, so I dont know what would this mean in regards of the "openness" of the firmware itself.

http://www.sony.net/Products/Linux/DI/category01.html



On a side note, how similar/different is developing for canon and for sony?


Thanks to whoever can help whit this.

Hey there guys! have you contacted www.lensrentals.com ??  they have access to pretty beaten up cameras, and often they do teardowns!
http://www.lensrentals.com/blog/2012/04/5d-iii-strip-tease
They are pretty daring and have a great spirit... a lot of good could come if they got interested on this proyect!