GH1 firmware research volunteers required

[Vitaliy Kiselev]

As for bricking risk it is not very low. Because now most companies do not separate loader and firmware (LG is good example here, in their TVs loader is placed even in different chip, so it is almost impossible to brick LG TV as loader allow serial cable update). Panasonic here is among bad examples.

We'll concentrate on bitrates and sound stuff, start from simple constants towards more changes.

[Oetam]

The GH1 has to be able to produce native 24p if it can produce native 25p, right?

I also think the 30min time limit and region frame rate differences have to be hardware specific, because there is no difference between the pal and ntsc firmware file. I'm pretty sure this differences are implemented though, that means this could be overridden with a proper firmware hack.

Upping the bit rate could be possible, but if the engine isn't able to process 17mb/s properly (MUD) the chance of mud would be much higher in something like 25mb/s, does that make sense?

[Vitaliy Kiselev]

The GH1 has to be able to produce native 24p if it can produce native 25p, right?

I also think the 30min time limit and region frame rate differences have to be hardware specific, because there is no difference between the pal and ntsc firmware file. I'm pretty sure this differences are implemented though, that means this could be overridden with a proper firmware hack.

Upping the bit rate could be possible, but if the engine isn't able to process 17mb/s properly (MUD) the chance of mud would be much higher in something like 25mb/s, does that make sense?

Try to read previous pages, please. They have parts that deals with 30min limit and regional settings. Both things are just stored is EEPROM areas (and are not touched up by upgrade, as many-many other calibration settings).

As for MUD :-), it is caused by lame codec at low bitrate. Such codec can't be compared to AVCHD camcorders as it is much simpler. Look at Canon bitrates, so if we could up bitrate for 1080p24 up to 40-50Mbps, image will be much better.
People who know how to make high quality footage have big need even for uncompressed HDMI output.
If GH2 could deliver this, it'll be good.
But I believe that it won't be the case as processor used in DSLR cameras was the same as in top compacts, and now this is Venus HD II or MN89521RF. As you might notice, this looks like minor revision, and according to compacts specifications it is.
Most probably bitrate will be up to 24Mbit in GH2 and codec will be slightly improved in other parts.

After working with my consumer HD cameras with zebras, adjustable audio (really like SET+AGC setting) with live meter and headphone output, it is hard to shoot on any current DSLR.

[Vitaliy Kiselev]

Seems that all function call mappings resolved.

[DanDOF]

@tester13

I will gladly donate for 720p24

[Vitaliy Kiselev]

@tester13

I will gladly donate for 720p24

I am pretty ok with donations that go to test bodies.

We must choose a reputable guy to hold donation paypal account and donation page.


P.S. If anyone interested, for the first time complete database for Canon 5D Mark II is now available, thanks to Trammel. Unfortunately 5D mark II firmware decrypton tools are not available (they only decrypt flasher, and not payload). And dumps are not public either.

[thisisapocalypse]

After working with my consumer HD cameras with zebras, adjustable audio (really like SET+AGC setting) with live meter and headphone output, it is hard to shoot on any current DSLR.

2nd system sound (which is superior for so many other reasons anyways) and using a light meter solves these problems - at least if you're doing narrative work, and even if you're not, it doesn't take *that* long to check things with a light meter.

[Vitaliy Kiselev]

2nd system sound (which is superior for so many other reasons anyways) and using a light meter solves these problems - at least if you're doing narrative work, and even if you're not, it doesn't take *that* long to check things with a light meter.

Zebra and light meters are just two absolutely different tools.
As for 2nd system audio recording.
Yes today it is quite common even at the idie level. But it won't be long until one company put good codec, op-amps and normally adjustable parameters with live monitoring. Such audio board with jacks for 4 channel mixing will cost less then $20 if mass produced. All else is software. And as we'll have touch screens on every DSLR in next two years, I don't see any problems.

[weismangroup]

Hi Tester:

Just want to say that I'm very excited about the work you are doing! I'm not a programmer, so unforunately not many ways I can help, but would gladly contribute $ to a fund to help support the project.

By the way, do you think that unlocking to be able to shoot in either 50p or 60p is in the realm of possibility? We shoot a lot overseas, and it would be a great to have user selectable PAL/NTSC serttings.

-mw

[Vitaliy Kiselev]

From current standpoint it seems that 50p/60p switching is not hard.
After we could find how to generate digest we could make two firmware versions for 50p or 60p (ignoring camera region).
User selectable options are not possible at this time as we need good compatible compiler, knowledge of OS thread functions, interface functions, exact RAM map, etc.

[JerryB]

I would also gladly donate and promote for others to donate to a paypal account if two things were able to be achieved or at least at the top of the list.

1) AGC disabled
2) Native 24fps in both 1080 & 720

It seems from past posts that both of these seem attainable in regards to the current hardware. Please let us know if things get brought up a step and you have TZ5 success along with a place to begin donating. There are 1,536 member of the GH1 group at Vimeo, if only 1/4 (high fraction but just for example) of those people donated $5 each to the cause, that would be $1875. Enough to buy a GH1 and TZ5.

If you think it's worth it, according to the information you already have, I hope you continue to crack the GH1 code!

[Vitaliy Kiselev]

As for AGC and native 24fps they could be hard.
We must know exactly used audio codec model. This is how ML become able to remove AGC.
I think that it is quite easy to remove few screws and look at Audio P.C.B. at the top.


Detailed analysys of native fps could be only possible after major video codec related things become clear. List all current native progressive modes, please (it is useful to provide detailed information from video%

[nobbystylus]

as far as i'm concerned i think boosting bit rate up to 24 mbps would be the priority for any firmware hack, and, from the sounds of things, one of the easier options..

[alignment1]

as far as i'm concerned i think boosting bit rate up to 24 mbps would be the priority for any firmware hack, and, from the sounds of things, one of the easier options..

+1

[Andreus]

Firmware research

Higher bitrate is great, but IMO the most important feature that the camera lacks right now (for cine usage) is some type of live external monitoring during record.

The camera already has live composite (Standard def) out from the viewfinder that disappears when you hit record. If it can stay on during record, that will open up a world of possibilities for this camera Obviously live HDMI out would be a filmmakers dream come true, but I imagine composite out would be easier given that it already exists on the camera in restrained form.

+1 on higher bitrate

I would be willing to donate orders of magnitude more than $5 to such a research effort. This is more than a lottery ticket, this is (hopefully) an attempt to really support the GH1 as a pro cine production tool, something Panasonic has had no interest in doing, and IMO this effort is worth its weight in gold. How about everyone on this board hoping to make movies with the GH1 donate say, $50-100?

If the research works and we get better bit rates, live monitoring, audio, etc.. your donation pays for itself in one gig that you wouldn't have otherwise gotten b/c you didnt have a camera that cost thousands of $$$s more. Sounds like a value to me!