GH1 firmware research volunteers required

[Vitaliy Kiselev]

The only hiccup I see is that the camera freezes when trying to play a native FHD 24p .mts file. But...

No one promized that it won't.
As player do not expect such surprizes.
We'll investigate further file structure and all flags.

I believe that you can try to lower GOP to slightly reduce mud.
By idea mud in native fps and low GOP must be much lower.

[billy fattey]

Look numbers above input incriment field, they show you exact number, this is hexadecimal number, so 1.2D also exists and works, you have about 65.000 version numbers to go :-)
Plus we have version revert patch in reserve.

I don't understand. I've tried putting in decimal and hexadecimal numbers as increments, but I get the same result. I've attached a screenshot. What number should I use next as an increment so that I have 65,000 more? Thanks.

[Car3o]

i'd like to add that a 3rd party battery patch would be nice for those who want to get extra batteries. i have the original 1.2 firmware that allows 3rd party batteries if needed.

[Vitaliy Kiselev]

Sorry, I do not understand your point.
0xFFFF is maximum version number :-)
Calculate that is equal 0xFFFF-0x121.

[Vitaliy Kiselev]

i'd like to add that a 3rd party battery patch would be nice for those who want to get extra batteries. i have the original 1.2 firmware that allows 3rd party batteries if needed.

I know, this is same low level stuff as AGC. But easear to spot.
Read my previous posts. I traced it back to main camera state structure and know hardware error number.
But it is almost impossible to find where it sets error state and code.

As for 1.21 (actual) firmware with battery support it is of little help, as it is wastly different from 1.22 (1.12 is much much closer to 1.22).

[e-steve]

Have you guys seen how the 5DMkII Magic Lantern guys were able to structure their development/testing, etc?

Was there something you unique they had so that they could create new functionality (I think they added audio level meters and some other new features)? Maybe the creator could give us some pointers? I don't understand it enough to know how their dev process was different, but this project seems to have similar to potential to theirs in terms of impact to GH1 users!

Here's more info on their firmware mods:
http://magiclantern.wikia.com/wiki/Magic_Lantern_Firmware_Wiki

[Car3o]

alright...good to know

[Vitaliy Kiselev]

e-steve

Thanks for old news :-)
ML is one man show, in fact.
Try to read my previous messages in this thread about ML, if you want to know more.
I don't like this project concept and I don't want to go their way, period.

[billy fattey]

Sorry, I do not understand your point.
0xFFFF is maximum version number :-)
Calculate that is equal 0xFFFF-0x121.

Ok, so you're saying everything is working how it should. I have upgraded firmware 35 times so the GH1 thinks I am upgrading to 1.4. That's normal, right?

I don't care that much, but if this whole experiment doesn't yield results then the only way to upgrade to whatever version Panasonic is coming out with in May (maybe 1.3?) is by using your emergency patch. Is that right?

[Vitaliy Kiselev]

No, you can use patcher for 1.3 to make it 14.56 :-)

[rawdod]

Hey, I just wanted to say, this thread rules. I hope we can really fix the codec and framerate limitations on this camera, I feel like the way this camera encodes is partially fucked up and just increasing the bitrate or even to all I frames prolly wont be enough, but maybe it would go pretty far. The sensor itself, the CPU and the hardware that writes to the SD cards, I hope are up to snuff. ASM with no comments is some pretty hard core stuff, I can code C but dealing with that ... good night. lol. I'll be blown away if yall can read the sensor data directly then jump to your own encoding function ;P How about red raw? :D

[g.l]

re. card speed, it's possible that the card isn't actually being written to 100% of the time, but rather in small bursts (because the camera is busy doing something else, like compression).

If that's true then don't expect that you can simply figure out the needed card speed from the video bitrate, you may need a certain percentage more.

[Camera Expert]

A Sandisk class 6 should have adequate speed for these tests.

[NURBS]

I give small notice'

New Ptool version have some problem from GOP, it wasn't stable, few same shot have different data. Testing more..

For example: (shoot stable picture, don't move or shake camera, duration 1:00)

Version change - X
30min removal - X
VA Bitrate Adjustment - 20xxxxxxx
VA Bitrate Adjustment 2 - 24xxxxxxx
VA Bitrate Top Limit - X
AV Native 24/25p attempt -X
VA 720p50 GOP Size - Change
VA 1080p25 GOP Size - Change
AU Audio encoding bps - 320000

720p GOP - 4 (14,7Mbps) and 1080p GOP - 4 (18,8Mbps)

change to

720p GOP - 4 (8,8Mbps) and 1080p GOP - 5 (18,2Mbps)

and change to

720p GOP - 3 (14,9Mbps) and 1080p GOP - 3 (7,5Mbps)

and change to

720p GOP - 3 (15,4Mbps) and 1080p GOP - 5 (18,0Mbps)

Regards
NURBS

[Vitaliy Kiselev]

PTool 3.14 Released
Native 24p/25p patch added,
problems with 24p fixed now

Ptool 3.13 - Died in infancy

[Vitaliy Kiselev]

NURBS

All I can say is that GOP patches are unchanged from your previous tests.
It is not possible to understand that caused your fluctuations.

Can you provide small (5s) clips for all video resolutions and all framerates (plus all settings of compression)?
Including ALL resolutions of MJPEG.

[NURBS]

Okey one moment.

[NURBS]

Upload link: http://rapidshare.com/files/381874574/video_example.zip.html

Setting for this video ;-) not too bad, like been tonight tests.. (maybe I was too tired) :badputer:

Ptoot 3.14 ver. ;-)

Version change - X
30min removal - X
VA Bitrate Adjustment - 20xxxxxxx
VA Bitrate Adjustment 2 - 24xxxxxxx
VA Bitrate Top Limit - X
Native 24/25p attempt -X
VA 720p50 GOP Size - Change
VA 1080p25 GOP Size - Change
AU Audio encoding bps - 320000

720p GOP - 3 (17,9Mbps) and 1080p GOP - 4 (17,2Mbps)

Regards
NURBS

[NURBS]

But interesting news. (same settings)

After 1 h. recording on SHD (720p mode, but very hi bitrate 20,1Mbps) last record on FHD (1080p mode) stoping after 10 sec. (video file is empty)
Maybe my camera slightly mad..

Testing more..

[tyampel]

I know, this is same low level stuff as AGC. But easear to spot.
Read my previous posts. I traced it back to main camera state structure and know hardware error number.
But it is almost impossible to find where it sets error state and code.

As for 1.21 (actual) firmware with battery support it is of little help, as it is wastly different from 1.22 (1.12 is much much closer to 1.22).
Batter support should be a very low priority.
The Chinese are already making batteries with chips inside that defeat the "origin" tests. I bough mine from BestBatt for $23.00, have the price of the genuine.
They also sell a battery for $16.00 that will only work with the early firmware.
So for $7 in savings it is not worth to waste Tester's time.
Let him tackle the mud issue solving which will elevate the GH1 to a totally new level.
He already solved two nuisance issues - the 30 minutes recording level for European markets and Japanese only restriction for the Japanese region.

Great job so far.

[Jalad]

I disagree. It's likely all these batteries will stop working when Panasonic release 1.3 - and I assume tester13 will be patching this firmware when it arrives.

[NURBS]

Test results: This settings is stable (for my camera). ;-) 40min rec. non stop for all mode.

Version change - X
30min removal - X
Native 24/25p - X
VA Bitrate Adjustment - 20xxxxxxx
VA Bitrate Adjustment 2 - 24xxxxxxx
VA Bitrate Top Limit - X
VA 720p50 GOP Size - 3
VA 1080p25 GOP Size - 10
AU Audio encoding bps - 320000

720p (19,1Mbps) and 1080p (19,0Mbps)

[spamrakuen]

Test results: This settings is stable (for my camera). ;-) 40min rec. non stop for all mode.

Version change - X
30min removal - X
Native 24/25p - X
VA Bitrate Adjustment - 20xxxxxxx
VA Bitrate Adjustment 2 - 24xxxxxxx
VA Bitrate Top Limit - X
VA 720p50 GOP Size - 3
VA 1080p25 GOP Size - 10
AU Audio encoding bps - 320000

720p (19,1Mbps) and 1080p (19,0Mbps)

Would be nice to see a clip with these settings in order to check the IQ.

[dc98]

Im realy interested too, please post some clips with these settings...some scenes with fast movin' and some with complex scenes (detailed...like grass, or forrest)

[NURBS]

Some video for this test: http://rapidshare.com/files/381932496/video_test_3_10.zip.html

plus MJPEG 24p

Regards
NURBS

[spamrakuen]

Some video for this test: http://rapidshare.com/files/381932496/video_test_3_10.zip.html


Thanks!



plus MJPEG 24p


MJPEG @ 24fps is working? :shocked:
Or there are still "jumps" when panning and so?

[e-steve]

I might suggest that the 3rd party battery issue also might wait until later, since we all can just spend 20 bucks more and get a first party version.

It is possible, that with such an addition to the firmware patch, that it could make a certain company unhappy with these efforts... so best to stick to things that don't impact their accessory sales until a little bit later? Just a suggestion.

[philiplipetz]

Some video for this test: http://rapidshare.com/files/381932496/video_test_3_10.zip.html

plus MJPEG 24p

Regards
NURBS

ERROR

This file is neither allocated to a Premium Account, or a Collector's Account, and can therefore only be downloaded 10 times.

This limit is reached.

To download this file, the uploader either needs to transfer this file into his/her Collector's Account, or upload the file again. The file can later be moved to a Collector's Account. The uploader just needs to click the delete link of the file to get further information.

[svecher]

Batter [sic] support should be a very low priority.
The Chinese are already making batteries with chips inside that defeat the He already solved two nuisance issues - the 30 minutes recording level for European markets and Japanese only restriction for the Japanese region.
Great job so far.
+1. In addition to available alternatives, there is an unknown liabilility concern. Can everyone applying patches sign a disclaimer promising not to go after tester13 (not likely it will have much effect anyway ;) the first time something "stupid" happens? Ultimately, T13 makes his own decisions, but my suggestion is not to touch this one.

[Jesse Brauning]

The progress on new firmware is definitely encouraging.
I would like to see everyone involved be as professional as possible about this, similar to the magic lantern stuff on the camera side.
Also, maybe if you could upload the video to VIMEO or something, that would be helpful.
Though the compression might defeat the purpose of the test footage.

[billy fattey]

Thanks!



MJPEG @ 24fps is working? :shocked:
Or there are still "jumps" when panning and so?

MJPEG at 24fps still skips every 5th frame. Not complaining, just letting everyone know.

I have been testing changes to 1080p24 GOP but without any change in quality.

PTool 3.14

Version Change X
Native 24p/25p X
MJPEG 30fps->24fps X
VA Bitrate adjustment 20000000
VA Bitrate adjustment2 24000000
VA Bitrate top limit (20->25mbps) X
VA 1080p24 GOP Size (I've tried 2,5,8,10)

It's stable and even the bitrate goes up to 2.5 MB/s (for GOP size 2 and 5), but the quality isn't noticeably different. Yet.

[billy fattey]

The progress on new firmware is definitely encouraging.
I would like to see everyone involved be as professional as possible about this, similar to the magic lantern stuff on the camera side.
Also, maybe if you could upload the video to VIMEO or something, that would be helpful.
Though the compression might defeat the purpose of the test footage.

I recommend Dropbox. It's easy, available for all platforms, no download limits, etc.

[tyampel]

+1. In addition to available alternatives, there is an unknown liabilility concern. Can everyone applying patches sign a disclaimer promising not to go after tester13 (not likely it will have much effect anyway ;) the first time something "stupid" happens? Ultimately, T13 makes his own decisions, but my suggestion is not to touch this one.

If something stupid happens due to the modified firmware the user that installed it after being properly warned about such risk is out of $660.00.

In the US it will cost you much more to say "hello" to a lawyer. ;)

[svecher]

If something stupid happens due to the modified firmware the user that installed it after being properly warned about such risk is out of $660.00.
In the US it will cost you much more to say "hello" to a lawyer. ;)
I was thinking along the lines of: bad battery installed, explodes, sprays user with melted lithium cobalt oxide, who then proceeds to sue Panasonic for not being dilligent enough in stopping anonymous hackers from luring innocent Panasonic customers into such dangerous risk-taking with dreams of mudless video, clean audio and who knows what else. So, yeah, my point is: don't touch it with a ten foot pole, especially a disassembler.

[Svart]

I was thinking along the lines of: bad battery installed, explodes, sprays user with melted lithium cobalt oxide, who then proceeds to sue Panasonic for not being dilligent enough in stopping anonymous hackers from luring innocent Panasonic customers into such dangerous risk-taking with dreams of mudless video, clean audio and who knows what else. So, yeah, my point is: don't touch it with a ten foot pole, especially a disassembler.



You wouldn't get sprayed with lithium. The largest risk of dangerous lithium precipitation is when the battery is charging. I should know, I designed battery chargers for LI-ION batteries for a couple of years.

[sammysammy]

This is a dumb question, I'm not into programing or anything like that..but what about the banding issue in underexposed areas (in low light situations), is that something that can be treated or its mostly hardware..thanks for all the work so far!

[Svart]

The fixed noise is a byproduct of the CMOS sensor. It's a hardware issue. However, the gain of the amplifiers/buffers between the CMOS and the processing add noise as well. Nobody knows for sure how much each plays into the fixed noise banding. It would take a serious hardware debugging to do so. I could do it but I'm not willing to take my camera apart at this point. Maybe when the warranty runs out I'll give it a go.

Most likely there won't be anything we can do about it.

Also, I'm not sure that the CMOS sensor itself is the limitation for the line scanning speeds that give us the infamous shutter skewing we know and hate. Since the sensor is technically a dumb device, a processor must be the one to do the work. The rolling shutter is probably an artifact of the speed ability of the processor to sample each line/pixel, process and output. It's likely a bandwidth limitation for the type/cost of the parts used. In short, I bet Pana used the cheapest parts they could use and still get decent video. Other recorders have used CMOS sensors without much or any rolling shutter(RED seems to have done it since the initial RED one came out with some accounts of rolling shutter which was fixed via firmware according to people who claim to have seen it). Panasonic's new pro camcorder also uses the same sensor as the GH1 and I would bet does not exhibit rolling shutter. The real question is: How much of the processing is software and how much is hardware? If it's software, pana or a firmware writer could essentially kill all the unused processes that were un-needed(such as still picture stuff) and then dedicate all of that CPU time to processing the video.

[kainekainekaine]

Do the GF1, G1, Olympus E-p cameras suffer from the fixed noise or is it just specific to the GH1?

[billy fattey]

The fixed noise is a byproduct of the CMOS sensor. It's a hardware issue. However, the gain of the amplifiers/buffers between the CMOS and the processing add noise as well. Nobody knows for sure how much each plays into the fixed noise banding. It would take a serious hardware debugging to do so. I could do it but I'm not willing to take my camera apart at this point. Maybe when the warranty runs out I'll give it a go.

Most likely there won't be anything we can do about it.

Also, I'm not sure that the CMOS sensor itself is the limitation for the line scanning speeds that give us the infamous shutter skewing we know and hate. Since the sensor is technically a dumb device, a processor must be the one to do the work. The rolling shutter is probably an artifact of the speed ability of the processor to sample each line/pixel, process and output. It's likely a bandwidth limitation for the type/cost of the parts used. In short, I bet Pana used the cheapest parts they could use and still get decent video. Other recorders have used CMOS sensors without much or any rolling shutter(RED seems to have done it since the initial RED one came out with some accounts of rolling shutter which was fixed via firmware according to people who claim to have seen it). Panasonic's new pro camcorder also uses the same sensor as the GH1 and I would bet does not exhibit rolling shutter. The real question is: How much of the processing is software and how much is hardware? If it's software, pana or a firmware writer could essentially kill all the unused processes that were un-needed(such as still picture stuff) and then dedicate all of that CPU time to processing the video.



I feel almost exactly the opposite. I would think the fixed noise/banding is a software problem and the rolling shutting is a hardware problem.

The fact that we can take a still picture that doesn't exhibit banding issues means to me that it's not the sensor, it's something between the sensor and the SD card. So it's either a hardware bandwidth issue or it's software. Considering Panasonic sells high end video cameras I would think purposely crippling a consumer product is likely.

The rolling shutter problem exists everywhere, even in film to some degree. CMOS sensors are known to exhibit it, and only the highest quality and latest CMOS sensors can limit it to a triviality. It's not software.

[Kholi]

The fixed noise is a byproduct of the CMOS sensor. It's a hardware issue. However, the gain of the amplifiers/buffers between the CMOS and the processing add noise as well. Nobody knows for sure how much each plays into the fixed noise banding. It would take a serious hardware debugging to do so. I could do it but I'm not willing to take my camera apart at this point. Maybe when the warranty runs out I'll give it a go.

Most likely there won't be anything we can do about it.

Also, I'm not sure that the CMOS sensor itself is the limitation for the line scanning speeds that give us the infamous shutter skewing we know and hate. Since the sensor is technically a dumb device, a processor must be the one to do the work. The rolling shutter is probably an artifact of the speed ability of the processor to sample each line/pixel, process and output. It's likely a bandwidth limitation for the type/cost of the parts used. In short, I bet Pana used the cheapest parts they could use and still get decent video. Other recorders have used CMOS sensors without much or any rolling shutter(RED seems to have done it since the initial RED one came out with some accounts of rolling shutter which was fixed via firmware according to people who claim to have seen it). Panasonic's new pro camcorder also uses the same sensor as the GH1 and I would bet does not exhibit rolling shutter. The real question is: How much of the processing is software and how much is hardware? If it's software, pana or a firmware writer could essentially kill all the unused processes that were un-needed(such as still picture stuff) and then dedicate all of that CPU time to processing the video.


This is true. Then, the M-X Sensor upgrade corrected it even more. So, I believe you're accurate, it can be both software and hardware yet not the responsibility of the actual sensor.

Barry_Green may know for sure.

Not an AF100 thread, but, this is why I have high hopes for improved rolling shutter, banding/FPN, etc in the AF100. Larger body to fit better processing, higher-priced item to justify cost to manufacture.

[Svart]

I feel almost exactly the opposite. I would think the fixed noise/banding is a software problem and the rolling shutting is a hardware problem.

The fact that we can take a still picture that doesn't exhibit banding issues means to me that it's not the sensor, it's something between the sensor and the SD card. So it's either a hardware bandwidth issue or it's software. Considering Panasonic sells high end video cameras I would think purposely crippling a consumer product is likely.

The rolling shutter problem exists everywhere, even in film to some degree. CMOS sensors are known to exhibit it, and only the highest quality and latest CMOS sensors can limit it to a triviality. It's not software.


FPN IS always present in a CMOS sensor. A clever use of post-processing could essentially get rid of it through different means.

http://en.wikipedia.org/wiki/Fixed-pattern_noise

Taking a picture is not the same as taking video. We would have to know exactly what Pana is doing when they take a picture anyway. They could very well be taking a full (global) shutter shot since they do indeed use a mechanical shutter as part of their process. The resources needed for a single picture is much less than would be needed for 24, 30 or 60 pictures a second, I.E. video.

Anyway, the CMOS sensor is just a matrix of pixels with wires connecting them in a big grid. An external device must scan each row and column in a specific pattern to read each pixel. An analog to digital stage takes the voltage and turns it into a datastream. The CPU then needs to turn all of the datastreams from each of the columns and rows into something it can process. This takes time and resources. Time and resources will be limited by the speed of the CPU, the amount of RAM, bandwidth of the communications, etc but this also includes all of the routines that the system is running.

So yeah you can say that the rolling shutter could be a hardware problem in that Pana likely used the slowest(cheapest) parts it could get away with, but it's more likely that their firmware is simply running too much stuff to spend a lot of time sampling the CMOS sensor. If they were to dedicate more CPU time to the sampling of the CMOS sensor they could likely reduce the rolling shutter. Global shutter would probably need a hardware redesign though.

[billy fattey]

FPN IS always present in a CMOS sensor. A clever use of post-processing could essentially get rid of it through different means.

http://en.wikipedia.org/wiki/Fixed-pattern_noise

Taking a picture is not the same as taking video. We would have to know exactly what Pana is doing when they take a picture anyway. They could very well be taking a full (global) shutter shot since they do indeed use a mechanical shutter as part of their process. The resources needed for a single picture is much less than would be needed for 24, 30 or 60 pictures a second, I.E. video.

Anyway, the CMOS sensor is just a matrix of pixels with wires connecting them in a big grid. An external device must scan each row and column in a specific pattern to read each pixel. An analog to digital stage takes the voltage and turns it into a datastream. The CPU then needs to turn all of the datastreams from each of the columns and rows into something it can process. This takes time and resources. Time and resources will be limited by the speed of the CPU, the amount of RAM, bandwidth of the communications, etc but this also includes all of the routines that the system is running.

So yeah you can say that the rolling shutter could be a hardware problem in that Pana likely used the slowest(cheapest) parts it could get away with, but it's more likely that their firmware is simply running too much stuff to spend a lot of time sampling the CMOS sensor. If they were to dedicate more CPU time to the sampling of the CMOS sensor they could likely reduce the rolling shutter. Global shutter would probably need a hardware redesign though.

I shouldn't have lumped fixed noise in with banding. Other than that, I stand by what I said.

[Vitaliy Kiselev]

Guys, please do no go to banding and rolling shutter things.
We can't fix them, so it is better to discuss in separate thread, especially then level of discussion is not very high.
From software point, preventing banding almost completely is entirely possible, but it requires costly calibration and either ADC adjustments or RAW video correction.
All current DSLRs do not have annough resources to do this.
Yet another approach is to remove line skipping and use full frame scaling using good algorithms.


I'am reading H264 encoder related stuff, especially VBR quality estimators.
We also know some audio related functions so, for AGC we can go from this.
Guys, no one still want to disassembly his GH1 and look for used CODEC?
As it can be best approach.

So, our targets:
Low level sensor setup routines, like scanning speed, fps, etc.
H264 encoder settings allowing to adjust quality estimator, I and P frames size allocation.
Trying to locate AGC (or limiter :-) ) related stuff.
Video output functions research.

[Vitaliy Kiselev]

To Anyone

Can you provide small (5s) clips for all video resolutions and all framerates (plus all settings of compression)?
Including ALL resolutions of MJPEG (including very low ones).
Use original firmware without any patches.
I need this for research purposes.
Camera must be mounted on tripod and show same static scene.

[Car3o]

as long as I've only changed the lang, would my video still be okay for testing? i can post by tonight/tomorrow.

[Vitaliy Kiselev]

as long as I've only changed the lang, would my video still be okay for testing? i can post by tonight/tomorrow.

Car3o, I believe that our new camera will be at your home today or tomorrow?

We also need high speed SD cards for me and our active testers. They'll be around $35 for each one.

Your video will be OK, yes, but we need AVCHD PAL ones also.

[Car3o]

Roger that. Camera is sched for tomorrow. If you want a card for testing please provide a link.

[Vitaliy Kiselev]

Roger that. Camera is sched for tomorrow. If you want a card for testing please provide a link.

http://shop.ebay.com/i.html?_nkw=sandisk+4Gb+Extreme+III&_sacat=0&_dmpt=Digital_Camera_Memory_Cards&_odkw=sandisk+4Gb+Extreme+III&_osacat=0&bkBtn=&_trksid=p3286.m270.l1313

We need 30Mb/s edition. Class 10 is not required, Class 6 is ok, as long as we have right edition.
I believe you can try to find it locally.

[Vitaliy Kiselev]

May be we'll be writing 422 Intra, who knows :-) :2vrolijk_08:

[tyampel]

[quote=tester13;1977757]Guys, please do no go to banding and rolling shutter things.
We can't fix them, so it is better to discuss in separate thread, especially then level of discussion is not very high.
From software point, preventing banding almost completely is entirely possible, but it requires costly calibration and either ADC adjustments or RAW video correction.
All current DSLRs do not have annough resources to do this.
Yet another approach is to remove line skipping and use full frame scaling using good algorithms.


I'am reading H264 encoder related stuff, especially VBR quality estimators.
We also know some audio related functions so, for AGC we can go from this.
Guys, no one still want to disassembly his GH1 and look for used CODEC?
As it can be best approach.



Do you want us to buy the Audio PCB board?
It is part number VEP54012A.
It is just $33 from here.

http://www.partstore.com/SearchResults.aspx?q=Panasonic+Vep54012a&x=46&y=12

Perhaps we have some spare cash in the fund for this.

Good luck.

[Vitaliy Kiselev]

At first, it is better to look at this.
And all we need is look.

Do they supply also Main Board?
As it is better to have one for backup purposes, so I can just replace it in case of real trouble.

[Vitaliy Kiselev]

It is also interesting to look at error codes (remember this in service manual?)
After we changed GOP settings, may be we could find something here.

[tyampel]

At first, it is better to look at this.
And all we need is look.

Do they supply also Main Board?
As it is better to have one for backup purposes, so I can just replace it in case of real trouble.

Yes. There are actually two versions according to the service manual - A and B.
The VEP56084A costs $264.
They don't seem to have the B version, which seems to be for Europe.

One of our testers - NURBS - has the service manual and knows how to read the error codes through diagnostic menus.

Good luck.

[Vitaliy Kiselev]

Thanks, if'll have enough fund we'll buy spare motherboard, as we need USA version.
But not yet.

Service manual links can be found at thread start.

[x_WOrPIG_x]

I have set up ptools for Mac, i successfully updated the patched firmware, but i'm not quite sure what to test. What program do you use to retrieve the clip data? I did a few test last night with MPEG 24p ( i see what you mean by skipped frames now ) and the 23.97p 720, but that was supper choppy. Let me know what I can do to help.

[Vitaliy Kiselev]

It seems that I got how it scan sensor

In quite amazing manner.
Read speed seems to be constant (according to my calculations).
It adjust reading window to accomodate different FPS.
So, in reality 1280x720p60 and 1280x720p50 use absolutely different sensor frames.
And 1280p50 horizontal resolution will be lower.

(SensorFrameW*SensorFrameH*fps)=constant.

[Vitaliy Kiselev]

I have set up ptools for Mac, i successfully updated the patched firmware, but i'm not quite sure what to test. What program do you use to retrieve the clip data? I did a few test last night with MPEG 24p ( i see what you mean by skipped frames now ) and the 23.97p 720, but that was supper choppy. Let me know what I can do to help.


Test if this can be used to make running ptool on Mac :
http://winebottler.kronenberg.org/ (http://winebottler.kronenberg.org/)

Most things are tested now, and we are moving to sensor related things.
VBR quality estimators, audio things, etc.

[Vitaliy Kiselev]

So, it make good answer if we can have 720p24.
It seems that we can't.
And MJPEG 30fps is just some form of 60fps processing.
This is my current understanding of situation.

[Vitaliy Kiselev]

To prove my theory we need to make controlled experiment with shooting
paper with carefully printed fine grid with numbers and camera on tripod.
So we could check if this is true.

This also answers why Canon don't have 720p24.
They can't :-)

[spamrakuen]

So, it make good answer if we can have 720p24.
It seems that we can't.
And MJPEG 30fps is just some form of 60fps processing.
This is my current understanding of situation.

Then... what about MJPEG @ 60fps? :happy:

[Vitaliy Kiselev]

Then... what about MJPEG @ 60fps? :happy:

50Mbps? I do not expect to MJPEG encoder and/or SD controller be able to do such work.
But this knowledge answers many stupid questions from Canon and Panasonic owners of why they do not provide 720p24.

May be after long research we could toggle frame combine feature and get 720p25 (out of 720p50) and 720p30 (out of 720p60).
But I don't see this as very useful.

[Car3o]

To prove my theory we need to make controlled experiment with shooting
paper with carefully printed fine grid with numbers and camera on tripod.
So we could check if this is true.

This also answers why Canon don't have 720p24.
They can't :-)


Make the chart and I'll shoot away ;)

[Vitaliy Kiselev]

Choose large and one you like:

http://images.google.com/images?um=1&hl=ru&tbs=isch%3A1&sa=1&q=graph+paper&aq=f&aqi=&aql=&oq=&gs_rfai=&start=0

We need to test in PAL mode (720p,1080p) and NTSC(720p,1080p) mode.
Camera must be tightly fixed as well as target.

[Vitaliy Kiselev]

And here is generator:

http://incompetech.com/graphpaper/plain/

[Isaac_Brody]

No 24P MJPEG? Ah, breaking my heart. :(

[Vitaliy Kiselev]

No 24P MJPEG? Ah, breaking my heart. :(

May be 25p later :-)
As 30fps=60fps/2, so 25fps will be just 50fps/2.
But we do not know that much to make it today.

And it looks that we have quite a headroom for FullHD mode.
2752750 pixels vs 2073600, but no width reserve :-)

[John Caballero]

But the Nikon D90 has 24p Mjpeg. How come?

[billy fattey]

And it looks that we have quite a headroom for FullHD mode.
2752750 pixels vs 2073600, but no width reserve :-)


Can you explain what that means in more detail?

[DanDOF]

But the Nikon D90 has 24p Mjpeg. How come?

I'm starting to miss my D90. Relatively small 720/24p files in .mov format... etc.

[JerryB]

Tester13, are you certain that this sensor can not record 24fps at 720p or does this theory need more testing or research? This would certainly be a disapointment if so. Great job on everything else so far though! I hope you keep up the dedicated pace. I'm sure you can see now from the quick fundraising that there are a lot of dedicated and hopeful GH1 filmmakers out there (including myself).

[John Caballero]

I really don't understand how the D90 records 720 MJPEG at 24p and the GH1 wouldn't. Especially with the much better sensor in the GH1.

[Ben_B]

If it's possible to push the motion JPEG to 60p instead of 30p that would be amazing because the 60p can be slow motioned or converted to 24p in post (which is actually pretty easy and can look quite good...certainly better than 30p.)

I realize this means twice the throughput which could mean half the bitrate, but if the bitrate can be increased too then 60p could be an ideal solution.

I get this probably sounds silly to those in the know..I am just operating on the assumption that 60p might be slightly more possible than 24p since things seem capable of 30p.

[spamrakuen]

May be 25p later :-)
As 30fps=60fps/2, so 25fps will be just 50fps/2.
But we do not know that much to make it today.


25fps MJPEG would be nice. :cheesy:

[Isaac_Brody]

Yeah I'd be happy with clean mud free 25P. :)

[TrueIndigo]

"No width reserve" -- does that mean we could get 2k CinemaScope (2048 x 858) then? Will it possible to custom set the region of interest on the sensor? And would there be a possibility to use a 1:1 pixel window area (no pixel downsampling), useful for C-mount lenses with the smaller image circle?

[rawfa]

From my understanding higher bitrates would mean better compression and would allow the footage to endure more intense CC without falling to pieces. If this is correct this would be the biggest advantage for me. How high do you thing the bitrate could go in FullHD mode?

[DJDecay]

From my understanding higher bitrates would mean better compression and would allow the footage to endure more intense CC without falling to pieces. If this is correct this would be the biggest advantage for me. How high do you thing the bitrate could go in FullHD mode?

I don't think more than 24Mbps will be possible due to the H.264 codec chip on-camera, I think it's already being fooled into doing High@4.0 instead of Standard@4.0, HiProfile@4.1 would be the next best (24Mbps). 4.2, 4.3 and 5.0 and 5.1 would most likely not be possible at all. (the latter have 4:2:2 support).

My money is on it being identical to the iPhone and other H.264 low power codec processors, El Gato Turbo 264 HD comes to mind.

[Vitaliy Kiselev]

Here is illustration:

http://www.dvxuser6.com/uploaded/60298/1272713592.jpg

Wider sensor window is for 1280x720p60, and other is 1280x720p50.
So, in reality horizontal resolution must be better in p60.
But both do not have 1280 horizontal pixels :-)
Plus, from used numbers I can assume that this is not real raw setup data.
As usable sensor size is about 4130x3000, so some kind of binning mode is activated.

We must shoot graph paper to understand better.

[Vitaliy Kiselev]

It does so at 24.00PsF No? not at the more preferable 23.976 though.
And at some weird resolution that needs de-staircasing to convert No?

D90 have absolutely different hardware.
5D Mark II added 24fps is more interesting case, I believe they also added scanning areas on bottom and may be on top.

[Vitaliy Kiselev]

One more addition, camera actually shoots at 720p59,94 and 1080p23,97
All scan time calculations prove it.

[Path]

tester13, that confirms what the manual states...the sensor output is 1080p24 and 720p60. Both are lazily referencing 23.97 and 59.94.

[Vitaliy Kiselev]

tester13, that confirms what the manual states...the sensor output is 1080p24 and 720p60. Both are lazily referencing 23.97 and 59.94.

Try to read my previous post, please!!!
They are not lazely referencing anything, they actually shoot at 23.97 and 59.94, period.

[Path]

take it easy man, the manual states exactly that in FHD mode "Sensor output is 24 fps". I am saying that Panasonic is being lazy in their manual by printing 24 when, in reality, the fps off the sensor is 23.97.

[Vitaliy Kiselev]

take it easy man, the manual states exactly that in FHD mode "Sensor output is 24 fps". I am saying that Panasonic is being lazy in their manual by printing 24 when, in reality, the fps off the sensor is 23.97.

I clearly said that actual fps are.
Manual is written by technical writers who know nothing about camera internal functions.
And sensor have no such thing as "fps" . At all. It have sensor read speed.
At least try to read thread you are posting to.

[hunter richards]

24fps= 23.976 in nearly every ntsc video camera, this isnt something new to the gh1. Most people just say 24fps but they really mean 23.976/23.98. Its a trickle down from when b+w tv became color.

From wikipedia:

"In January 1950, the Committee was reconstituted to standardize color television (http://en.wikipedia.org/wiki/Color_television). In December 1953, it unanimously approved what is now called the NTSC color television standard (later defined as RS-170a). The "compatible color" standard retained full backward compatibility with existing black-and-white television sets. Color information was added to the black-and-white image by adding a color subcarrier (http://en.wikipedia.org/wiki/Subcarrier) of 4.5 × 455/572 MHz (approximately 3.58 MHz) to the video signal. To reduce the visibility of interference between the chrominance signal and FM sound carrier required a slight reduction of the frame rate (http://en.wikipedia.org/wiki/Frame_rate) from 30 frames per second to approximately 29.97 frames per second..."

[billy fattey]

Tester13, you guys are saying the same thing. We have a language barrier. I think we're all trying to exhibit some patience, and I hope you will do the same.

By the way, is somebody getting that graph paper footage you requested? If not, I can do it later today. I would have it ready about 6 hours from now. Just let me know.

[Vitaliy Kiselev]

Tester13, you guys are saying the same thing. We have a language barrier. I think we're all trying to exhibit some patience, and I hope you will do the same.

By the way, is somebody getting that graph paper footage you requested? If not, I can do it later today. I would have it ready about 6 hours from now. Just let me know.

We are no saying same thing.
As from sensor scan calculations you could clearly get actual fps.
So, no such term as "fps" exists for sensor, only read speed and used sensor window.
By enlarging window you are dropping "fps".
I do not know if it is possible to change read speed at all, but in video shooting they use constant speed.

Graph paper is still actual, make one using online generator link I provided, print it out, and shoot in tightly controlled manner (as you'll need to change firmware with NTSC->PAL or PAL->NTSC patch, but camera must be fixed and it is not easy). In extreme case, just align top left corner to be on big grid point.

[Rabot]

I'm doing this right now.

The only patches I applied for both was Version Change / 30min removal / (pal/ntsc patch)
Shooted 1080 on a random grid, tripod, 35mm focal lenght extly the same setting.

did you need the original Private sctructure or can I make a vimeo version or something?


EDIT:

The patch worked since they really asked me to format the card when I wanted to record 25fps footage (while I had 24p footy on card)

[Vitaliy Kiselev]

You can provide just individual frames here.

And I am more interested in 720p.
But 1080p is also good.
Grid must be fine enought.

[Rabot]

Hum you'll tell me once I show them witch is in about 5 mins.
And just by checking them i can tell you that they're pretty much the same.

[Rabot]

http://old.akwd.com/video/TESTGH1PALNTSC.mp4

original 1080p file. I suggest downlaoding it, it's quite small.
I didn't remove pulldown, since FCP handle it correctly in a 29.97 timelime, the pal version is wierd, you'll see what i mean..

[jeracravo]

wierd it is...

[Vitaliy Kiselev]

We don't need video, only individual frames. Finer grid.
We need four frames shot for 1080p24, 1080p25 (just extract them)
and 720p50 and 720p60. All must be shot from exact same position.
As for 1080p24 and 1080p25 having same FOV it is good as we have vertical cropping on both and horizontal on 1080p25 (so start sensor position is the same).
We also need to compare 1080p24 and 720p60 FOV and FOV of both 720p rates.

[Rabot]

doing this, but so fa rno diffrences..

[Rabot]

http://old.akwd.com/video/7201080palntsc.zip

Photo sucks but I hope it's enough.

[Vitaliy Kiselev]

OK thanks.
May be it is possible to make them JPG (they are some type of QuickType format?) and upsize 720 ones (to 1920x1080)?

[Vitaliy Kiselev]

And one more thing.
It'll be very good to have TZ5 body for experiments (as it is very cheap at arount 90-130 at ebay). MJPEG settings and MJPEG engine overall seems to be similar.

[Rabot]

http://old.akwd.com/video/sizejpeg.zip

here you go

[Vitaliy Kiselev]

Interesting. Thanks.
Later we must shoot more precise versions with more markings and keeping center stready.
So, in 1080p25 we have slightly different RAW windows and cropping later. It is OK.
But different RAW windows for 720p must have effect.
Or they just sacrifice resolution in 720p mode.
Next time we must shoot good resolution test charts also.
And compare results.

[Car3o]

I believe Barry has reso charts up on the web already. don't know what setting he shot it, if it was 720 or 1080.

regarding chart. if you point to exactly to the one you want, i will shoot (NTSC) for you. That way I'm not printing 5 different versions.

Camera is skedded to arrive on Tuesday. I was working today and missed delivery. Apparently fed-ex is closed till Tuesday. Lame.....

[Ben_B]

I believe Barry has reso charts up on the web already. don't know what setting he shot it, if it was 720 or 1080.

regarding chart. if you point to exactly to the one you want, i will shoot (NTSC) for you. That way I'm not printing 5 different versions.

Camera is skedded to arrive on Tuesday. I was working today and missed delivery. Apparently fed-ex is closed till Tuesday. Lame.....

He did both but just NTSC. I haven't been following here too closely but wasn't there a thought that the PAL and NTSC versions resolved 720p differently because they used different parts of the sensor?

[tyampel]

I believe Barry has reso charts up on the web already. don't know what setting he shot it, if it was 720 or 1080.

regarding chart. if you point to exactly to the one you want, i will shoot (NTSC) for you. That way I'm not printing 5 different versions.

Camera is skedded to arrive on Tuesday. I was working today and missed delivery. Apparently fed-ex is closed till Tuesday. Lame.....

Do you have enough funds to buy a TZ5?

[alignment1]

Camera is skedded to arrive on Tuesday. I was working today and missed delivery. Apparently fed-ex is closed till Tuesday. Lame.....


MISSED THE DELIVERY!!!!!:crybaby::crybaby:

C'mon Car3o you're killin us--work?? screw work, we've got serious code to crack and you're worried about paying bills??!!?

Sheesh- we're all handing you money!! just take a cut, pay your light bill and GET THE CAMERA TO TESTER13!!!!!!:2vrolijk_08::2vrolijk_08::2vrolijk_ 08:

Just playin-- tuesday it is:Drogar-Happy(DBG):

Z

[Ian-T]

lol

[Car3o]

haha, yeah yeah......I'm off Tues so all is good there. Beside, we're still waiting on that adapter to be delivered, so no biggie. Which is ordered and on the way. As far as funds go. There is still $60.00 in the piggie and no sd card has been ordered yet. I'm looking locally first and then going to discuss whether or not to buy online. i'll keep everyone posted on what's going on.

[Camera Expert]

I think to play it safe with all the testings, it's probably best to get a Sandisk Extreme III, class 6 card. B&H has an 8 gig for $45 and although it's slightly cheaper at Amazon, a lot of those cards are coming from 3rd party vendors unless you know which one is reliable.

[seanmcleod]

Hey Guys, I'm sorry I'm so late to party, kept seeing the post title but didn't ever check into it, now I see so many pages and feel like something awesome is going on... anyone mind giving me a nutshell answer of what is exactly going on and where it stands?

I gather you're trying to develop an independent firmware update to include some of our wishlist wants for this camera?

[Camera Expert]

Another camera of interest is the LX3 since it's MJPEG 720p runs at 24 frames per second although it's a bit more expansive that the TZ5.

I'll have to contribute a little something myself but I'll probably take a few days or more until I get some spare change.

[Car3o]

Hey Guys, I'm sorry I'm so late to party, kept seeing the post title but didn't ever check into it, now I see so many pages and feel like something awesome is going on... anyone mind giving me a nutshell answer of what is exactly going on and where it stands?

I gather you're trying to develop an independent firmware update to include some of our wishlist wants for this camera?

first page. everything there is to know is just about covered there.

[seanmcleod]

first page. everything there is to know is just about covered there.

Hey man yeah I read that first page already, and still kind of confused... these are things that you've accomplished? 24fps Mpeg 720 recording etc... I guess Im just kind of looking for what the end goal you guys are hoping for... is that everything listed in the AVailable patches that you hope to accomplish or have accomplished? Are there other things you're hoping to achieve?

[Isaac_Brody]

Hey man yeah I read that first page already, and still kind of confused... these are things that you've accomplished? 24fps Mpeg 720 recording etc... I guess Im just kind of looking for what the end goal you guys are hoping for... is that everything listed in the AVailable patches that you hope to accomplish or have accomplished? Are there other things you're hoping to achieve?

Improved compression to reduce or eliminate mud from fast pans in 1080 mode.

Pal/NTSC switchable for region free shooting.

MJPEG 24P for totally mud free shooting and native 24P shooting.

1080 24P avchd to avoid pulldown removal in post.

AGC disable.

Basically the goal is to make this camera function better than it does since Panasonic isn't going to release a firmware that accomplishes this. Even if it's just a matter of enabling native 24P in 1080 to avoid the annoying hassle of pulldown that's still a huge improvement.

[Chris Light]

Isaac...i read your post, then i read the quote in your sig.

WWOD?

[Vitaliy Kiselev]

Hey man yeah I read that first page already, and still kind of confused... these are things that you've accomplished? 24fps Mpeg 720 recording etc... I guess Im just kind of looking for what the end goal you guys are hoping for... is that everything listed in the AVailable patches that you hope to accomplish or have accomplished? Are there other things you're hoping to achieve?

Try to read all below "Major issues for today are" on first page.

[Vitaliy Kiselev]

Improved compression to reduce or eliminate mud from fast pans in 1080 mode.

Pal/NTSC switchable for region free shooting.

MJPEG 24P for totally mud free shooting and native 24P shooting.

1080 24P avchd to avoid pulldown removal in post.

AGC disable.

Basically the goal is to make this camera function better than it does since Panasonic isn't going to release a firmware that accomplishes this. Even if it's just a matter of enabling native 24P in 1080 to avoid the annoying hassle of pulldown that's still a huge improvement.

I also siggest you to read first page.
As number 2 and 4 are done.

[Vitaliy Kiselev]

haha, yeah yeah......I'm off Tues so all is good there. Beside, we're still waiting on that adapter to be delivered, so no biggie. Which is ordered and on the way. As far as funds go. There is still $60.00 in the piggie and no sd card has been ordered yet. I'm looking locally first and then going to discuss whether or not to buy online.

This looks confusing. Are we down to $60 on funds?

[Essami]

Try to read all below "Major issues for today are" on first page.

I think it might be a good idea to start a thread discussing the user end of the firmware patch and keep this thread reserved for the work and testing of the patch?

[Vitaliy Kiselev]

As for sensor related things.
We are yet to understand this, as constant WxHxfps suggest constant read rate and it can not be coincidence. But we can have some internal binning algorithm and reading of whole line scaled to suggested resolution.
So, more shoots are required, including 16:9 photos and video test charts.


P.S. LX is costly and it is not an option due to encrypted firmware.

[Vitaliy Kiselev]

I think it might be a good idea to start a thread discussing the user end of the firmware patch and keep this thread reserved for the work and testing of the patch?

I don't like this idea.
What exactly do you expect from "user oriented" thread?
With time we'll move to dedicated forum and such threads will be ok, but today they are just nor very useful.

[rawfa]

I don't like this idea.
What exactly do you expect from "user oriented" thread?
With time we'll move to dedicated forum and such threads will be ok, but today they are just nor very useful.

I understand what he is saying. Even though this thread shows enormous potential it can be very technical for some people at times. But at the moment I think Tester13. Lets just let the man do his work and discuss the benefits latter.

[schakal.]

Hi, short question...

I use a PAL version and have fw update 1.21 on it, which version increment should i type in in your ptool to make an new fw update?

1.21.1 or something?

Another question, what does the 24/25p native patch exactly for my pal camera?

i have read all sites in this thread because it´s very exciting what happens here, but my english is not so good to understand all youve written here, maybe anyone could write the first page with all infos also in an german version?

[Vitaliy Kiselev]

Google translation can be used to make english to deutch translation :-)
Use default increment (1)
Native is native, 25p video instead of 50i.

[schakal.]

thanks tester13 :) sry, i know, you are doing a very very good job, but sometimes also google translation could not do it´s best ;)

[Essami]

I don't like this idea.
What exactly do you expect from "user oriented" thread?

Just the fact that people want to speculate about many different aspects of the GH1 hack but not all is pertinent to the research done here. You seem to indirectly suggest it yourself, this was my motivation for my suggestion.


Guys, please do no go to banding and rolling shutter things.
We can't fix them, so it is better to discuss in separate thread,

[Vitaliy Kiselev]

I don't know how banding or rolling shutter are related to hack.
Expecially considering usual level of such discussions.

[schakal.]

So i have tried a patch on my pal version..

Version change - 1
30min limit removal
native 24p/25p


recording without problems

but

connect the camera to usb, camera freeze... no connection.. battery remove helps...
turn on camera an press the play button, it freezes the camera too.... also battery remove helps...


has anyone the same problem?

edit:

i made a second fw only with 30min lim. removing, all is working fine, so there is a problem with 25p native?

[Vitaliy Kiselev]

connect the camera to usb, camera freeze... no connection.. battery remove helps...
turn on camera an press the play button, it freezes the camera too.... also battery remove helps...

i made a second fw only with 30min lim. removing, all is working fine, so there is a problem with 25p native?

Try to delete all native footage from SD card.
Camera can't play native 25p videos, so this can be only reason for freezing.
Native patch changes only encoding settings and can't affect any USB connection parts.
Use SD card reader next time.

[Essami]

I don't know how banding or rolling shutter are related to hack.
Expecially considering usual level of such discussions.

Yes, exactly.

[schakal.]

okay, so, the camera can work with native 24p but not with native 25p ?

So do i have to change the camera from pal to ntsc to work with native files on camera?

[Vitaliy Kiselev]

Yes, exactly.

And?
Go and discuss this thing in other thread, as I am not interested in such low level discussion.

[Vitaliy Kiselev]

okay, so, the camera can work with native 24p but not with native 25p ?

So do i have to change the camera from pal to ntsc to work with native files on camera?


Can you read my posts carefully?
24p and 25p makes no difference.
Camera can't play 24p/25p video files, hence it freeze.

[schakal.]

i read it carefully but you have written


Try to delete all native footage from SD card.
Camera can't play native 25p videos, so this can be only reason for freezing.


So i was thinking 24p will work..

sry......

[Essami]

And?
Go and discuss this thing in other thread, as I am not interested in such low level discussion.

hehehe, man, thats exactly what Im saying :D Obviously some difficulties in understanding, Ill leave it here, sorry for the bother!

[Isaac_Brody]

I also siggest you to read first page.
As number 2 and 4 are done.


Ok, so NTSC 24P native works properly then? Can someone upload a clip so I can get my pixelpeep on? I want to test the post workflow and see how much easier it is.

Thanks. Somehow that slipped through without my noticing.

[Solomon Chase]

Ok, so NTSC 24P native works properly then? Can someone upload a clip so I can get my pixelpeep on? I want to test the post workflow and see how much easier it is.

Thanks. Somehow that slipped through without my noticing.

seconded.

[Car3o]

This looks confusing. Are we down to $60 on funds?

Paypal balance $758.56

Gh1.............$659.99
Gh1/Pentax Mount..............$34.68

Total.........$63.89


so yeah, we're down to $60.00 and change.
as requested. i'll post the paypal screen grabs with everyone's info blocked out.
just to have it public and out there

[tyampel]

Paypal balance $758.56

Gh1.............$659.99
Gh1/Pentax Mount..............$34.68

Total.........$63.89


so yeah, we're down to $60.00 and change.
as requested. i'll post the paypal screen grabs with everyone's info blocked out.
just to have it public and out there

I vote to go ahead and buy the Sandisk Extreme 8Gb SD card.

[Vitaliy Kiselev]

OK, and you'll just ask US post to send camera for free. Fine :-)

As I just figured out we had major error in donation button code.
This resulted in no more money donated to project.

[monton1999]

OK, and you'll just ask US post to send camera for free. Fine :-)

As I just figured out we had major error in donation button code.
This resulted in no more money donated to project.

I have donated $40 just minutes ago and apparently everything went fine with the transaction.

[Vitaliy Kiselev]

I changed button code text back to normal about 20 minutes ago, so all can be ok already.

[Car3o]

Yeah, good catch on tester13's end. i guess when i created the button, or since when people started using the button to donate, instead of link, i put richb.hibner@gmail.comm instead of richb.hibner@gmail.com....so if you've donated after April 27th, 2010, you're donation most likely didn't come through. i'll see if paypal has that balance in their system somehow. will advise. also, if you did donate check to see if the funds even went through. sorry guys, damn typo on these mac keyboards at work screwed everything up!

[tyampel]

Yeah, good catch on tester13's end. i guess when i created the button, or since when people started using the button to donate, instead of link, i put richb.hibner@gmail.comm instead of richb.hibner@gmail.com....so if you've donated after April 27th, 2010, you're donation most likely didn't come through. i'll see if paypal has that balance in their system somehow. will advise. also, if you did donate check to see if the funds even went through. sorry guys, damn typo on these mac keyboards at work screwed everything up!

Do you see my donations on April 22 and April 25?

Thanks.

[Car3o]

update: no email is registered under richb.hibner@gmail.comm, so all funds sent, didn't get sent anywhere. that's good. however, if you donated after the 27th, the funds didn't get sent to tester13's hard effort.

[Vitaliy Kiselev]

Yeah, good catch on tester13's end. i guess when i created the button, or since when people started using the button to donate, instead of link, i put richb.hibner@gmail.comm instead of richb.hibner@gmail.com....so if you've donated after April 27th, 2010, you're donation most likely didn't come through. i'll see if paypal has that balance in their system somehow. will advise. also, if you did donate check to see if the funds even went through. sorry guys, damn typo on these mac keyboards at work screwed everything up!

Reason of error really was not your button only, but domain forwarding limit, then I moved to hosted page and just pasted button text :-)

Anyway, all is good now.
I also hope that we could get TZ5 body, especially considering their low price, as MJPEG compression experiments are much better to do on this one and not GH1.
I am now digging compression settings and behaviours with some progress already. But without actual body tests are not very useful, as I need quite a number of them in very controller fashion and very systematic approach.

[Car3o]

Do you see my donations on April 22 and April 25?

Thanks.

yes

[tyampel]

[quote=tester13;1979192]OK, and you'll just ask US post to send camera for free. Fine :-)

I did not expect the postal service to deliver the package for free.
I was hoping to cover shipping cost with another donation if not enough money is left in the fund.

[monton1999]

I changed button code text back to normal about 20 minutes ago, so all can be ok already.

OK. I cancelled the previous transaction and donated again to the correct account.

[Car3o]

Check your names on the list. I still haven't received any new donations. Tester13, please up old link till we get the button fixed.

http://docs.google.com/leaf?id=0B5hXpZ5tKLfTZWI1NjM0M2UtYTZlNi00MWE0LThiM jctYTVmYWY5MDM0OTcx&hl=en


http://docs.google.com/leaf?id=0B5hXpZ5tKLfTMjQ1YWFiZDYtNzM5Ny00M2M3LWIxN mQtZmZmZTE3ZDdiNmFh&hl=en

http://docs.google.com/leaf?id=0B5hXpZ5tKLfTZjZjMzExNjMtZGUwNi00M2NhLWI0O WUtOTUxODlmMGI1Mjc2&hl=en

[Vitaliy Kiselev]

I already fixed button code on page.
All must work now.
Any people donating can check their mail for PayPal receipts.

I also believe it is better to hide all names, and show sum instead. Or delete pictures at all.
As we have some people who wanted to remain anonymous.

[Car3o]

well if the button is working i'm still not receiving :)

if i block all names, leave amount open, how would someone know they sent??

[Vitaliy Kiselev]

well if the button is working i'm still not receiving :)

if i block all names, leave amount open, how would someone know they sent??

I hope they can use their Paypal account history page for sent money :-)

P.S. I also changed all account letters to lower case recently.

[monton1999]

well if the button is working i'm still not receiving :)

if i block all names, leave amount open, how would someone know they sent??

If you examine the transaction on your Paypal account you should see something like this:

http://j.imagehost.org/0255/PaymentOK.jpg

Note that user name is Rich Hibner and email address RichB.Hibner@gmail.com ending with ".com" and not ".comm". Also note that transaction status (last line) should be "Completed". If transaction status shows "Unclaimed" instead of "Completed" you should cancel the the transaction and try again.

As for why you still don't see my donation on your Paypal account balance... I have no idea.

[e-steve]

I already fixed button code on page.
All must work now.
Any people donating can check their mail for PayPal receipts.


Are we saying that the PayPal button code is on-par in complexity as Panasonic GH1 firmware?

:nads:

Kidding... keep up the good work guys.

I'll send a donation in today.:thumbsup:

[Car3o]

Donations are back to working, just received.

[TurboTwo]

So how much much has actually been donated but hasn`t shown up or been credited to the fund?

[weismangroup]

I am about to apply NTSC>PAL Conversion. This is my first firmware change, so just want to make sure I've got this right. I've got the following options checked:

Version Change (increment 1)
30 Minute Removal
NTSC>PAL Conversion

Everything else is unchecked. I saw in a set of previous posts around page 70 that one user had their GOP set wrong (15 instead of 13) - Do I need to make any other changes (like to GOP setting), or will the proper GOP settings be set by default with the above 3 boxes ticked?

Thanks,

mw

[Car3o]

PayPal balance: $121.67 USD

This is with the 2 newest donations. If you feel like you have donated check the 3 files I posted and it will show you if you have or not.

[Car3o]

I have noticed that when I post the amount, the forum gets buggy.

[TurboTwo]

I have noticed that when I post the amount, the forum gets buggy.


What does that mean?! :-)

[tyampel]

What does that mean?! :-)
There seems to be a bug in the script for this forum.
This bug shows up when the current page is full.
The page number seems to be getting incremented but the last page cannot be accessed.

Referred in some posts as end of page issue.

[alignment1]

broken donate link?! yikes! Old transaction cancelled-- money re-sent, ( I hope ):lipsrseal
Z

[Cavemandude]

That forum "next page" bug has been around for ages on this website and it happens in other forums here not just GH1 forum so you don't need to keep bringing it up every time there is a new page here.

Randy

[Vitaliy Kiselev]

That forum "next page" bug has been around for ages on this website and it happens in other forums here not just GH1 forum so you don't need to keep bringing it up every time there is a new page here.

I believe it is just some fun as bug is quite unusual :-)
And forum owners could report this bug to developers long time ago to have it fixed.

[Car3o]

That forum "next page" bug has been around for ages on this website and it happens in other forums here not just GH1 forum so you don't need to keep bringing it up every time there is a new page here.

Randy


It's so easy even a cavemandude can do it

[AdrianF]

I cancelled my donation, so is the donate button now working?

[Car3o]

So far I've received 3 donations since the error, so that's a yes.

[aczelkri]

Hi,

I hope it is not a double post - is there a chance to change the audio sampling rate of MJPEG movies from 16k to 48k? The codec is not that important but 16k is audibly worse than 48k.

Br,
Kristof

[jeracravo]

it's showing 94 pages on this thread, but it only goes 'till 93rd...
that's a bug

[Vitaliy Kiselev]

Hi,
I hope it is not a double post - is there a chance to change the audio sampling rate of MJPEG movies from 16k to 48k? The codec is not that important but 16k is audibly worse than 48k.


I'll try to do it in near releases.
Panasonic is doing this intentionally, as they save bitrate for video and 48k PCM audio consume much more space than compressed 192kbps in AVCHD.

[Vitaliy Kiselev]

I am still digging in compression things, so I need more tests.

1) VA Bitrate Adjustment - do not touch anything else and lower it initially to 12000000, then to 9000000, and then to 6000000. Shoot short clips in 1080p.
2) VA Bitrate Adjustment 2 - do not touch anything else (including 1) ) and lower it initially to 14000000, then to 9000000, and then to 6000000. Shoot short clips in each mode, including MJPEG 1280p30.

Note. You must shoot something with many details, like grass.

[ProjX v2.0]

Not sure if this was covered in the last 94 pages but might it be possible to increase the frame rate to over 60fps, maybe even up to 120fps?

[Vitaliy Kiselev]

Not sure if this was covered in the last 94 pages but might it be possible to increase the frame rate to over 60fps, maybe even up to 120fps?

May be it is possible by lowering resolution and FOV. But required major firmware rewrite. As many privately developed software products it is not universal, so it have tight connection to available resolutions, fps rates, etc.
So, it is better to just buy compact or specialized camera if you need this.

[hunter richards]

Just thought of this last night;

Would it be possible to tap into the 1920x1080 jpeg "engine" or whatever is allready used for 16:9 small jpegs and send the 24fps image stream to be recorded in a series of jpegs?

Im not talking about using the downscaler (sensor probably cant come close to handling that at 24fps), but just sending the frames over to what already looks like some sort of built in hardware compressor with it being 1920x1080 and all.

With a very fast SD card, you should be able to handle 24frames persecond of 1920x1080 jpegs (at least for a period of time) and the quality would be amazing.

[g.l]

MJPEG is actually that, a series of independent jpegs stored in a movie file wrapper. So the best MJPEG bitrate is the best you can get for real-time JPEG video compression.

[Vitaliy Kiselev]

With a very fast SD card, you should be able to handle 24frames per second of 1920x1080 jpegs (at least for a period of time) and the quality would be amazing.

I do not believe that JPEG compressor can do 24 FullHD frames, and 50MBits/s require also much larger buffers . SD controller write speed limit can be about 30-40Mbits/s, we don't know yet.
Firmware is not build like setting resolution and next universal engine obtain RAW frame, process it, move to designated compressor, etc.
It is highly specialized, uses codes for each resolution and fps settings, and this things are used in many many places, so most things are hardcoded to predefinded resolutions.
Sensor setup also seems to be not universal.

[rawfa]

FullHD 24/25p at 40mbps would be incredible! If you can make it work I'm sure dvxuser.com will build you a statue :)

[Vitaliy Kiselev]

FullHD 24/25p at 40mbps would be incredible! If you can make it work I'm sure dvxuser.com will build you a statue :)

Don't be sure about it. :-)
We can be just few months ahead of GH2 with normal codec implementation.
As that we have in GH1 is problems of implementation, as AVCHD with I and P frames and no CABAC can be worse than MPEG2.
We need normal encoder chip from TM700 and I believe it is main difference in new 4/3" Panasonic camcorder.
And yes, with current encoder we need about 40-50Mbit, as Canon one is similar and show such bitrates.
Problem is that hardware can be not capable of it, or memory mapping can be restriction (buffer sizes can be limited).

[Vitaliy Kiselev]

Moving to low level stream analizers I see that something is wrong with Panasonic guys.
Really wrong, and they must fix it quick.
It is definitely that encoder settings are wrong (also VBR quality estimators for P frames are wrongly setup).
Bitrate selected is too low.
Wrappers must be removed and player adjusted to play native files (today it can't do it).
I don't talk here about idiotic ZS3/7 (compacts) frame duping in 720p60.

Plus low GOP results (plus bitrate patches) in weird bitrate allocation, see below:
http://www.dvxuser6.com/uploaded/60298/1272908357.jpg

Big surges in some P frames :-)

Please provide me links for GH1 original (with wrapper) files and similar 7D/550D files.
It is interesting to compare at low level.

[hunter richards]

Tester- so you are saying panasonic made a mistake in the settings in their firmware for avc encoding? i.e. That its possible to "fix" it and still shoot 17mbps with out too much mud?

[Vitaliy Kiselev]

I believe that at least 24-28Mbit is possible for this hardware.
So, quality estimators must be changed in next firmware, bitrate settings must be changed also.
Low level adjustment can be made.
Compression artifacts won't be gone completely (as you can see it on normal 24Mbit comcorders streams).
But will be less.
Error can even be in adapting 720p encoder (as chip used was originally indended for ZS3, as I think) to 1080p.
Problem can be as low as motion vector estimation or some related math (as lower fps are much more affected).
They clearly need few men and good systematic tests to adjust it.
Hope someone related to GH1 project read this. :-)

[tyampel]

Moving to low level stream analizers I see that something is wrong with Panasonic guys.
Really wrong, and they must fix it quick.
It is definitely that encoder settings are wrong (also VBR quality estimators for P frames are wrongly setup).
Bitrate selected is too low.
Wrappers must be removed and player adjusted to play native files (today it can't do it).
I don't talk here about idiotic ZS3/7 (compacts) frame duping in 720p60.

Plus low GOP results (plus bitrate patches) in weird bitrate allocation, see below:
http://www.dvxuser6.com/uploaded/60298/1272908357.jpg

Big surges in some P frames :-)

Please provide me links for GH1 original (with wrapper) files and similar 7D/550D files.
It is interesting to compare at low level.

Found some original MTS files at DPReview. I hope they are useful to you.

http://www.dpreview.com/news/0905/09051503panasonicgh1preview.asp

Good luck.

[Vitaliy Kiselev]

Thanks.

Look at this picture:
http://www.dvxuser6.com/uploaded/60298/1272911954.jpg

As you can see spike in P frame is actually just wasted space.
Useful info (motion vectors+transformation) is still the same.
So, we hit here wall of quality estimators.

[Vitaliy Kiselev]

Just looked frame by frame on DPReview examples (in 720p) and they are not too bad.
So, best way is to try in 1080p for now longer GOP, may be even about 60.
We could see how P frames size is adjusting to available bitrate.
Plus try to up bitrate a little (up to 20-21 must be safe).
At least for now I believe that we have some quality estimaton wall.
First thing Panasonic must do is to up bitrate to about 28-30Mbps and make it almost constant by adjusting VBR estimators and making bitrate setting to be limit, not estimator values.

But without actual testers it is waste of time.

[farfromsubtle]

Hey there tester13,

I wonder, since they are extremely similar camera's would it be very difficult to modify the GF1 as well? Specifically, manual control over aperture and shutter speed while video recording. If these modifications were made, the GF1 would basically be a cheaper, smaller, GH1. That would be pretty awesome!

[JerryB]

Tester13 & Car3o,

Have you received your GH1 and adapter yet, if not, when is it scheduled to get to you. I imagine this will speed up your testing/progress quite a bit..

[Car3o]

The camera, according to Fed-Ex, is skedded to arrive tomorrow. As it's previous delivery I was at work and missed it. The adapter is being sent USPS so I can' t track it. However it was sent 4/30, should be here soon. I'll keep you guys updated. Once the adapter comes I'll mail it all out together.

[kaplanfx]

First thing Panasonic must do is to up bitrate to about 28-30Mbps and make it almost constant by adjusting VBR estimators and making bitrate setting to be limit, not estimator values.

I don't believe this will be possible. The AVCHD spec maxes out at "up to 24Mbps VBR". I believe there are only two rated speeds for 1080, "up to 17Mbps" and "up to 24Mbps". Panasonic would have to choose one of those to remain within spec.

-kap

[tyampel]

The camera, according to Fed-Ex, is skedded to arrive tomorrow. As it's previous delivery I was at work and missed it. The adapter is being sent USPS so I can' t track it. However it was sent 4/30, should be here soon. I'll keep you guys updated. Once the adapter comes I'll mail it all out together.
Please order a fast SD card. Does not have to be big, 4GB should suffice.

[Chibs]

I'm just posting to say I won't be able to test any firmware for the following week as I'm on a holiday. Keep up the good work! :)

[Car3o]

Pentax Mount came in today.

[Path]

Plus low GOP results (plus bitrate patches) in weird bitrate allocation, see below:
http://www.dvxuser6.com/uploaded/60298/1272908357.jpg

Big surges in some P frames :-)


tester, this is interesting indeed. I assume this is from an MTS file, is it 720 or 1080 and at what frame rate? Also, are the values on the y-axis bytes?

If this output is from modified firmware with the bit rate increased, it seems the increased rate is obviously not spread equally among p-frames. Can you comment on whether this may be related to not having all the bit rate routines discovered in the firmware?

and, one final question...how much of the firmware have you dissassembled? I have spent some time dissassembling the 1.22 firmware in IDA Free 4.9...are you sharing your IDA files?

thx

[Vitaliy Kiselev]

tester, this is interesting indeed. I assume this is from an MTS file, is it 720 or 1080 and at what frame rate? Also, are the values on the y-axis bytes?

It is 1080p with changed GAP and bitrate.
And, of course, it is frame size on y axis.


If this output is from modified firmware with the bit rate increased, it seems the increased rate is obviously not spread equally among p-frames. Can you comment on whether this may be related to not having all the bit rate routines discovered in the firmware?

Please try to read my recent posts carefully. They have all the answers.



and, one final question...how much of the firmware have you dissassembled? I have spent some time dissassembling the 1.22 firmware in IDA Free 4.9...are you sharing your IDA files?

1) It is not about disassembling, this project is about reversing. Disassembling is just obtaining assembler instructions and have no meaning without actual reversing.
2) About sharing IDA files. First, it is illegal, I could provide here many points, but won't do it. Second, it requires expirienced reverser on other side, who is owner of GH1.

[Path]

P.P.S. If someone have GH1 (is the owner) and want to look at IDA code you can follow blog posts download all tools, install IDA Free and request IDB file. It is completely legal and it is not that hard. MN103 code is easear to understand even then x86, not to say RISC assemblers. I won't provide file for people who don't own GH1.

As a GH1 owner, I was simply responding to your offer here. I have the tools and IDA 4.9, just need the IDB so my ida project would look like yours. Are you saying it is illegal to share the IDB file?

[Vitaliy Kiselev]

As a GH1 owner, I was simply responding to your offer here. I have the tools and IDA 4.9, just need the IDB so my ida project would look like yours. Are you saying it is illegal to share the IDB file?

In fact, yes, I was wrong initially. I had few conversatitons about this since then. IDB file contains all original firmware code, so it is illegal to distribute, and if you are in US or UK it can be still illegal if we manage to remove all code somehow.
I want to keep this project as legally clean as possible. I am inverstigating now how we can remove Panasonic code from IDB files (and add it later on your side).

[Vitaliy Kiselev]

Testers wanted!!!


Shoot this resolution chart using
1) 16:9 photo mode (at maximum resolution)
2) 16:9 photo mode (at 1920x1080)
3) Movie mode (1080p24 or p25)
4) Movie mode (720p60 or p50)
5) Keep whole 16:9 chart tightly within frame.
6) Level chart and your camera.
7) Use tripod.

www.pentax-hack.info/firmware/pan/ISO_12233-reschart.pdf (http://www.pentax-hack.info/firmware/pan/ISO_12233-reschart.pdf)

[bumkicho]

I wish I was home testing this instead of at work. Tester13, just to clarify it for whoever is going to do the test shoot, should it be done after patch or before? Perhaps it doesn't matter.

BTW, just donated a few dollars to the project. Can you or Car3o confirm it please? Just want to make sure the money went to the right account.

[tyampel]

I wish I was home testing this instead of at work. Tester13, just to clarify it for whoever is going to do the test shoot, should it be done after patch or before? Perhaps it doesn't matter.

BTW, just donated a few dollars to the project. Can you or Car3o confirm it please? Just want to make sure the money went to the right account.

My understanding is he needs this without the patch to study the camera's behavior.
Only Car30 can confirm receipt of funds.
Thanks for your contribution.

[Vitaliy Kiselev]

My understanding is he needs this without the patch to study the camera's behavior.
Only Car30 can confirm receipt of funds.
Thanks for your contribution.

Thanks for donation.
You must shoot without any patch, or with PAL->NTSC(or reverse) patches if you also want to shoot at different framerates.

[billy fattey]

How do you want the footage? original .mts or quicktime?

[bumkicho]

Thanks for donation.
You must shoot without any patch, or with PAL->NTSC(or reverse) patches if you also want to shoot at different framerates.

Tester13, this is very exciting and I thank you for all the work you are putting into this project!!

[Vitaliy Kiselev]

How do you want the footage? original .mts or quicktime?

Only original files.

[alignment1]

:2vrolijk_08:don't miss the fed-ex guy Car3o!!:Drogar-BigGrin(DBG)

[billy fattey]

how big does the chart need to be. Can I just print it on a 8.5X11 piece of paper or does it need to be bigger?

[Car3o]

haha, yeah. i made sure to sign the piece of paper and stick it to the door. Good news.

1) Gh1.......delivered
2) Gh1 mount..........delivered
3) Gh1 3rd party battery...........delivered

Everything is here minus the card. Tester13, do you want to wait for card to be ordered or send what I have now?


PayPal balance: $149.80 USD

[Car3o]

I wish I was home testing this instead of at work. Tester13, just to clarify it for whoever is going to do the test shoot, should it be done after patch or before? Perhaps it doesn't matter.

BTW, just donated a few dollars to the project. Can you or Car3o confirm it please? Just want to make sure the money went to the right account.

Donation received.

[Vitaliy Kiselev]

how big does the chart need to be. Can I just print it on a 8.5X11 piece of paper or does it need to be bigger?

It is ok, just use good resolution printer, that's all.

[Essami]

Testers wanted!!!


Shoot this resolution chart using
1) 16:9 photo mode (at maximum resolution)
2) 16:9 photo mode (at 1920x1080)
3) Movie mode (1080p24 or p25)
4) Movie mode (720p60 or p50)
5) Keep whole 16:9 chart tightly within frame.
6) Level chart and your camera.
7) Use tripod.

www.pentax-hack.info/firmware/pan/ISO_12233-reschart.pdf (http://www.pentax-hack.info/firmware/pan/ISO_12233-reschart.pdf)

Here's my effort with a PAL camera, no hack.

http://dl.dropbox.com/u/3241184/GH1res.zip

Sami

[tyampel]

haha, yeah. i made sure to sign the piece of paper and stick it to the door. Good news.

1) Gh1.......delivered
2) Gh1 mount..........delivered
3) Gh1 3rd party battery...........delivered

Everything is here minus the card. Tester13, do you want to wait for card to be ordered or send what I have now?


PayPal balance: $149.80 USD
Looks like a good (working) condition TZ5 is available on Ebay for $130.
Does not leave much for shipping, but I am willing to cover the shortage.
Do you know how much do you need to ship this stuff to Tester13?

[Vitaliy Kiselev]

Looks like a good (working) condition TZ5 is available on Ebay for $130.
Does not leave much for shipping, but I am willing to cover the shortage.
Do you know how much do you need to ship this stuff to Tester13?

Lets first ship first batch. I gave all instructions.
Later, if funds will be enougth go for TZ5.
It is not much hurry, as even MJPEG engine is not entirely the same on TZ5.
I plan to use it for compression experiments and for screen graphic drwaing experiments (this is very similar considering same screens used).
But this also requires large amount of reversing work to be done.

Btw. Very useful thing will be to have some utility to display icons, pictures and fonts inside firmware. If you have good C knowledge you can contact me using PM for details.

[kainekainekaine]

Hey tester, any chance you will be improving the Still functions of the GH1.

[TurboTwo]

Tester,just how much difference will having the GH1 body mean to the hack and the progress you can make?

Thanks again for all your work so far and if you need more funds for things please post,my paypal finger is still willing.

[Vitaliy Kiselev]

Here's my effort with a PAL camera, no hack.

http://dl.dropbox.com/u/3241184/GH1res.zip

Sami


Thanks.
So we have same exactly FOV (unless you adjusted it during shooting).
But simple scaling is very noticeable on short clips.
Try to shoot 1080p using native patch.
May be even PAL->NTSC+native.
So you could also shoot 720p60 in same conditions :-)

Important addition - try to shoot photo of max resolution in 4:3, etc, keeping same camera position and zoom, so we can know how 16:9 is cropped from the sensor.

[Oedipax]

Tester, would it be possible to offer alternate aspect ratios for video capture like with stills mode? Would love to have a native 2.35:1 mode and also 1.33:1 (and I guess 1.66:1 if the other two are easily doable). Thanks again for all the hard work and I apologize in advance if this was already covered somewhere in the last 32 pages.

[Vitaliy Kiselev]

Hey tester, any chance you will be improving the Still functions of the GH1. I would love the bulb mode to be unlmited or at least longer than the maximum 4 minutes. Cheers.

I'll try to look at this as soon as I'll have body. But can't promize anything.

[bumkicho]

Not that I have any saying in this, but if it were up to me, I would rather see Tester13 focus on getting higher bitrate 24fps video in all three video modes (1080p and 720p AVCHD, 720p MJPEG). To me, that gets the highest priority.

[Vitaliy Kiselev]

Tester,just how much difference will having the GH1 body mean to the hack and the progress you can make?

Thanks again for all your work so far and if you need more funds for things please post,my paypal finger is still willing.

It is hard to say.
We'll just have more low level experiments.
May be they'll help to get some results in our areas of interest.

[Vitaliy Kiselev]

Tester, would it be possible to offer alternate aspect ratios for video capture like with stills mode? Would love to have a native 2.35:1 mode and also 1.33:1 (and I guess 1.66:1 if the other two are easily doable). Thanks again for all the hard work and I apologize in advance if this was already covered somewhere in the last 32 pages.

I think it is easy for Panasonic, but very hard for us.
Due to fact that too much constants are hardcoded for 1980 and 720.

[bumkicho]

Tester, would it be possible to offer alternate aspect ratios for video capture like with stills mode? Would love to have a native 2.35:1 mode and also 1.33:1 (and I guess 1.66:1 if the other two are easily doable). Thanks again for all the hard work and I apologize in advance if this was already covered somewhere in the last 32 pages.

If shooting at 2.35:1 ratio is possible, let's say GH1 only needs to process 1920x820 image instead of 1920x1080, would it free up GH1 cpu resource to process images better? just speculating...

[Vitaliy Kiselev]

Not that I have any saying in this, but if it were up to me, I would rather see Tester13 focus on getting higher bitrate 24fps video in all three video modes (1080p and 720p AVCHD, 720p MJPEG). To me, that gets the highest priority.

You already can adjust bitrate (and I really need tests with lower settings - see my previous posts for complete explanations!).
But encoder also have parameters for I and P frames, so adjusting bitrate up do not lead to quality improvement.

[Essami]

Thanks.
So we have same exactly FOV (unless you adjusted it during shooting).
But simple scaling is very noticeable on short clips.
Try to shoot 1080p using native patch.
May be even PAL->NTSC+native.
So you could also shoot 720p60 in same conditions :-)

Important addition - try to shoot photo of max resolution in 4:3, etc, keeping same camera position and zoom, so we can know how 16:9 is cropped from the sensor.

Camera was fixed on the tripod between each shot, no adjustments made whatsoever.

Sorry, I need my camera daily for very important shots for the coming weeks and I don't want to do the patch just to be 100% safe. Maybe someone else can go on from here?

I can do the max resolution shots tomorrow if no one else does it by then.

Sami

[Vitaliy Kiselev]

I can do the max resolution shots tomorrow if no one else does it by then.


In this case try to find mode with maximum FOV (view viewfinder) and make one shot at maximum resolution (aligning chart to 16:9 boundaries, of course). Must be fast :-)

[Ben_B]

Keep in mind that the GH1's 4:3 does not use the full sensor it is also a crop....the full sensor is larger than the image circle the lenses produce in the corners. This is so in 16:9 it can actually get a little wider and lose less resolution. There is no picture setting on the GH1 that uses those far corners of the sensor.

[Vitaliy Kiselev]

Keep in mind that the GH1's 4:3 does not use the full sensor it is also a crop....the full sensor is larger than the image circle the lenses produce in the corners. This is so in 16:9 it can actually get a little wider and lose less resolution. There is no picture setting on the GH1 that uses those far corners of the sensor.

I know. We just want to find crop positions.

[Ben_B]

Is there a lens we should be using for this? Ideally a 50mm because it would have no distortion?

[TrueIndigo]

Bumkicho: "If shooting at 2.35:1 ratio is possible, let's say GH1 only needs to process 1920x820 image instead of 1920x1080" -- actually, even 2K CinemaScope (2048 x 858) is a smaller area than full HD (1920 x 1080). I would love that to be possible with this little camera. Interesting idea that using just less than FH pixel area might help image quality even more (when combined with increased bitrate and AVCHD optmisation). Even if the codec is improved with the GH2, it's unlikely they will offer an enthusiast ratio like 2k CinemaScope -- this sort of unique possibility (among other things) would extend the useful life of the GH1 mod even after the GH2 is out.

And while we're on about changing frame shapes, earlier I wondered if it could be possible to set the region of interest to be a 1:1 area on the sensor (so no sensor downsampling), again, maybe in CinemaScope ratio, which would be useful for C-mount lens users because of the smaller image circle?

Cheers, inspiring thread...

[Vitaliy Kiselev]

Is there a lens we should be using for this? Ideally a 50mm because it would have no distortion?

I think any non-fisheye will do :-)
As we are looking only at approximate cropping.

[alignment1]

page turn......didn't work

[Cavemandude]

page turn......didn't work
Unless you have OCD, who cares?

Randy

[tyampel]

page turn......didn't work
Known bug.
Perhaps it will now with more posts.

[alignment1]

Known bug.
Perhaps it will now with more posts.

well aware of it:beer: I was trying to bump it to the next page with a post...then modified my post with 'didn't work'

cheers

[e-steve]

It's possible that this bug appears on super long threads... let's hope it will sustain another 100 pages... by next week :)

[krass66]

It's possible that this bug appears on super long threads... let's hope it will sustain another 100 pages... by next week :)

Few more weeks and this thread will compete with Obama's healthcare bill.

[Vitaliy Kiselev]

PTool 3.15 released
VA Bitrate adjustment 3 patch added (for 1080p only)
1920 frame height patch added

[Isaac_Brody]

It strikes me as amazing that you can figure this out without a camera to test with. It's like shooting ducks in the dark.

[John Caballero]

It strikes me as amazing that you can figure this out without a camera to test with. It's like shooting ducks in the dark.

Just wait until he gets the camera in his hands! Pow!

[Vitaliy Kiselev]

It strikes me as amazing that you can figure this out without a camera to test with. It's like shooting ducks in the dark.

It is interesting to know that good hunter do not calculate shooting position and timing (otherwise he will be always late), he just recognizes and reconstruct it from expirience. And many hunters shoot knowing distance and hearing noise of flushing duck. So, sometimes thay can shoot in the dark with success.

[Vitaliy Kiselev]

Most interesting part is to test new bitrate adjustment (and we must start by lowering value, not upping!!!)

And 1920 height, it can work, but most probably it won't (you must try to use divisible by 16 height value).

[Cavemandude]

If you think this "Page Bug" issue is only with large threads then look at the one I started that only has 10 posts on it.

http://www.dvxuser.com/V6/showthread.php?t=206365&page=1

It ain't the size of the thread so give up on it, let it be and we won't have to keep reading about it on every other post here.

Randy

[Vitaliy Kiselev]

Here is for comparison parameters of Panasonic consumer HD camera, 1920 60i:

Bitrate declared/actual - 16.8/15.6Mb (GH1 is almost the same)
CABAC - Yes (GH1 - No)
Reframes - 4 (GH1 - 1)
GOP Size - 15 (GH1 - same 0.5sec GOP)
I Frame Size - 290Kb (GH1 - 225kb)
P Frames in GOP - 4, each about 90kb (GH1 - 55kb)
B Frames in GOP - 10, each about 37kb (no B frames in GH1)

Here we have smart encoder using bitrate as it must be used.
And reason of mud in GH1 can be seen clearly. P frames size is really bad.
Plus we must take in account that degradation is also due to no B frames and 1 Reframe only.

So, to compare quality with TM700 encoder at 24Mbit we need about 55-60Mbit, may be even 70Mbit, as encoding is actually worse than MPEG2.

[Tameside]

how does it compare with the 720p implementation?, is it deliberately "hobbled-deliberately" or is it due to the available processor power, bad engineering or just lack of knowledge by the camera division, surely they share the know-how around the buisness?

Fantastic work Tester13 and at least we are now starting to understand the reason 1080p has mud, more than Panasonic ever told us. Is it a matter of waiting for the GH2 in reality with properly implemented compression?

Many thanks

[TurboTwo]

I`d love to know what Panasonic make of this endeavour and what is being discovered..to be a fly on the wall in their office would be very interesting indeed...(obviously a multilingual fly)

[Tameside]

I dont get why they even put the 1080p into the camera if its not capable of it? as it stands its a very lame implementation and they should be very very ashamed of themselves, It makes you want to reconsider spending any more cash with these guys, they could of done alot better am sure than this.

[NURBS]

Hello'

Testing new ptool ver.

Change only this mode.

VA 1920 frame height: 1080 -- change to --> 820 (result. camera can't record video and show text message: NO ADDITIONAL SELECTIONS CAN BE MADE).

VA 1920 frame height: 1080 -- change to --> 1180 (result. camera can't record video and show text message: NO ADDITIONAL SELECTIONS CAN BE MADE).

Testing more.

Regards
NURBS

[NURBS]

Next test:

VA Bitrate adjustment 3: 3753 --change to--> 7506 (Can't write more 5sec.) - Frame rate : 23.947 fps wow :evil:

VA Bitrate adjustment 3: 3753 --change to--> 5000 (Can't write more 9sec.) - Frame rate : 25.191 fps wow :cheesy:

VA Bitrate adjustment 3: 3753 --change to--> 4000 - 3800 (Can't write more 28sec.) - Frame rate : 25.000 fps

VA Bitrate adjustment 3: 3753 --change to--> 2000 (Can't write flie)

Regards
NURBS

[Tameside]

even though you are using the new patched f/w? I want to wait until the new f/w because I want to update the new fw for the 45mm PL lens too which is coming!

[tyampel]

Next test:

VA Bitrate adjustment 3: 3753 --change to--> 7506 (Can't write more 5sec.) - Frame rate : 23.947 fps wow :evil:

VA Bitrate adjustment 3: 3753 --change to--> 5000 (Can't write more 9sec.) - Frame rate : 25.191 fps wow :cheesy:

VA Bitrate adjustment 3: 3753 --change to--> 4000 - 3800 (Can't write more 28sec.) - Frame rate : 25.000 fps

VA Bitrate adjustment 3: 3753 --change to--> 2000 (Can't write flie)

Regards
NURBS


Can you check if there are any diagnostic messages when you go to the service menu. You know how to do this as per your GH1 service manual post.

[Vitaliy Kiselev]

Hello'

VA 1920 frame height: 1080 -- change to --> 820 (result. camera can't record video and show text message: NO ADDITIONAL SELECTIONS CAN BE MADE).



Try to use 816, not 820.
Anyway it looks we'll need to adjust this.

[Vitaliy Kiselev]

I dont get why they even put the 1080p into the camera if its not capable of it? as it stands its a very lame implementation and they should be very very ashamed of themselves, It makes you want to reconsider spending any more cash with these guys, they could of done alot better am sure than this.

It is capable.
But their marketing department specified AVCHD and bitrate available.
It is hard to do anything with this hardware and such restrictions.
Best thing is to introduce 1080 film mode that removes all AVCHD restrictions and uses maximum possible bitrate.

[Vitaliy Kiselev]

Next test:
VA Bitrate adjustment 3: 3753 --change to--> 7506 (Can't write more 5sec.) - Frame rate : 23.947 fps wow :evil:

VA Bitrate adjustment 3: 3753 --change to--> 5000 (Can't write more 9sec.) - Frame rate : 25.191 fps wow :cheesy:

VA Bitrate adjustment 3: 3753 --change to--> 4000 - 3800 (Can't write more 28sec.) - Frame rate : 25.000 fps

VA Bitrate adjustment 3: 3753 --change to--> 2000 (Can't write flie)


Can it write 720p with this settings?
It will be good to upload this short clips somethere.

[Tameside]

when the 1080p is good its very very good, awesome in fact. If we had stable reliable footage it would truly rock.

Thanks T'side

[Vitaliy Kiselev]

when the 1080p is good its very very good, awesome in fact. If we had stable reliable footage it would truly rock.


You can see resolution charts shooting in this thread to understand sensor potential. Look at static 1920x1080 and loo at short video clip.
Of course, most degradatition is due to binning used.