PDA

View Full Version : GH1 firmware research volunteers required



Pages : 1 2 3 4 5 6 7 8 [9] 10 11 12 13

GrgurMG
06-03-2010, 10:12 AM
No, if you mean firmware hack, and quite easy by normal firmware modification.
This limit is not MJPEG related it is FAT file system related.
Due to weird licensing issues flash cards still use FAT system instead of normal ones.

Not entirely true... though maybe half true. The overall FAT32 file limitation on the GH1 is 4GB. Someone mentioned the GH1 spanning 2GB ACVHD clips and this is not the case.. it spans 4GB (3.99 GB to be exact) clips.. at least according to all the footage I have taken.

But you may be half right... as you mentioned there being alot of old code... perhaps the MJPEG limitation is there due to old FAT16 code they never bothered updating... though then again, it could be some entirely other reason we have not thought about.

Isaac_Brody
06-03-2010, 10:29 AM
Oh, I just remembered that while we call this mjpeg, it's really photojpeg which is much more efficient.

This is worth reading to see just how efficient. Click on 4:2:2C (4:2:2 Compressed Codec Formats) to read about the difference between mjpeg and photojpeg.

http://codecs.onerivermedia.com/

And does anyone know what colorspace the GH1 shoots in? Is there an option for Adobe RGB within camera? Or are there no settings in the menus?

tyampel
06-03-2010, 10:58 AM
http://codecs.onerivermedia.com/

And does anyone know what colorspace the GH1 shoots in? Is there an option for Adobe RGB within camera? Or are there no settings in the menus?
[/quote]

Adobe color space option is available on the menu. Page 122.

Isaac_Brody
06-03-2010, 11:02 AM
Thanks, is this what most people shoot in? Even just rerendering in after effects to Adobe RGB flattens that over contrasty image making it look more natural. I'd definitely like to see a shot with contrast dialed all the way down and set to Adobe RGB. Hopefully it's possible to flatten image more and overcome the baked in look.

alignment1
06-03-2010, 11:54 AM
Thanks, is this what most people shoot in? Even just rerendering in after effects to Adobe RGB flattens that over contrasty image making it look more natural. I'd definitely like to see a shot with contrast dialed all the way down and set to Adobe RGB. Hopefully it's possible to flatten image more and overcome the baked in look.

I always thought that this was for stills only, as you need to put the dial on 'M' to make adjustment to colorspace.

This is great! I just did a quick test- Adobe RGB gave me much more detail in what would have been blown out whites in sRGB---hence I think it's flatter!!:shocked:


We should take this conversation to the 'Baked In" thread
http://www.dvxuser.com/V6/showthread.php?t=212460

Cheers thanks Isaac!

Svart
06-03-2010, 12:05 PM
I tried this on my GH1 some time ago after reading that the 5dMK2 also had a setting for AdobeRGB vs. RGB and the person who wrote the article claimed that Canon told him to use AdobeRGB for better color.

I could not tell much, if any, difference personally.

alignment1
06-03-2010, 12:14 PM
I tried this on my GH1 some time ago after reading that the 5dMK2 also had a setting for AdobeRGB vs. RGB and the person who wrote the article claimed that Canon told him to use AdobeRGB for better color.

I could not tell much, if any, difference personally.

Actually I did too- I think it was a rep from NAB or something- anyhow...

just did a couple more tests---difference for SURE!!! Adobe RGB is the way!!!!

I'm in a red room- (My office)- Huge difference- Mind you, I'm shooting 'Nostalgic' (supposed least contrasty setting) Cont. -2 / Sharp +1 / Sat. 0 / NR -2.

But now w/ Adobe RGB it's even flatter!! It makes my sRGB look like a canon:evil:

Yoohooo this cam gets better and better! Can't wait for these FW's to be concrete!

mimirsan
06-03-2010, 12:22 PM
just wondering...can we stay on topic?

Essami
06-03-2010, 12:27 PM
just did a couple more tests---difference for SURE!!!

I can confirm that theres a difference too. Before getting too excited we should look into what actually happens and if there really is more info there or is it something that a simple color space change can do in post as well, ie. is there actually more info or is our computer or NLE showing it differently so it appears its flatter.

Sami

PS. good idea to start a new thread on this one!!! Maybe OP should do it?

alignment1
06-03-2010, 12:39 PM
just wondering...can we stay on topic?

nice first post- welcome to the forum

yes- I apologize let's

Svart
06-03-2010, 01:37 PM
just wondering...can we stay on topic?

We were discussing the usefulness of Tester13 taking time away from his busy schedule and continued GH1 work to look into changing the color space on the GH1. This includes the thought that simple changes in setup might circumvent his need to divert from his current work. How is that NOT on topic?

Oleg2d
06-03-2010, 02:41 PM
Excuse me for my English - it is terrible!
I tested many versions bitrate in AVCHD, but to me I not seemed that to a type of a difference. Some of pictures look superb, but some pictures awfully - like mobile phone camera. AVCHD patch is a dark horse. But patch for MJPEG it very well looks. Thanks for your work Tester13, do not surrender!

mimirsan
06-03-2010, 03:05 PM
nice first post- welcome to the forum

yes- I apologize let's

Hehe Thanks! :engel017:
yeah I just realised after posting!
Im a noob so forgive me! The points you guys made were actually not completely out of topic!
Have been reading through the thread and its hard to keep track how the progress is going sometimes.

I think what the tester is doing is great stuff...just hope that he can get around the main issues with AVCHD codec (or making Mjpeg even better as I use it most nowadays) and that things dont go too off track.

The "mud/breakup" issue is the only real negative on the gh1...fixing this would be perfect.

Isaac_Brody
06-03-2010, 04:18 PM
Can someone running the mjpeg/photojpeg hack do me a favor?

Since I don't have a camera I want to get a hold of mjpeg footage shot AdobeRGB and with contrast and sharpness dialed all the way down to zero or however far they go. I just need to see something well exposed, outside or inside at a higher bitrate. ( Whichever high bitrate is stable) I'm just doing some post tests and could use the neutral footage. You can PM me directly if you don't want to take this thread further off topic.

:dankk2:

/BACK ON TOPIC.

PappasArts
06-03-2010, 08:10 PM
Did some more testing with the high bit rate Nurbs posted. Got 107Mbits. However my Extreme class 10 just can't go beyond 5 secs on that. We need class 20....... Whenever that will be.. I notice that 90Mbits is to much as well. There just isn't enough head room with these cards for that. However clips that are 70ish Mbits seem to go with out data stop by camera. Good thing is cards get better, and there might be even higher end pro level ones with sustained data writes at 15+.

I need a setting that has a lower top end bit rate like 70ish. Any suggested ones?

Here is a frame from footage shot today in CDM Newport Beach with Anamorphic at 19mm; and the file was at a 70Mbits data rate.





Michael Pappas
http://www.pbase.com/Arrfilms
Anamorphic GH1 Frames with high bit codec
http://www.dvxuser.com/V6/showpost.php?p=2006938&postcount=1956

John Caballero
06-03-2010, 08:22 PM
Here is a frame from footage shot today in CDM Newport Beach with Anamorphic at 19mm; and the file was at a 70Mbits data rate.


That was mjpeg?

PappasArts
06-03-2010, 08:56 PM
That was mjpeg?

Yes, that was 720 mjpeg.

These were from yesterday.
http://www.dvxuser.com/V6/showpost.php?p=2006938&postcount=1956


.

Car3o
06-03-2010, 09:02 PM
looks great. if this gets native 24p, it'll crush mk2. of course that's my opinion.

Tesselator
06-03-2010, 10:16 PM
Hi All, Hi Tester13...

I've lived in Japan for the past 22 years. Originally from So.Cal. USA. I'm basically in early retirement and have lots of time on my hands. My (pertinent) background in chronological order is:

8-bit video game development (C=64),
16-bit video game development (Amiga),
Import/Export computer/pro-video hardware/software,
CG artist (a little famous in Japan) TV/Movies,
Lightwave 3D engineering (NewTek). My name is in the about box of all 9.x versions,
Nose-picker and navel gazer for the past 2 years. Err, I mean Gentleman in early retirement. ;)

Count me in on the helping out scene! Although currently I don't know very much. I know that the GH1 is selling here in Japan (new, in-box, body only) for $350 ~ $400. Exports to the US/EU end up being about $500 after everything is considered. And I know the idiotic dumb-arses in Panasonic's marketing group have ruined the natural distribution of their "G" product line by not including the appropriate Language options natively. Oh, and since my hobby is still photography (my first SLR - purchased new at the time (http://tesselator.gpmod.com/AsahiFlex_SLR/)) and part of my work experience is/was video related I know some things about cameras and lenses - but not an incredible amount. Last thing, I also know that I hate forum nazis and forum police. :P

You can e-mail me from this site or PM me. Tell me what I can do utilitarian style, to help out. I've got lots of free time. ;)


Nice to meet you all and I look forward to active participation here on this forum and testing/contributing behind the scenes as much as I may.

John Caballero
06-03-2010, 10:31 PM
Yes, that was 720 mjpeg


If only 24p could be made to work properly! Mjpeg would definetely be my only codec of choice.

Vitaliy Kiselev
06-03-2010, 10:39 PM
Hi All, Hi Tester13...
I've lived in Japan for the past 22 years. Originally from So.Cal. USA. I'm basically in early retirement and have lots of time on my hands. My (pertinent) background in chronological order

It seems impressive.
Having developed some pretty amateur things on Sinclair ZX Spectrum (including heavy asm usage) I can say that it is pretty different from current game development.
May be you can be head of our communication team?
As we want to get in touch with Panasonic department to get some information. May be even firmware object files or source parts.
For other things see second post on first page, please.


Count me in on the helping out scene! Although currently I don't know very much. I know that the GH1 is selling here in Japan (new, in-box, body only) for $350 ~ $400. Exports to the US/EU end up being about $500 after everything is considered. And I know the idiotic dumb-arses in Panasonic's marketing group have ruined the natural distribution of their "G" product line by not including the appropriate Language options natively.

As we already don't have language problem it is ok.
We'll use your information in case if our body will be bricked.

Do you have GH1 yourself?


You can e-mail me from this site or PM me. Tell me what I can do utilitarian style, to help out. I've got lots of free time. ;)

We are pretty open, so unless something privite comes along or talk becomes very technical I suggest to remain on this thread.

Oleg2d
06-04-2010, 12:58 AM
I need a setting that has a lower top end bit rate like 70ish. Any suggested ones?

+1 I need settings like this too

Oedipax
06-04-2010, 01:17 AM
This is so cool to watch as it comes together a day at a time. The MJPEG stuff looks amazing... so glad I didn't sell my GH1 ;)

rundavids
06-04-2010, 01:38 AM
This is off topic. Sorry. But, I've been reviewing my 1080p footage on camera and computer and I found that on all the footages, if you zoom in to the entire lower bottom are lines of rainbow-colored artifacts going on. Anyone have this problem too?

PappasArts
06-04-2010, 02:20 AM
+1 I need settings like this too

Also a stop point for lower bit point. Shadows - low key areas in scenes in low light environments would benefit if the variable bit rate didn't go below 50Mbits just cause the scene is dark.

Is it possible to control the VBR on the low too?

I'm testing this one as of now from "jobless" So tomorrow I will head out and see what happens.
Quality settings (E1 to E4) - 720, 412, 352, 300
Table settings (E1 to E4) - 24, 24, 24, 24
Indoor shooting
Bit rate : 74.5 Mbps
Bits/(Pixel*Frame) : 2.695



Nurbs setting was amazing looking today. However when it went over 80Mbits- Recording stopped in very high detail scenes that sky rocketed to 107Mbits, which is astounding looking. It's the cards that are limiting this I believe

Can someone link me to where I can read up to understand what this Quality settings (E1 to E4) mean and how these numeric adjustments relate to the codec.


Here's another frame from "Nurbs" settings. This footage is at 107Mbits. 107Mbits is so smooth and clean. The Dynamic range of the GH1 is so obvious in this shot as the previous frame posted. The camera stopped around 6 seconds. Class 10's Extreme cards can't cut it. However it would be nice to have a SD dummy card with a cable that could go into a CF card that could. Hell, I don't think that's possible. However it would be nice.

As the others, this was Anamorphic 720 upped to 1920x812.


Pappas

Vitaliy Kiselev
06-04-2010, 02:44 AM
PTool 3.32 Released
GH1
AVCHD Compression patches added (for testers only!).
GF1
Almost all useful patches ported to GF1,
including latest MJPEG settings.
G1
Interface patch ported.
Third-party battery patch ported.

Testers - please see notes on first page, second post.

Adventsam
06-04-2010, 03:00 AM
The G2 is SDXC compliant and in theory runs exFAT, therefore wouldn't be any restriction on file size and it could be ramped up very very fast in mjpeg. Would the patch work for that too? its a better cam than the GF1 due to GH1 form factor imo plus mic stereo input.

Also, Nurbs said he had it runnning at 25p, ino dropped frames, is this correct.

Many thanks

Vitaliy Kiselev
06-04-2010, 03:15 AM
The G2 ... Would the patch work for that too? its a better cam than the GF1 due to GH1 form factor imo plus mic stereo input.

If Panasonic will provide us unencrypted firmware it can be made :-)
As I understand for now we don't have any.


Also, Nurbs said he had it runnning at 25p, ino dropped frames, is this correct.

No.
Only 24fps mode exists, with dropping frames.

If you need 25fps, shoot AVCHD 720p. It have no dropping frames.

Adventsam
06-04-2010, 03:18 AM
G2 firmware update will come, hopefully soon.

e-steve
06-04-2010, 03:36 AM
I'm not sure what else we are needing that costs $$, but since this thread has well over 200,000 views at this point, I hope everyone reading it has taken the time to donate, if just 5 or 10 dollars.

If we have 500-1000 unique readers that amounts to a lot of money to help this effort - so remember even a small amount will add up to a lot of $ to help this effort!

(this should now include GF1 owners at this point, as you can see Tester13 is supporting this cam too)

Tesselator
06-04-2010, 05:01 AM
It seems impressive.
Having developed some pretty amateur things on Sinclair ZX Spectrum (including heavy asm usage) I can say that it is pretty different from current game development.

Yup, very different. When I was game programming it was 1 to 3 guys on the job plus one musician or artist occasionally. Now it's 20 to 50 guys on a team with massively more complex code all cooperating (hopefully) with the OS (whereas we used to kick out the OS all together. ;))


May be you can be head of our communication team?
As we want to get in touch with Panasonic department to get some information. May be even firmware object files or source parts.

I can try! I got pretty good at that sort of human finessing during my import/export company adventures. I'm out of practice but it certainly can't hurt to try. ;) And these days it's by e-mail and typing I assume - which I like doing.


For other things see second post on first page, please.

As we already don't have language problem it is ok.
We'll use your information in case if our body will be bricked.

Will do,
good to know,
let's hope I get that far. ;)


Do you have GH1 yourself?

Just ordered one the day before yesterday. $370 (new). It should be here in less than 5 days. I'm currently still shooting (digitally) on the Konica Minolta A2 - a 7 year old 8mpx bridge camera which like the GH1 of today was feature rich and a little ahead of the curve. I still get good results with it but I want to use some of the hundreds (actually about 50) of lenses I own and play with video so I chose the GH1. Here's some of the most recent A2 samples if noone minds the in-line images :Drogar-KnockedOut(D





http://tesselator.gpmod.com/Flower_Session_99/content/bin/images/large/PICT1971.jpg






http://tesselator.gpmod.com/Flower_Session_99/content/bin/images/large/PICT2047.jpg






http://tesselator.gpmod.com/Flower_Session_99/content/bin/images/large/PICT2035.jpg






http://tesselator.gpmod.com/Flower_Session_99/content/bin/images/large/PICT2062.jpg






http://tesselator.gpmod.com/Images/_People/PICT3872_sm.jpg



Very recent BIFs are here:
http://tesselator.gpmod.com/Beer_With_Ken_and_Hiro/

More flowers and stuff:
Nikon F5: http://tesselator.gpmod.com/Chickens_On_Film/
Cluck: http://tesselator.gpmod.com/A_Walk_In_The_Park/
Lunch: http://tesselator.gpmod.com/River_Lunch/
Some Micro Lenses: http://tesselator.gpmod.com/Microscope_Objectives/


We are pretty open, so unless something privite comes along or talk becomes very technical I suggest to remain on this thread.

OK, will do!
Thanks for taking the time to reply!
Hopefully I can be of some use!

,

,

.

Chris Light
06-04-2010, 05:27 AM
fantastic colors Tesselator!

MatzeB
06-04-2010, 05:28 AM
However it would be nice to have a SD dummy card with a cable that could go into a CF card that could. Hell, I don't think that's possible. However it would be nice.


Just found this Micro SD Slot Adapter:
mobidapter
http://www.elandigitalsystems.com/adapter/mobidapter.php
http://www.iostore.co.uk/products/showdetails.asp?start=0&id=43&subid=0&pid=221

With a SD to Micro SD adapter you can plug it into the camera. There are no writing speed specs on the website. Maybe it could work with a USB to CF Card Adapter. Propably would need some kind of power supply.

Vitaliy Kiselev
06-04-2010, 05:50 AM
Just found this Micro SD Slot Adapter:
mobidapter
http://www.elandigitalsystems.com/adapter/mobidapter.php



First. Stop searching high speed card. As I already said this is camera limit, not your SD card.

jobless
06-04-2010, 05:59 AM
Avchd TEST:

AVCHD compression 1

Overall bit rate : 15.1 Mbps
Maximum Overall bit rate : 18.0 Mbps
Bit rate : 14.3 Mbps
Maximum bit rate : 16.6 Mbps
Bits/(Pixel*Frame) : 0.288


AVCHD compression 2

Overall bit rate : 4 552 Kbps
Maximum Overall bit rate : 18.0 Mbps
Bit rate : 4 171 Kbps
Maximum bit rate : 16.6 Mbps
Bits/(Pixel*Frame) : 0.084

AVCHD compression 3

Overall bit rate : 15.2 Mbps
Maximum Overall bit rate : 18.0 Mbps
Bit rate : 14.4 Mbps
Maximum bit rate : 16.6 Mbps
Bits/(Pixel*Frame) : 0.290


AVCHD compression 2 & 3

Overall bit rate : 7 964 Kbps
Maximum Overall bit rate : 18.0 Mbps

Bit rate : 7 448 Kbps
Maximum bit rate : 16.6 Mbps
Bits/(Pixel*Frame) : 0.150

AVCHD compression 2 & 1

Overall bit rate : 4 567 Kbps
Maximum Overall bit rate : 18.0 Mbps
Bit rate : 4 185 Kbps
Maximum bit rate : 16.6 Mbps
Bits/(Pixel*Frame) : 0.084


AVCHD compression 1 2 3

Overall bit rate : 8 053 Kbps
Maximum Overall bit rate : 18.0 Mbps
Bit rate : 7 533 Kbps
Maximum bit rate : 16.6 Mbps
Bits/(Pixel*Frame) : 0.152

Download link:
http://www.nikolicnemanja.com/test.rar


In rar file are include all 6 versions
Sorry for choppy pans.... I have really bad tripod...

Vitaliy Kiselev
06-04-2010, 06:08 AM
Avchd TEST:
In rar file are include all 6 versions
Sorry for choppy pans.... I have really bad tripod...

Choppy pans don't matter.
Thanks for very useful test.
We'll be slowly improving our understanding of AVCHD encoder.
I'll look at low level info.
Many iterations of similar patches and tests are required to get result.

RandyQ
06-04-2010, 06:13 AM
After seeing all these high-bitrate MJPEG tests using NURBS settings - does this mean these new MJPEG modes are ready for real world use or still premature? I'd like to see if I can start using it in an upcoming shoot.

Vitaliy Kiselev
06-04-2010, 06:16 AM
After seeing all these high-bitrate MJPEG tests using NURBS settings - does this mean these new MJPEG modes are ready for real world use or still premature? I'd like to see if I can start using it in an upcoming shoot.

Only problem that can arise is you can come to very high detailed shot and have record stopped.
So, test your settings with your card before this.
Plus I suggest really good card, freshly formatted. As any delay at such speed and you'll have buffer overrun and card speed message.

Tesselator
06-04-2010, 06:30 AM
fantastic colors Tesselator!


Thanks.
--

On color space discussed previously See: http://www.dvxuser.com/V6/showthread.php?t=212482

Vitaliy Kiselev
06-04-2010, 06:40 AM
In rar file are include all 6 versions
Sorry for choppy pans.... I have really bad tripod...

I need also same original footage (without any hacks, etc).
And 1&3 both checked is absent.

Vitaliy Kiselev
06-04-2010, 06:44 AM
Thanks.
On color space discussed previously

Is it possible to scale pictures down a little?
As posts looks huge and are a little offtopic here.

Tesselator
06-04-2010, 06:55 AM
Is it possible to scale pictures down a little?
As posts looks huge and are a little offtopic here.

OK. Sorry about that. :oops: :shocked:

Sounds like someone needs to donate you a normal sized monitor! :violin:

Vitaliy Kiselev
06-04-2010, 06:57 AM
OK.
Sounds like someone needs to donate you a normal sized monitor! :)

Right now I am working on 32" monitor :-)
I am ok if you donate me 40" UltraRes one :-)

Tesselator
06-04-2010, 07:02 AM
Ha! You're already bigger than me!

Wait, everyone is bigger than me.

Wait wait, that's a totally different topic... nevermind. ;P

thisisapocalypse
06-04-2010, 07:28 AM
This is off topic. Sorry. But, I've been reviewing my 1080p footage on camera and computer and I found that on all the footages, if you zoom in to the entire lower bottom are lines of rainbow-colored artifacts going on. Anyone have this problem too?

Do you find it in the bottom 8 pixels only? Mine has that issue, I've learned to live with it and crop it out.

jobless
06-04-2010, 07:36 AM
AVCHD compression 1 & 3

Overall bit rate : 15.3 Mbps
Maximum Overall bit rate : 18.0 Mbps
Bit rate : 14.5 Mbps
Maximum bit rate : 16.6 Mbps
Bits/(Pixel*Frame) : 0.291

AVCHD without patches

Overall bit rate : 15.2 Mbps
Maximum Overall bit rate : 18.0 Mbps
Bit rate : 14.4 Mbps
Maximum bit rate : 16.6 Mbps
Bits/(Pixel*Frame) : 0.231
Frame rate : 29.970 fps

http://www.nikolicnemanja.com/test2.rar

Wait a minut for download 'cause still uploading...

e-steve
06-04-2010, 07:52 AM
I would appreciate it if you recropped your avatar photo... higher (maybe 60% higher)

jobless
06-04-2010, 07:59 AM
I would appreciate it if you recropped your avatar photo... higher (maybe 60% higher)

Sure...

Jasketti
06-04-2010, 08:30 AM
GF1
Almost all useful patches ported to GF1,
including latest MJPEG settings.


Amazing job tester13 and all other testers!

I will definately make a donation once I get my next paycheck :) my main interest however is to see the HDMI live output come true on the GH1 or even at GF1. Hope you can make it. Though it looks like the MJPEG highbitrate is becoming a fantastic option too!

I already own a GH1 and I am looking forward to get GF1 for multicam projects. How close is the GF1 720P MJPEG picture quality compared to GH1 with this patch? Is it exactly the same? Do you think GF1 could be hacked to shoot 1080P?

Btw. GF1 body only cost 599.90 here in Finland... :( any hints where to get it for better price.

Vitaliy Kiselev
06-04-2010, 08:58 AM
my main interest however is to see the HDMI live output come true on the GH1 or even at GF1.

You must understand that it is very hard.
Almost impossible to do without documentation or at least object files.


I already own a GH1 and I am looking forward to get GF1 for multicam projects. How close is the GF1 720P MJPEG picture quality compared to GH1 with this patch? Is it exactly the same? Do you think GF1 could be hacked to shoot 1080P?

I expect maximum bitrate to be the same.
All else is sensor dependant.
So, no 1080p.


Btw. GF1 body only cost 599.90 € here in Finland... :( any hints where to get it for better price.

Look on ebay for Japanese GF1 bodies.
You can also ask any reputable guys here who live in Japan.

rundavids
06-04-2010, 10:28 AM
Hi thisisapocalypse. Yes, the artifacts are there as you say they are. I'm glad I'm not the only one. Yeah, I'll be cropping my footages from now on.

Slice
06-04-2010, 10:32 AM
Tester13, great job. I find myself checking this thread every couple hours.

Have you found anything relating to MJPEG and it's 4:2:0 color space. It would be a great for us VFX artists if we could switch it to 4:2:2 or 4::4:4.

Thanks for the time and effort your putting into this. You sir, are awesome.

Isaac_Brody
06-04-2010, 10:43 AM
Have you found anything relating to MJPEG and it's 4:2:0 color space. It would be a great for us VFX artists if we could switch it to 4:2:2 or 4::4:4.

At 75 percent quality photojpeg is 4:2:2. Not to mention that it's intraframe. That should be plenty for vfx. I have totally given up on the AVCHD component for this camera. It's not going to compete with the mjpeg.

nikgid
06-04-2010, 10:48 AM
PTool 3.32 Released
GF1
Almost all useful patches ported to GF1,
including latest MJPEG settings.


Yay, hurray!
Thanks so very much! Will check it out asap.

Btw. is there any chance of patching the minimum shutter-speed (for stills) below 1/4000s? Would be very convenient for shooting wide-open (for DOF) with fast lens (like the f1.7/20mm or old legacy lenses) during daylight.

craion
06-04-2010, 11:03 AM
Amazing job tester13 and all other testers!

I will definately make a donation once I get my next paycheck :) my main interest however is to see the HDMI live output come true on the GH1 or even at GF1. Hope you can make it. Though it looks like the MJPEG highbitrate is becoming a fantastic option too!

I already own a GH1 and I am looking forward to get GF1 for multicam projects. How close is the GF1 720P MJPEG picture quality compared to GH1 with this patch? Is it exactly the same? Do you think GF1 could be hacked to shoot 1080P?

Btw. GF1 body only cost 599.90 here in Finland... :( any hints where to get it for better price.

I bought from this seller,GF1 kit with 20mm lens for around 580 :
http://stores.shop.ebay.ie/eTech-Maniac__W0QQ_sidZ55451216?_nkw=panasonic&submit=Search&LH_TitleDesc=1

Vitaliy Kiselev
06-04-2010, 11:11 AM
At 75 percent quality photojpeg is 4:2:2.

I believe that this is not correct.
YCC422 or YCC420 selection depends on JPEG compressor implementation.
It is better to perform simple color resolution tests.


I have totally given up on the AVCHD component for this camera. It's not going to compete with the mjpeg.

May be you gave up too soon :-)

Isaac_Brody
06-04-2010, 11:16 AM
I definitely did. Unless you've got a miracle up your sleeve you haven't posted yet. :)

jobless
06-04-2010, 11:18 AM
What are you Bodyfobic?
But it's ok.... Flower is nice... :)

rotrier
06-04-2010, 11:29 AM
I love the look of the film, I know this is off topic but I need to know where I can find those Anamorphic lens. Thanks.

andyjar
06-04-2010, 12:51 PM
Strait off the cam, AVCHD looks far better then MJPEG. While it's great we can modify it so much, why modify the lower quality option to be higher when we could change the high quality option to be higher?

JPEG is infamous for being a very artifacty compressed format.

Isaac_Brody
06-04-2010, 12:53 PM
Strait off the cam, AVCHD looks far better then MJPEG. While it's great we can modify it so much, why modify the lower quality option to be higher when we could change the high quality option to be higher?

JPEG is infamous for being a very artifacty compressed format.


If that were true there wouldn't be countless threads complaining of mud and the "rembrandt" effect. Are we looking at the same footage? Show me where this is true.

Take a look at the footage from this post with comparison footage between mjpeg and avchd.

http://www.dvxuser.com/V6/showpost.php?p=2005410&postcount=1855

http://www.dvxuser.com/V6/showpost.php?p=2007277&postcount=1972

GrgurMG
06-04-2010, 01:39 PM
If that were true there wouldn't be countless threads complaining of mud and the "rembrandt" effect. Are we looking at the same footage? Show me where this is true.

Take a look at the footage from this post with comparison footage between mjpeg and avchd.

http://www.dvxuser.com/V6/showpost.php?p=2005410&postcount=1855

http://www.dvxuser.com/V6/showpost.php?p=2007277&postcount=1972

Well, I personally think it's fruitful to pursue both paths and get the most out of each. As I mentioned before, if we can fix some of the H264 encoding flaws of the GH1 in addition to raising the bitrate, the AVCHD should look better or as good as the MJPEG while taking up less space on card... providing there isn't very heavy full frame motion in the scene.

I'm not sure how the AVCHD processes work inside the camera, but if very high bitrates with shorter GOPs are achievable with the AVCHD then if there's quality or levels of clean footage to be had past our current MJPEG results, then AVCHD will achieve it in many scenes. For quick multiclip applications like filmaking, the MJPEG is managable, but for other applications like event videography, documentaries, etc etc.. where recording times and card space are more important convenience factors, MJPEG doesn't cut it in all workflows.

I think if both movies modes are tweeked to their full potential, there would be practical uses for each.

spamrakuen
06-04-2010, 01:55 PM
I think if both movies modes are tweeked to their full potential, there would be practical uses for each.

Sure, but people wants to get rid off mud, and now we're soooo close. Only the problem with the 30->24fps hack remains in order to have the holy grial for many mud-frustrated GH1 owners... :violin:

PappasArts
06-04-2010, 02:00 PM
I have totally given up on the AVCHD component for this camera. It's not going to compete with the mjpeg.


Isaac, the Mjpeg just handles even over/under exposures entirely different. Reminds me of my S16 scans, however no grain.

This was shot a day apart, however it demonstrates in an unscientific manner between Mjpeg and AVCHD at 17Mbits from the GH1


Pappas

bumkicho
06-04-2010, 02:18 PM
I think if both movies modes are tweeked to their full potential, there would be practical uses for each.

There are some people claming that they have given up on AVCHD modes (FHD, SH) on GH1, but the good news is tester13 hasn't given up on them. And when tester13 starts making progress on improving AVCHD Mbps up to its potential, people will come around.

If tester13 says he is giving up on AVCHD modes and there is nothing he can do to improve, then well.. I hope that day never comes. Perhaps you can facilitate the process by testing out AVCHD variables in the patch.

jobless
06-04-2010, 02:31 PM
there are some people claming that they have given up on avchd modes (fhd, sh) on gh1, but the good news is tester13 hasn't given up on them. And when tester13 starts making progress on improving avchd mbps up to its potential, people will come around.

If tester13 says he is giving up on avchd modes and there is nothing he can do to improve, then well.. I hope that day never comes. Perhaps you can facilitate the process by testing out avchd variables in the patch.

+1

thisisapocalypse
06-04-2010, 03:05 PM
Isaac, the Mjpeg just handles even over/under exposures entirely different. Reminds me of my S16 scans, however no grain.

This was shot a day apart, however it demonstrates in an unscientific manner between Mjpeg and AVCHD at 17Mbits from the GH1


Pappas

It does appear that the MJPEG has a better grip on dynamic range, but the lighting in those two shots is so different you really can't compare them, unfortunately. The AVCHD looks sharper to me, but I like the way the MJPEG seems to really have a better dynamic range...but I could be wrong on that and the lighting may have been really flat and uncontrasty that day and that's why that image looks that way, by comparison the AVCHD one has very harsh highlights and shadows.

Are you using the Panasonic LGA7200 for your anamorphics?

d1rockwild
06-04-2010, 03:56 PM
yeah with the lighting conditions so different its really hard to tell which one is a better image.

Isaac_Brody
06-04-2010, 04:00 PM
I see more detail in the mjpeg clips posted, detail which just becomes a smudge in the AVCHD mode. Of course I'd be happy with a better AVCHD mode. Bring it! :)

kainekainekaine
06-04-2010, 06:50 PM
Panasonic have blocked four thirds Olympus teleconverters ec-14 and ec-20 from working with G cameras but they work with E-p cameras although it is not recommended by them.

Barry_Green
06-04-2010, 06:53 PM
I think what Isaac and I (and perhaps others) are saying is: we're not asking for the moon, and we don't expect tester13 to move heaven and earth. We'd be happy with a mud-free razor sharp codec. So if tester13 wants to focus his efforts on one approach, rather than trying to keep them all in play, that's fine with us. Whichever is more feasible and easier to accomplish: great MJPG at 720/24fps, or great MJPG at 1080/24fps, or great AVCHD at 1080/24pN.

So far it seems like we have great MJPG at 720/30p, and the beginnings of a possible 1080p MJPG, but really not much in the way of improvement in the AVCHD mode (other than the excellent 24pN conversion, which eliminates the goofy pulldown!)

So, there's three different modes, and any one of them would be ideal, if we could get any of them "finished" -- meaning, the great bitrate improvements of the MJPG mode but also at 24fps, or the 24pN of the AVCHD mode, but also with the quality improvements that the MJPG mode has shown.

So far we don't have a finished version of any of the three potential paths, and yes of course we'd love to have all three. But I'm saying "hey, whichever's easiest, if you want to focus on that then please do, because we really only need one of 'em!"

Vitaliy Kiselev
06-04-2010, 07:21 PM
So far we don't have a finished version of any of the three potential paths, and yes of course we'd love to have all three. But I'm saying "hey, whichever's easiest, if you want to focus on that then please do, because we really only need one of 'em!"

No one knows which is the easiest one.
AVCHD way is good as it have all problem solved except bitrate values.
As for mud - take a look at latests tests video. Even static shots are complete mud at low bitrate. Problem here is that AVCHD is more like afterthought for this models, so we can have many limits on each stage.
MJPEG have famous 2Gb limit, size and framerate must be changed.
This involves sensor setup. And this is very risky parts, as any error in this part and you have bricked camera due to fact that GH1 do not have separate play mode on power up (like TZ5).

So here is small plan:

We'll go for both MJPEG and AVCHD routes.
For MJPEG we'll try to make limit at 4Gb.
For AVCHD we'll make many small AVCHD patches to understand encoder behaviour.

Our problems

We need compilers and, especially, service software.
Service software is invalueable, as all we have is interface strings.
And interface part is highly separated from functional in Panasonic cameras.
Even understanding of service mod (with three letter addreviations can help us).

PappasArts
06-04-2010, 07:25 PM
two shots is so different you really can't compare them, unfortunately. The AVCHD looks sharper to me, but I like the way the MJPEG seems to really have a better dynamic range...but I could be wrong on that and the lighting may have been really flat and uncontrasty that day and that's why that image looks that way, by comparison the AVCHD one has very harsh highlights and shadows.Are you using the Panasonic LGA7200 for your anamorphics?



I didn't like the difference in lighting either. So I just went down the street and shot them both again, at the same time...

http://i33.photobucket.com/albums/d94/PappasArts/th_Picture33copy.jpg (http://s33.photobucket.com/albums/d94/PappasArts/?action=view&current=Picture33copy.jpg)

http://i33.photobucket.com/albums/d94/PappasArts/th_Picture34copy.jpg (http://s33.photobucket.com/albums/d94/PappasArts/?action=view&current=Picture34copy.jpg)

http://i33.photobucket.com/albums/d94/PappasArts/th_Picture35copy.jpg (http://s33.photobucket.com/albums/d94/PappasArts/?action=view&current=Picture35copy.jpg)





Michael Pappas
http://www.pbase.com/Arrfilms
Anamorphic GH1 Frames
http://www.dvxuser.com/V6/showpost.php?p=2006938&postcount=1956

Vitaliy Kiselev
06-04-2010, 07:30 PM
Tesselator and other development guys

How about making a tool to find pictures and icons inside firmware?
I want small tool that allow you to travel on binary firmware image and see its contents in various modes (1bpp, 2bpp, 4bpp, 8bpp, 16bpp, 24bpp, 32bpp) on screen (plus all information, like offset, etc).
With ability to shift viewing window, go to specified offset, etc.
Window must have ability to change size as required.
Looks pretty similar to old games development.

Note that old firmware extraction tool link is www.pentax-hack.info/firmware/pan/ptool_old.zip (http://www.pentax-hack.info/firmware/pan/ptool_old.zip) .

thisisapocalypse
06-04-2010, 07:32 PM
I didn't like the difference in lighting either. So I just went down the street and shot them both again, at the same time...

http://i33.photobucket.com/albums/d94/PappasArts/th_Picture33copy.jpg (http://s33.photobucket.com/albums/d94/PappasArts/?action=view&current=Picture33copy.jpg)

http://i33.photobucket.com/albums/d94/PappasArts/th_Picture34copy.jpg (http://s33.photobucket.com/albums/d94/PappasArts/?action=view&current=Picture34copy.jpg)

http://i33.photobucket.com/albums/d94/PappasArts/th_Picture35copy.jpg (http://s33.photobucket.com/albums/d94/PappasArts/?action=view&current=Picture35copy.jpg)




Wow, okay - now we're talking - there's no question the MJPEG mode outshines AVCHD here. More detail, better DR (if only slightly and may be due to the additional detail and not really DR - I don't know for sure)

The colors in the MJPG file appear more vivid - were your exposure settings identical?

Now we just need 24p!

Cavemandude
06-04-2010, 07:41 PM
In the first two comparisons, I saw more detail in the AVCHD stills. In the motion comparison it wasn't even close, MJPEG wins.

Randy

Vitaliy Kiselev
06-04-2010, 07:49 PM
In the first two comparisons, I saw more detail in the AVCHD stills. In the motion comparison it wasn't even close, MJPEG wins.

Randy

I really suggest starting new thread and moving all this stuff concerning bitraites, comparisons, etc to it.

PappasArts
06-04-2010, 08:13 PM
In the first two comparisons, I saw more detail in the AVCHD stills. In the motion comparison it wasn't even close, MJPEG wins.

Randy


Well I decided to put on some of my good Nikon glass. Now, with no Panny anamorphic to muck it up. Here's some frames of 720P.




Michael Pappas
http://www.pbase.com/Arrfilms
Anamorphic GH1 Frames
http://www.dvxuser.com/V6/showpost.php?p=2006938&postcount=1956

PappasArts
06-04-2010, 08:15 PM
I really suggest starting new thread and moving all this stuff concerning bitraites, comparisons, etc to it.

I agree. This will cause the thread not to be focused on the research.. If I haven't said it, thank you for what your doing "tester13".



Pappas

Malko
06-04-2010, 10:40 PM
Hey, been silently following this thread since it was less than 20 pages. Great work so far! If there is a way I can help in the form of contacting folks or shooting specific test footage let me know.

Just gave the latest ptool a try in order to remove the 60i wrapper. Just two questions about it (tried using search, but no dice so sorry if this is a repeat). I noticed that when bringing the new 24p footage into final cut, the format inspector says that:

1. Field is still Upper (Odd) rather than progressive and
2. The bit rate for similar footage was 14.1 rather than 17.2 (I assume this is an average since it's just 1 number in a VBR clip?)

So I wonder, should I deinterlace this footage still? Or is it truly progressive but a bad tag?

Vitaliy Kiselev
06-04-2010, 11:10 PM
Hey, been silently following this thread since it was less than 20 pages. Great work so far! If there is a way I can help in the form of contacting folks or shooting specific test footage let me know.

See first page for all possible help (second post).



1. Field is still Upper (Odd) rather than progressive and
2. The bit rate for similar footage was 14.1 rather than 17.2 (I assume this is an average since it's just 1 number in a VBR clip?)


I can see clips using low level tools and they are all progressive if you use native 24p, without doubt.
We can still have some flag set in header, but all players play such footage fine.

To compare bitrate you need very controlled experiments.

Vitaliy Kiselev
06-04-2010, 11:28 PM
PTool 3.33 Released
MJPEG Size and fps patches added.

See tester notes on first page-second post , please!

MJPEG Size and fps patches are important to understand how MJPEG part works.

Malko
06-04-2010, 11:31 PM
Yeah, just shot something again similar with the patch and it turned out 13.8 but looked fine when stacked up next to the first one. I'll try something to max it out tomorrow just to see if the encoder's capability is the same and will report if it's not good. I'll also throw it on a timeline within a progressive sequence with XDCAM EX > ProRes to see if any issues arise with the actual picture.

Thank you so much! You have saved so many future logistical headaches with even just this single feature.

Vitaliy Kiselev
06-05-2010, 01:10 AM
Looks like 1280->848 works, but reverse produce heavy artifacts.

P.S. Problem is identified and fixed.

I am able to produce 1920x1080p30 working MJPEG (most probably just unscaled 1280 for now).
Do not see any problems with this.
New patches will be available very soon.

Isaac_Brody
06-05-2010, 01:43 AM
Looks like 1280->848 works, but reverse produce heavy artifacts.


I don't understand, are you changing the resolution down and then immediately adjusting it from 848 back to 1280? Is it just to see if you can change it?

Vitaliy Kiselev
06-05-2010, 02:04 AM
PTool 3.33 Updated without version change

MJPEG 1280m Encoder settings patches added.

How-to get working 1920p30:
MJPEG 1280p width - input 1920
MJPEG 1280p height - input 1080
MJPEG 1280p Encoder width - input 2096
MJPEG 1280p Encoder buffer - input 4821120

P.S. Resolution will be like upscaled 1280. But we need more tests on this.

Vitaliy Kiselev
06-05-2010, 02:26 AM
I don't understand, are you changing the resolution down and then immediately adjusting it from 848 back to 1280? Is it just to see if you can change it?

It was to understand some encoder parameters and LSI behavious. All I did is change size fo different modes.
And it helped already.
We know that we have "free"(no resources consumed) scaler before encoder.
So, it uppes our chances to have both 1920p24 and 1280p24 if we could switch sensor mode.
And least from MJPEG encoder side I don't see problems.

NURBS
06-05-2010, 02:41 AM
Hey Guys'

Testing again'

MjPEG 30fps - Xfps - input 25
MJPEG 1280p width - input 1920
MJPEG 1280p height - input 1080
MJPEG 1280p Encoder width - input 2096
MJPEG 1280p Encoder buffer - input 4821120
MJPEG 848p width - input 1280
MJPEG 848p height - input 720

1280p mode:

Duration : 10s 80ms
Bit rate : 25.7 Mbps
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate : 25.000 fps
Bits/(Pixel*Frame) : 0.496

Working well

848p mode:

Duration : 10s 80ms
Bit rate : 9 454 Kbps
Width : 1 280 pixels
Height : 720 pixels
Display aspect ratio : 16:9
Frame rate : 25.000 fps
Bits/(Pixel*Frame) : 0.410

Picture have a problem :) need hi buffer or something.

Regards
NURBS

Vitaliy Kiselev
06-05-2010, 02:45 AM
Will it be possible, to have MJPEG in 720/50p as well with `the better bitrate?
Or enhanced codec with better bitrate for AVCHD 720 50p?


And?
Something prevents you to use already existing patch to change MJPEG framerate?
Try it. Not 50fps, but 60fps can work.

As for AVCHD - I am working on it.
AVCHD is very complicated thing (plus requires very good hardware), and it looks that it is very tightly programmed for certain bitrates.
Plus it looks that it have some bugs in bitrate allocation resulting in sudden spikes (and this spikes can bre real cause of freezing).

Vitaliy Kiselev
06-05-2010, 02:48 AM
Hey Guys'
1280p mode:

Duration : 10s 80ms
Bit rate : 25.7 Mbps
Width : 1 920 pixels
Height : 1 080 pixels


You can use quality adjustments :-)


848p mode:
Picture have a problem :) need hi buffer or something.


Look at tester notes - I updated them.
Use 848p mode changes to test lower resolutions (change to 640x480, 320x240, etc).
As for picture problem in 848m I know how to adjust buffer for this mode also, but it have no use in this moment.

AKED
06-05-2010, 02:50 AM
And?
Something prevents you to use already existing patch to change MJPEG framerate?
Try it. Not 50fps, but 60fps can work.

OK, I will try, but here in PAL country, 60p is not ideal.
Ayway I give it a try.


As for AVCHD - I am working on it.
AVCHD is very complicated thing (plus requires very good hardware), and it looks that it is very tightly programmed for certain bitrates.
Plus it looks that it have some bugs in bitrate allocation resulting in sudden spikes (and this spikes can bre real cause of freezing).

Good luck for it. If anybody, you can do it.

Vitaliy Kiselev
06-05-2010, 02:52 AM
As for 1920 mode.
Even in 1280p upscaled version it have very good use - keying work.
As upscaling is prior encoding, so you have much better color resolution.
I believe someone asked improvements in this field not long time ago.

spamrakuen
06-05-2010, 02:56 AM
OK, I will try, but here in PAL country, 60p is not ideal.
Ayway I give it a try.

I also think that 60fps may work with current patch.
About PAL countries, 60fps will give us good slow-mo, and 60->25 conversion, although not perfect, gives far better results and motion than 30->25.

So, even not ideal, it would be very useful.

NURBS
06-05-2010, 02:57 AM
Test for fun: same settings and quality settings.

Quality settings (E1 to E4) - 720, 512, 428, 400
Table settings (E1 to E4) - 24, 24, 24, 24

Duration : 2s 880ms
Bit rate mode : Variable
Bit rate : 85.0 Mbps
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate mode : Constant
Frame rate : 25.000 fps
Bits/(Pixel*Frame) : 1.639

Working well but camera stop recording after 2 sec. ;-) quality is superb!!!!

Regards
NURBS

Vitaliy Kiselev
06-05-2010, 03:05 AM
Test for fun: same settings and quality settings.

Quality settings (E1 to E4) - 720, 512, 428, 400
Table settings (E1 to E4) - 24, 24, 24, 24



Yes it is too much.
Camera top bitrate is about 75-80Mbit (not average!).
And this settings lead to spikes in more detailed (or noisy) frames.

Try to use 1fps setting :-)
For timelapse like footage.

NURBS
06-05-2010, 03:12 AM
MjPEG 30fps - Xfps - input 60
MJPEG 1280p width - input 1920
MJPEG 1280p height - input 1080
MJPEG 1280p Encoder width - input 2096
MJPEG 1280p Encoder buffer - input 4821120
MJPEG 848p width - input 640
MJPEG 848p height - input 480

1280p mode:

Duration : 6s 0ms
Bit rate : 52.8 Mbps
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate : 60.000 fps
Bits/(Pixel*Frame) : 0.424

Working well but stop recording after 6 sec.

848p mode:

Duration : 9s 500ms
Bit rate : 21.0 Mbps
Width : 640 pixels
Height : 480 pixels
Display aspect ratio : 4:3
Frame rate : 60.000 fps
Bits/(Pixel*Frame) : 1.141

Working well ;-) next tes 120fps :) like RED

Vitaliy Kiselev
06-05-2010, 03:18 AM
It is not required to test 120fps, as it is encoder framerate, not sensor.

Test only 1280 mode with 30fps->60fps patch (do not use all else initially).
And check that it is indeed 60fps. Using some motion.

It is also fun to look how scaler work.
It gets full source picture and scales to defined resolution (look at 640x480 squeezed look).

NURBS
06-05-2010, 03:25 AM
Test for fun :bath:

MjPEG 30fps - Xfps - input 120
MJPEG 848p width - input 640
MJPEG 848p height - input 480

1280p mode:

Duration : 500ms
Bit rate : 103 Mbps
Width : 1 280 pixels
Height : 720 pixels
Display aspect ratio : 16:9
Frame rate : 120.000 fps
Bits/(Pixel*Frame) : 0.929

Working well but stop recording after 0.5 sec. :violent5:

848p mode:

Duration : 2s 500ms
Bit rate : 45.6 Mbps
Width : 640 pixels
Height : 480 pixels
Display aspect ratio : 4:3
Frame rate mode : Constant
Frame rate : 120.000 fps
Bits/(Pixel*Frame) : 1.238

Working well but stop recording after 2.5 sec. :happy:

NURBS
06-05-2010, 03:40 AM
Try to use 1fps setting :-)
For timelapse like footage.

1fps - Can't working

Camera freezing when i want to stop recording, maybe I not patient :) but two minutes too long time for me. :Drogar-Happy(DBG):

Vitaliy Kiselev
06-05-2010, 03:47 AM
1fps - Can't working

Camera freezing when i want to stop recording, maybe I not patient :) but two minutes too long time for me. :Drogar-Happy(DBG):

5fps must work in this case :-)
Anyway it is not too important.
Compare 60fps and 30fps 1280 footage.
See if motion is actually shot at 60fps.
This is priority task.

bumkicho
06-05-2010, 05:22 AM
MJPEG - 60fps

Quality settings (E1 to E4) - 352, 220, 200, 184
Table settings (E1 to E4) - 24, 24, 24, 24

Motion is actually 60fps. Stops recording at 3 seconds. I use class 4 card. I saw NURBS was able to record 6 seconds of 60fps. Perhaps he is using a card with faster write speed.

jobless
06-05-2010, 05:29 AM
Sorry guys No real 60p :(
I tested 720 and 848 at 60 p and it's just speedup footage...
So no slomotion is possible...

Link:
http://www.nikolicnemanja.com/test3.rar

NURBS
06-05-2010, 05:37 AM
Very interesting thing.

848p mode:

MJPEG 848p width - input 848
MJPEG 848p height - input 320

The Picture little squeezed :)

bumkicho
06-05-2010, 05:40 AM
Sorry guys No real 60p :(

My footage looked like 60fps in after effects. Can someone download the footage from below link and tell me if it is real 60fps or not?

12315420

NURBS
06-05-2010, 05:56 AM
Very interestin thing too:

MJPEG mode let do change resolution ;-)

Duration : 10s 80ms
Bit rate mode : Variable
Bit rate : 28.9 Mbps
Width : 2 048 pixels
Height : 854 pixels
Display aspect ratio : 2.35:1
Frame rate mode : Constant
Frame rate : 25.000 fps
Bits/(Pixel*Frame) : 0.661

Like 2K SUPER 35 / SUPER TECHNISCOPE :bath:

But have some problem ;-)

bumkicho
06-05-2010, 06:04 AM
I like what I am seeing from

MJPEG fps tesing - 24fps
Quality settings (E1 to E4) - 352, 220, 200, 184
Table settings (E1 to E4) - 24, 24, 24, 24


However, I would like someone to download the footage and test it to confirm.

12315734

NURBS,

I was going to do the aspect ratio test next, but you keep at it and let us know. Great work testing!!

jobless
06-05-2010, 06:11 AM
24p and 25p still skipping frames...

spamrakuen
06-05-2010, 06:15 AM
24p and 25p still skipping frames...

Strange behaviour. I thought that 60fps would duplicate frames, but it didn't. Hope tester13 now understands better how this works, thanks to these tests.

spamrakuen
06-05-2010, 07:00 AM
Sorry guys No real 60p :(
I tested 720 and 848 at 60 p and it's just speedup footage...
So no slomotion is possible...

Link:
http://www.nikolicnemanja.com/test3.rar

Your camera is in PAL or NTSC mode?

jobless
06-05-2010, 07:03 AM
your camera is in pal or ntsc mode?

ntsc

Rabot
06-05-2010, 08:08 AM
Hi people, I tried something.

MjPEG 30fps - Xfps - input 60
MJPEG 848p width - input 1280
MJPEG 848p height - input 544

For the MJPEG quality; I doubled every value and for the table I wrote 24 on the four.

848p mode:

Duration : 8s 0ms
Bit rate : 21 Mbps
Width : 1280 pixels
Height : 544 pixels
Frame rate : 60.000 fps

Everything was ok execpt this;

http://old.akwd.com/video/P1160986.MOV
(for fast people, please let it load)

(also I used QT for the info, if you need something more specific I will load it, but I'm a mac user)


DID SOME OTHER TEST, IT IS JUST SQUEEZED, I DOESNT NOT REALLY SHOOT ANAMORPHIC OR ANYTHING.

sunsurfh2o
06-05-2010, 09:30 AM
Definitely on to something, here. Looks to me like the real deal. Only camera stops recording in less than 3 secs. For my purposes 1080p24 in either format, 720p60 in MJPEG (well 1080 would be nice here too!) would be golden. I shoot fast action most of the time, with pans when necessary to follow the athletes and AVCHD just doesn't hold up there. Now if we can just find quality/table settings that optimize without premature cessation we'd be golden!

MJPEG - 60fps

Quality settings (E1 to E4) - 352, 220, 200, 184
Table settings (E1 to E4) - 24, 24, 24, 24

Uploaded orig files and rendered slomo here:

http://rapidshare.com/files/395591141/MJPEG_60.zip.html

Only 10 downloads available according to Rapidshare so better get on it if it's important to you! I'll try and upload elsewhere, too.

jeracravo
06-05-2010, 09:47 AM
I've tried:

MjPEG 30fps - Xfps - input 25
MJPEG 1280p width - input 1920
MJPEG 1280p height - input 1080
MJPEG 1280p Encoder width - input 2096
MJPEG 1280p Encoder buffer - input 4821120
MJPEG 848p width - input 1280
MJPEG 848p height - input 720

and when I stoped recording the camera froze and asked me to turn it off then on again. after this, it rendered an useless MOV file.

I tried with a SanDisk Class 2 Ultra II and with a Class 6 Patriot both had the same problem and I was recording for about a minute.

Any ideas?

jobless
06-05-2010, 10:01 AM
Definitely on to something, here. Looks to me like the real deal. Only camera stops recording in less than 3 secs. For my purposes 1080p24 in either format, 720p60 in MJPEG (well 1080 would be nice here too!) would be golden. I shoot fast action most of the time, with pans when necessary to follow the athletes and AVCHD just doesn't hold up there. Now if we can just find quality/table settings that optimize without premature cessation we'd be golden!

MJPEG - 60fps

Quality settings (E1 to E4) - 352, 220, 200, 184
Table settings (E1 to E4) - 24, 24, 24, 24

Uploaded orig files and rendered slomo here:

http://rapidshare.com/files/395591141/MJPEG_60.zip.html

Only 10 downloads available according to Rapidshare so better get on it if it's important to you! I'll try and upload elsewhere, too.

You are wrong here. You got slomo from 30p to 24p.
As I said 60p doesn't work.
Try to shoot same scene with Avchd 720 60fps and you'll see the difference.

sunsurfh2o
06-05-2010, 10:37 AM
You are wrong here. You got slomo from 30p to 24p.
As I said 60p doesn't work.
Try to shoot same scene with Avchd 720 60fps and you'll see the difference.

but my timeline/clip duration went from 7 sec to 18, roughly... 2.5x consistent with math on 60p -> 24p. What am I missing?

PDR
06-05-2010, 10:42 AM
I believe jobless is correct. I think what's happening here is the time scale is wrong

The clips you uploaded are actually 30fps and 6 seconds long, not 3 seconds long. The only thing that is changed is the clip header for FPS field

So although there are 180 unique frames per clip, it should be over 6 seconds = 30FPS . When the header says 60FPS , your media player/software is instructed to playback at 60FPS, that's why it looks "sped up"

It wasn't as clear in the example with the kid, because that video was just bobbing motion. It's more clear in your dog example.

GrgurMG
06-05-2010, 12:43 PM
The ability to scale footage resolution is very interesting. Like tester13 said, being able to upscale before encoding is nice stepup for certain purposes... although note your also squeezing a higher resolution into the same available bitrate... so it's give and take.

Most should realize at this point the scaling function applied toward aspect ratio changes seems of limited use.. because all it is is scaling.. not cropping or letting you use more sensor or anything like that. Possible legit uses however would be to pre-encode-stretch out footage if your actually using an anamorphic adapter... also the 1440x1080 might have some use if you plan on mixing the footage in a mainly HDV workflow/timeline, as it would save time and some NLEs don't scale well by default.. That's assuming the GH1 scales decently... I haven't analyzed the test footage that well.

sammysammy
06-05-2010, 05:33 PM
Ok i just tested the 60p mpeg patch with these settings , i have a class 4 16gb toshiba card ..

MJPEG - 60fps

Quality settings (E1 to E4) - 300, 225, 175, 150
Table settings (E1 to E4) - 24, 24, 24, 24

result- the gh1 records only for 4 seconds on the gh1 (but in reality its 8 seconds,meaning for each sec the gh1 shows it recorded,in real life time its 2 seconds)
-played the file back, it was 4 seconds speed it up, for normal results the file should be 8 seconds with normal speed play back
-sony vegas sees the file as 60p

here is the original clip (straight from gh1, 60p at 125ss )
http://www.sendspace.com/file/730b8a

here is the file i edited in vegas (to get 24p from 60p....instead of .4 play back, i had to use .2 because the original file was 8 sec not 4 sec)
http://www.sendspace.com/file/nmu2mw

Vitaliy Kiselev
06-05-2010, 09:22 PM
I suggest to move all posts not related to testing according to tester notes to separate thread.
As I already said our target is systematic testing.
And most posts here have no use for me (but can have some use for guys who want to find some settings for their task).

We already miss some important things due to flood of messages.
Like AVCHD Compression testing in previous verion was not complete, 1920 testing was done correctly, but 1280 in all quality settings had been missed.
Next important thing is that you need to perform RESET operation from menu after each firmware update. Because some settings are stored in EEPROM (this includes some AVCHD settings).

PappasArts
06-05-2010, 10:07 PM
I suggest to move all posts not related to testing according to tester notes to separate thread.
As I already said our target is systematic testing.
And most posts here have no use for me (but can have some use for guys who want to find some settings for their task).

We already miss some important things due to flood of messages.
Like AVCHD Compression testing in previous verion was not complete, 1920 testing was done correctly, but 1280 in all quality settings had been missed.
Next important thing is that you need to perform RESET operation from menu after each firmware update. Because some settings are stored in EEPROM (this includes some AVCHD settings).


I agree, I went to read up on your post that 1080p 30/new Ptool, and it was a flood in here.

This thread I set up is a perfect area for footage-analysis discussions that are not related to the actual R&D you and the testers are doing.

http://www.dvxuser.com/V6/showthread.php?p=2008984#post2008984

Pappas

Isaac_Brody
06-05-2010, 10:11 PM
I've cleaned up a bunch of posts. I will be deleting posts that are off topic here. The more people post random questions, the more results get buried, and the slower progress is made. So create new threads if you want to discuss results.

EDIT: Use Pappas thread for results and discussion/requests.

:dankk2:

Vitaliy Kiselev
06-05-2010, 10:22 PM
1920p25 and 1920p24 sensor patches are upcoming.
Due to risky nature I won't be testing them myself.
But risk here is much smaller compared to our early third-party battery patches.
I'll change sensor mode setting from 30fps to 1920 25fps and 2397fps (no 2400fps setting is present).

PappasArts
06-05-2010, 10:28 PM
Keep the 30P 1080 still there in the Ptool. There a lot of uses for 1080 @ 30FPS. IMHO it's the way the industry should go. 30 is excellent for television work.

Car3o
06-05-2010, 10:32 PM
1920p25 and 1920p24 sensor patches are upcoming.
Due to risky nature I won't be testing them myself.
But risk here is much smaller compared to our early third-party battery patches.
I'll change sensor mode setting from 30fps to 1920 25fps and 2397fps (no 2400fps setting is present).

why is so much risk you're not willing to try on a camera that was bought for testing purposes?

Vitaliy Kiselev
06-05-2010, 10:50 PM
PTool 3.33 Updated without version change

MJPEG Sensor mode patch added.

Vitaliy Kiselev
06-05-2010, 10:53 PM
why is so much risk you're not willing to try on a camera that was bought for testing purposes?

Sorry, I don't get that you are trying to say or ask here.

Yes, camera was bought for testing purposes, but not for risky patches.
And this was clearly indicated from the start.
Because bricking or malfunction induce very big delay in our project.

Car3o
06-05-2010, 11:15 PM
Sorry, I don't get that you are trying to say or ask here.

Yes, camera was bought for testing purposes, but not for risky patches.
And this was clearly indicated from the start.
Because bricking or malfunction induce very big delay in our project.

Alright, I mean it's tougher to have people who use their camera's for business or leisure activities vs a camera that's used primarily for testing to install a patch that may brick a camera. I suppose there are some out there who have nothing to lose.

Vitaliy Kiselev
06-05-2010, 11:25 PM
Alright, I mean it's tougher to have people who use their camera's for business or leisure activities vs a camera that's used primarily for testing to install a patch that may brick a camera.

First, we have third-party battery patch that went through few iterations.
And this patch was 10x times more risky (as it changes system startup code).
MJPEG Sensor patch can be risky, yes.
But bricking probability is very low generally.
I played with some sensor settings (changing 1920 mode) and nothing happend.
I believe that you sometimes forget about our projects status.
This is not work project, this is free will and community project.

Ozpeter
06-05-2010, 11:43 PM
I get the impression that tester13's geographical location may not enable any kind of quick repair to the 'project GH1' if bricked. Others may be more favourably placed. But of course, if testers are depending on daily use of a GH1 to earn a living, that's another problem.

PappasArts
06-06-2010, 12:02 AM
Sorry, I don't get that you are trying to say or ask here.

Yes, camera was bought for testing purposes, but not for risky patches.
And this was clearly indicated from the start.
Because bricking or malfunction induce very big delay in our project.

No sense in damaging your camera. This GH1 is vital for your research. 30FPS will work wonderfully, until a safe passage to 24FPS mode can be established.

That said 30FPS format is one of my favorite and is very important; as I worked with Iwerks, and 60FPS showscan. 30 always felt right. So no matter what, this is a win win situation.

Tester, could you please send me info so I could read up on how the Mpeg Quality settings (E1 to E4) Table settings (E1 to E4) work. I would like to stop the VBR on low end. so, no matter if the cap was on the lens it wouldn't dip below 45Mbits. A range of 45 to 80mbits top is very good. The lower would lend to very good compressed shadows and dark areas not having it fall below 40.


Pappas

Vitaliy Kiselev
06-06-2010, 12:17 AM
That said 30FPS format is one of my favorite and is very important; as I worked with Iwerks, and 60FPS showscan. 30 always felt right. So no matter what, this is a win win situation.

Please, stop mentioning 30fps mode as it is dying.
Use new patch and look it help hint.
It is still in place and will remain so.


Tester, could you please send me info so I could read up on how the Mpeg Quality settings (E1 to E4) Table settings (E1 to E4) work.

Sorry, I can't you send this info, as I don't have it in the form you request.

Vitaliy Kiselev
06-06-2010, 01:01 AM
Testing news.

After some more analysys I tested sensor patch.
1920 settings,
MJPEG 24fps
Plus sensor mode changed to 7.
For me it looks like FullHD MJPEG at 24fps is working, but I not analyzed it carefully, as main target was to test risk.

You can test it yourself at any time.
With zero risk (status updated to "Tested").

Kholi
06-06-2010, 01:10 AM
Time to get a GH-1 AGAIN.

Amazing work, guys. I feel like a bottom feeder, but stellar.

Already beastly camera made even more beastly. SD out is going to really be nuts.

Exciting.

Lexcalcin
06-06-2010, 01:17 AM
I tried mpeg 720p50 at 83 mbits on a moving fan, its works only for 6s, but the framerate can be increased without damage, though it can't be used for serious purpose.

P1050781.MOV (http://larochebay.free.fr/repaire/P1050781.MOV) 06-Jun-2010 08:59

right click + save as

Vitaliy Kiselev
06-06-2010, 01:20 AM
I tried mpeg 720p50 at 83 mbits, its works only for 6s, but the framerate can be increased without damage, though it can't be used for serious purpose.

Please, read other posts before posting.
Your 6s limit is due to too high bitrate. It is known.
And about framerate also.

Plus all posts not related to sensor patch testing will be moved to other thread.

Kholi
06-06-2010, 01:32 AM
I have a question, Tester:

What is the highest class card that would handle the bitrate? And, is it beneficial at all to spend the money just to run a high bitrate 1080/24P?

Thanks for all of your hard work?

What's next on the list? Working on AVCHD? Video out?

Vitaliy Kiselev
06-06-2010, 01:50 AM
What is the highest class card that would handle the bitrate? And, is it beneficial at all to spend the money just to run a high bitrate 1080/24P?


Camera itself have limit at around 80Mbit (and this is not average bitrate!).
If you look at GH1 from MJPEG side, you'll see that Panasonic need only to improve data throughput of SD controller (I say about 4x times), MJPEG encoder (2x times) and increase buffer size (3x). All is easy and cheap to do.
And they'll have dream of many guys (200-250Mbit MJPEG is quite good :-) ).


What's next on the list? Working on AVCHD? Video out?

I provided plan some time ago.
Today we need information and tools.
I wrote about it before, please search posts by user name.
Otherwise we'll hit the wall.

Vitaliy Kiselev
06-06-2010, 01:58 AM
Found interesting consequence of sensor patch.
Manual zoom enlargement do not work (weird picture).
Looks like they use high resolution of RAW data in this mode.
Funny.

spamrakuen
06-06-2010, 02:04 AM
Hi tester13,

One question about new sensor patch. If I'm not wrong, previous 1080p tests were result of scaling a lower resolution image. With this new sensor patch, source image is same resolution as in AVCHD FHD? It is "true" 1080p?

Thanks,

Vitaliy Kiselev
06-06-2010, 02:09 AM
Hi tester13,

One question about new sensor patch. If I'm not wrong, previous 1080p tests were result of scaling a lower resolution imagen. With this new sensor patch, source image is same resolution as in AVCHD FHD? It is "true" 1080p?



You can set sensor output to be same as in all AVCHD modes.
But serious and systematic tests are required.
Including resolution tests.
As RAW resolution must be good even in 30fps mode.

P.S. Anyone noticed that we have special FHD icon for MJPEG in playback mode? :-)

sammysammy
06-06-2010, 03:01 AM
yes!wow..

Kholi
06-06-2010, 03:08 AM
You can set sensor output to be same as in all AVCHD modes.
But serious and systematic tests are required.
Including resolution tests.
As RAW resolution must be good even in 30fps mode.

P.S. Anyone noticed that we have special FHD icon for MJPEG in playback mode? :-)

Can you clarify a bit more on 1080P? Is it an actual 1080 or upscaled from 720? I suppose if it looks good it doesn't even matter.

So far the 720 is looking amazing. The 1080 grabs I've seen do look that much more resolute coming straight out of camera.

I suppose this should make testing PL glass a bit more fun.

NURBS
06-06-2010, 03:11 AM
Testing new ptool: :bath:

MJPEG sensor mode:

6 - 50p working well (focus assistant can't work)

8 - 25p working well (focus assistant can't work)

I don't see any difference for old mode.

PS: Please guys who can see difference at this pach, test it (it's not risky).

Regards
NURBS

spamrakuen
06-06-2010, 03:18 AM
Testing new ptool: :bath:

MJPEG sensor mode:

6 - 50p working well (focus assistant can't work)

8 - 25p working well (focus assistant can't work)

I don't see any difference for old mode.

Regards
NURBS

One question, if you switch to AVCHD, focus assistant works?
Or in other words, can this be a workaround for focusing;
1 - Enable AVCHD and focus.
2 - Then enable MJPEG and record.
3 - Stop recording
4 - Go to step 1

NURBS
06-06-2010, 03:25 AM
No, focus assistant can't work on all camera modes. It's no good, because focus assistance very important for manual lens.

Vitaliy Kiselev
06-06-2010, 03:35 AM
6 - 50p working well (focus assistant can't work)
8 - 25p working well (focus assistant can't work)
I don't see any difference for old mode.


We need really systematic tests.
Words "don't see any difference" are not enough.
You must shoot smooth pans.
Or some falling objects.
So skipped frames can be seen clearly (and please, do not touch quality patches!).
On my small test footage I do not see skipped frames on 1920p24, but I can be wrong.

Vitaliy Kiselev
06-06-2010, 03:36 AM
No, focus assistant can't work on all camera modes. It's no good, because focus assistance very important for manual lens.

Please, be patient.
Of course it don't work in all modes as we change mode inside sensor setup procedure.
With time we'll be able to change it on much higher level.

NURBS
06-06-2010, 03:39 AM
Very interesting thing.

When you recording in AVCHD mode - TV out show brown color screen. :-)

PS: All right I'm testing afresh this modes ;)

jobless
06-06-2010, 03:45 AM
Guys something strange happening with my camera and new patch....
I applied num 7 23,976 sensor output and mjpeg resize to 1080...
My camera stops recording after 1 sec , but after that when I restarted the camera LCD behave strange. And I can't go to play mode any more... so I can't rollback to previous firmware...
Tester please HELP ... I think I broke my GH1....

NURBS
06-06-2010, 03:47 AM
Format you SD card or delete all files. And try again.

PappasArts
06-06-2010, 03:48 AM
Can the 24fps patch be applied to the 720P as well, and does the zoom feature not work with this too.

Does the 1080 30p patch mode also not have the zoom feature as well?


.

Vitaliy Kiselev
06-06-2010, 03:49 AM
Guys something strange happening with my camera and new patch....


Most probably you forget encoder patches (width and buffer).
Delete all contents from your SD card and all must be normal with play mode.
My camera working fine with 7 mode patch.

jobless
06-06-2010, 03:52 AM
I applied Width and buffer.... But I increase bitrate for mjpeg. So it's possible that bitrate causing the problem...
Howevar I formated my card and now I can go to play mode...
So no problem for now ...
Sorry...

Vitaliy Kiselev
06-06-2010, 03:53 AM
Very interesting thing.

When you recording in AVCHD mode - TV out show brown color screen. :-)

PS: All right I'm testing afresh this modes ;)

It looks that we still have frame skipping (in 24fps mode), but screen looks quite weird even during smooth pans as 24fps.
It can mean only that MJPEG is actually 60fps mode and not patched 30fps mode.
So 50fps and 60fps must be also tested.

spamrakuen
06-06-2010, 04:01 AM
It looks that we still have frame skipping.

Maybe sensor mode is changing when record button is pressed, so the sensor patch only is working while not recording.

NURBS
06-06-2010, 04:01 AM
I'm testing again this mode, no changing.

Vitaliy Kiselev
06-06-2010, 05:07 AM
I did few more tests.
And indeed it is wrong mode (this one is used for magnification).

Vitaliy Kiselev
06-06-2010, 05:37 AM
PTool updated without version change.

Added sensor 3-8 modes adjustments.

Note.
Here I try to describe why this patches can be risky and why we need to test them.

We are patching sensor setup function. Normally it is called upon each mode change.
Some of this modes is default liveview mode (must be 60fps), I believe it is 0 mode not included in patches list. But it can be also 3-5 modes. Worst part that they can be used for MJPEG.

spamrakuen
06-06-2010, 05:45 AM
Here I try to describe why this patches can be risky and why we need to test them.


Maybe risky patches should be tested first by one tester only. This way, only one camera may brick, and money can be collected in order to buy a new one for that tester.

Vitaliy Kiselev
06-06-2010, 05:49 AM
Maybe risky patches should be tested first by one tester only. This way, only one camera may brick, and money can be collected in order to buy a new one for that tester.

Just write here before you start testing particular sensor patch (1920 patches must be pretty safe anyway).

spamrakuen
06-06-2010, 05:53 AM
Just write here before you start testing particular sensor patch (1920 patches must be pretty safe anyway).

That's a good idea. People should post here before testing one of those risky patches (I don't have a GH1, anyway).

NURBS
06-06-2010, 06:22 AM
Tester how to do change in these modes? Please write some example. Thanks.

GrgurMG
06-06-2010, 07:01 AM
First page... 2nd post (for testers). Tester13 frequently updates those two posts and you should check their frequently for answers to many questions.

Vitaliy Kiselev
06-06-2010, 07:10 AM
Tester how to do change in these modes? Please write some example. Thanks.

Just look at help hints for codes. Patch change one mode (3rd for example to 7th) to other. I suggest to change code to 7 at first.
You can apply multiple patches to save time (and increase risk :-) ).

Vitaliy Kiselev
06-06-2010, 08:23 AM
So this applies to 24p as well as 30p?


It is frame encoder settings, so framerate have no meaning here.


Set sensor mode 7 AND encoding settings (above) for best results?

I don't know that you mean under "set sensor mode to 7".
As in latest version have Sensor Mode 3 - Sensor Mode 8 patches.
And this is all risky unetsted patches, with goal to find MJPEG mode.

RandyQ
06-06-2010, 11:24 AM
Very interesting thing.

When you recording in AVCHD mode - TV out show brown color screen. :-)


What do you mean by this, NURBS? Is it just a flat brown color, or a brown tinted video output of the footage you are shooting?

Car3o
06-06-2010, 03:36 PM
is there no way to make MJPEG CBR instead of VBR?

Isaac_Brody
06-06-2010, 04:20 PM
Mjpeg is constant bitrate. You can even check yourself in quicktime when playing and enabling movie inspector.

sammysammy
06-06-2010, 04:40 PM
here is a small test clip of the sensor patch (7) 24p with slow pans .. i dont see skipped frames but, to me the image quality takes a hit, i see artifactscreen while recording in this mode...thanks tester for your hard work!

http://www.sendspace.com/file/ftmwxt

srry for not posting that info, i checked the following version change ,third party battery,mpeg encoder 30p to 24p,mpeg sensor mode (entered 7),1280m width 1920,1280m height 1080,encoder width2096,encoder height 4821120

i hope i tested for the right settings

Isaac_Brody
06-06-2010, 05:26 PM
Just a reminder to keep this thread geared for testing and send all questions here:

http://www.dvxuser.com/V6/showthread.php?t=212263&page=2

And if you want to discuss results or post stills:

http://www.dvxuser.com/V6/showthread.php?t=212609

Vitaliy Kiselev
06-06-2010, 09:14 PM
Important request.

I want to remind you that we need tools to move forward at good pace.

First tool.
Small utility with good interface with ability to view firmware image and find icons and other images inside (visually showing images and offsets).
It do not requare any special knowledge, just C/C++.

Second.
We really need Panasonic compilers. They are available as parts of AM3 (MN103S) boards.
So, if you work on any electronics related company you can order this items.

Third.
Panasonic service software can be very helpful.
Go and ask your service center is they have it.
Most of the service software leaks are made exactly this simple way.
We are not intended to make it public.
So, if you want to provide it and remain anonymous you can be sure that it'll be used for development only.

GrgurMG
06-06-2010, 09:21 PM
I would make it a point to reiterate the fact that when it comes to what certain settings may or may not be actually doing (for instance the sensor settings)... Tester13 may not even be the best person to ask directly because he's looking for those here to methodically test these different settings and help extrapolate what these settings are doing based on the footage. That is the whole point of this thread, to carefuly test things and report back... as appose to him releasing a new patch and having to figure for himself and explain to us what it's doing.

That being said... if people want to start comparing real resolution vs fake in the possible sensor modes... might I suggest we start getting a bit more technical and shoot things like resolution charts to look for differences. Shooting plant life is great.. but if your simply looking for differences in fidelity between footage.. it's much more straighforward to look at the lines on a test chart. If anything.. I should know within a couple day if I'm exchanging my GH1 or not and either way will be able to contribute to testing myself shortly thenafter.

Here's a link to the kind of chart I mean, for those who are unsure....

http://www.dpreview.com/news/0510/05101404newreschart.asp

http://www.screencapturenews.com/wp-content/uploads/resolution-chart-original.jpg

Vitaliy Kiselev
06-06-2010, 09:25 PM
We already have good chart:
www.pentax-hack.info/firmware/pan/ISO_12233-reschart.pdf (http://www.pentax-hack.info/firmware/pan/ISO_12233-reschart.pdf)

I'll add link to first page.

Isaac_Brody
06-06-2010, 09:32 PM
I looked up the service center, looks like there's only one authorized in the US.

http://img405.imageshack.us/img405/2973/picture2xw.png

Who feels lucky enough to call. :happy:

Might have better luck with Europe.

GrgurMG
06-06-2010, 11:10 PM
We already have good chart:
www.pentax-hack.info/firmware/pan/ISO_12233-reschart.pdf (http://www.pentax-hack.info/firmware/pan/ISO_12233-reschart.pdf)

I'll add link to first page.

Yup, that works. Similar to the one in my first link. The dpreview chart is actually cleaner and higher resolution.. but for video resolutions the pentax-hack.info one should suffice.


I looked up the service center, looks like there's only one authorized in the US.

I have a feeling anyone we'd get on the phone would either be clueless or worried about getting in trouble for forwarding us to someone whose not clueless. It's worth a shot though.. someone with a really good poker fa..c.... erm... voice. Maybe I'll take a crack at it tommorrow if I have time.

We already have the actual service manual, right?

spamrakuen
06-07-2010, 12:16 AM
here is a small test clip of the sensor patch (7) 24p with slow pans .. i dont see skipped frames but, to me the image quality takes a hit, i see artifactscreen while recording in this mode...thanks tester for your hard work!


I can see skipped frames :(
Anyone else can?
Which sensor modes did you changed?
All of them? :shocked:



srry for not posting that info, i checked the following version change ,third party battery,mpeg encoder 30p to 24p,mpeg sensor mode (entered 7),1280m width 1920,1280m height 1080,encoder width2096,encoder height 4821120


You didn't changed IQ settings (Quality and Table), that can explain the artifacts.

John Caballero
06-07-2010, 01:31 AM
Just a thought and I have no idea if it will make sense but the same way tester13 is trying to upscale mjpeg to 1080p can the AVCHD 1080p24 with no pulldown patch could be downscaled to 720p24? makes sense?

sammysammy
06-07-2010, 03:53 AM
spamruken, i didnt know that i need to patch the Quality and Table settings for the sensor patch)7) ,what values should i test ?

spamrakuen
06-07-2010, 04:58 AM
spamruken, i didnt know that i need to patch the Quality and Table settings for the sensor patch)7) ,what values should i test ?

I have no idea. You can check lastest tests (a few pages ago in this thread) for reference values, or you can also ask here;
http://www.dvxuser.com/V6/showthread.php?t=212609

We aren't sure how these values work, that's why testing them are for.

Which sensor modes did you changed?

axetota
06-07-2010, 06:47 AM
Hello. I tested sensor modes 3-5. I first enabled MJPEG 30fps->24fps and then sequentially and individually switch modes 3-5 to mode 7. They all skip frames. When switching mode 4 to 7 I get a funky picture during recording. It is a cropped, with a purple bar at the bottom. some frames do not even have a picture.

Also, I tried switching mode 3 to 7 but instead of 30fps->24fps patch, I used the one below it (30fps->xfps) and entered the value as 24. In this case the recording is duplicating frames, not just skipping them.

I did not touch anything that has to do with frame resizing or quality settings.

Isaac_Brody
06-07-2010, 09:48 AM
Are any of these sensor modes 24P with pulldown added? Has anyone tried these sensor modes without using a 24P patch and trying to remove pulldown from the 30P material?

Vitaliy Kiselev
06-07-2010, 10:35 AM
Hello. I tested sensor modes 3-5. I first enabled MJPEG 30fps->24fps and then sequentially and individually switch modes 3-5 to mode 7. They all skip frames. When switching mode 4 to 7 I get a funky picture during recording. It is a cropped, with a purple bar at the bottom. some frames do not even have a picture.

Also, I tried switching mode 3 to 7 but instead of 30fps->24fps patch, I used the one below it (30fps->xfps) and entered the value as 24. In this case the recording is duplicating frames, not just skipping them.

Please, do systematic testing for each mode.

As I understand mode 4 is actually MJPEG mode.
Please, provide resulted MJPEG movie files.
You also need to test all AVCHD modes with 4->7 patch.
Next, try to change 4 to other modes (not 7 and 4, of course).

Vitaliy Kiselev
06-07-2010, 10:36 AM
PTool 3.34 Released
AVCHD 1080p24/25->720p24/25 patch added (tested and working).
Sensor mode patches added (officially :-) ).
MJPEG Size 1280m->1920m and MJPEG Encoder 1280m->1920m patches added (for use friendly way to set 1920x1080 MJPEG mode).

alignment1
06-07-2010, 10:44 AM
MOVED

http://www.dvxuser.com/V6/showthread...=212263&page=2 (http://www.dvxuser.com/V6/showthread.php?t=212263&page=2)

Isaac_Brody
06-07-2010, 10:51 AM
Just a reminder to keep this thread geared for testing and send all questions here:

http://www.dvxuser.com/V6/showthread...=212263&page=2 (http://www.dvxuser.com/V6/showthread.php?t=212263&page=2)

And if you want to discuss results or post stills:

http://www.dvxuser.com/V6/showthread.php?t=212609

The more questions you ask here, the slower progress is made. So please leave this thread for testing results and related discussion, otherwise I will delete your posts.
__________________

Vitaliy Kiselev
06-07-2010, 10:53 AM
So is this true AVCHD 1080p24 w/ no pulldown?
and are we still not able to view in playback on cam?

It is true progressive 720p24 (or 25 in PAL).
And it if viewable on camera without problems, just tested.
So, it looks that 1080p24 viewing may be fixable.

tyampel
06-07-2010, 10:56 AM
So is this true AVCHD 1080p24 w/ no pulldown?
and are we still not able to view in playback on cam?

Great achievements Tester13!!:2vrolijk_08:
I am afraid it is too early to open the wine bottles ;)
What tester13 meant is that his single step patch is working, so testers won't have to go through multiple set up steps.
The camera works as before.

tester13 edit: guys - please read my previous message carefully. Otherwise you mislead many thread readers.

spamrakuen
06-07-2010, 11:10 AM
Sensor mode 6 without MJPEG Encoder frame rate adjustment, causes no difference to the image. MJPEG mode seems to ignore sensor mode. What does sensor mode affect, AVCHD only?

Axetota tested mode 4 and got a funky picture, so ins't ignored by MJPEG.

alignment1
06-07-2010, 11:13 AM
I am afraid it is too early to open the wine bottles ;)
What tester13 meant is that his single step patch is working, so testers won't have to go through multiple set up steps.
The camera works as before.

tester13 edit: guys - please read my previous message carefully. Otherwise you mislead many thread readers.

now I'm confused- can people come over here and explain:

http://www.dvxuser.com/V6/showthread.php?t=212263&page=2

cheers

Vitaliy Kiselev
06-07-2010, 11:20 AM
Sensor mode 6 without MJPEG Encoder frame rate adjustment, causes no difference to the image. MJPEG mode seems to ignore sensor mode. What does sensor mode affect, AVCHD only?

Sorry, but I do not understand anything in such form.
Please report all patches used and your exact testing steps, pllus results.
Plus, look at early posts by other users - Sensor mode 4 patch seems to be match for MJPEG mode change.

spamrakuen
06-07-2010, 11:38 AM
Sorry, but I do not understand anything in such form.
Please report all patches used and your exact testing steps, pllus results.
Plus, look at early posts by other users - Sensor mode 4 patch seems to be match for MJPEG mode change.

If sensor mode 4 is the sensor mode used for MJPEG, and since mode 7 (1080p@24fps) doesn't seem to work, an interesting test could be setting 6 (720p@50fps) for sensor mode 4.

Mode 6 is the same resolution than mode 4, but has 50 fps. Only fps changes. If it works, we could have a 720p@50fps MJPEG mode. Also, if then we set FPS for MJPEG encoder to 25 instead of 50fps, we can have a 720p@25fps mode.

Maybe the only reason mode 7 isn't working in Axetota test is because of the 1080p resolution.

Vitaliy Kiselev
06-07-2010, 11:39 AM
I want to clarify some things.

Any questions concerning newly added patches or old patches in testing state go to this thread.
Any general questions to sticky FAQ topic.
And other footage discussion to "high bitrate" topic.

Vitaliy Kiselev
06-07-2010, 11:42 AM
Mode 6 is the same resolution than mode 4, and has 50 fps. Only fps changes. If it works, we could have a 720p@50fps MJPEG mode. Also, if then we set FPS for MJPEG encoder to 25 instead of 50fps, we can have a 720p@25fps mode.

And why do you think that it is same resolution?
Yes, after scaler resolution is the same.
But as far as I understand (today) RAW data resolution is different.
This is like 720p60 and new 720p24 AVCHD, second looks like same resolution, but in reality it is not so simple as it is scaled from much higher RAW resolution.

spamrakuen
06-07-2010, 11:45 AM
And why do you think that it is same resolution?
Yes, after scaler resolution is the same.
But as far as I understand (today) RAW data resolution is different.
This is like 720p60 and new 720p24 AVCHD, second looks like same resolution, but in reality it is not so simple as it is scaled from much higher RAW resolution.

Argh! You're right. I forgot that (many pages ago you told us).
Sorry, maybe isn't a good idea as I though :lipsrseal

Vitaliy Kiselev
06-07-2010, 11:51 AM
New 720p24 mode also shows us real mud reason in 1080 modes.

http://www.dvxuser6.com/uploaded/60298/1275932955.jpg

Notice how P frame size (blue bars) increases during movement.
Even here it looks like it hits some wall.

But in 1080p even still P frame size is already near top limit.
So, any movement causes hit to the wall and big overcompression.
Looks like idiotic reason to comply to 17Mbit standard rate caused all this mess.

abasfly
06-07-2010, 12:00 PM
I know not everyone is interested but tried the patch on the gf1, the 30min removal worked as expected :) but in 720 mode the gf1 is different as the gh1, the gf1 does not do 50fps it does 25 (pal) the thing is that it is wrapped in 50 frames, I dont know if it is interlaced or what but my timeline is 25p works perfect.
so I am not clear what the 720p 50 to 25 patch is doing.

Im buying some batteries to test the patch.

keep up the good work cant wait for recording in M mode

JerryB
06-07-2010, 12:00 PM
With ptools ver3.34 Tester13 added native 24p in both 1080 and 720 AVCHD and states it is a new patch and is "tested and working" and plays back in cam....

Does this mean that we have true native 24p without skipped frames??? Has this patch been tested by any one else to verify this???

Vitaliy Kiselev
06-07-2010, 12:04 PM
so I am not clear what the 720p 50 to 25 patch is doing.


GF1 uses frame doubling from actual 25fps or 30fps footage.
All this patch is doing is saving you some disk space by removing such doubling. Bitrate also drops.

Vitaliy Kiselev
06-07-2010, 12:17 PM
Or hardware limit?

They design this LSI :-) So, it does not matter.
Plus I do not believe in such hardware limit as I see P frames size spikes during AVCHD patches testing. And each such spike is times more than current size.
In reality P frames have some constant top size value (and this value is adjusted for lower bitrates, btw).
If you have big bitrate setting, sometimes all remaining bitrate go to last P-frame in GOP sequence.
So, if we could crack this hard nut, it'll be good.

spamrakuen
06-07-2010, 12:21 PM
Maybe the scaler is our problem. Maybe that function/process/hardware wants a concrete input frame size that we can't change. Maybe if we can skip the scaling process and output a MJPEG with the same size than the sensor mode, it will work...

Or maybe there are some parameters that tell the scaler the input frame size,... :-m

Vitaliy Kiselev
06-07-2010, 12:25 PM
Maybe the scaler is our problem. Maybe that function/process/hardware wants a concrete input frame size that we can't change. Maybe if we can skip the scaling process and output a MJPEG with the same size than the sensor mode, it will work...

Please, stop posting fantasies :-)
Current sensor patches are aimed at one thing only - find MJPEG sensor mode, test if it is used for AVCHD 720p mode also.
And changing MJPEG mode to other mode get appropriate movie files (with weird picture) and provide me link to them.

spamrakuen
06-07-2010, 12:27 PM
Please, stop posting fantasies :-)


Sorry :lipsrseal
Just trying to give some ideas.

John Caballero
06-07-2010, 12:31 PM
Sorry
Just trying to give some ideas.

LOL! You are doing great guys. Great input by all the testers. Cheers!

Jason Ramsey
06-07-2010, 01:02 PM
Hello, again Spike.... For someone who hates this place so much, you sure do register here a lot....

mimirsan
06-07-2010, 01:04 PM
Hi Great work!
I tried the Mjpeg size/encoder patch it was very excited to see the higher resolution in mjpeg.
Then after shooting for about a sec I had a warning about my card speed...do I need something faster than class 6 sdhc?

Barry_Green
06-07-2010, 01:27 PM
Okay, a note about card speed, for those who don't know -- first, not all card manufacturers are equal, and second, there are definitely bogus knockoff cards out there. But thirdly, the way the speed ratings work is that the rated speed is a *minimum* but it is not a maximum.

So it's entirely possible that a Class 4 card could actually end up being faster than a Class 10 card!

How? Because you have to recall that the Class designations have grown; there was a time (recently) when there wasn't even a Class 10 designation. So if a card can sustain writing 4 megabytes *or faster* per second, it qualifies as a Class 4 card. If a card can sustain writing 6 megabytes per second (or more) then it is eligible for a Class 6 designation (but could also be classified as Class 4!) In order to earn Class 10 designation, it would have to be able to sustain at least 10 megabytes per second, but could also be classified as Class 6 or even Class 4 (or, heck, even Class 2, if they wanted to.)

So what if you had a really fast card back when Class 4 was the maximum designation? What if you had a card that could sustain 20 megabytes per second? It'd still be rated as Class 4, because that's the highest they had.

So is it possible that a Class 4 card could handle 20 megabytes a second, and a Class 10 card could handle only half that? Yes, that's possible. Not really all that likely, but it's definitely possible. So you can't just assume that a higher class rating will definitely mean a faster card. It SHOULD work that way, but it's not guaranteed. The only guarantee you have is that the Class 4 will support at least 4 megabytes (32 megabits) per second, and the Class 6 will support at least 6 megabytes (48 megabits) per second, and the Class 10 will support at least 10 megabytes (80 megabits) per second.

But any of those classes could be substantially faster. Or, if you're using some ripoff/knockoff card, your card may not meet those speeds at all.

The only way to know is to download a speed-testing program and actually try your cards out.

axetota
06-07-2010, 02:59 PM
Just ran another quick test. I set sensor mode 4 to 7 only. AVCHD in 720p mode records normally with this setting. I think sensor mode 4 is definitely the MJPEG mode. It is the only one that when I change it, there is a funky picture. All the other modes when changed, this far, produce the same frame skipping picture (like if you don't change sensor modes but turn on 24p MJPEG) .

And on a side note, when I up MJPEG to 1920x1080 and switch sensor mode 4 to 7, there is still a funky picture.

NURBS
06-07-2010, 03:38 PM
Hey Guys'

Testing again'

Setting: AVCHD 1080p25/24 -> 720p25/24

Working but no resize picture ;) it mode crop picture 720p from 1080p ;-) like digital zoom.

Next test:

MJPEG mode sensor test.
(turn on: MJPEG Encoder 30fps -> Xfps: - 25)

Setting: Sensor mode 3: 3 -> 8

Working well, focus assistance too ;-)

Setting: Sensor mode 3: 3 -> 7

Working well, focus assistance too ;-)

Setting: Sensor mode 3: 3 -> 6

Working well, focus assistance too ;-)

Setting: Sensor mode 3: 3 -> 5

Working well, focus assistance too ;-)

Setting: Sensor mode 3: 3 -> 4

Working well, focus assistance too ;-)

(Next modes test and Files will be soon).

Files link: http://rapidshare.com/files/396443657/Sensor_Mode_3_TEST.zip

Regards
NURBS

spamrakuen
06-07-2010, 03:52 PM
(Next modes test and Files will be soon).


Cool! :2vrolijk_08:

Mode 4 is mode used by MJPEG, next one to be tested?

NURBS
06-07-2010, 03:54 PM
Next test:

MJPEG mode sensor test.
(turn on: MJPEG Encoder 30fps -> Xfps: - 25)

Setting: Sensor mode 4: 4 -> 8

Working fail (see on the stop frame), focus assistance working well ;-) (AVCHD mode too working well no changes)

Setting: Sensor mode 4: 4 -> 7

Working fail - Same as 8

Setting: Sensor mode 4: 4 -> 6

Working fail - but not same as 7 and 8 picture below ;-)

Setting: Sensor mode 4: 4 -> 5

Working well, focus assistance too ;-)

Setting: Sensor mode 4: 4 -> 3

Working fail - but not same as 6, 7 and 8 picture below ;-)

Files link: http://rapidshare.com/files/396457160/Sensor_Mode_4_TEST.zip

(Next modes test and Files will be soon).

Regards
NURBS

spamrakuen
06-07-2010, 03:57 PM
NURBS, can you test setting sensor mode 4: 4 -> 6, and then upload the file, please? (setting also 25fps and/or 50fps)

(also some motion needed to check if frames are dropped!)

Thanks!

NURBS, I think the images you are posting are crucial to understand what's not working (a too small buffer for 1080 picture, maybe?)

NURBS
06-07-2010, 04:14 PM
It's 720p not 1080p ;-)

spamrakuen
06-07-2010, 04:27 PM
It's 720p not 1080p ;-)

I know. And I'm using wrong numbers. I don't know the true sensor sizes. That's why I talk about 1080 and so.

I think that pixel count for 1080 > 720 PAL > 720 NTSC

There should be a buffer (in the scaler?) designed for 720 NTSC pictures that needs to be bigger for 720 PAL and 1080 sensor frame sizes.

I think, that's why there is that purple bar, and it is bigger on 1080 mode.

I'm (almost) sure tester could fix this. Anyway, even with artifacts the framerate may be ok, and without frame drops! So I'm hoping to see these files!:2vrolijk_08:

Tester13, forgive me if I'm fantasizing again, sorry.

NURBS
06-07-2010, 04:31 PM
Files is available :bath:

NURBS
06-07-2010, 04:42 PM
Next short test:

MJPEG mode sensor test.
(turn on: MJPEG Encoder 30fps -> Xfps: - 50 and MJPEG Encoder 1280m ->1920m)

Setting: Sensor mode 4: 4 -> 6

Same as old result ;-)

File link: http://rapidshare.com/files/396460185/P1010125.MOV

Next tests will be tomorrow ;) Good night!

Regards
NURBS

Isaac_Brody
06-07-2010, 05:15 PM
Next short test:

MJPEG mode sensor test.
(turn on: MJPEG Encoder 30fps -> Xfps: - 50 and MJPEG Encoder 1280m ->1920m)

Setting: Sensor mode 4: 4 -> 6

Same as old result ;-)

File link: http://rapidshare.com/files/396460185/P1010125.MOV

Next tests will be tomorrow ;) Good night!

Regards
NURBS

Actually, this setting doesn't look like a dropped frame, at least to my eyes. I mean there's the artifact at the bottom but I'm not seeing the jerk in movement like the other clips.

Someone else look at this, it doesn't look like a dropped frame to me. Even when i scrub frame by frame. What's weird is the other clip in sensor test 4, Mode_4_to_6 definitely has a skipped frame.

spamrakuen
06-07-2010, 05:16 PM
Not sure (didn't tested all of them), but I think motion isn't working yet (frame skipping). Anyone else, please, can confirm?

Tested only P1010125.MOV (fast motion) and mode_4_to_6.mov (skipping)

:cry:

Isaac_Brody
06-07-2010, 05:23 PM
Not sure (didn't tested all of them), but I think motion isn't working yet (frame skipping). Anyone else, please, can confirm?

Tested only P1010125.MOV (fast motion) and mode_4_to_6.mov (skipping)

:cry:

P1010125.mov doesn't look like a skip, but mode_4_to_6.mov definitely looks like a skip to me. I'm not sure if it's because P1010125 isn't panning as fast or what though so perhaps I'm not seeing the skip?

I think Tester's right that sensor mode 4 is MJPEG though.

Sean Jeong
06-07-2010, 06:25 PM
If we convert the 1080 24/25 to 720 24/25 would we have to de-interlace the footage?

Rabot
06-07-2010, 06:55 PM
The headphone clip (P1010125.mov) look really good on the motion.

RidgeShark
06-07-2010, 09:07 PM
In clip P1010125.mov, I am seeing duplicates of frames 23 and 48, but all other frames appear to be duplicate free. Strange that there are such duplicates, but overall the clip has very nice motion. It would be good to have a constantly moving shot to determine if other duplicates are occurring.

I am using QTInput in Avisynth and viewing in Virtual Dub.

Vitaliy Kiselev
06-07-2010, 11:23 PM
Yeah, 720p23,97 AVCHD is clipped from full 1920 picture, must be fixed later:-)
But best thing that we can play this progressive 24fps footage in camera without problems.
Good question is why we can't play 1080 progressive clips and got freezing instead.

Vitaliy Kiselev
06-07-2010, 11:51 PM
Setting: Sensor mode 4: 4 -> 5
Working well, focus assistance too ;-)


Thanks for testing.
Modes 4,5 share almost all setup routines.
So it was expected.

I'll add mode 4 adjustment patches soon, to adjust most sensor setup parameters individually.

Vitaliy Kiselev
06-08-2010, 03:36 AM
PTool v3.34 updated without version change.

I added MJPEG and AVCHD 720p mode constants adjustment (Sensor m4).
Please, test one at a time. Large constants can be changed more.
Change small values only slightly.
Report must contain picture (from MJPEG HD).
Sometimes very strange results could appear (especially in 800-900 range constants).
Do not shoot in such modes for more than 3-4 seconds, as some settings looks to be related to amplification.

P.S. Remember to switch camera to NTSC if you want to check that same mode is used actually for 720p60 and MJPEG 30fps (and it is the same :-) ).

NURBS
06-08-2010, 11:02 AM
Next test:

MJPEG mode sensor test.
(turn on: MJPEG Encoder 30fps -> Xfps: - 25)

Setting: Sensor mode 5: 5 -> 8 (AVCHD 1080p -> 720p same crop. no changes)

Working well, focus assistance too ;-)

Setting: Sensor mode 5: 5 -> 7 (AVCHD 1080p -> 720p same crop. no changes)

Working well, focus assistance too ;-)

Setting: Sensor mode 5: 5 -> 6 (AVCHD 1080p -> 720p same crop. no changes)

Working well, focus assistance too ;-)

Setting: Sensor mode 5: 5 -> 4 (AVCHD 1080p -> 720p same crop. no changes)

Working well, focus assistance too ;-)

Setting: Sensor mode 5: 5 -> 3 (AVCHD 1080p -> 720p same crop. no changes)

Working well, focus assistance too ;-)

(Next modes test and Files will be soon).

Files link: http://rapidshare.com/files/396746012/Sensor_Mode_5_TEST.zip

Regards
NURBS

NURBS
06-08-2010, 11:30 AM
Next test:

MJPEG mode sensor test.
(turn on: MJPEG Encoder 30fps -> Xfps: - 25)

Setting: Sensor mode 6: 6 -> 8 (AVCHD SHD mode 720p mode have a white line in right side)

Working well, focus assistance too ;-)

Setting: Sensor mode 6: 6 -> 7 (Same situation)

Working well, focus assistance too ;-)

Setting: Sensor mode 6: 6 -> 3 (AVCHD SHD mode 720p50 mode have a lot of damages on the picture)

I thing this setting for SHD 720p50 mode.

(Next modes test and Files will be soon).

Files link: http://rapidshare.com/files/396754647/Sensor_Mode_6_TEST.rar

billy fattey
06-08-2010, 11:57 AM
Is everyone having the same problem I am with rapidshare? I get to download one file as a free user, but then when I try the next it says I'm over my limit. Am I missing something?

Car3o
06-08-2010, 12:26 PM
that's how rapid share works.

NURBS
06-08-2010, 12:33 PM
Next test:

MJPEG mode sensor test.
(turn on: MJPEG Encoder 30fps -> Xfps: - 25)

Setting:

Sensor mode 4: 4 -> 6
Sensor m4 adju 1: 856 -> 880

Working well, don't have a picture damages problem :shocked:

File link: http://rapidshare.com/files/396774846/test_880.mov

spamrakuen
06-08-2010, 12:37 PM
Is everyone having the same problem I am with rapidshare? I get to download one file as a free user, but then when I try the next it says I'm over my limit. Am I missing something?

Use jDownloader.
http://jdownloader.org/
It will retry automatically

NURBS
06-08-2010, 12:42 PM
Next test:

MJPEG mode sensor test.
(turn on: MJPEG Encoder 30fps -> Xfps: - 25)

Setting:

Sensor mode 4: 4 -> 6
Sensor m4 adju 1: 856 -> 880
Sensor m4 adju 2: 870 -> 912
Sensor m4 adju 3: 10 -> 11
Sensor m4 adju 4: 838 -> 848
Sensor m4 adju 5: 52 -> 54

Result below on the picture
Video link: http://rapidshare.com/files/396777560/P1010138.MOV

spamrakuen
06-08-2010, 12:55 PM
Sensor mode 4: 4 -> 6
Sensor m4 adju 1: 856 -> 880

Working well, don't have a picture damages problem :shocked:


That's so cool!
But I think there is still frame skipping :cry:

svecher
06-08-2010, 01:02 PM
Use jDownloader.
http://jdownloader.org/
It will retry automatically
I believe that RapidShare has switched to limit 10 downloads per file some time ago for free accounts. It is best then for most people to refrain from the urge to pixel-peep these test files to make sure that those who really need to see them (tester13) actually do get to see them. Or buy a paid rapidshare account for Nurbs since he is so involved with testing.

billy fattey
06-08-2010, 03:03 PM
Dropbox is free and easy.

butler360
06-08-2010, 03:06 PM
If you split your files up into 50MB segments with something like 7zip, Skydrive has 25GB free storage with no limits on downloads.

bwwd
06-08-2010, 03:56 PM
you can throw 700 mb at megaupload.com for free

IanB
06-08-2010, 08:06 PM
Ok, here is the first test for me:
Settings were changed for an attempt at AVHCD quality improvement

This is the only thing I changed except 'native' 24p setting set

Video Buffer : 25xxxxxx
Video Bitrate Adjustment: 22xxxxxxx
Overall Bitrate Adjustment: 24xxxxxxx

Good:
Camera Zoom worked

Bad:
Camera would not record video after an attempt to record was tried.
Picture Images could not be taken either after the video attempt.

Result:
I am re-verting to a non-edited bitrate firmware right now. Maybe others will have better luck?

Vitaliy Kiselev
06-08-2010, 09:53 PM
Special note.

Any discussion of Andrew related stuff will be removed from this thread.
This especially concerns Andrew himself under different nicknames.

Vitaliy Kiselev
06-08-2010, 10:14 PM
Next test:
Setting:
Sensor mode 4: 4 -> 6
Sensor m4 adju 1: 856 -> 880


Please do not touch Sensor mode 4 patch.
As it makes all other adjustments useless.
Change one thing at a time.
As for 800sh constants you can go from about 600 to about 1300.
Try to start from sw and sh patches (last two).

Vitaliy Kiselev
06-09-2010, 12:24 AM
AVCHD mud explained

Look at this strange picture:
http://www.dvxuser6.com/uploaded/60298/1276064182.jpg

It shows block types and motion vectors for P frame.
Scene is specially very simple, overexposed and out of focus for easy viewing.
On simple scene and careful motion you have enough interframe blocks and some intraframe blocks (things with small divisions).
Lets look at high detail scene. Without motion, interframe blocks are able to maintain quality of I frame. But as soon as motion kicks in, you have almost only intraframe blocks.
As P frame size hits limitation, intraframe blocks compression became insane (3x-6x more size reduction). So, you have something like overcompressed I frame than P frame.

park25
06-09-2010, 02:05 AM
Ok, here is the first test for me:
Settings were changed for an attempt at AVHCD quality improvement



It's due to too high Video Bitrate Adjustment value.
More than 20xxxxxxx value will crash the camera.
My bitrates combination for AVCHD is 40xxxxxxx:20xxxxxxx:24xxxxxxx.
And if you are able to review the pictures, I mean to avoid freezing the body when touching the Play button, make sure the first shot is a picture, not an AVCHD movie.

Adventsam
06-09-2010, 02:30 AM
AVCHD mud explained
It shows block types and motion vectors for P frame.
...
As P frame size hits limitation, intraframe blocks compression became insane (3x-6x more size reduction). So, you have something like overcompressed I frame than P frame.

Dont Panasonic know what they are doing?

Vitaliy Kiselev
06-09-2010, 03:14 AM
Mayor breakthrough coming. Stay tuned.

As for Panasonic. No, they don't know. How they can make P frame size larger than I frame and have worse quality? This is kind of the hint to that's coming.

Vitaliy Kiselev
06-09-2010, 03:59 AM
PTool 3.35 released

Limiting bitrate patch added. Now you can adjust AVCHD bitrate (with use of other three patches that we have already).

Tested settings:



Video buffer - a) 33.500.000 b) 40.000.000
Video bitrate adjustment - a) 22.000.000 b) 32.000.000
Overall bitrate adjustment - a) 24.000.000 b) 35.000.000
Limiting bitrate adjustment - a) 26.000.000 b) 40.000.000


a) Gives you 22Mbit bitrate with some spikes up to 24Mbit.
b) Gives you 32Mbit bitrate with some spikes up to 34Mbit.

For 720p60 actual bitrate can be slightly lower (I get 27Mbit at b) ).

Top bitrate and best settings will be found by testers later.
Possibility is it can be as high as MJPEG top bitrate for FullHD mode.

Q: Such high bitrate clips are playable on camera?
A: Yes, no problems during my limited testing. All plays fine.

Jack Daniel Stanley
06-09-2010, 04:04 AM
PTool 3.35 released

Limiting bitrate patch added. Now you can adjust AVCHD bitrate (with use of other three patches that we have already).


whoa ...

Adventsam
06-09-2010, 04:30 AM
My battery is on charge, have to give this one a go. Tester13 you are incredibe if you solve this, thank you for keeping the faith with avc.

Paul Shields
06-09-2010, 04:55 AM
Adjustable bitrate AVCHD? Playback in camera? This is awesome news. Wish I was at home rather than at work so I could try this right now.

I hope everyone who is so far benefiting from these updates is also donating :).

jobless
06-09-2010, 05:08 AM
PTool 3.35 released

Limiting bitrate patch added. Now you can adjust AVCHD bitrate (with use of other three patches that we have already).

Tested settings:



Video buffer - a) 33.500.000 b) 40.000.000
Video bitrate adjustment - a) 22.000.000 b) 32.000.000
Overall bitrate adjustment - a) 24.000.000 b) 35.000.000
Limiting bitrate adjustment - a) 26.000.000 b) 40.000.000


a) Gives you 22Mbit bitrate with some spikes up to 24Mbit.
b) Gives you 32Mbit bitrate with some spikes up to 34Mbit.

For 720p60 actual bitrate can be slightly lower (I get 27Mbit at b) ).

Top bitrate and best settings will be found by testers later.
Possibility is it can be as high as MJPEG top bitrate for FullHD mode.

Q: Such high bitrate clips are playable on camera?
A: Yes, no problems during my limited testing. All plays fine.

I confirm. MUD free image with b settings !!!
I'll upload the test footage in a moment...

update:
download link:
http://www.nikolicnemanja.com/test4.rar
I started panning after 5 second of the footage...

kiguar
06-09-2010, 05:34 AM
obviously a great progress is made.
are there any restrictions? can it be used for let say production?

Vitaliy Kiselev
06-09-2010, 05:38 AM
From my testing they can be used for profuction already.
As for restrictions - testers will make it clear in uncoming days.
We'll have patches with presets for AVCHD and MJPEG later.