GH1 firmware research volunteers required

I have good news and bad news.

First off, these settings work. However, class 6 speed is NOT enough for recording longer then a couple of seconds. After about 30 seconds I get the message that Recording has been cancelled due to the writing speed of the disk. However, I'm still left with a short clip of pure awesome. :) I tried to induce mud, but none appeared. I will no try it again in my over-detailed garden, to see if we're truly rid of any mud.

Would be nice to see these clips, and other ones with default camera settings. So we can see what you have achieved :thumbsup:
 
I have good news and bad news.

First off, these settings work. However, class 6 speed is NOT enough for recording longer then a couple of seconds. After about 30 seconds I get the message that Recording has been cancelled due to the writing speed of the disk. However, I'm still left with a short clip of pure awesome. :) I tried to induce mud, but none appeared. I will no try it again in my over-detailed garden, to see if we're truly rid of any mud.

This is great progress guys! Can you Chibs, Nurbs or DrBlaz pls post clips on vimeo/youtube? Posting .MTS files does nothing if we have nothing to preview with. I preview files via Log&Capture window in FCP- but that entails having entire folder structure not just MTS file by itself.

On second thought does 'Clipwrap' allow you to preview MTS by itself??

Thanks and nice job--

No Mud=sweet cam!!
 
The bitrate and fps mods are looking very promising, great work everyone testing them.

I'd also like to see a few different resolutions/aspect ratios, eg. Cinemascope widescreen (which would allocate the bitrate better rather than having to crop in post), and 3:2 / 4:3 modes. Sometimes those aspects are actually quite useful. And of course it would be awesome if the entire sensor area could be recorded. Having extra pixels to crop in post can often be handy, especially when shooting in 3D where you usually need to crop the frames a bit for various reasons.
 
+1
Get a class 10 card to tester13 and NURBS as soon as possible.
\/ \/ and Chibs

Definitely.

Edit: ... be careful though, the class ratings can be very vague. As this guy says:

"For example, according to Rob Galbraith's tests for SD cards in the Nikon D90, the sustained JPEG write speed for Class 6 cards runs from about 3.6MBps (Fail!) to as fast as 18.5MBps. But by the Class metric, you can theoretically stick both in your HD camcorder and have no performance worries. "

It's worth reading that link, seems you really need to research the speed of these cards, there are class 6 cards that exeed the class 10 minimum speed.
 
Last edited:
I have good news and bad news.

First off, these settings work. However, class 6 speed is NOT enough for recording longer then a couple of seconds. After about 30 seconds I get the message that Recording has been cancelled due to the writing speed of the disk.

how big is the 30 sec. clip? what class 6 (brand/modell/size) card do you have?
i remember some write speed test recently with different cards, it was about checking if the card itself is the bottleneck (and could of course say nothing about whether an particular cam has the chance to use the full speed or would have any restricting bottleneck itself)
i'd like to do some math to see if class6 has any chance...

+1 get Chibs/NURBS an fast card (determine first how fast it has to be) from the funds, they are the heavy testers
 
But of course! I have uploaded some samples. If you backtrack in this thread, you can get some of the same sort of clips from my garden without modification, and with a first try. They are all full of mudd, as you can see. Quick sweeps, pans, and shakes.

Please read below if you're going to download these:
http://www.megaupload.com/?d=5UCOHVTI

With the bitrate settings as above, there is no mud at all to be found. I tried my hardest, both on 720p as on 1080p, as you can see in both clips. The downside?
At 1080p I can record for only about ten seconds each time, before getting the message that the recording has been cancelled due to writing speeds.
With 720p I can continue for a while and have not yet run into this issue.
This was done with an el-cheapo Class 6 Transcend (I have bad experiences with these card, two failed within weeks of purchase. I would not recommend, but it's the fastest I have).

However, both clips show quite some framedrops during the large sweeps. I sincerely hope this is due to the slow card, but I'm afraid there's no way for me to test this. :( NURBS, do you have ANY framedrops in your video if you make quick sweeps and pans in detailed shots? Please scrub through your video frame for frame to be sure. What are your last working settings that worked without any problems?
Here are the mediainfo analysis data:

720p: (Please note, I did not change the audio quality, so I didn't list them)

File size : 80.0 MiB
Duration : 32s 775ms
Overall bit rate : 20.4 Mbps (Wow!)
Maximum Overall bit rate : 24.0 Mbps

Video
ID : 4113 (0x1011)
Menu ID : 1 (0x1)
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L4.0
Format settings, CABAC : No
Format settings, ReFrames : 1 frame
Duration : 32s 420ms
Bit rate : 19.4 Mbps
Width : 1 280 pixels
Height : 720 pixels
Display aspect ratio : 16:9
Frame rate : 50.000 fps
Original frame rate : 100.000 fps
Standard : NTSC
Resolution : 8 bits
Colorimetry : 4:2:0
Scan type : Progressive
Bits/(Pixel*Frame) : 0.422
Stream size : 75.1 MiB (94%)

1080p:

File size : 26.0 MiB
Duration : 12s 920ms
Overall bit rate : 16.8 Mbps (Perhaps due to writing problems with the card?)
Maximum Overall bit rate : 24.0 Mbps

Video
ID : 4113 (0x1011)
Menu ID : 1 (0x1)
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L4.0
Format settings, CABAC : No
Format settings, ReFrames : 2 frames
Duration : 12s 760ms
Bit rate : 15.9 Mbps
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate : 25.000 fps
Standard : NTSC
Resolution : 8 bits
Colorimetry : 4:2:0
Scan type : Interlaced (This seems incorrect?)
Scan order : Top Field First
Bits/(Pixel*Frame) : 0.308
Stream size : 24.3 MiB (93%)
 
This is great progress guys! Can you Chibs, Nurbs or DrBlaz pls post clips on vimeo/youtube? Posting .MTS files does nothing if we have nothing to preview with. I preview files via Log&Capture window in FCP- but that entails having entire folder structure not just MTS file by itself.

On second thought does 'Clipwrap' allow you to preview MTS by itself??

Thanks and nice job--

No Mud=sweet cam!!
For simple playbacks of MTS files without data structures I use Splash, of which a simple version can be downloaded for free at http://www.mirillis.com/splash.html :)
 
The bitrate and fps mods are looking very promising, great work everyone testing them.

I'd also like to see a few different resolutions/aspect ratios, eg. Cinemascope widescreen (which would allocate the bitrate better rather than having to crop in post), and 3:2 / 4:3 modes. Sometimes those aspects are actually quite useful. And of course it would be awesome if the entire sensor area could be recorded. Having extra pixels to crop in post can often be handy, especially when shooting in 3D where you usually need to crop the frames a bit for various reasons.

Fully Fully Fully AGREE - would love a cropped 2.35:1 option to which we could allocate the full bitrate.
 
I agree on card issues, I believe few very good Class 10 cards can be useful.

It seems that GOP is especially useful.
But we not done with it yet, as I and P frame size restrictions are here still.
Reducing GOP, you reduce P number and hence increase average frame size.
 
But of course! I have uploaded some samples. If you backtrack in this thread, you can get some of the same sort of clips from my garden without modification, and with a first try. They are all full of mudd, as you can see. Quick sweeps, pans, and shakes.

Please read below if you're going to download these:
http://www.megaupload.com/?d=5UCOHVTI

With the bitrate settings as above, there is no mud at all to be found. I tried my hardest, both on 720p as on 1080p, as you can see in both clips. The downside?
At 1080p I can record for only about ten seconds each time, before getting the message that the recording has been cancelled due to writing speeds.
With 720p I can continue for a while and have not yet run into this issue.
This was done with an el-cheapo Class 6 Transcend (I have bad experiences with these card, two failed within weeks of purchase. I would not recommend, but it's the fastest I have).

However, both clips show quite some framedrops during the large sweeps. I sincerely hope this is due to the slow card, but I'm afraid there's no way for me to test this. :( NURBS, do you have ANY framedrops in your video if you make quick sweeps and pans in detailed shots? Please scrub through your video frame for frame to be sure. What are your last working settings that worked without any problems?
Here are the mediainfo analysis data:

720p: (Please note, I did not change the audio quality, so I didn't list them)

File size : 80.0 MiB
Duration : 32s 775ms
Overall bit rate : 20.4 Mbps (Wow!)
Maximum Overall bit rate : 24.0 Mbps

Video
ID : 4113 (0x1011)
Menu ID : 1 (0x1)
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L4.0
Format settings, CABAC : No
Format settings, ReFrames : 1 frame
Duration : 32s 420ms
Bit rate : 19.4 Mbps
Width : 1 280 pixels
Height : 720 pixels
Display aspect ratio : 16:9
Frame rate : 50.000 fps
Original frame rate : 100.000 fps
Standard : NTSC
Resolution : 8 bits
Colorimetry : 4:2:0
Scan type : Progressive
Bits/(Pixel*Frame) : 0.422
Stream size : 75.1 MiB (94%)

1080p:

File size : 26.0 MiB
Duration : 12s 920ms
Overall bit rate : 16.8 Mbps (Perhaps due to writing problems with the card?)
Maximum Overall bit rate : 24.0 Mbps

Video
ID : 4113 (0x1011)
Menu ID : 1 (0x1)
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L4.0
Format settings, CABAC : No
Format settings, ReFrames : 2 frames
Duration : 12s 760ms
Bit rate : 15.9 Mbps
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate : 25.000 fps
Standard : NTSC
Resolution : 8 bits
Colorimetry : 4:2:0
Scan type : Interlaced (This seems incorrect?)
Scan order : Top Field First
Bits/(Pixel*Frame) : 0.308
Stream size : 24.3 MiB (93%)

100fps!!, whats going on here, 720p at 100fps would be nice?
 
I agree on card issues, I believe few very good Class 10 cards can be useful.

It seems that GOP is especially useful.
But we not done with it yet, as I and P frame size restrictions are here still.
Reducing GOP, you reduce P number and hence increase average frame size.
What is this restriction you are talking about? Does it cause the frame skipping?
 
PTool 3.10 Released
MJPEG 24p still is research, back to 24fps
PAL and NTSC slightly modded
Added "VA Param 7", "VA Param 1080p 1A" and "VA Param 720p 1A"

It is interesting to see how they affect frame sizes and/or bitrate.
"VA Param 7" have also simple code patch (it is in sme video setup part, so only freeze risk)

More testers can be good, as VA settings are very low risk things, located in video encoding setup library function.
 
Last edited:
What is this restriction you are talking about? Does it cause the frame skipping?

We have somewhere some I,P frame size restriction or I/P frame size ratio.
So, mud is not caused by long GOP, it is caused by extremely low P frame size coupled with big problems in motion vector estimation routines.
 
http://sourceforge.net/projects/cardtest/ - software for testing cards - might be relevant?? Note that card speed can vary from one 'end' to the other and so even if you get (say) five minutes of smooth running, a real test is to fill the card, and see if it can keep up (without either the card or the camera melting, too!).

Maybe high bitrates in this camera will be like putting a Merlin jet engine in some old car - it'll go very, very fast... until the wheels fall off.
 
PTool 3.10 Released
MJPEG 24p still is research, back to 24fps
PAL and NTSC slightly modded
Added "VA Param 7", "VA Param 1080p 1A" and "VA Param 720p 1A"

It is interesting to see how they affect frame sizes and/or bitrate.
"VA Param 7" have also simple code patch (it is in sme video setup part, so only freeze risk)

More testers can be good, as VA settings are very low risk things, located in video encoding setup library function.
Could you explain the added parameters?
I think people are not hesitant because of bricking risk, but because being unable to upgrade to any new Panasonic original firmware later on, as a firmware revert function is not yet built.
 
Reverting to original firmware.

Reverting to original firmware.

Could you explain the added parameters?
I think people are not hesitant because of bricking risk, but because being unable to upgrade to any new Panasonic original firmware later on, as a firmware revert function is not yet built.

By definition the tool will revert to original 1.22 level albeit under another number, if you don't check any boxes or alter any values. Perhaps this should be better explained on the first page.
 
By definition the tool will revert to original 1.22 level albeit under another number, if you don't check any boxes or alter any values. Perhaps this should be better explained on the first page.
I am well aware of this, but as you said yourself, the version number is changed. This means there's a limited amount of updates possible at this time (Although that amount is so large it isn't a problem (80000 at least?), but still.
 
Back
Top