75Mbps GH1 Peak Performance Patch

anyone seeing bitrates up towards the 75Mbps peak? been playing with FHD (pal) averaging around 40Mbps, it doesn't seem to get any higher tho? even on the highest detailed scenes.

has anyone else found this?
 
anyone seeing bitrates up towards the 75Mbps peak? been playing with FHD (pal) averaging around 40Mbps, it doesn't seem to get any higher tho?
The Peak Performance Patch limits the average bitrate to about 50Mbps, and I've found it to range somewhat lower in PAL mode. Peak bitrate occurs only during key frames, and is limited to a bit less than 75Mbps. This is a very high peak bitrate, and it enables the AVCHD codec to encode high quality key frames even when the subject matter is highly detailed.

The bitrate reported by various AVCHD players is only the average bitrate. The only tool I know of that reports the peak bitrate is cbrandin's Stream Parser.
 
thanks for the info lpowell stream parser doesn't install on my mac, been using vlc, looks like a live bit rate readout, when the captured footage moves onto low detail the read out lowers and high detail the read out jumps up.

anything over the stock 17 is a winner with me :smile:
 
anyone seeing bitrates up towards the 75Mbps peak?
Here is one that is very similar to the early ducks swimming on waters with ripples that failed with v1 or v2. VLC reports more than 70Mbps in the first 2 or 3 seconds, and it is above 50Mbps for most of the clip. Clearly lpowell should feel very proud of the Peak performance Hack. I have many video clips without a failure. The following clip was done with Creative Movie, the Pan 14-45 lens, early morning today. Film mode is Dynamic with -2 NR, -1 Saturation, widest aperture... Al
Edit: Vimeo processing does not do justice to the actual file, by the way. If you want to see the stunning video clip you need to download the very large 120MB file
 
Last edited:
I have got some pulsing clips too

I have got some pulsing clips too

I still find the pulsing more than a little annoying on some of my vids when hooked up to my TV though, but since no one else has raised the issue, I guess I'll have to live with it.

I have tried the initial Peak Performance settings with 25p Native checked.

I got some very low quality, major pulsing video clips. They are like water-colors flowing.

Black areas start as black and then get some (more) colors and thus "detail".

It seems to me, apart from the very low quality in my video, the I-frame is just as it is and then the P-frames slowly catch up for the detail (quality) not captured by the I-frame. It looks like, for some reason, the I-frame is very limited in bitrate and the P-frames may just show the same.
 
I have tried the initial Peak Performance settings with 25p Native checked.
I can virtually guarantee you that the Peak Performance Patch will not work properly with 24 or 25p Native checked. ;)

Also take care when recording in 720p50 mode to set your shutter speed to 1/50 or faster (1/60 in 720p60 mode). If you use a shutter speed that's slower than your framerate, the AVCHD encoder's bitrate will drop and it will produce extremely low-quality results.
 
Perhaps shutter speed was too slow, but other settings seemed to deliver quality

Perhaps shutter speed was too slow, but other settings seemed to deliver quality

I can virtually guarantee you that the Peak Performance Patch will not work properly with 24 or 25p Native checked. ;)

Also take care when recording in 720p50 mode to set your shutter speed to 1/50 or faster (1/60 in 720p60 mode). If you use a shutter speed that's slower than your framerate, the AVCHD encoder's bitrate will drop and it will produce extremely low-quality results.

OK, that may have been the problem. I am not sure, but I can imagine the image needed slow shutter speed. I was filming a show on tv.

You can get a clue here. The clip is only a couple of seconds. But the low quality (and) "smearing" is surely visible.

http://tinypic.com/r/14e57h3/7

However, I applied other Native 25p settings I know to be not totally reliable after this test and I got good quality then. I wonder if the need for slow shutter speeds was gone then. The scene was more or less the same.
 
Last edited:
I recently got a 32GB Sandisk Extreme Pro 45MB/s UHS-1 SDHC card. I benchmarked it on the built-in SD card reader of my HP Pavilion Elite HPE Intel i7 workstation and tried it out on my GH1 with my 75Mbps Peak Performance Patch. I then compared its performance to a 16GB Sandisk Extreme 30MB/s Class 10 SDHC card.

Verdict: On my equipment, the 32GB Sandisk Extreme Pro UHS-1 was no faster than the 16GB Sandisk Extreme Class 10 SD card.
 
Continuing to test this patch in all three modes. Had bright sun two days ago so tried recording MGPEG at the beach with fast moving/swimming dogs. Excellent results....no recording errors.

AVCHD 1080 recording with full sun and bright overcast with moving dogs and high details in open field and wooded setting also quite successful but with a couple recording failures at 5-9 seconds.

AVCHD 720 continues to be problematic with more than half my test clips failing. I have uploaded to Vimeo an example of a video that failed at a little over 6 seconds. Media Info shows 48.0 Mbps average for the clip. http://vimeo.com/19571359

Another clip with the same type of detail failed at 4 seconds but when immediately restarted continued to record for 34 seconds. Both these clips showed 48.3 Mbps average.

Constant for 720p=518400 in versions 3 & 4 of this patch. In the Fast Action 50 Mbps AVCHD 720p30 3-frame GOP patch, Constant for 720p=561600. I am wondering if it would be possible to increase to the 561600 number in the Peak Performance Patch while still maintaining stability.

Thank You again lpowell for all your patches! You are greatly appreciated & respected.
 
Constant for 720p=518400 in versions 3 & 4 of this patch. In the Fast Action 50 Mbps AVCHD 720p30 3-frame GOP patch, Constant for 720p=561600. I am wondering if it would be possible to increase to the 561600 number in the Peak Performance Patch while still maintaining stability.
Thanks for posting your results, ngw. I've also seen isolated instances of an initial recording stopping in the first few seconds, while the second attempt continued without failure. I suspect there is occasionally a greater processing load at the start of a recording that causes this fault, but I haven't been able to identify any particular conditions that consistently provoke it.

As you noted, I've adjusted the Constant for 720p=518400 value specifically to handle the 60p frame rate of this patch. However, it was necessary to choose a value that worked well for both NTSC 60p and PAL 50p recordings. Feel free to experiment with this setting, I'd be very interested if you find a value that improves the reliability for highly detailed subject matter.
 
I have been playing around with your patch with native 24/25p checked and it has held up quite well indoors. I had a go on the star chart and it only failed twice out of 6 attempts.

I was shooing in 1080/24p mode

Will take it outside and see what happens or is it not worth it?

I am using 8gb transcend class 10
 
Last edited:
GH1 purchased used from B&H

GH1 purchased used from B&H

Hey guys I've been trying to read as much about the Hack patches as I can before actually taking the plunge. Since I received my GH1 from B&H (from their used department), I've tried to set the video settings in order to get good video like everyone is talking about. I'm using the settings for Creative motion picture mode: AVCHD, FHD, Manual, standard I. Exposure. I'm using the Transcend 16gb class 10 memory chip. I've updated the fireware from the 1.2 as shipped too the current 1.32 from the Panasonic site.

I've tried several times to video outside with both Manual and Auto settings. The video looks very bad. It's hard to explain as I've never seen this before. When taking a moving - panning shot of my back yard, parts of the grass seems to break up, not anything like I've seen on this board.

I've been shooting video for over 20 years. I currently have the Sony EX3 and have added the GH1 and soon to receive the GH2. I plan to shoot a cooking show with the three of them and edit in Adobe Premiere CS5.

I think there is something wrong with the camera, but wanted to see what you guys thought or if you've seen this before. Maybe I've set something wrong. I'm very familiar with professional camcorders, but these DSLR cameras are a little different.

Would appreciate any suggestions as to what you think.

Dan
 
Hello,
To Vitaly or anyone who thinks they can be of help. I am one of the "stupid" ones who did not do his research before purchasing a GH1. The camera I purchased is a version 1.34 with a WEOG serial. I have not been able to hack it. Any pointers? Is a hack in the works? Is there any place where I can donate to the cause towards a hack? Also, can anyone tell me if there is a hack or any way to increase the MB bit rate/sec on the LUMIX FZ 100? Is it possible? Please help!! I am desperate for answers and guidance. Thanks.
 
Hello,
To Vitaly or anyone who thinks they can be of help. I am one of the "stupid" ones who did not do his research before purchasing a GH1. The camera I purchased is a version 1.34 with a WEOG serial. I have not been able to hack it. Any pointers? Is a hack in the works? Is there any place where I can donate to the cause towards a hack? Also, can anyone tell me if there is a hack or any way to increase the MB bit rate/sec on the LUMIX FZ 100? Is it possible? Please help!! I am desperate for answers and guidance. Thanks.

Skenti unfortunately the camera is unhackable as of now, and it seems Vitaly and company are moving onto the GH2. The hackable GH1s, or "gh13s" as we all know them now have a premium body price over the unhackable ones that you have.
 
Back
Top