AVCHD high bitrate tests- MUD FREE footage

Car3o said:
i don't want to post inclusive results. Everyone has varying results at the moment. once we get something more accepted across the board from a large number of users, i'll update it with the settings. right now jobless is posting settings on the first page that seem to be working for him and different users. i can maybe link to here.
Click to expand...
I just reread jobless's and timmyjoe's posts and they are actually both recommending one and the same set of values for AVCHD adjustment. For some reason, the first time I read jobless's post I thought he recommended applying all values labeled as "c" in p-tool. That didn't work for me. His recommended settings are actually spelled out in the middle of the post.

Video bitrate adjustment 38000000
Overall bitrate adjustment 40000000
Limiting bitrate adjustment 42000000

For these settings avg. bitrate in 1080 MODE is 34.4 Mbps and for 720 34.5 Mbps. There are no problems with recording of any kind either with 1080 or 720. No MUD with this setting also. I should say that these are stable settings and I would recommend them for normal users that want to shoot with 1080 and 720.
Click to expand...
The way TJ spelled them out was more concise for me, similar to the way you've organized the FAQ post, with screenshots and all. Perhaps you can incorporate these AVCHD settings then, since they are tested, stable and all. Thanks!
 
Just got a Panasonic Class 10 card. Tested it with Crystal Disk Benchmark, and it's awesome -- 20mB/sec read, 17.8mB/sec write.

Still can't record high-bitrate 1080p MJPG. Fails after about two seconds.
 
jobless said:
Your shutter speed must be 60 or faster ( 100, 125, etc)
If you put shutter at 50, Gh1 duplicate frames and create problems for codec with these settings.

As for shooting 720 60p you should use shutter speed 125 for nicer motion blur regardless of settings you use. (even with original firmware)
Click to expand...

I just tried it with the shutter at 60 and it works great. Thanks! Time to shoot some stuff and donate!
 
Barry_Green said:
Just got a Panasonic Class 10 card. Tested it with Crystal Disk Benchmark, and it's awesome -- 20mB/sec read, 17.8mB/sec write.

Still can't record high-bitrate 1080p MJPG. Fails after about two seconds.
Click to expand...

So if you can't use a class 10 for that, what would you use for 80mbit MJPEG?
 
Kholi said:
Not all cards are created equal is the issue. Pappas here is running pretty heavy PhotoJPEG settings on an Extreme Class 10, but Barry can't seem to run the same settings on his Class 6 Lexar if I read right.

So, cards are very important. I know I can't run the same settings as everyone else on my crap Transcend Class 6 16gb card. But, apparently it's working on faster cards or solid Class 6 cards.

We'll know when more people get Class 10 Extreme's or Panasonics.
Click to expand...


What I'm saying guys is that we're collectively doing this backwards. Instead of finding cards that fit our firmware setting *guesses* we should be finding settings that work with cards so that there aren't any problems when you go and pick something out.

We *know* that the card class is the guaranteed write speed of the card. We should be finding settings that max these speeds out but do not go over. That way we will also *know* that certain settings will work for all cards in the specified class.

Some cards can sustain higher than stated speeds and we can address those cards too but just because one card can do it, doesn't mean another card of the same brand/model can do it too, thus the reason why a card is usually rated under what it actually can do.

I design electronics and this is called derating. A common analog is computer overclocking. Some chips can be overclocked easily, while others cannot even when they are of the same batch. They are tested and derated depending on their result.

That's all I saying guys. I'm going to start a spreadsheet of some kind to follow my testing. I will be starting with the stock settings and incrementing one of the settings at a time. This will take some time and once I get the plan together I'll be asking for others to do specific tests and report back in order to get it done faster.

The other thing is that as we go along, Tester13 is very likely to get into parts of the firmware that will shed light on these settings and how they work with each other. As he keeps saying, we need scientific results, not wild guesses. This might make my testing, and that of other people, completely useless but who knows at this point?
 
Svart said:
What I'm saying guys is that we're collectively doing this backwards. Instead of finding cards that fit our firmware setting *guesses* we should be finding settings that work with cards so that there aren't any problems when you go and pick something out.
Click to expand...


Exactly!

We need to concentrate on the Quality and setting tables. Even the buffer numbers etc might be an issue.

What is crazy is I have shots, mind you not long at 105mbits on other experimental settings. However if i can do 105mbits on my card at 720, then on a different formula I can't get my 720s to peak over 60 without stopping. So something else is limiting or causing the spike to shut down the filming. Another setting will give me 113.28Mbps for a few seconds. Basically we need to figure out what and how these number relate. The numbers I posted in the other thread are stable for the most part and the Mbits peak at around 83.01mbits... Knowing what these numbers mean are very important though.

Barry, you said that one of my settings I sent you had no strange artifacts and one did. Which one was that?

I'm noticing that these numeric formulas also relate to that issue.

Pappas
 
Barry_Green said:
Just got a Panasonic Class 10 card. Tested it with Crystal Disk Benchmark, and it's awesome -- 20mB/sec read, 17.8mB/sec write.

Still can't record high-bitrate 1080p MJPG. Fails after about two seconds.
Click to expand...

I installed this program and can't get it to work right. It takes like 30 minutes to report anything and it's always the same results regardless of card tested.

Do you know of any other benchmarking software for SD cards? Searching online turns up thousands of benchmark results and very few actual programs.
 
Well guys, I just hacked my firmware for the first time. I used the 'c' settings in Ptool for AVCHD, which I understand to be the 'd' settings, I believe from the FAQs. I'll be shooting some high detail stuff tomorrow at the harbor, but for tonight I just wanted to have a look at FPN...just in case as Barry has noted.

So I shot the worst nightmare imaginable for FPN...a white wall in the pretty dark corner of a room with one overhead light at night, 50mm at F1.4 ISO 640,30 shutter, Nostalgic mode, -2 on everything.

I won't post grabs because the results were almost identical, maybe even a hair worse for the patch.

But here's the catch, with playback in Toast...it showed the original non-patched firmware clip as having around 14.45 mbps, and the hacked firmware starting at around 12, climbing to just over 19 only if I change focus or pan the camera slightly. Somewhere else I read this, that it might be possible to force the bit-rate to a certain place...I wonder if that would make any difference? Because on low-detail static shots the bit-rate falls quite low. Seems to me if it could be kept higher, it might help with the noise, if nothing else, but I'm not an expert, for sure.

Also, a still photo of the same shot has much less FPN...so I have hope for the mjpeg mode. Looking forward to detail and daylight in the morning, with a faster card!
 
jobless said:
Video bitrate adjustment simplified Test:

_______________TRIPOD TEST_________________

Same scene, same lighting, same F stop and shutter speed

As for test with pans and hidetailed scene... Everything looks the same at low level. I notice that with 'Video bitrate adjustment simplified' avg bitrate is little lower then with 'Video bitrate adjustment' .... 2-3 Mbps for same settings ( that could be due to different light conditions). Same settings that I used with 'Video bitrate adjustment' create same problems with 'Video bitrate adjustment simplified'
Click to expand...

Sorry, I do not understand your test results completely.
From your tests bitrate is exactly the same, it can be so for static scenes, especially low detailed.
Try to make high detail static test and high detail panning test.
And post low level results here.
 
Hey guys,

Thought I'd mention something I noticed while i was testing D settings with my cheap-o Transcend card: camera stopped recording on 720p after 4sec, but if I turned AF off it kept going for twice that much (on the exact same scene with exact same panning motion). I could repeat the effect every time...
 
For Updates: MIKOS ARTS on Twitter
http://PAPPASARTS.WORDPRESS.COM
http://TWITTER.COM/PAPPASARTS
http://MIKOSarts.wordpress.com
http://MIKOSarts.wordpress.com/page-2-of-blog
http://mikosarts2.wordpress.com/blog-page-2
http://mikosarts2.wordpress.com
http://MIKOSarts.COM
http://MIKOS.info


Still can't record high-bitrate 1080p MJPG. Fails after about two seconds.[/QUOTE]

Barry try the 720P mode. See how long you get. Also are you using the 1440X1080 setting for shooting 1080P?

A week ago when I was testing various settings for 720P, I would get very high bit rates. Ones even higher than this. I was filming extreme stuff just to see the break point on the GH1. Later I resorted to the charts for this, since I could go even farther. Which makes me wonder, if the GH1 can push through these and some that were 110Mbits+. The bandwidth in the GH1 is there, just something may not be set right with buffer or etc-Something possibly!

Here are some examples of what I got from testing a week ago.

Picture104.jpg

Picture107.jpg

Picture105.jpg

Picture106.jpg

Picture108.jpg

HERE ARE A COUPLE OF 1080P DONE TONIGHT IN THE SQUEEZE MODE 1440X1080P
Picture111.jpg

Picture110.jpg



MIKOS ARTS
http://PAPPASARTS.WORDPRESS.COM
http://TWITTER.COM/PAPPASARTS
http://MIKOSarts.wordpress.com
http://MIKOSarts.wordpress.com/page-2-of-blog
http://mikosarts2.wordpress.com/blog-page-2
http://mikosarts2.wordpress.com
http://MIKOSarts.COM
http://MIKOS.info


[size=+0]BlackMagic Digital Camera Info Thread:[/size] http://www.bmcuser.com/showthread.p...mera-Technology-Announcements-amp-Latest-Info
 
Last edited:
If the link won't be public and is just for a select few to view/analyze I could give Pappas an FTP username through my webhosting (VPS I use for me and my clients, thus the concerns with a public link to massive files, although I won't get complaints from the host like on a shared host and have 10TB bandwidth/month). I think I have at least 10GB of disk space available (probably could make more room) and I know it's a fast link (supposedly up to 1gbps burst).
 
butler360 said:
If the link won't be public and is just for a select few to view/analyze I could give Pappas an FTP username through my webhosting (VPS I use for me and my clients, thus the concerns with a public link to massive files, although I won't get complaints from the host like on a shared host and have 10TB bandwidth/month). I think I have at least 10GB of disk space available (probably could make more room) and I know it's a fast link (supposedly up to 1gbps burst).
Click to expand...

No, we need public server.
So, traffic will be huge as well as load.
 
I've just been trying out the higher bitrate AVCHD and I've got an odd result.
Overall extremely happy. Beautiful footage and amazingly little mud, even when I do crazy whip pans.

But looking back at my footage I've found one odd moment where the codec seems to hiccough for no obvious reason.

Here's a good frame:
1276426785.jpg




and here's the immediate following frame.


1276426857.jpg


As you can see there's a sudden loss of detail and artefacting all across the frame.
Although its wide and sharp so there's lots of details there's very little movement and
nothing obvious to make the codec lose it. The image recovers over a couple of frames
and then exactly 13 frames later the same things happen. And again 13 frames later.
After that it seems fine. In total I shot about 20 minutes and have only noticed this on
the one occasion.

My settings were

Video buffer 40000000
Video Bitrate Adj 32000000
Video Bitrate Adj Simp 32000000
Overall Bitrate Adj 35000000
Limiting Bitrate Adj 39000000
FHD using kit lens
using the latest 3.36 ptools.
I've got a 16gb Transcend class 6 card.

The clip was about 7 minutes long and the fault happened near the end -
so is this some kind of overheating/overloading problem?
 
nathankw said:
My settings were

Video buffer 40000000
Video Bitrate Adj 32000000
Video Bitrate Adj Simp 32000000
Overall Bitrate Adj 35000000
Limiting Bitrate Adj 39000000
FHD using kit lens
using the latest 3.36 ptools.
I've got a 16gb Transcend class 6 card.
Click to expand...

Don't use Video Bitrate Adj 32000000 and Video Bitrate Adj Simp 32000000 together...
Go to first page and read first two posts.
 
Maybe in due course the tool could include some features to check for 'sane' values - in other words if you put value 'n' in box 'b' then it won't let you put value 'm' in box 'c' if that wouldn't be sensible.
 
You must log in or register to reply here.
Back
Top