AVCHD high bitrate tests- MUD FREE footage

Car3o said:
the VIDEO BUFFER isn't doing anything.
Click to expand...

is it a known fact?because like explained early today higher video buffer settings gave choppy video, while all the other values stayed the same..so it must be doing something..i could be wrong..

i just started reading the replys, i got lost on one page..i guess its different for each one depending on the card.
 
Svart said:
Tested one of the alternate "D" settings

VIDEO BUFFER..................................84000000
VIDEO BITRATE ADJUSTMENT)............60000000
OVERALL BITRATE ADJUSTMENT.........56000000
LIMITING BITRATE ADJUSTMENT.........60000000

no card error recording however the files have encoding artifacts, large macroblocking and pulsing areas.

Tested on both Patriot LX class 10 and AData class 6. All files have same problems.
Click to expand...

tested those settings while fast panning through a really dense and detailed scene in my backyard.
Tried 720p and 1080p had no trouble at all, recorded flawlessly and played back flawlessly.

Used a lexar class 10 8gb SDHC 133x
 
i posted what he said. it's 100% dependent on cards, however, nobody wants to post their card information, only how fast it. and it's important to run your cards through a write/read program as well. i only quoted what tester and jobless said, but he also said more testing required.
 
For starters, thanks to tester13, jobless, Car3o and everyone else who has worked on this. I am really impressed.

Took the plunge and patched my firmware. Wanted to see how it effected the image quality in 1080p AVCHD when in motion, which is what I normally shoot for the narrative work I do.

Before I did the patch, I shot a scene with the camera on a tripod with the following settings, then repeated that scene as closely as possible to see the difference.

GH1 in Standard Setting
FHD
ISO 100
1/50th second shutter speed
Cooke Speed Panchro 18mm at T 11

Here is a compressed version of the clip:

http://www.lilfolksproductions.com/BeforePatch.htm

I then did the patch with the following settings:
-Native 24p/25p
-Video Buffer - 40000000
-Video Bitrate Adjust - 38000000
-Overall Bitrate Adjust - 40000000
-Limiting Bitrate Adjust - 42000000

Then I attempted to shoot the same scene with the same settings:

GH1 in Standard Setting
FHD
ISO 100
1/50th second shutter speed
Cooke Speed Panchro 18mm at T 8 (it had gotten darker)

Here is a compressed version of the clip:

http://www.lilfolksproductions.com/AfterPatch.htm

The compressed clips are too small to tell much, but I wanted to post them for reference.

I took two frame grabs of each clip, in an attempt to show the difference. To make the frame grabs manageable I had to compress them slightly.

Looking at the tree going past, you can see the difference in the following two frame grabs. Also, after the patch there is more detail in the house in the background.

Before the Patch:
BeforePatch1a.jpg


After the Patch:
AfterPatch1a.jpg


And there is more noise in the dark area of the porch post as it goes by (I know there is a technical name for that, but it escapes me at the moment). Also the details of the shingles on the brown roof are clearer after the patch.

Before the Patch:
BeforePatch2a.jpg


After the Patch:
AfterPatch2a.jpg



I was using a pretty slow SanDisk Ultra SDHC Card, 8GH, class 4 and I did not have any problems with the camera stopping during the recordings. What is shown above is only part of a 45 second clip which I shot a number of times and had no issues.

I think that about covers it.

Best,
-Tim
 
Video bitrate adjustment simplified Test:

_______________TRIPOD TEST_________________

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

results:

tripod.jpg



With these settings 1080 won't record....
Video bitrate adjustment simplified 600000000
Overall bitrate adjustment 62000000
Limiting bitrate adjustment 70000000


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' And limit for both 1080 and 720 on my card is with this settings:

Video bitrate adjustment simplified 38000000
Overall bitrate adjustment 40000000
Limiting bitrate adjustment 42000000
 
Last edited:
Another thing. If you want, you could PM me with your settings and explanations (problems, cards, links to files, etc), so that I can put all settings on first page...
 
my GH13 makes me feel like a boy with new shoes. thanks tester an testers...

****ers at paypal froze my account. :kali: i'll be getting my bro to make a donation for me as soon as i can get to the bank.

dunno if it helps?
Native 24p/25p
Video Buffer 84,000,000
Video Bitrate Adj 60,000,000
Overall Bitrate Adj 56,000,000
Limiting Bitrate Adj 60,000,000
16GB kingston class 4/ 16GB integral ultimaPRO class 6. high detail foliage footage looks like a Vang Gogh both on PC an playback in camera.
changed to C setting. ALL IS SWEET
thinking of a panasonic class 10?
 
Car3o said:
i posted what he said. it's 100% dependent on cards, however, nobody wants to post their card information, only how fast it. and it's important to run your cards through a write/read program as well. i only quoted what tester and jobless said, but he also said more testing required.
Click to expand...


I'll be testing my cards for sure but I'm not convinced it's cards. i think the card speed is an indicator that we are not using good setting choices. I think randomly trying to force settings that we "want" to work is a recipe for failure.

Besides this, the math shows that even a class 6 card (going on speed rating, 6MB/s (that's Mbyte)) should be able to handle up to 48Mb/s(that's Mbit). The other thing is that when I get artifacts, I get it on both my class 10 and class 6 cards exactly the same.. Which would mean it's not the card speed.

I'm working on thinking up a testing scheme that will help quantify my thoughts and hopefully yield some insight.

My theory is that the "video" bitrate is exactly what it says, the "overall" bitrate is video+audio and the "limiting" is what sets the compression routine's absolute maximum output is allowed to be.

In what I've seen today, if I set the Video bitrate close to the Limiting bitrate I start getting macroblocking.

Most people have also been testing in fairly quiet environments. I would like to see someone who's camera has been working OK with the "D" settings test those settings in a very noisy/dynamic sound environment.

I've run out of time for testing today so I'll try to work on my testing ideas and post them when ready.
 
Last edited:
Car3o said:
bating for info like that won't yield any results to your learning. if you've got a question ask or check out the FAQ. these threads are mainly geared toward testers who have experience. once things are settled and we have a better understanding how these patches work, they'll be available for people who don't understand as much.
Click to expand...
Where are these infamous FAQ?
 
Svart said:
I'll be testing my cards for sure but I'm not convinced it's cards. i think the card speed is an indicator that we are not using good setting choices. I think randomly trying to force settings that we "want" to work is a recipe for failure.

Besides this, the math shows that even a class 6 card (going on speed rating, 6MB/s (that's Mbyte)) should be able to handle up to 48Mb/s(that's Mbit). The other thing is that when I get artifacts, I get it on both my class 10 and class 6 cards exactly the same.. Which would mean it's not the card speed.

I'm working on thinking up a testing scheme that will help quantify my thoughts and hopefully yield some insight.

My theory is that the "video" bitrate is exactly what it says, the "overall" bitrate is video+audio and the "limiting" is what sets the compression routine's absolute maximum output is allowed to be.

In what I've seen today, if I set the Video bitrate close to the Limiting bitrate I start getting macroblocking.

Most people have also been testing in fairly quiet environments. I would like to see someone who's camera has been working OK with the "D" settings test those settings in a very noisy/dynamic sound environment.

I've run out of time for testing today so I'll try to work on my testing ideas and post them when ready.
Click to expand...

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.
 
Car3o said:
top of the thread.
Click to expand...
First post? But that's why I've asked for something clear. That post is for advanced users. If you come here for the first time, you have no clue about anything!

Where do I have simple answers like:

How does this work?

Should I download anything? Where?

I'd like to see something like the Magic Lantern. Very informative.
 
Caesar said:
Where are these infamous FAQ?
Click to expand...
I used information from the first post in this thread for patching AVCHD bitrate, but it [bitrate] has not actually gone up in that case (Avg 18Mbit). Then I applied the slightly modified settings posted above by timmyjoe (http://www.dvxuser.com/V6/showpost.php?p=2016622&postcount=364) and then I could see the improvement in IQ and bitrate (24 Mbit). Thanks TJ! I'm using 32 Gb Sandisk Extreme Class 10 card.

Car3o, could you please add those settings (at least) to your main FAQ, as recommended newbie settings? There is so much stuff posted, we need to start separating wheat from the chaff.
 
that post is for all users. if you don't understand the instructions right now, wait and read some more stuff and ask around in the FAQ thread.
 
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...

and I can't run the same settings as Barry on his class 6 on my Ultra II class 4. so yeah, definitely cards are the issue. pretty soon, lexar, panasonic and lexar's stock is going up.
 
A note on cards...I just drove to Best Buy and was forced into their only class 10 option, a 8gb Scandisk extreme...for $80...this is double the price of Amazon! However, only now, upon getting back to the house, did I realize there is a Fry's in the area with confirmed class 10 lexars for quite a bit less, so I'll be returning the best-buy card.

I'm in San Diego at the moment and will be doing some tests tomorrow down on Shelter Island of the AVCHD patch. If anyone wants to join in, and hang-out on an 80 foot sailboat and do some shooting, just drop me a line. A pretty girlfriend that we could shoot on the boat would make the test more interesting....
 
Kellar42 said:
A note on cards...I just drove to Best Buy and was forced into their only class 10 option, a 8gb Scandisk extreme...for $80...this is double the price of Amazon! However, only now, upon getting back to the house, did I realize there is a Fry's in the area with confirmed class 10 lexars for quite a bit less, so I'll be returning the best-buy card.

I'm in San Diego at the moment and will be doing some tests tomorrow down on Shelter Island of the AVCHD patch. If anyone wants to join in, and hang-out on an 80 foot sailboat and do some shooting, just drop me a line. A pretty girlfriend that we could shoot on the boat would make the test more interesting....
Click to expand...

whoaa....this sounds a bit more than shooting "video"....if you get what i mean..giggity.
 
I did my second attempt today using the same settings as Timmyjoe. I have a PAL camera, and having done the firmware change twice, I still only get 24fps on the MJPEG footage, and not the AVCHD. When I transfer the AVCHD footage into Final Cut Pro as Pro Res, it comes up as 25fps, and the data rate is 178.37 Mbit/s. Also my MJPEG wouldn't record to Class 4 card always. Often froze and came up with message as per pic. Any ideas why I don't get 24fps?

4694365711_a732118a5f_b.jpg
 
svecher said:
I used information from the first post in this thread for patching AVCHD bitrate, but it [bitrate] has not actually gone up in that case (Avg 18Mbit). Then I applied the slightly modified settings posted above by timmyjoe (http://www.dvxuser.com/V6/showpost.php?p=2016622&postcount=364) and then I could see the improvement in IQ and bitrate (24 Mbit). Thanks TJ! I'm using 32 Gb Sandisk Extreme Class 10 card.

Car3o, could you please add those settings (at least) to your main FAQ, as recommended newbie settings? There is so much stuff posted, we need to start separating wheat from the chaff.
Click to expand...

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.
 
You must log in or register to reply here.
Back
Top