Native 24P/25p settings/problems discussion

[richtrav]

Guys, I'm just having no luck at all with the C (or surfer) settings with the 24p wrapper enabled. It still crashes the camera fairly quickly and still gives me the muddy I frames.

Now what I HAVE had some luck with - when 24p is on - is going back to the A settings for video and limiting bitrate, setting the preset bitrates both to 234, then cranking up the limiting bitrate way up into the 60-90 range. It did crash on me once so it's not bulletproof, but I have not seen any I frames turn to complete mush even when the camera was set on a table (don't have a tripod right here right now). It'll pulse a little but not horribly and I don't see those telltale dropped P frames right before any I frames. Subjectively I still think there's a little more detail when running the C settings with the 24p wrapper disabled but there's not that much of a difference in static scenes (haven't tried panning or motion yet, and have to go to work)

 

[PappasArts]

Cool, I'll get on it.

Has Vitaliy told us what he changed in the native 24p to get this to work? Not to sound unappreciative of his work, I do wish he was more open about what changes he is making. I think it would help us sanity check the changes before we go through lots of testing.

I wish we had more info as well-

Do we select both Preset Birate 1 & 2?- or just one of them. I notice both are 234-

Guys, I'm just having no luck at all with the C (or surfer) settings with the 24p wrapper enabled. It still crashes the camera fairly quickly and still gives me the muddy I frames.

Me too!

Pappas

 

[svecher]

Has Vitaliy told us what he changed in the native 24p to get this to work? Not to sound unappreciative of his work, I do wish he was more open about what changes he is making. I think it would help us sanity check the changes before we go through lots of testing.

I wish we had more info as well-

Questions to that effect were asked in development thread. If I understand the response correctly, observed stability improvements are due to Bitrate Patch(es) tweaking some parameters of AVCHD encorder (apparently successfully).

I am currently [sporadically] real-world testing Set "D" settings posted earlier http://www.dvxuser.com/V6/showpost.php?p=2045981&postcount=252

 

[PappasArts]

Questions to that effect were asked in development thread. If I understand the response correctly, observed stability improvements are due to Bitrate Patch(es) tweaking some parameters of AVCHD encorder (apparently successfully).

I am currently [sporadically] real-world testing Set "D" settings posted earlier http://www.dvxuser.com/V6/showpost.php?p=2045981&postcount=252

svecher,

Are you selecting/ticking both pathches 1 & 2, or just one of them?

Pappas

 

[svecher]

How I learned to stop worrying and love the bomb

How I learned to stop worrying and love the bomb

Are you selecting/ticking both pathches 1 & 2, or just one of them?

Both, see full settings below. Static shots with both patches checked produce files with I frames of uniform size, which makes sense. When only the first patch is checked I frame sizes vary. When shooting non-static scenes you still get unpredictable frame size variance, including rogue frames, in either scenario. Not knowing any better I decided to stick with settings that produce "logical" results in at least one type of scene (high detail static shots). Also, I have yet to see a crash or card speed error (which I used to get all the time with exact same settings before Preset bitrate patches were introduced).

[Information]
Comment=Surfer AVCHD settings with matching preset bitrates @ 500
SD_Card=Sandisk Extreme Class 10
Camera=GH1 v1.32
[Settings]
Prevent version compare=Checked
Native 24p/25p=Checked
Video Bitrate FHD/SH=50000000
Video Bitrate H=50000000
Video Bitrate L=32000000
Overall Bitrate=56000000
Limiting Bitrate=60000000
[B]Preset bitrate=500
Preset bitrate 2=500
[/B]

 

[rambooc1]

Thanks i,m going to try the preset1 and 2 trick see if it makes a difference to my preset1 only setting result.

What prompted you to use 500.

R

 

[svecher]

What prompted you to use 500.

Vitaliy said to match preset rates to "Video Bitrate FHD/SH". Note that values are specified in different units of measurement. Video Bitrate is specified in bits, so Video Bitrate of "XYZABCDEF" equals ~"XY.Z" Mbits. Preset Bitrates of "XYZ" also equals "XY.Z" Mbits.

 

[Park Edwards]

Both, see full settings below. Static shots with both patches checked produce files with I frames of uniform size, which makes sense. When only the first patch is checked I frame sizes vary. When shooting non-static scenes you still get unpredictable frame size variance, including rogue frames, in either scenario. Not knowing any better I decided to stick with settings that produce "logical" results in at least one type of scene (high detail static shots). Also, I have yet to see a crash or card speed error (which I used to get all the time with exact same settings before Preset bitrate patches were introduced).

[Information]
Comment=Surfer AVCHD settings with matching preset bitrates @ 500
SD_Card=Sandisk Extreme Class 10
Camera=GH1 v1.32
[Settings]
Prevent version compare=Checked
Native 24p/25p=Checked
Video Bitrate FHD/SH=50000000
Video Bitrate H=50000000
Video Bitrate L=32000000
Overall Bitrate=56000000
Limiting Bitrate=60000000
[B]Preset bitrate=500
Preset bitrate 2=500
[/B]

what's your card to compare with others.

 

[richtrav]

Both, see full settings below. Static shots with both patches checked produce files with I frames of uniform size, which makes sense. When only the first patch is checked I frame sizes vary. When shooting non-static scenes you still get unpredictable frame size variance, including rogue frames, in either scenario. Not knowing any better I decided to stick with settings that produce "logical" results in at least one type of scene (high detail static shots). Also, I have yet to see a crash or card speed error (which I used to get all the time with exact same settings before Preset bitrate patches were introduced).

[Information]
Comment=Surfer AVCHD settings with matching preset bitrates @ 500
SD_Card=Sandisk Extreme Class 10
Camera=GH1 v1.32
[Settings]
Prevent version compare=Checked
Native 24p/25p=Checked
Video Bitrate FHD/SH=50000000
Video Bitrate H=50000000
Video Bitrate L=32000000
Overall Bitrate=56000000
Limiting Bitrate=60000000
[B]Preset bitrate=500
Preset bitrate 2=500
[/B]

These are the exact settings I used (minus H and L adjustments, which I don't use) and they froze up the camera after 4-15 seconds on a static high detail scene and Panasonic Class 10 card. Upping the preset also didn't seem to help. 60i has been pretty solid so it stays on as default. I'd just much rather have 24p and playing with it has become kind of a hobby when I don't have anything better to do

 

[rambooc1]

what's your card to compare with others.

The card is listed in the settings snapshot (blue screen)

 

[rambooc1]

Ok thanks, must have missed the matching bitrate/preset advice somewhere along the line

 

[rambooc1]

Svecher, those surfer settings with 500/500 and 25pN enabled i can confirm have made a huge difference in stability. This is the first time i have been able to record long (2min+) clips in SHD and SH modes with Native 25p enabled, all this on a class 6 16gb card. Very happy.

Actual average bitrates have not changed much 14mbps- 42mbps depending on detail etc but man the stability is heaps better. Just need to do some controlled testing now and try break it on Pappas chart. But for tripod and handheld this is the bee's knee's.

Please more people test this one and report your findings.

R

 

[PappasArts]

Svecher, those surfer settings with 500/500 and 25pN enabled i can confirm have made a huge difference in stability. This is the first time i have been able to record long (2min+) clips in SHD and SH modes with Native 25p enabled, all this on a class 6 16gb card. Very happy.

Actual average bitrates have not changed much 14mbps- 42mbps depending on detail etc but man the stability is heaps better. Just need to do some controlled testing now and try break it on Pappas chart. But for tripod and handheld this is the bee's knee's.

Please more people test this one and report your findings.

R

Rambooc1,

The chart footage you posted the other day- the grids look larger then they should be. Just in case you got the half res version- here is the full res version. If your filming the chart right, the red text will be visible at the very bottom edge of the frame.

Here is a link to the 2K chart: http://www.sendspace.com/file/5f0if7

Pappas

 

[svecher]

These are the exact settings I used (minus H and L adjustments, which I don't use) and they froze up the camera after 4-15 seconds on a static high detail scene and Panasonic Class 10 card. Upping the preset also didn't seem to help. 60i has been pretty solid so it stays on as default. I'd just much rather have 24p and playing with it has become kind of a hobby when I don't have anything better to do

That's strange, as Panasonic Class 10 Cards have shown similar performance in previous tests if I remember correctly.

Not that I don't trust you, but could you please post your settings in the same ini file format? Sometimes you need a different set of eyes to spot a problem. We've got the tools now, so let's use them..

 

[svecher]

Svecher, those surfer settings with 500/500 and 25pN enabled i can confirm have made a huge difference in stability. This is the first time i have been able to record long (2min+) clips in SHD and SH modes with Native 25p enabled, all this on a class 6 16gb card. Very happy.

Class 6? That's a new high! :thumbsup:

 

[rambooc1]

Ohhh yeah Pappas, that full res chart beats me. I can't record that for any more than 5 - secs. Real world stuff no problems. If i had a class 10 card i would like to try see the difference to the class6

MY settings Surfer same as svecher but PAL (DELETE THE .TXT FROM THE END)

 

[rambooc1]

Class 6? That's a new high! :thumbsup:

Yes just one real world fail (card speed error, not lockup so no need to power down). It was with a 20mm wide FD "L"" lens lots of detail static shot. Moving no problems.

I think you are getting close to max regardless of card.

 

[svecher]

Yes just one real world fail (card speed error, not lockup so no need to power down). It was with a 20mm wide FD "L"" lens lots of detail static shot. Moving no problems.

I think you are getting close to max regardless of card.

Well, let's see what that bitrate is. Since cam didn't freeze it should have saved the file. What does StreamEye show?

 

[rambooc1]

Well, let's see what that bitrate is. Since cam didn't freeze it should have saved the file. What does StreamEye show?

Ummmmmmmmmmm.... maybe a faster card might handle those I frame spikes after all. ...Damn big arn't they?

 

[richtrav]

That's strange, as Panasonic Class 10 Cards have shown similar performance in previous tests if I remember correctly.

Not that I don't trust you, but could you please post your settings in the same ini file format? Sometimes you need a different set of eyes to spot a problem. We've got the tools now, so let's use them..

Hehe, save them? Doh, yeah I forgot about that feature. They're history now, I'll attempt to recreate the scene of the crime this weekend if I get a little time. It was just my simple casual "backyard test". I really would like this 24p patch thing to work out, I'm not a fan of 60i at all.