Blackout-Powell Native 24p Patch v2 shows field dominance

N

naavt

Well-known member
I've just hacked my GH1 with Blackout-Powell Native 24p Patch v2, shot a couple of seconds from my window and imported the footage over to FCP to see the clip properties.

It shows upper filed dominance when the Native 25p is checked on PTool 3.61d FW.

Isn't this odd? It should show "No field" dominance when it's true progressive right?

If someone throws some light here I'll be very appreciated.

Thanks
 
bump!

72 views and not one fellow member to comment on this :(

Anyone? I'm sure that probably someone out there has faced the same problem! Or is just me?
 
My problem: too many posts about the same problem and posters not reading

My problem: too many posts about the same problem and posters not reading

naavt said:
bump!

72 views and not one fellow member to comment on this :(

Anyone? I'm sure that probably someone out there has faced the same problem! Or is just me?
Click to expand...

While you perhaps have been sitting and watching, waiting for the view counter to cross this magical 72 barrier, you might as well have been actively searching for any related problems on these forums.

Of course, you may have had other things to do. So may others.
If you aren't patient, then don't use the internet passively.
If you want attention, undress and start running and screaming down the street.

My problem is I am seeing people being impatient.

Not a day goes by without a complaint that their needs haven't been satisfied.

I see multiple messages posted in several threads dealing with just a single problem of a single person.
Apparently, they do think they are SO important that almost the complete community should divert their attention to them.

I even see multiple threads being started about the same problem. HELP!!!!

H E L P ! ! ! H E L P ! ! !

WTF! What is so important in your life?
I guess food is, as a primary need. Most other things don't really matter that much.
And we should be glad we have food.
All else that we get, from others who are willing to contribute to our comfort in life, you should see that as a bonus.
While you get it, enjoy it.

If you don't get it, don't complain.

I also see posters saying they don't (want to) take the time to read information that is already present.

Why? Why would you just sit back and wait, expecting others to put effort into what you apparently just can do at the time?

I really don't understand much of the world today. Well I do, these are primitive needs, as well, that cause this behavior. However, these are primitive in another way. These are more like pre-human. They lack the reason that we separates us from the (other) animals.
 
Dear Fool4UAnyway,

Sorry if my "active searching" on this forum came up with zero results. Perhaps you're a better searcher, or simply just a clever guy than I am, but FWIW I can assure you that I've searched with no avail.

Nonetheless, one thing strikes my inferior mind! What makes an illuminated person like you spend so many words about NOTHING REALLY besides your own complain, when you could have put your brilliant mind to work and help a fellow member with less words than you have spent???
 
get real

get real

Isn't there a mud-slinging sub-forum somewhere so you guys could have-it-out until both are satisified?

Ok. To the question, a basic question: HOW are you importing to FCP... if you are relying on log-and-transfer, and transcoding to ProRes, well you have simply un-done what LPowelll and others have spent hours doing for you, namely NATIVE progressieve, not Psf...

I have stopped using FCP and may be mistaken, but that program might just be creating progressive segmented material from true progressive, in which case, you really have nothing to worry about.

and
Are you shooting with a PAL or NTSC camera?
 
ninetto said:
HOW are you importing to FCP... if you are relying on log-and-transfer, and transcoding to ProRes, well you have simply un-done what LPowelll and others have spent hours doing for you, namely NATIVE progressieve, not Psf...

I have stopped using FCP and may be mistaken, but that program might just be creating progressive segmented material from true progressive, in which case, you really have nothing to worry about.

and
Are you shooting with a PAL or NTSC camera?
Click to expand...

Yes, I'm importing trough FCP Log & Transfer window to ProRes.

Using PAL

So for your answer I assume my whole workflow is wrong right?
 
@Naavt- This was a known issue with the original hack on the GH1(3)


ninetto said:
if you are relying on log-and-transfer, and transcoding to ProRes, well you have simply un-done what LPowelll and others have spent hours doing for you, namely NATIVE progressieve, not Psf...
Click to expand...
what?!
What else would you transcode to? Isn't it's native codec h.264? Impossible to edit w/ in FCP. ...well not impossible, but certainly not fun.
 
what?!
What else would you transcode to? Isn't it's native codec h.264? Impossible to edit w/ in FCP.
Click to expand...


Please READ my entire post before screaming at me with exclamation marks. The rest of my post reads:
I have stopped using FCP and may be mistaken, but that program might just be creating progressive segmented material from true progressive, in which case, you really have nothing to worry about.
Click to expand...

What I was describing was my understanding of FCP -which I no longer use- and its work flow in wrapping 24(25)p as 50i, it is all described here:
http://en.wikipedia.org/wiki/24p
 
Last edited:
I've looked at the frames produced by setting progressive mode on the GH1. It turns out that the encoding is a little different than what you would expect with progressive footage. Normally you would see all macroblocks in an I frame encoded using "intra" encoding. With the GH1 progressive mode you see half the lines (interleaved) encoded using "inter" encoding based on the previous, interleaved line. Field dominance determines which lines, even or odd, contain the reference macroblocks used for the inter encoding. In other words, progressive footage on the GH1 is actually a hybrid of interlaced and progressive, even when the progressive flag is ticked. An editor will need to know (or figure out) which lines are dominant to be able to correctly decode the images. It still looks progressive because both parts of the frame were captured at once - this is just an encoding issue.

Chris
 
cbrandin said:
I've looked at the frames produced by setting progressive mode on the GH1. It turns out that the encoding is a little different than what you would expect with progressive footage. Normally you would see all macroblocks in an I frame encoded using "intra" encoding. With the GH1 progressive mode you see half the lines (interleaved) encoded using "inter" encoding based on the previous, interleaved line. Field dominance determines which lines, even or odd, contain the reference macroblocks used for the inter encoding. In other words, progressive footage on the GH1 is actually a hybrid of interlaced and progressive, even when the progressive flag is ticked. An editor will need to know (or figure out) which lines are dominant to be able to correctly decode the images. It still looks progressive because both parts of the frame were captured at once - this is just an encoding issue.

Chris
Click to expand...
This explains why Corel Visual Studio X4 could not render with one of these files inserted into the timeline. I thought at the time (about a month ago) I posted this problem that there was some bug somewhere... Al
 
cbrandin said:
I've looked at the frames produced by setting progressive mode on the GH1. It turns out that the encoding is a little different than what you would expect with progressive footage.

Chris
Click to expand...

This is something I was not aware and seems pretty damn important.
The whole basis, as I understood it, for native 24(25)p was that the AVCHD encoder works BETTER with full frames than interlaced (PsF). This info seems to contradict that, although I guess that depends on your definition of "better".

I too had posted when the Blackout/Powell 24p hack was first released, that although the results looked excellent, my Decklink Blackmagic card using a Adobe CS5 editing station would not preview these files. I was forced to use the adobe player which was much more CPU-intensive and worse quality. Which is why, at some point, I gave up on native progressive I have intead been shooting with LPowell's 75Mbps hack, which is not native progressive, and listed hier:
http://www.dvxuser.com/V6/showthread.php?254789-75Mbps-GH1-Peak-Reliability-Patch-V2-update

The files created with this hack do not seem to break the norm rules in a way that offends the Blackmagic card, i.e. they can be played and edited fluently.
My 2cents on this progressive gordian knot...

best, ninetto
 
I may have misspoken in my previous post. I revisited the streams with native 24p checked, actually looking at the macroblocks themselves, and they are indeed all intra encoded I frames. However, the headers are a little strange. I noticed that the GH1 chops 1080 frames into multiple slices, whereas the GH2 does not, and 720p modes are not on either camera. Another interesting thing - the wrapped progressive footage (the 24p over 60i) is not actually just progressive in a 60i wrapper - the I frames are actually encoded as interlaced, as I explained in the previous post. I no longer have a GH1 so I can't test this any more. Sorry if I caused any confusion.

Somebody (I think it was Barry) discovered that color-subsampling still produce the results that are expected with interlaced streams, not what you would get with true progressive streams. I think the native 24p streams sit somewhere between true progressive and interlaced. Probably because some of the encoding steps still operate assuming interlaced output.

It doesn't surprise me that some decoders work with these streams and some don't. Some information about streams is repeated at different levels of headers. Depending on which representations of information are used, decoders will interpret things differently - especially if information is inconsistent. For example, the TS header reports type 4 streams (1080i, it should be type 6) for streams with native 24p checked, but the AVC headers show 1080p. These types of inconsistencies are more common than you might suppose.

Chris
 
Last edited:
For someone who doesn't understand field dominance and intraframe stuff but would like to use Ipowell's native 24p patches, log and transfer with FCP and convert to ProRes, is there anything I need to know about for editing/exporting/etc? Or would someone mid breaking this down in more laymen terms as to what this means? Thanks guys.
 
You must log in or register to reply here.
Back
Top