Personal View site logo
Make sure to join PV on Telegram or Facebook! Perfect to keep up with community on your smartphone.
Please, support PV!
It allows to keep PV going, with more focus towards AI, but keeping be one of the few truly independent places.
Hacked GH1 and Final Cut Pro
  • gohanmjgohanmj
    I am starting this thread because I have done a lot of research on the GH1 hack but I cannot find any information on what should be a common real-world problem with the hack. Here is the problem and all relevant information:

    1. I have hacked my previously unhackable GH17 with all of the popular patches and a couple of my own modified patches using ptool 362.

    2. Of all patches I have tried, lpowell's max latitude 100mbps patch seems to have the highest average bit rate according to Media Info.

    3. Media Info says the max latitude patch has a max overall bit rate of 108 Mbps and averages 20.8 Mbps

    4. Media Info says that footage from my unhacked GH17 has a max overall bit rate of 18 Mbps and averages 11.8 Mbps

    5. Once my footage is transferred into Final Cut Pro 7 using Pro Res 422, the data rate in FCP7 of my unhacked footage averages 17 MBps while the data rate of my hacked footage averages about 13.5 MBps.

    6. This discrepancy makes no sense to me. And moreover it would seem that if you edit in FCP7 that you are better off not hacking the GH1. Is this assumption correct? Can someone share more information on this issue?
  • 6 Replies sorted by
  • nomadnomad
    ProRes is not constant bitrate, it is mildly adaptive. So, check with the same content and make sure that you have the same settings, not like ProRes 422 and ProRes HQ in another case. Might be the reason for discrepancies.

    I had 90 mbps with the 240 GOP1 version from Driftwood on GH2 (very high detail) and it translated to around 135 in ProRes. No need to worry, the hacked footage should look better after transcoding too.

    But you should have a look at Premiere Pro, which can handle the footage natively. All that transcoding is a major PITA in my opinion.

    And finally: trust your eyes!
  • gohanmjgohanmj
    Thanks for the reply Nomad. All settings are the same in FCP7 for both the hacked and unhacked clips. So I still don't understand why the data rate goes up in FCP for unhacked clips and down in FCP for hacked clips.

    Just to add a little more spice on the situation, I took a few more test clips today under identical conditions with 1 exception, I did not tick the 24/25fps box in the max latitude 100mbps patch. The result? Pro res still decreased the data rate, but this time it dropped to 17.2 MBps on all clips instead of 13.5 MBps. This represented a .1MBps improvement over the unhacked clips (17.2 MBps vs 17.1 MBps).

    The eye test revealed a little less fuzz in the shadows of the hacked clips vs the unhacked clips.

    FCP is also showing a field dominance of "upper odd" for all hacked clips rather 24/25fps is ticked or not in ptool 3.62. The fps are correct for hack clips in FCP, (23.98) when 24/25fps is ticked in ptool but field dominance should not be "upper odd". So my only guess is that ptool is not correctly deinterlacing or something is going wrong in the conversion with pro res. Either way, it does not seem like the 24p hack and editing in FCP 7 will work together.

    Any Thoughts????
  • nomadnomad
    PTools is not responsible for deinterlacing at all. This is a known bug in FCP, it always assumes 1080 to be interlaced. You can change that with one right-click, it's only a flag.

    I got around 17.6 MB/s with the latest GOP1 patch on very high detail and ProRes HQ increased the size by about 20%. So, that looks quite right to me, given the fact that both are Intra and that the quality of the original footage is stunning already. Only shows how far Driftwood got with his patch.

    I really don't understand what is going wrong for you, I've never seen bitrate decrease after ProRes HQ encoding.

    But I left FCP behind (after ten years!) for PPro anyway…
  • gohanmjgohanmj
    Just to clarify Nomad. Your saying that the fcp7 readout of "upper (odd)" in the field dominance column of all my hacked GH1 clips is erroneous? And that my clips are truly 23.98 fps with no crappy frames?

    Pro res HQ does not decrease my bit rate. I said that in Media Info the Mbps reading on my hacked clips seems to average around 20 Mbps but when I log and transfer the same clips into FCP7 using Pro res 422, the data rate column in FCP's browser shows average data rates of about 13.5 MB/sec. Conversely, the Media Info reading on my unhacked clips averages around 12 Mbps but once those same clips are transferred to FCP using 422, the data rate reading in FCP averages 17 MB/sec. I don't understand that.

    Could you direct me or post a link to the latest GOP1 patch you referenced?
  • nomadnomad
    Yes, FCP is telling you BS. Change the setting to "None", display at 100% and just step though your footage. If you don't see interlace artifacts in fast motion, it's progressive.

    The latest by Driftwood is here: http://personal-view.com/talks/discussion/1002/official-low-gop-topic-series-2/p5 (somewhere after the first third).

    And ignore your bitrates, just enjoy the images ;-)
  • ghanz301ghanz301
    Since it is unhacked, it might be that FCP7 is registering the void interlaced lines as pixel-data due to the wrapper.
Add a Comment