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.
Official Low GOP topic, series 2
  • 1022 Replies sorted by
  • @driftwood
    Thanks mate!
    If I think from the result of my restrictive test, and used with 24p, I will think that this is very stable, but if scenery like many leaves which move violently is recorded, the former Q=16 version may be safe.
    Since I would like to use 80%mode, I like Q=16 version.

    Thank you for your liking my DIY steadicam.
    But I think that Merlin is easier-to-use. ;-)
  • Driftwood: "I want to buy one off you!"

    Me too!
  • Hi Guys....first of all a big Thank You to all those who have contributed...be it developing to testing and sharing results, you truly have changed the game.

    Following on from @Driftwood comment "I think the bigger the card in some respects could also be the problem."
    This is something I have realized after extensive tests on Transcend cards.

    My 32gb Transcend card persistently fails after 8 seconds when using "QuantMeBaby" and "reAQuainted" and sometimes quicker when using the death chart.
    However I decided to try my 16gb version of the same card and I cannot get it to fail at all in the same scenes or any scene to be exact.

    I found this odd myself but I do hope it sheds light for Transcend Card users....

    Try The 16gb Transcend cards!!!!
  • @bkmcwd yeah when you gonna release that beast...
  • @paglez
    Thanks for beeing my 2nd customer! :-)

    @PerryWilson
    Of course, I will... ;-)
  • I just join this topic, thank so much for all the info. As this is already series 2 can any one please tell me where I find the latest guide/download link to hack the GH2 with 50mbp/s GOP 3
  • @bkmcwd
    I need to test your Q15 setting. I bet shadow detail looks great due to larger i-frames in lower light.

    Does it pass Stray's Extreme chart at ISO3200, 1/1600 shutter speed using panasonic lens?
    I know this is an extreme test, but it certainly gives us an idea on stability.
  • I've been shooting the whole day with the reAquainted patch on a sandisk 30mb/s
    The only occasion when I managed to get the camera stop recording was when shooting in 80% mode, after about 10 seconds.. Buffer overflow, I suspect.
    (I did low light, highly detailed scenes with and without telecrop.)
    Works like a charm! The biggest issue so far has nothing to do with the patch and everything to do with how premiere cs5 handles the export / what settings to use for best results.
    I'm going to try my 45mb/s sandisk card sometimes, and see how it fares with similar settings (both are low level formatted)
  • @RRRR thanks for the information!
  • @proaudio4
    Thanks always! :-)
    In my test, it has passed the test which moving Stray's chart finely at ISO3200,1/800 using pany 14-140mm as screen shot.
    I think that it cannot pass in use of 80% mode.
    When you cannot be satisfied, you can also try Q16 version.
  • I am trying to create a hybrid patch between Driftwoods GOP 1-176 reAQuainted and 44MB C_Brandin's 12GOP, so that the the 720p is all C_Brandin's (for spanning) and the 24p is all you're reAQuainted set for my cinema stuff but Im a tad confused about one part

    Should I change the video to buffer to 300000 like it is set in the Driftwood patch,
    Or leave it at the default like the C_Brandin's patch?????????

    ....Again--My goal is to only change my 24p modes to GOP1 176 and keep all my 720p at the 32Mb span settings from C Brandin's 44MB patch.
  • @No_SuRReNDer
    I have been considering trying this out too but stumbled upon the same problem. If you figure it out, please do post.
  • You should be good to go changing the video buffer to 3000000 but try it :-)
  • @RRRR Try the 100M INTRA for 80% (30fps) patch. Its called 'Aquamotion' to be found
    http://personal-view.com/talks/uploads/FileUpload/c3/4180ac568d008828627fd3fd110db5.zip
    Forget all the other settings - only the 1080p24 100M for 30fps is the one to use (set your cam to 80%)
  • @Driftwood and all those using reAQuainted patch

    What are your experiences with noise and this patch? The attached .png's show noise level in blue channel from some video I shot today using reAquainted. ISO was 160, cant exactly remember aperture, but outside it was around f4-5.8 and inside f2-2.8. Lens was voigtlander 25mm.

    The noise cleans up with MB denoiser without softening the image noticeably, and although I'm an amateur it definitely feels like there is more scope for cc and grading with this patch. I guess I just want reassurance that this is normal with the reAQuainted patch. Also I shot some stuff indoors at night time, the only lighting being crappy energy saving light bulbs. ISO was under 1600, lens minolta MD 50mm at F2. Again there was a lot of noise in blue channel and there were also bands/lines of noise in some areas (they do not stretch across the whole image, just parts).

    Cheers
    Blue channel Noise_00535.png
    1920 x 1080 - 1M
    Blue Channel Noise01_00216.png
    1920 x 1080 - 1M
  • @bkmcwd
    Thanks for the information!
    I will test and let you know.
  • Well, trying to export into any of the compressed codecs like MPEG-4, H.264, etc seems a bit fruitless for showing quality comparisons, so I'm uploading to FTP three versions of an exterior clip I shot. There will be the original .mts file plus 2 ProResHQ clips, 1 that is very lightly graded, as in barely and one that has a Kodachrome-like look. These are big files, so hopefully you have a fast connection. They should be up by tomorrow morning. I'll post the link to the files then.

    As for my candlelight test with the Driftwood patch, I'll need a bit more time with that as I want to do the same type of thing where I provide original files plus a normal grade, plus a noise reduced grade. I abandoned the RED MX idea as I think others were right, what's the point. I thought it was more important to try different GH2 look settings and compare those, so that's what I did. I'll post a link as soon as those are ready.
  • Please note, topic will be closed soon, and new series started.
  • Is the new series a continuation of the most recent Low GOP patches and tests? Is it a way to clean up threads? Just haven't been around long enough to know how this works here. :-)
  • @smsjr , I think it would be ok to post some png files of the scene for those that can't download it and it would be faster also. Perhaps I did misunderstand your above post, did you abandon the test between the the gh2 and your Red MX. If so I think it is an error, for many reasons.

    I think the hacked gh2 is in need of recognition, what I mean is that until now is that many people either don't know about it or some just dish it because it is not a Canon. without any comparative benchmark we just have to speculate. The gh2 is a true DSLR 2, true high definition, no moire/aliasing, proper monitoring and autofocus. Now with the hack is is even much better. But many are still thinking that DSLR technology will have these ineherent flaws because the Canon do. It is as if only the Canon exist. Look how many are or starting to claim the 1 DX is going to be the second generation dslr.

    If we can have some good test of the gh2 against one of the industry standard camera, we would have ground to change the perception towards that little wonder cam. Do I expect the gh2 to beat the redmx in terms of resolution or dynamic range, hell no. But if it approaches it in any way I think it will be no mean feat.
  • @driftwood: Cheers! I haven't tried it out yet - I downloaded it so it's definately on my mind. I just thought I'd see how your high end patch makes my camera behave in different settings (even ones that are not intended for the patch in question). I think it could shoot 7-8 seconds or so of 80% without issues - I'll try to do some really short takes to see how it works out.
  • @driftwood

    This was shot with voigtlander 25mm in EX tele mode, reAQuainted patch, ISO 320, f.2, WB set to sun, smooth -2-2 -1 -2. I was using sandisk extreme 32GB 30mb/s

    There are bands of noise that flicker over the plant plot like tv interference. Is it a bad idea to use EX tele with reAQuainted or do I have a problem?

    Many thanks to anyone that can help
    EXtele_Bluechannel_noise_00367.png
    1920 x 1080 - 2M
  • @ignatius

    I haven't yet tried reAQuainted but some ISOs on the GH2 (320 included) seem to be buggy (http://personal-view.com/talks/discussion/1280/gh2-iso-noise-bug-at-320-640-and-1250). If you get there from the immediate upper ISO level (400 in this case) there seems to be no problem. Get there from another ISO and then: Ugly Noise galore! EX tele is sampling the central part of the sensor making things even worse. This could well explain the bands of noise in the cup. Are the other two pics (the ones on your previous post) taken from an EX tele shoot as well?
  • @driftwood, any led way on a slow mo patch that works? I am shoot some stuff next weekend in a boxing gym, will love to have something that I can get some good slow mo shuts
    Any one find/have something that works?

    Thanks.
  • @HillTop1, look above - driftwood posted a comment with a link to the aquamotion patch - which is designed for 80% shooting.
This topic is closed.
← All Discussions