Personal View site logo
Driftwood - Experimental Series 2: Low Rider, Cluster v8, V9, Intravenus II, GH3onaGH2, AN, Boom
  • 1008 Replies sorted by
  • @driftwood. Yes. SH fail was in NTSC. All my testing is in NTSC.

  • Thank you @Zaven13 for your impeccable testing. Lookout for another surprise shortly, Cluster 'Spizz' ...

    PS Wassit SH mode fail in NTSC?

  • @driftwood. Nick I just tested the CL X Nebula 6GOP with Panny 14-140mm and Sandisk 64GB 95MB card. As you said, the best HBR/FSH yet in any of the Cluster settings. Quality is great and the bitrates are high. The only mode that failed was SH but H mode works just fine. It works in IA mode also. Thanks. Great job.

    Update: I just did a Streamparser analysis of all modes and these are the highest frame sizes I have seen in any of the settings. You have outdone yourself again.

  • @hay... CM night had a blueish cast

  • *** NEW ***

    Here it is: The first trial of Cluster X NEBULA 6 GOP primed for NTSC. Is this the best HBR/FSH yet? Let me know testers. PAL editing of FSH/HBR to do...

    UPDATE: PAL modes tuned a tad for FSH HBR. NTSC users also use Trial 2 as its the same as Trial 1.

    Cluster X 'NEBULA' 6 GOP Trial 1 - setg.zip
    1K
    Trial 2 - PAL tuned - Cluster X 'NEBULA' 6 GOP - setf.zip
    1K
  • @driftwood, and hope of higher audio bitrates / better audio in 2k MJPEG mode?

  • Low Rider v1 footage snippets, anamorphic lens attached. Details outlined in the Vimeo description. Thank you @driftwood @Vitaliy_Kiselev, et al.

  • Just passing along the results of my test with Intravenous 2: the camera cut out after about a minute because of speed limitations. The memory card used was a Transcend Class 10 32 gb car. I was shooting 24 fps, FSH quality, and I'm fairly sure I was shooting at 6400 ISO, so I may have overtaxed it a bit. I hope this info helps with the development of the Intravenous patch.

  • How many minutes in average are you all getting on a 64gb card for:

    DREWnet clusters

    Intravenus

    Thank you all

  • forgot to mention - all on 24h 24p smooth -2-2-2-2

  • @driftwood - I tried cluster x trial 3 and cluster v9 boom today. Cluster x trial 3 failed while zooming with 14-42 panny at iso200. cluster v9 boom failed after 1 minute into the recording when panning. v9 boom was with tokina 11-16 at iso1600. i used sandisk extreme 64GB 45MB/s.

  • @driftwood is there a patch or possibility of a patch that would be a hybrid of aq1 & intravenus? Basically something that has the detail/super high bitrate of sedna with the same kind of noise as intravenus?

    sorry if this has already been asked, experimented with, or already exist

  • When driftwood says: "Generally, for 24p I use two of my settings when filming; e.g. Boom or Intravenus for mid-in tight shots and any of the same setting with Q set to default (Untick all Q related stuff) for high detail wide shots."

    Is the reference to Q here mean, in PTool, unchecking all 4 boxes under "Patches for testers" --> "Quantizer"?

  • @jhero Different GOPs require less bitrate. The bitrate required for Long GOP 12/15 doesn't need to be huge. At 32-40Mbps the new Cluster variations look pretty good. Try and take a look.

  • Hey apologies if this is an obvious answer but I can't seem to remember which setting yielded the best HBR output. Any one know off the top of their head? Thank you

  • *** NEW ***

    Cluster v8 DREWnet 'cbrandin Cinema Smooth matrix' and v9 'DREWnet' 'Boom' 4444 matrix 12/15 GOP variations are out on page one. X is still in development. Nebula versions of each coming.

    I am looking at ways of maximising HBR/FSH in Cluster X. This will probs be my last ever Cluster setting for the GH2.

  • Been in the Keys and Miami for a week shooting hotel and resort promo stuff all in IV2 and Sanity5.1. I'll post some screen grabs as soon as I can. Probably not till next week when I get back. So far so good. Everything is shot on Samyang Primes and the SLR Magic 12mm.

  • @rajamalik Did you shoot anything you'd be able to post? I haven't had a chance to try this out, and it seems there isn't much shot with it on here.

  • Field testing is officially rained out. As the GH2 isn't "splash proof" -- no point in risking it. They say tomorrow will be 72 degrees and sunny -- so it should work.

    Early this morning I used version 2 with talking head studio shots -- and everything looked outstanding. That's about as low on codec demand as you can get -- but everything worked and the 24ph files spanned.

    Again, I'm really liking it.

    I'm now going to download trial 3 and will report back tomorrow.

  • @driftwood tested intravenus 2 yesterday. Thank you for this patch. Shall I use -2-2-2-2 std profile for this??? is it good to use intravenus 2 in wide landscape shots?? The cinema smooth matrix is so filmic. I felt I am using a mini arri. Thank you once again. I hope there will be more of it soon.

  • @Zaven13 Shorter GOPs drive bitrate higher and of course as you aware Longer GOPs means more prediction and more effeciency on the bitrate. Most Cluster 12/15 GOP FSH/HBR modes should be around 35-40Mbps for equivalent 80-90Mbps for Intra. Anything beyond that is normally unstable.

    When I finish Nebula you will notice the ratio for bitrates can go slightly higher.

    Frame Limits often come down and bitrates go up as the GOP decreases (and i frames per second go up!).

    Thanks for testing @Mariella

  • Cluster X seems to be a superb setting. With higher FSH bitrates this setting would be one of my favorites!

  • @driftwood. I tested Cluster X trial 3 with your latest tweaks using my usual test scenarios and all modes worked fine including zooming with 14-140 lens and IA mode. The only thing that bothers me is the low bitrates of HBR and FSH which is mid to high thirties. I get 20MB higher bitrates for HBR and FSH with CL7 Nebula Sharp2 at mid to high fifties. May be bitrate is not a big factor in image quality but if someone can explain how the combination of bitrates, GOP length and frame rates effects the image quality and the fact that 20 MB higher bitrates may not make a big difference depending on other factors, then I will feel more comfortable. I am learning and striving for the best. Cheers.

  • @gravitate, the MJPEG is a VERY friendly codec to edit.

    I just looked at two interviews, one after the other, one shot with 2K HD (MJPEG) and the other shot in 1080p30 HBR AVCHD. The HBR was "clinical" and even too sharp (this, with sharpness at -2 using Standard Profile) with the tack sharp 20mm 1.7 lens on the GH2. Women don't want their "talking head" videos to reveal every pore and grey hair. :)

    The MJPEG lets me: a) access the clips with the stills b) drag & drop all my footage & not have to use some import utility c) no transcoding to a mode edit-friendly codec - no duplicate source files / extra hard drive space wasted d) are quite good video quality e) are larger than Full HD so I can do a subtle push from 94% to 100% on the source clip in post, which I did. And I can execute this push at the exact right moment in the video. I can zoom even more in a 720p delivery project.

    I just need the audio to be much better than the low bitrate it is currently set at.

    I am curious as to the "patches made for my situation." Can you elaborate on that? I admit, I am new to hacking & patches so there is a LOT I do not know.