Personal View site logo
Wiki discussion
  • 40 Replies sorted by
  • @balazar, great work, love the charts with comparison of patches.

    Is it possible to also add media times to chart as to each recommended card? I shoot a lot of long timelapse and look for stable spanning for shots as long as I can until card is full.

  • Thanks, @jeffryalan. A simple bit of math should suffice to convert bit rates to recording times.

    143 divided by the bit rate in Mbps = minutes per gigabyte

  • So, is that a variable affected by minimum/maximum read/write throughput rates?

    For example, the Sandisk Extreme Pro 16Gb 45MB/s has a sequential transfer rate for reading at the top end of 44.8 MB/s and a low of 37.2 MB/s, but the reason that it fails spectacularly for the GH2 is that it's WRITE throughput goes from 36.6 MB/s to a miserable 800 KB/s. The speed on the card does not necessarily reflect the actual average write throughput.

    We could assume a potential media allowance through simple calculation but will that be only a rough estimate or are benchmark tests needed to find the actual length of media? It would be easy then to compile a chart with min/max times per recommended card in relation to the significant patches.

    (ASQ) Another Stupid Question: is 143 a constant for all calculations?

    http://www.tomshardware.com/reviews/sdxc-sdhc-uhs-i,2940-12.html

  • 143 is a constant. Say you are shooting with one of Driftwood's 146 Mbps patch settings. 143 / 146 Mbps = 0.98 minutes per gigabyte. A 32-GB card holds about 30 GB, so that gives you 30 * 0.98, or about 29 and a half minutes.

    Card performance is another issue altogether. There are many dimensions of card performance, so it is hard to predict which cards will work with which patch settings, even if you had detailed benchmarks. Card recommendations are based on testing of actual cards while shooting video in the GH2.

  • That is a pretty cool factoid, 143 divided by bit rate, multiplied by card space in GB = time left for shot.

    Thanks for that @balazar

  • Great charts in the new wiki.

    Is it possible to add a new field in the "effective bitrate/spans/vmm80% 30p/ETC" chart that states whether a patch+card combination allows for in-camera playback or not?

    It is quite important for many collaborative shoots and many people might want to choose a lower bitrate patch to be able to keep this functionality working.

  • If someone wants to figure out which patch settings support in-camera playback, that can be added to the wiki. I don't think the memory card is a factor.

  • I don't have the camera at the moment, I've been borrowing one from a friend. I'll test the next time I get hold of it, but I only have access to 30Mb/s Sandisk cards.

    Depending on the kind of shoot it's a feature that you might need, so knowing it a concrete patch disables it is very useful to help choose the right patch.

  • Where do we put information about the undocumented hack parameters? Let's say I would like to share something like: "Quantizer for 1080 modes" - Regulates the size/quality of I-Frames in 1080 modes. Does not seem to affect P/B-Frames.

  • The "Quantizer for 1080 modes" setting most definitely affects P and B-frames.

    As for a wiki, there is a wiki page at linked from the tooltips of several of the PTool settings, but the page is essentially empty and it's asking for a password to edit it. I'm all for having a wiki page somewhere to consolidate findings about the various PTool settings. There's also an Encoder Research page in the new wiki, but I don't have permission to edit it.

  • See? That's the kind of info that shouldn't be "reserved" for the "elite". ;)

    I'm really not complaining because when I was trying to build my patch I asked around by PM and I eventually got all the answers by you, by driftwood and lpowell, yet I believe it would be an asset to the community if that info was organised and readily available.

    As for that parameter, I run a few (uncontrolled I should say) tests yesterday and got the impression that the only influence on P-Frame size was the Quantizer Tables. Perhaps shooting a static scene might have something to do with it. Anyway it's a pity that this information is so spread and that the wiki is empty regarding these matters... :(

  • @balazer : Been looking a bit more deeply at the wiki and it has a lot of useful and well organised information, but just not for patchers. I don't know how many of us are doing it on our own and need more info on the Patch for Testers parameters but I'd really like to see it.

  • I could create a new wiki page for GH2 testers, but I think we should put the question to Vitaliy, since it looks like he already had some idea of where the information should live and who should be allowed to update it.

    There is a lot of information locked up in individual testers' heads that they haven't bothered to share in the forum. It would be great to consolidate that info.

  • We'll certainly have such section. Just wait for new version a little, ok?

    As we'll need to move many things out of developer section to amke it available for active settings authors.

  • Great to hear that! Thanks.