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 GH2 "Stalin" hack development topic
  • 939 Replies sorted by
  • >It's probably from the sensor's higher native frame rate (60fps) which gets cut into 24p in firmware (correct?) whereas the GH1 has a sensor with 24p natively.

    I think that this is incorrect.
    Both GH1 and GH2 have 24p native sensor modes as I know.
    If you feel that motion blur is somehow different you need to perform scientific test.
  • ok, I don't know for sure. there has been some comparison and there seems to be a difference in the way the motion looks but it's a lot of opinion rather than scientific evidence. Can you suggest a good test for this?
  • > Can you suggest a good test for this?

    Mount a GH2 on top of a GH1 with the same settings and pan.

    It's already been proven that there's NO difference in 'strobiness' with the GH2 vs GH1. Any strobing people are seeing is merely from a purer, sharper image, when the settings are wrong (not 180 shutter).
  • Also sharpening can make edges look more strobey (especially on LCDs which hate high-contrast in nearby pixels), try turning it all the way down.
  • There is also the b frames factor
  • Yeah No sdi on the sf100 what a weird deal:(
  • >There is also the b frames factor

    And what difference do they make apart from better encoding efficiency? No, strobing won't be and ISN'T caused by b-frames, and there's no actual difference in strobe between the GH1 vs GH2. Like _gl said, it's mostly likely to be a bad monitor in combination with a lot of in-camera sharpening. Anyway, this is the hack thread so we should probably continue discussing it elsewhere. ;)
  • Hi Vitaliy, could you publish high resolution images of your's "dump" gh1/gh2 bodies. It's very interesting to analize audio and remote controle related hardware....
    and do you have service manual for gh2?
    Thanks in advance!
  • >It's very interesting to analize audio and remote controle related hardware....

    They do not have such "hardware" as modern LSI is so much integrated :-)

    >do you have service manual for gh2

    I won't provide comment on this.
  • It affects the sense of motion.
    The continuity of micro detail.
    Large scale strobe is on the altar of 24/25p.
  • "Mount a GH2 on top of a GH1 with the same settings and pan."

    Watched all the videos and did it myself. Looks different. I'm asking for a scientific approach, not another subjective study where half of the viewers see the difference and the other half don't.
  • Whatever setup one uses to compare GH1 and GH2 seems subjective. Interpreting results is subjective, too. In other words... don't waste your time on it that has little or no value which is subjective statement by itself.
  • Here's a very interesting report that everyone who runs a hacked GH should read:
    http://www.tomshardware.com/reviews/sdxc-sdhc-uhs-i,2940.html
  • At first glance, one might conclude that the USH-I cards are the way to go. But look closely at the chart "Write Throughput: h2benchw 3.13" on page 12. It looks like the Sandisk Extreme (16GB, Class 10) would be best for our purposes.
  • I was so happy reading about native 24p for GH2, but now after I used GH2 in 24p mode with glidetrack and glidecam I understand that it's very bad for this kind of videos. For static shots it's ok... 1080 50i is pretty good, hope soon we will have possibility to use this with 24Mb bitrate. 24p mode for me is very strong... no motion blur between frames...
  • >At first glance, one might conclude that the USH-I cards are the way to go...
    UHS-I is a new standard, so if the hardware of the GH1/GH2 does not support the differences will be subtle. Care must be taken to make this type of comparisons.
  • >....But look closely at the chart "Write Throughput: h2benchw 3.13" on page 12. It looks like the Sandisk Extreme (16GB, Class 10) would be best for our purposes.

    IMHO the only important test for our purposes is the sequential write. The critical moment is precisely that of writing, and in any case never happens to write and read simultaneously.
  • Min write speed seems like the most important parameter.
  • Hi Vitaliy,
    (about audio hw :)
    >They do not have such "hardware" as modern LSI is so much integrated :-)
    hardware - I mean all that isn't firmware :)
    I'm interesting in audio path through camera: in GH1 audio path
    for ext. mic (most interesting):
    "remote pcb" through ext.mic pcb through audio pcb
    for int. mic: mic pcb through audio pcb.

    as I guess all audio processing is limited to this pcbs (including AGC): remote pcb, ext.mic pcb, mic pcb, audio pcb.
    Schematic Diagrams and Parts Lists provided for all this pcbs, excluding audio pcb.
    and this pcb is most interesting - there is main audio controller here...
    as you see all audio lays out of main processing board, all this board has quate clear interfaces. why not dream about simple modifications... like:
    - ext. mic with AGC to simple line level in
    - internal speaker to headphone out
    - unbalanced ext. mic to balanced mic or line level in
    -...

    some questions remain if I'm not misstaken about remote controle: separate remote control and ext. mic input.
    unfortunately from remote pcb Schematic Diagrams there are questions (which probably direct to firmware and hence to patching):
    is realy remote pcb not connected to flash pcb (from which contacts from main shutter botton -"shutter full" and "shutter half" - goes to main processor board?
    in this way main cpu make direct and separate decision about activation of shutter from remote controle by analizing resistance between contacts on jack (this mean thet this analisis should be done main cpu loop as looking to shutter full/half signals from flash pcb)!
    ...this is why I ask you about detailed images from yours dump cameras.
    Thanks for your time!
  • @iandrew
    I already answered to your question, sorry.
    Any info copy&pasted from any SM and not in restricted development forum will be deleted.
    I think that you know the reason for this.
  • Is it possible to imagine some kind of crop 1.1 mode to make a 2k or more. That would be great if we need to make some picture improvement with AE or something...
    Why can't we choose the size of this crop mode?
  • >Why can't we choose the size of this crop mode?

    Most probably because sensor can't do it :-)
  • Vitaliy (or anyone) just out of interest, when using the eyepiece viewfinder (I usually use the LCD) I've noticed that it seems to strobe through different colours. Is that my imagination or does it use a rotating colour filter wheel or something? Just curious, really - nothing to do with hack!
  • Noticed this as well. It seems to happen due to the small LCD and the lenses that focus it into your eye. If you look side to side in the VF quickly, you'll see the images separate.
  • >Noticed this as well. It seems to happen due to the small LCD and the lenses that focus it into your eye. If you look side to side in the VF quickly, you'll see the images separate.

    No it is not due to small LCD or lenses.
    It happens because you have b/w LCD and cycling colors backlight, as I understand.
    Look at any paper about DLP projectors - this works on same principle, and you can see same artefacts.