Personal View site logo
GH4 4K Panasonic video camera, User reviews and opinions
  • 1273 Replies sorted by
  • AdamTAdamT

    I'll say it again: IMO time stretching FHD 60p to 96fps looks a lot nicer than 96fps VFR.

    Here's another example:

    I have no idea what's causing that color weirdness in the VFR shot since I was shooting by window light only.

    btw, to get 96fps from 60fps on a 24fps timeline, set time stretch to 400%.

  • KobKob

    In

    the guy compares various scenes shot by GH4 vs. BM 4K.

    I find the BM4K results more pleasant to the eye - less "metalic" and less artificial highlighting. Also, in the last white flower scene, the BM4K shows a bit better resolution.

    Are the differences might be attributable to the different lenses used, to the camera setup or are they in the camera technology itself?

  • kevin_kirchmankevin_kirchman

    That video above is kind of ridiculous. They are using a 42,000 dollar lens on the BM4k and a 200 dollar kit lens on the GH4. That being said, its still respectable how it looks.

  • KobKob

    @kevin lol :-) I did not realize that the Canon lens is so high-end...

  • chauncychauncy

    that lumix lens is like the worst lumix lens available.

    I'd like to see blackmagic vs gh4, all of them bmpc, bmcc, pocket. I own the bmcc and have seen nothing to convince me the gh4 is capable of a better image. Then again gh4 is ridiculously versatile, not bm's strong suite.

    But early days, raw is king, but will be interesting to see 4:2:2 gh4 4k when it comes along.

  • VeskuVesku

    Has anyone tested properly and carefully how much GH4 1080P is better than GH3 1080P. I have not yet seen any very good comparison frames. Many has said that GH4 fullhd is not much better than GH3. Sony RX10 for example has cleaner fullhd and better resolution than GH3. And ancient Panasonic camcorders (TM700) has more resolution too.

    GH3 must do some line skipping or color reduction because of slower sensor.

    Everyone is saying that 4k downsampled to 1080P is much better than 1080P from camera. Is the GH4 scaler still so bad that 1080P is not good enough despite of very high bitrate and full sensor scan. 1080P It is still important if one wants wider video with full sensor width or 60P for sports or slow motion. I know that from 4k we can have perfect 4:4:4 fullhd but GH4 fullhd should be very good too.

    In my tests with GH3 1080P video is about 85-90% of ideal fullhd resolution and colors are not perfect either. GH4 should be much better straight from camera.

  • Vitaliy_KiselevVitaliy_Kiselev

    GH3 must do some line skipping or color reduction because of slower sensor.

    It does not do any skipping or "color reduction". It just combines pixels, same as GH4 is doing.

    And yes, rescaled 4K will be always better as algorithms for such rescaling is better than just simple ones used for on sensor scaling.

  • VeskuVesku

    @Vitaliy_Kiselev

    Gh3 can read whole sensor (4:3) 12bit RAW data about 6 times/second. In 60P it must read sensor 10 times faster. Do you know how it combines pixels because so much has to trough away? Algorithm is good because GH3 60P is very good indeed. I dont think it has time to read all the sensor pixels like RX10 does. GH3 has some aliasing like it skips something. I wonder if GH4 reads sensor better because it has faster sensor and more horsepower to do scaling. Many still says that GH4 fullhd looks the same as GH3.

  • Vitaliy_KiselevVitaliy_Kiselev

    Gh3 can read whole sensor (4:3) 12bit RAW data about 6 times/second. In 60P it must read sensor 10 times faster.

    Huh, sensor is not so dumb thing. It is not faster (as datarate is similar) it is just smaller resolution at higher framerate.

    GH3 has some aliasing like it skips something. I wonder if GH4 reads sensor better because it has faster sensor and more horsepower to do scaling. Many still says that GH4 fullhd looks the same as GH3.

    It does not skip anything. And aliasing just happens due to pixel combining used.

  • VeskuVesku

    @Vitaliy_Kiselev

    Sorry, this is going offtopic but is interesting. I think the sensor must read all the 12 million pixels before it can combine them. Could GH3 or even GH4 read all these 12 million pixels 60 times per second. Obviously GH4 can not read them all at 96fps because 96P is so much softer than 60P. There is no 4k 60P by same reason in Gh4.

    But back to main question. Is the GH4 fullhd better than GH3? I will use it anyway with sports and wide shots.

  • Vitaliy_KiselevVitaliy_Kiselev

    I think the sensor must read all the 14 million pixels before it can combine them.

    Usually no.

    It is best to make new topic.

  • aldolegaaldolega

    Is it possible that the only reason 96fps is softer is just that it's way more frames to pack into the same datarate as lower framerates? Meaning less space for each frame (or GOP I guess)? I guess that wouldn't explain the increased aliasing though.

  • Vitaliy_KiselevVitaliy_Kiselev

    I guess that wouldn't explain the increased aliasing though.

    It can be just different sensor mode, like one that actually produced 800-900p resolution that is scaled back to fullhd using main LSI.

  • VeskuVesku

    From http://www.slashcam.de/

    GH3 and GH4 fullHD samples looks so similar that one could think that they use same sensor sampling. Could it be so.

    gh4-c100-gh3.jpg
    1984 x 1058 - 538K
  • theconformisttheconformist

    Hey all,

    So received my GH4 one week ago and have been pairing it with the Panny 12-35 & 35-100.

    All I can say is this is THE camera for me, and I suspect 99% of shooters out there. There are ways it can be improved slightly, but with so many new and great features for video, I expect the GH4 will be the standard for years to come, the other guys are so far behind it's not even funny.

    Obviously the VBR mode is a BIT lacking (seems just shy of the clarity of 1080p) but it's the highest quality 96fps mode I am aware of anywhere near the price point.

    As for the 4K - just incredible. The detail blows everything else a filmmaker without $20K+ to spend on a camera would consider.

    I was on the fence between the GH4 and BMPC (once FPN issues are resolved) but I know I made the right choice. I don't need 4K raw. I'll grab the Shogun when it comes out for 422 and call it a day.

    There is really only ONE thing I would change about the GH4, and it's that I would have put it in a cinema camera body. I'm not a hybrid shooter, and I suspect most people who are interested in this camera aren't either. I hate having to have separate devices for monitoring, sound, etc. If Panasonic made a dedicated video camera with these specs, they'd have a real winner.

  • SeekingHeartwoodSeekingHeartwood

    @AdamT thanks for the heads up about time stretching 60fps to 96, substantially better, and i've also had that weird light flickering when shooting in 96fps, i usually have to change the shutter speed to get rid of it

    do you have a good suggestion for shutter speed when shooting in 60fps with the intent to stretch to 96?

  • SuperSetSuperSet

    @AdamT Is there an equivalent Speed/Duration in Premiere Pro to 400% AE's time stretch?

    @Kob That BM4K image definitely looked more filmic, for sure.

  • mpgxsvcdmpgxsvcd

    Here is a video I shot last weekend. This is mostly 4K @ 30 FPS but there is some 96 FPS and 60 FPS slow motion. I didn't do any color correction at all. Just slapped some clips together and set it to music.

    Ironically the guy next to me had the GH4 but he had no clue what it was capable of. He set it to its lowest quality and bit rate and filmed a few static shots.

  • rNeilrNeil

    @SuperSet ...

    In PrPro, either right-click on a clip and choose Speed/Duration or highlight (select) a clip and use the main menu bar, Clip->Speed/Duration.

    Or simply click on the clip, use the hot-keys " Ctrl/Cmd -> R ". The same dialog box pops up any way you choose it ...

  • SuperSetSuperSet

    @rNeil Thanks, man. What's the equivalent speed duration in PPro to 400% AE's time stretch? I think PPro works in reverse from AE's time stretch so it's actually a lower number.

  • AstroAstro

    @Kob @chauncy Have a look at this test from an experienced filmaker that owns 3 cams (5D Mk3, Red Epic and GH4) I have seen BMCC comparisons too. the BMCC does do a nice image, but it's not the cam for me anyway.

    Amazingly the Gh4 renders quite a lot more detail than even the Red Epic (have a look at the last 300% stills) this is not just sharpening ...but actual detail that both the Red Epic and 5D Mk3 simply dont catch and render

    Also this short film

    You can also download an app that links to youtube clips and downloads them in full 4K the clips still has a lowered youtube bitrate, but its full 4K size. http://www.4kdownload.com/download And he also posted a link to the original files. Cheers Astro

  • anti12anti12

    "Amazingly the Gh4 renders quite a lot more detail than even the Red Epic"

    @Astro thats because the Epic is focused on the tree not the roof

  • v10tdiv10tdi

    @joesh

    for x in * ;do echo -ne \x33 | dd conv=notrunc bs=1 count=1 oseek=890 of=$x; done

    I executed this script with no errors that I could detect, however LR v5.4 was still unable to preview the RW2 files.

    I downloaded EXIFTOOL and changed the model to DMC-GH3 and Lightroom was able to work with the files.

    cd directory with RW2 files

    then type

    exiftool -Model=DMC-GH3 *.RW2

  • rNeilrNeil

    @SuperSet

    You are right ... as PrPro looks at is as to how fast (relatively speaking) you want it to become compared to the 'normal' speed. So ... thought about it for a couple seconds, thinking that as 400% is a double-doubled 100%, going the other way ... a half-halfed setting would be 25%. Tried that, looks about right I think. Good enough for starters or "to be getting on with" ...

  • joe1946joe1946

    GH4 4K UHD 2160p test of the Voigtländer Nokton 25mm f/0.95 close focus @ F2.8.

Add a Comment