Jump to content

deezid

Members
  • Posts

    840
  • Joined

  • Last visited

Posts posted by deezid

  1. 3 hours ago, AaronChicago said:

    The biggest surprise for me with this camera is the outstanding color reproduction. Skin tones are spot on. This matches up much better with my Red Dragon than the BMPCC4k. Would love to see some kind of 12 bit raw implemented.

    5.8K - 4000 ISO - 24p frame grab:

     

     

    ElaineTightJPEG.jpg

    Color on this camera is fantastic. Even by just using the standard V709 conversion. Can't wait for the ProResRAW update which would fix the NR issue anyway. We still don't know if it's 10 bit log or 12 bit linear...

  2. 4 hours ago, Mike Mgee said:

    @deezid Panasonic updated the S1 firmware to fix the issues you addressed. Can you please have a look and report back to us?

    Saw that, unfortunately only have the S1H here which doesn't have a new firmware announcement yet. 
    @Panasonic, if you read this: NR OFF Switch. lol

     

    Nick Driftwood is testing the beta version of the S1 firmware at the moment.

    6 hours ago, Lux Shots said:

    Kinda unrelated, but related. The latest update to Premiere Pro CC plays 5.9K HEVC/H265 footage at 100% quality like butter! I tested this with V-Log footage graded with a Varicam LUT with minor tweaks to taste.

    Pleasantly surprised ?!

    Nice. That would mean finally proper Cuda support like Resolve.
    Would love to see some benchmarks and tests.

  3. 22 hours ago, helium said:

    Here's Mr. Deezid's grade -- or something or other, no idea why he offered the shot here as he does  -- of an S1 v. BMPCC 4K shot, kindly provided by another participant.  See that brown potato quality of the BMPCC 4K  and the wonderfully "organic" quality of the S1?

    https://www.eoshd.com/comments/topic/34763-panasonic-s1-v-log-new-image-quality-king-of-the-hill/page/25/#comments

    That was my Lut for the Pocket 4K which for some reason is quite close to a V709 conversion which I used on the S1 clip. 

    AlexaX-2, IPP2 Rec709, V709 and probably other Rec709 compatible conversions have many things in common:

    • Great color definition
    • Great tonality
    • No or a very light color cast
    • Smooth roll-off
    • Even colorful lights are rolled off properly

    The standard BMD V4 Extended conversion on the Pocket 4K at least looks entirely different. And inside ACES the camera performs even worse.
    Tried to make P4K samples (around 100) look like corrected using my Lut but can only come relatively close. Can't really match it entirely, since there are too many hurdles to begin with.

    Maybe if I invest more time into developing it...

  4. 2 hours ago, Video Hummus said:

    I would give the S1H some time and a few firmware updates. They've added major codec upgrades and offerings in past cameras might not be different here.

    So close to a new project and nothing heard yet. ?
    Even borrowed a Pocket 4K from a friend in the meantime for other projects so I don't have to deal with ghosting. 
    Hopefully sooner than later.

    Quote

    offering full frame 4Kp60 as more competition comes out, or improve tracking AF.

    That's something that even the FX9 won't be able to do.
    Seems like FF sensor readout nowadays is still a bit too slow sadly.

    Edit: The EOS C500 MKII will do it.

  5. 20 minutes ago, thebrothersthre3 said:

    Its due to the difficulty he's having with the aces workflow for the Pocket as well as the brown cast. 

    Like @CaptainHook said, going straight to Rec709 on a yRGB timeline it's possible to fix all the issues mentioned. Also waiting for the latest work from @Sage, which looks like he can even go further, but that wouldn't fix any of the issues I get from working with ACES and the Pocket cameras atm and that would create issues with my upcoming project. Really on the fence right now...

    Also the Pocket 6K looks great after downscaling to 4K when Resolve scaling algorithm is set to smoother.
    The footage looks natural and basically unprocessed, the issues seen on the Pocket 4K like (moire/spatial nr becoming visible, sharpening halos) seem to be gone due to the larger sensor size and more important resolution.

    @crevice
    I agree the old BMCC looked more organic out of the box. No processing at all, really pleasant color and no clipping.

    Also forwarded many samples to Panasonic and I think they may reduce the issue like on the GH5 but not really fix it entirely which would mean by providing a NR OFF switch like on their EVA-1 and Varicam line which would also mean more noise and worse dynamic range ratings I don't care about anyway.

  6. 5 hours ago, Rinad Amir said:

    Yah same issue here blue clipping on S1H i hope Panasonic engineers are aware!  

    Rec709.tif 7.92 MB · 3 downloads V-Log.tif 7.92 MB · 5 downloads

    ehm, ok?
    There's nothing wrong with the source file. No oversaturation or even clipping in your example.

    4 hours ago, Jonathan Bergqvist said:

    With the wrong settings, Premiere will do that for you.Try doing in/out on the one frame you want to export in the timeline, file->export->media...->format: PNG -> be sure to select "Render at Maximum Depth" -> export 

    Did it fix the banding?

    Also in sequence settings. At least in CC2019 the issue with both settings being set should be gone.

    19 hours ago, CaptainHook said:

    That's an oversight on the Resolve teams part, I alerted it to them last week when a user reported that some options are also missing from the ACES ResolveFX plugin too.

    We are looking into it. It's not specific to the Pocket 4K, is happens with the G2 and other cameras. Its how gamut mapping is done - I've seen the same artefact on footage shot with ARRI/Red/Sony/etc from test files with saturated red highlights clipping and also seen it on publicly broadcast tv shows and movies with huge budgets shooting Alexa etc where the gamut mapping is not handled by the colourist (its very common on TV series on Netflix, HBO, Amazon, etc). FYI, Arri Wide Gamut is very similar in size and location of primaries as BMD Wide Gamut Gen 4.

    The issue is its a non-linear transform to address the problem so if you apply that correction to footage it's not easily reversible anymore in standard colour science workflows like ACES. So if you applied it on camera to ProRes footage and then transformed it to another colour space "technically" it would be wrong. Same if it were an option in Blackmagic RAW decode and you took that output to VFX workflow etc that requires linear or some other transform. So the user has to be careful about when to use this. But we are looking into it to make it easier. Otherwise for problem shots people can decode into another space and handle the gamut mapping themselves (I personally think this is preferable when possible so it can be tailored to each shot and target gamut but it does require the user to have a certain amount of knowledge and time to address it in post which is not reasonable to assume).

    That's great news.

    Still not sure about keeping my S1H.
    Postives: Great color, works perfectly fine in ACES, no clipping issues, great DR and roll-off, IBIS and battery life, ease of use
    Meh: Mediocre Codec choices which may cause issues, but all good so far
    Bad: Temporal Noise reduction ruining many shots - barely usable honestly. Even recorded externally to ProRes HQ (tried an Atomos Shogun and Odyssey 7Q+)

    Pocket 6K:
    Positives: Organic looking image With grain and no artifacts, good dynamic range, great screen, great codec choice with flexibility and no macro blocking nor banding which is also speedy on older machines
    Meh: build quality, will put it into a cage anyway
    Bad: Battery Life, no viewfinder - needs rigging and can't really use it in ACES atm due to clipping issues and color looking - one dimensional and brown?

    On a yRGB timeline I can fix many issues by myself but it's quite time consuming and the next big project coming up soon will rely on many VFX shots, so ACES would be great - like it has been on my other projects before.

    On the other hand the temporal noise reduction resulting in ghosting and smearing on the S1H made tracking a nightmare already in another project I'm working on atm. Basically have to do everything manually because the tracker gets confused. :(

    Maybe I should just switch back to the GH5 lol

  7. 1 hour ago, CaptainHook said:

    Incorrect.

    And we provide our colour science data to 3rd party apps, post houses, and studios like Netflix on request so they have the data to transform to ACES outside of Resolve if needed.

    But why is there no IDT or lets call it automatic transformation anywhere else but when using cDNG/Braw in Resolve? E.g. ProRes footage cannot be used inside an ACES workflow at all - even in Resolve - since no input transformation for BMD Film V4 is provided.

    Also is it possible to get the saturated red highlights fixed in the future?
    Color cast honestly isn't an issue using the UMP/G2 in BMD film V4 at least. My complaint is spefically about the Pocket 4K using the ACES transformation.

  8. 1 hour ago, Steven.Shen said:

    I met the same question.

    It seems like the bug of Adobe Premiere Pro cc 2019.

    I tried on cc 2018. The exposure seems right.

    You could create some kind of input Lut to convert to Rec709.
    https://cameramanben.github.io/LUTCalc/LUTCalc/index.html

    Wondering if that one works. Also change Sequence settings to max. bit depth to avoid further color shifts.

    Rec2100HLG_Rec2100-Rec709800_Rec709.cube

    2019-10-29.lutcalc

  9. 28 minutes ago, helium said:

    You have made *one* valuable contribution here:  reminding everyone, once again, how silly it is to take seriously the "technical" arguments of people who don't know what they're talking about.

    Then again, you're all over the web, spreading disinformation wherever you go, so it's kinda hard to ignore, duh.  But I'll certainly try, from this point out.

    Yeah, whatever.
    Netflix should do HD because your theater can't do more than 700 lines and also shouldn't care about missing ACES specs and future proofing their own content.

    And people should buy different cameras because Premiere introduces banding in their footage. ?

    2 hours ago, helium said:

    strange preferences (GH5??)

    Won some cinematography awards with this camera. The GH5 is still heavily flawed and so is every (consumer and prosumer) camera and many could be fixed by firmware updates... GH5 issues were: V-Log L using Rec709 matrices (so no real ACES compatibility either), weird color, small dynamic range, strong oversharpening etc. and even ghosting/smearing which was fixed with later firmwares after people kept complaining.

    Still the best choice for me at that time.

    So what's your point?

  10. 1 hour ago, helium said:

     

    1) who said a word about the resolution of film negative?  The issue is, the resolution available in film release prints with typical theatrical exhibition.   Do you not listen or is this some sort of baiting attempt? 

    And, really, you don't need any measurements to know how little actual resolution is present in release prints.    Nobody, and I mean nobody, sees as much detail in a movie theater with a 35mm release print, as she does watching a bluray at home.  Maybe you've never actually seen a film print?

    2) yes, BMD converts braw/cndg to log or other color space format, inside Resolve.  And how else?  If the system doesn't debayer and convert to a color space, there won't be anything to look at.  The issue with ACEs is, the lack of an input color space transform for the BMPCC 4K/6k, which is supplied (for example) with Resolve Color Management.  Whether this is decisive in your case, accounting for your strange preferences (GH5??), nobody can say.   Your fling out these charges, which change from week to week, and who knows what it's all based on?  Whether BMD provides documentation is beside the point.  And with its supposedly "potato" color science, who cares?  Why do you even want it?

    3) okay, so now there are *new* charges against Premiere, your original one having been discredited above.  I don't care -- no Premiere user here -- but does this never stop?  Every week, something else to dismiss as crap?  Is this some sort of game?

     

    1) We have 4K laser projection everywhere even here in Germany. And if your theater keeps playing inferior 16th gen 35mm film copies, that isn't Netflix fault.

    2) BMD doesn't provide any IDT or even documentation for implementation into an ACES workflow outside of Resolve and Fusion nor is BMD Film V4 properly calibrated on the Pocket 4K. Look at the Terra 4K and GH5s and Z2: The IMX294/299 is capable of producing nice color when done properly. Also Z-Cam provides documentation about their colorspace and even IDTs for usage inside OCIO and therefor basically any compositing tool - it's a must especially for Netflix who wants their originals edited like that.

    3) Reason: People keep blaiming their cameras for banding issues - which aren't caused by their cameras but Premiere Pro.

    Quote

    Is this some sort of game?

    A very specific one: The make @helium mad game, duh.

  11. 1 hour ago, crevice said:

    I get around 2 hours on one battery

    same here.

    2 hours ago, sanveer said:

    The Netflix Certification, and this, should push sales atleast 5 fold, if not more. 

     

    https://***URL not allowed***/panasonic-s1h-lab-test-dynamic-range-and-rolling-shutter-results/

    Class leading dynamic range. Or way beyond any other ILC hybrid out there. 

    hmm yh, by cheating with strong temporal noise reduction creating all kinds of artifacts when there's any kind of motion in low contrast areas.
    It really needs a fix, even if that means lower real dynamic range.

  12. 18 minutes ago, Brian Williams said:

    Thanks for this- I think I will stick to the SSD recording, that 8-bit noise is no joke compared to the 12-bit.

    yep, also 4K seems to look slightly cleaner with way more resolution, but when you start zoom in you'll see banding artifacts

     

    Verdict:
    I want to see more 12 Bit CineDNG files in 4K. Thanks ?

  13. 7 minutes ago, paulinventome said:

    To be expected, but have you had a chance to compare 10 bit, with a decent curve that could be as good as 12 and has more frame rate options

    Those new videos above show some horrific highlight clipping - i wonder what the workflow was with them...

    cheers
    Paul

    Noticed as well. But at least no weird red (BMD) or blue channel (S1) clipping.

  14. 27 minutes ago, helium said:

    Dude, you're talking about everything and anything.  Working backwards:  700 lines is about all you'll get in a movie theater, from a 35mm print  There's a white paper online you can search for, if you're interested, among other sources.  It's not a controversial observation.  "LOL".

    There is no ACES transform for the BMPCC 4K in Resolve, so it's not surprising you don't like the results.  Again, your "color science is weak", etc. is typical internet-speak, where unsubstantiated claims vie for attention.  One week it's the BMPCC 4K which is plastic and over-smoothing, then it's the S1H which blurring detail.  Next month, it's the BMPCC 4K which is too sharp, unless you prescription for downsampling is followed.  Or Premiere is 8 bit and Resolve is 32.  Until not.

    This stuff never stops, I guess it keeps the forums going.

    Film can resolve more detail than modern 4K cameras are able to, especially native 4K sensors. But keep going...

    When shooting Braw and cDNG on the Pocket 4K, 6K, UMP, URSA G2 etc... Davinci Resolve applies its own transformation (IDT) for BMD Film V4. Even worse BMD doesn't provide any documentation nor any proper IDT so it can be integrated into Baselight, OCIO, Nuke etc. don't you think? 

    Color science is another factor. It's just not good. It's quite bad actually and horrible to deal with. I'm a colorist myself and even prefer the GH5 over it - which has many other flaws and reasons why it hasn't been considered by Netflix. 

    Only the Ursa Mini Pro and G2 provide good color science in case you prefer shooting with BMD cameras. Great tonality and definition but are still plagued by the infamous red channel clipping issue.

    Premiere can be forced to higher bitdepth decoding but still crawls and shifts color around and doesn't have any color management. But maybe you like that. Idk?

  15. Dude, were not even talking about that.

    Netflix chooses cameras which match with their workflow and allow for a 4K output with HDR.
    The BMD Pocket 4K isn't usable in their preferred ACES (to Rec709 and Rec2020 ODT) workflow because color science is weak and the IDT only exaggerates the issues this camera has - it's a brownish nightmare with a clipping red channel on top and missing any fine nuances in hue which I tried to stretch but sometimes ended up in artifacts despite having a 12 bit codec to begin with.

    The Pocket 6K is way better in that regard but still cannot compete with even the UMP and G2. The codec is isn't the issue here either as shown on the G2.

    Also 35mm resolves way more than just 700 lines lol
    Still - wouldn't fit into Netflix's workflow.

  16. 3 minutes ago, helium said:

    A) nobody knows what Netflix's actual "quality" threshold is, after the camera meets the more or less arbitrary technical requirements.  

    B) unless everyone here is under contract with Netflix, the announcement doesn't mean a whole lot.  Why is the S1H on the list, and the BMPCC 4k/6K aren't?  Anything to do with actual image quality or for reasons of no consequence to anyone likely to buy any of the 3?  Nobody knows.

    C) And please, no "future proof" nonsense, in any but a marketing sense, which Netflix is itself promoting.  My 4K is bigger and better than your HD!  Actual audiences don't know or care, until they're told.

    B) dynamic range is slightly better on the S1H but color science is another league. Especially the Pocket 4K has the tonal range of a potato in comparison and the ACES IDT is horrendously bad.

    C) actual audiences pay for 4K.

×
×
  • Create New...