kye

Now I'm beginning to doubt myself. If I take the film stock ARRI was talking about and a super-big DR film stock and expose both of them to a XYLA 21-stop test chart they're not going to both have the same density of exposed film at the +3 and -3 patches. The high-DR film stock has more total DR, so the DR in the scene has to be represented as lower differences in the density of the stock. ie, this image has the stops quite far apart (GH1 DR test) and this image has the same stops much less far apart (GH6 DR test) The number of stops between two points is absolutely about the scene and nothing else, however the number of stops of light required to change the exposure value on a given negative or camera from 2% to 90% has everything to do with the film/sensor and nothing to do with the scene. Right?

ARRI film cameras are in no way "worse" cameras, it's the combination of various factors that makes this label.. they seem to correlate with "affordable" though! I'm on the fence about the GX85 being a bad camera - the 4K is pretty nice when viewed at 4K and when downsampled to 1080p it's a very nice 1080p image indeed. I don't know what it's DR of video is, but IIRC I did an under/over test some time ago and it was pretty flexible - much more than I would have thought. In terms of an update to the GX85? Give it h265 12-bit support and I'd be happy!

Nah, just as you get close you'll arbitrarily move the goal posts again to 6K or 8K or some other thing that makes very little difference except making you re-buy all your equipment again and be forced to choose between cameras that are severely stretched and limited in many ways. .......just like everyone did when they went from 1080p to 4K, mostly for very little reason except to sell more TVs.

I think it might be helpful if you'd read the link I posted? The figures are about the Characteristic Curve of the film, which is how exposed the film is (density) when it's exposed to certain amounts of light by the scene. 9 stops! 🙂

The image looks indistinguishable from my XC10, although I haven't edited the files from it in a while. Maybe with a much more contrasty grade the 10-bit advantage of this camera would be more visible.

Great tests and thanks for sharing. One option you might test is doing the gyro stabilisation and then the normal stabilisation on top of that. I suspect that Resolve will only let you apply one type of stabilisation, so you'd have to render out a clip with the gyro built in and then import and stabilise that clip in a second pass, although in reality if you wanted the extra stabilisation then you could just do a pass of all the gyro shots in a project and then replace the original clips with those processed ones to edit your final project. I suspect that would take it to gimbal levels of stabilisation, although still not giving the shutter angle support of a real gimbal. Personally, I absolutely detest the look of a hand-held gimbal - it's the look of camera that is completely stabilised in terms of where it's pointing but is bobbing around all over the place in 3D space, it just screams amateur to me. In this sense you'd want to limit the gyro stabilisation so that it's not dissonant with the camera physical movement, and in this kind of limited stabilisation the gyro looks easily up to the task. Very promising!

Fixing things in post is a real mixed-bag. Lenses with simple (ie, not complex) distortions can be fixed in post to make the lines relatively straight, but that's not going to cut it for things like large groups of people. The fundamental problem on wider lenses is that the world is angled and sensors are not, so objects on the edges of the frame will appear larger than ones in the middle of the frame. You can adjust the image to straighten lines (as you would do for architecture) but this makes the relative sizes between the edges and middle even worse. Or you can choose to adjust it so that the sizes are relatively similar but lines will not be straight. There's no way to get both in the same image unfortunately, and lenses that are rectilinear (have straight edges) will have the worst corner stretching. Here's what is happening: Notice how the corners get stretched? Things look even stranger when you move the camera. The 14mm f2.5 isn't too bad, but is still quite curved. IIRC it can be corrected pretty easily. I have it and can do a quick test..

Best wishes for a full and speedy recovery!

Interesting, that's a lot more forgiving than I would have thought, being that 5-stops over is pushing things right up into the highlight rolloff. Maybe I should have said 7 stops over! Just goes to show the relative "bit-depth" of analog and what kind of subtleties are there, should they be expanded out to be more visible. I guess that really emphasises the importance of bit-depth, and suggests why cameras like the ARRI, the Canon 5D with ML, etc have such great and flexible images. In a way, the 9-stop DR of film that ARRI was talking about is 9-stops combined with an almost limitless bit-depth - such a great combination for colour subtlety and richness. Every time I try and emulate the look of film in a project (not in a technical way but using pre-existing LUTs or transforms) I learn a lot about image aesthetics and colour and contrast and tone. I really need to get back into shooting more little tests and grading them just to get more practice.

Good points, but I think perhaps you took a few things out of context. Firstly about the DR figures.. your quote was correct - I did say "colour negative film has 5.5 stops of DR between 2% and 90%", and I did mean that, however, I absolutely did not mean "colour negative film has 5.5 stops of DR", which is a completely different statement and one that is factually incorrect, as you point out 🙂 (As you said that you didn't read the page I linked to, ARRI actually said "Current color negative materials can capture a dynamic range of 9 stops" and then had a summary table that outlined the exposure values at various points, of which the 2% was three stops below mid-grey and 90% was 2.5 above mid-grey). The reason I quoted stops of DR (either for the 2%-90% range for film or 5.2 stops for rec709) was to establish an absolute measure for how much contrast in a scene would translate to how much contrast in a given image. What I mean by this that when you adjust the image so that the clipping point of your input file is at 100IRE and your noise floor is at 0IRE, you can then adjust the contrast knob and you decrease the amount of stops that fit between (say) 10IRE and 90IRE, without losing the total DR (because it compresses that in the rolloffs). The point of mentioning these things was to make the comparison between the DR captured by these "worse" cameras and the level of contrast in a properly lit/exposed film pipeline (eg in theatres), or a properly lit/exposed 709 (eg video broadcast). This means that in these lesser cameras, by the time you shoot something in their 709 profiles (and maybe add some contrast or rolloff in post) then you're getting a similar level of contrast in the final image to these reference pipelines. The purpose of that comparison was to say "these cameras have a similar level of contrast to get the looks that are desirable and even nostalgic". I would go further than this actually, I think that given the (perhaps) 7-8 stops of DR in the video mode of these "worse" cameras it's easier to create an image for non-proficient colourists than it is to do so with a camera that captures more DR. I know that when I was grading the images from the XC10, I had trouble with its ~11 stops of DR, because I wanted to keep everything it captured, all the way from the texture in the clouds to the subtlest shadow details. I don't know what it was about that challenge that kept me from getting it right, if it was that when I added enough contrast I would lose any contrast in the parts of the image that got pushed into either rolloff, or if I didn't know how to balance the level of contrast with saturation (converting the XC10 C-Log into rec709 using the CST or Canon LUTs gave crazy saturation that I didn't know how to deal with at the time). Regardless, it was essentially too much for me to handle and I struggled. Had I been given an image with those levels of contrast built-in, I think I would have just taken what I was given and did what I could with it. There's also an element of the paradox of choice. The more choice we have with something, the more that we get anxiety around not having chosen the best option. The optimum amount of choice is (of course) not zero, but actually the peak where anxiety hasn't yet overtaken the experience is quite low in the overall spectrum, certainly lower than people would think, especially in todays world where manufacturers are always pushing us to think that more options is better.

Not at all. Their AF will ruin your life. Not a single Panasonic camera has ever created a single frame where a single item was in focus. Not even if you filmed in a forrest - it would pick a focus distance where no single trunk, branch, twig or leaf was in focus. It's so bad that it's probably the cause of global warming and religious extremism!

I agree they have an image problem. My point was that it's based on ignorance, which cannot be cured.

I meant the old P6K, as I was responding to @PPNS saying "dont really get the point of the p6k anymore since the release of the pro anyway, i just would've discontinued it" Now with the new P6K being the same size as the pro, yeah, that size advantage is gone. I heard someone say that you don't need the internal NDs if you prefer to use external ones, potentially as they might be higher quality than the internal ones (I think they were referring to people who use filters in a matte box). FrameVoyager also said they had problems with their P6K Pro NDs breaking because the camera was dropped by luggage handlers so the NDs are another moving part that can fail. Seems like a typical camera industry thing, as: You get some upgrades from the previous model and combined with the same price point it's a "better" value for money You pay for it with down-sides that you wouldn't have wanted and aren't welcome for a range of users That's basically how the camera industry rolls right?

I'm reminded of Werner Herzog saying that his cinematographers need to know their equipment and their especially their camera/lens so that they could shoot without looking through the viewfinder but would still know the framing. It was his minimum criteria for a cinematographer to be considered competent IIRC. It is one of the things that they talk about with street photography and using only a single prime lens - you get to the point of not needing to look because you already know the lens and the compositions you'd get. ...and yes, the grass is always greener with other cameras, but apart from resolution and increased DR, the 5D ML combo is right at the top of what is available, basically regardless of price.

They must rectify the chatter based on nothing? Good idea. Once they've done that they can teach the governments of the world how to rectify conspiracy theories, the science community how to rectify things like climate denial and creationism, sociologists how to rectify extremism and fundamentalism, etc. There's probably a lot of money in that, but even if it turns out there isn't, we'd still be left with a world that makes more sense, so that's probably a reasonable runners-up prize 🙂

In the best conditions I think the 5D with ML images are right up there with the Alexa. No, it's not equal, but within the 5Ds DR limitations, it really is an excellent image. Canon sensors and colour science are famous for good reason.

Absolutely. That takes the count to two people with calm, reasonable, and grounded in experience options about Panasonic AF. Still dozens to go. I've never said that AF wasn't useful, wasn't valid, or was somehow lesser than manually focusing. My issue is with people who endlessly suggest that the AF isn't usable in any situation by anyone ever (or simply jump straight to implying that Panasonic washing machines and microwave ovens are all going to evaporate since the quality of the AF on the first version of the GH5 will somehow bankrupt the entire multinational consumer brands entire existence).

The non-pro still has a number of advantages over the Pro model - size being a significant one I think. Of course, you could make the case that once a camera is as large as the P4K it doesn't matter if it's 50% larger, and there's some logic to that, but you're really getting into the territory where you can't hand-hold it indefinitely. I use a GH5/lens/mic hand-held and it's borderline for being able to hand-hold it indefinitely and on larger shooting days I've ended up with sore wrists the next day. Sure I'm not a professional camera operator so I don't have the strength there, but I'm also a relatively physically capable person so I'd say I'm somewhere in the middle. For me the major issue with the 6K range is that it's not a mirrorless mount. I understand that you can adapt m42 lenses to EF, but there are a number of DSLR lens mounts that aren't compatible, so it kind of eliminates maybe half or two-thirds of all FF/S35 vintage lenses from consideration. I like having MFT where I can just assume there will be an adapter and sensor coverage and perhaps even a speed booster option.

Yeah, you kind of assume it'll be less, but getting any info on the DR of the video mode of early cameras seems to be impossible when I've gone looking. Having a few stops less certainly isn't ideal, even to emulate lesser film stocks, but I do think there's sufficient scope for getting a nice amount of contrast. Good points and you're right, although DR is getting better with each new sensor generation. Smartphones are getting smarter through taking multiple exposures in super-quick bursts which works a lot of the time (but not all the time) and of course are benefiting from improvements in sensor tech too. I do think there's a point where things are enough though. One of the main challenges I have found is fitting in the stops of DR that you get from a modern camera into the limited DR of 709 - it's a question of what parts of the luma range to throw away and/or which to compress, and in making those decisions you're deciding about which things should have contrast and which won't. I guess part of my point with this thread is that the lower DR from 'worse' cameras forces you to choose that when filming, and also makes you have a "full" amount of contrast with most scenes, which I think is a fundamental aspect of nice images, except in very-low DR scenes of course.

If you want more GX85 nerdery and discussions then look no further.....

Welcome to the club - these things are great little cameras!! What sort of subjects are you going to use it for, and what lenses etc are you planning on using? I bought mine to be a second/time-lapse/backup camera to my GH5 for travel and for at-home for my "Go Shoot" project which is just trying to get out of the house and go shoot little adventures / days out / etc. In terms of lenses, I think there's a few ways I'm intending to use it. The first is with the 15mm F8 lens-cap lens for something truly portable but this only works outside during the day, or the 14mm F2.5 for a small but more capable setup. The second is with a modern zoom for flexibility of getting shots as they present themselves, like my 12-35mm f2.8 or the 12-60mm f2.8-4 which I'd love to buy (for the extra reach). The third is to go with primes and shoot more intentionally, like the 7.5mm F2, 17.5mm F0.95, 50mm F1.4 trio. The fourth is to go a bit vintage with a single prime, like a 28mm F2.8 with a speed booster (making it a 43mm FF equivalent lens). The fifth is to go for a vintage zoom, using my Tokina 28-70mm F3.5-4.5 with speed booster (making it a 43-110mm F5.5-7 FF equivalent lens). I do like the idea of using vintage glass on it because it softens the 4k which is a bit sharp for my tastes. The idea of all the various lens options etc was to make me excited to go out and shoot but also to use the outings as a test-bed for styles of shooting and lens choices and experimentation in editing and colour grading etc. The only way to get good at making films is to make films, and to make lots of them, and to try stuff and learn from what works and what doesn't. The Cine-D hack is really good too - if you haven't got that in there already it's highly recommended. The CineD profile is actually quite reminiscent of the colours that an Alexa would give you. Not the same, of course, but they both push a 'correct' image in the same sorts of directions and do the same sorts of desirable colour things. If you're not keen on the hack, the Natural profile is actually quite similar to the CineD profile in that it has a different flavour but is still employing the same nice colour treatments that the CineD and Alexa colour science do, and makes a great base to tweak in post if you're keen to customise the colours.

What if I sew a pillowcase onto the chest part of my jumper?

Yeah, there's lots to talk about in there, but I think the reality is that it's just different. Digital highlights are spectacular quality until they clip and quality goes instantly to zero, whereas film has a rolloff that just gets lower and lower quality, almost forever. Sure, you can bring the highlights down a couple of stops and there isn't nothing there, but try exposing skintones at 5 stops over and then bring them down by 5 stops - that will show that film doesn't have the DR of an Alexa, and not in a particularly kind way either! Shadows for both are similar for film and digital in that they both descend into a noise floor, and for that you have to take some sort of signal:noise ratio as the threshold, but of course we're aware that this can be gamed and that the practical and aesthetic attributes of this don't line-up to the maths of testing software and the 'enthusiasm' of camera PR departments. My point was that by the time you take a 10-stop digital image and effectively flatten the top couple of stops into a rolloff and the bottom couple of stops into a rolloff then you're not going to be missing much. Sure, film might have had another few stops compressed into the highlight rolloff and maybe you could recover a bit of stuff from them, but when presenting that rolloff in a final grade where it's up close to 100IRE it's barely-perceptible data with almost zero contrast, so any differences between the two in that final image aren't going to be that visible. I'm also aware that Kodak went on an all-out offensive in defence of film as digital was taking over and they were fighting for their lives. I remember going to a lecture by a Kodak technician in about 1995 and it was just 90 minutes of them talking about how film had more DR and that it meant you didn't have to light as accurately and so it paid for itself. He may as well have just got onto his knees on the stage and begged us to keep using film so they wouldn't go bankrupt. I remember leaving the lecture and just thinking 'but the tech will continue to get better' - it was almost embarrassing. Besides, rec709 broadcast had a similar DR and that's what we have seen for out entire lives apart from HDR and movie theatres.

Cool - one person did their own testing and evaluations in their own particular scenarios with their own lenses. Only about 50 more critics left to see if they're talking from experience or hype 🙂 I mean, I shoot MF so I really don't care, but it's pretty obvious that most critics are remembering the AF of the GH5 v1 firmware and haven't actually fact-checked themselves in about half-a-dozen cameras and dozens of firmware updates.

Is it a small battery though? Or is it a normal sized battery with an enormous and very bright screen? The 6K Pro has a 1500 nit 5 inch FHD touchscreen display - with any other camera that would be a separate monitor and would require a separate battery (or two). You can't get something for nothing, despite how much we want to!