kye

@ghostwind Your approach sounds sensible. I appreciate the buy-once-buy-well approach, and it sounds like you probably have a good enough idea on what your requirements are and how you work to be able to determine what will work for you, both creatively and from a return-on-investment perspective, and considering that the image you deliver may even be secondary for some clients compared how how impressive your camera looks or the specs you can advertise (like shooting in 4K). I don't know what platforms you're creating for, but 4K would give you the ability to deliver 4K 16:9 and also >1080p 9:16 vertical video from the same shot. We've spoken on these forums about vertical video and I recall a lifestyle/branding/advertising film-maker suggesting that filming vertical video for social and also for portrait-mounted TVs was becoming more in-demand. The framing may be different for different aspect ratios, but I would imagine not having to have a 90-degree adapter to flip the camera would be a plus on set, as for each shot you could just reframe and do another take, and then be able to produce a vertical and horizontal version of the same video without much extra time on set. The only other piece of advice I have is to get your hands on a few options and see how you like them. Even if you assume that Canon knows how to make a decent cinema camera (which is pretty well established) there may be little personal elements that would work or not work for you specifically. Ergonomically, conceptually, and for your workflow too.

So.... how did it turn out? Talk about ending your post on a cliff-hanger!! These writers, always trying to make sure you'll read their next post!

All valid points, but it makes me wonder how much you want to spend for a "what if" scenario. An alternative to this is a C100 and if the client wants 4K then hire a C200 for that project. At the point that clients routinely ask for 4K then you can consider how long a purchase would take to pay for itself based on data rather than a general concern. "What if" is the enemy of photography because it leads you down a path where you buy so much equipment that you don't have time to learn how to use it and you take so many things with you that you can't actually go anywhere or shoot anything. True wisdom is knowing how much is enough.

also, someone else just posted this excellent example of exposure tests.....

If you're asking me how I do it personally, then I just let the camera do it. I've played with my exposures enough to know that I'm ok within a certain range and can get what I need. I think this is personal. Have a read of these: https://www.hurlbutacademy.com/film-education-arri-alexa-vs-canon-c500/ https://www.hurlbutacademy.com/film-education-arri-alexa-vs-canon-c500-part-2/ If you do the tests and absolutely love it at a certain exposure and hate it at anything else then expose it like that. If you do the tests and discover you don't care, then let something else guide you. We all see differently, we all have different preferences, you will notice things I won't and vice-versa. I think there is a progression of knowledge and technique: Level 1: Full-auto go go go Level 2: I'm now very aware that I don't know how to do this right, I need help! Level 3: I've read a bunch and am experimenting and I've worked out all the settings and all the numbers and all the ratios and all the things Level 4: I don't bother measuring, I just adjust it until it looks right The Level 4 people can adjust it until it looks right because they know two things: what to adjust, and what looks right. To get to this level you have to go through the previous steps first. No shortcuts. My comments above about Level 4 apply here too. Read everything you can get your hands on, do all kinds of tests, learn everything you can, then just expose it so it looks right. I'm reminded of how professional colourists talk about grading skin tones. Level 3 people ask questions about the skin-tone line in the Vectorscope and the Level 4 people just laugh and ask "How would you grade these?"

Try everything. My suggestion above is a lot of work but think about it this way, you have two options: Spend an afternoon or two doing my test, learn how your camera responds, learn things about image quality, and build a new skill Shoot ETTR because someone on the internet told you to, and maybe years later you learn that it wasn't the best way and you could have been making nicer footage all those years I mean, Internet forums always have better advice than what professional cinematographers do themselves, right?

I understand your point, but I suspect we have a philosophical difference here. My end goal is to make art. Out of the many ways I could do this, I have chosen video as the medium, and for this purpose I own a camera. If I could stick a USB stick in the side of my head and render my memories to it then I wouldn't bother with any of the opto-electric-trinkets that we talk about here. So, when we talk about spending many thousands of dollars on equipment instead of $40 for a 10-year-old point and shoot on ebay, we are doing so because we think that expensive equipment gives us a nicer image. For me, and I suspect most others, nicer typically means that the end product looks far less crunched when compared to what we pointed the camera at than a cheaper camera would do. This is why our discussions talk about everything in the image pipeline. Walking through it, we talk about filters, lenses, the sensor stack, the sensor, the digital processing (colour science), the codec, editing, grading, delivery codecs, distribution, and final projection. We care about everything. I understand that not everyone wants to have perfectly realistic imaging, there is this concept of a 'cinematic' image and we talk about how certain distortions have a pleasing effect. Swirly bokeh lenses are an example of this - not everyone wants the modern look. However, when we're talking about what we want vs what we don't want, not everything is desirable. For example CA isn't something most people want. It's definitely not something I like the look of. There are other things we sometimes want to get rid of. ..and here we get to the crux of the problem - if there is a problem with something in the image pipeline then we have the choice of having everything else in the pipeline have high-quality and therefore ruthlessly reveal the issue, or we can deliberately lower the quality of something in order to effectively hide that issue. It's a question of how clearly do we want to see the best part of the image, knowing that we will also make it clear to see the worst part of the image too. For me, it's about spending the money on the worst part of the image pipeline. If someone said they had an Alexa and PL cine lenses but used a scratched $10 ND filter and were thinking of upgrading to an ARRI LF because they wanted to get better image quality we'd think they were crazy because they should replace the ND filter instead as that's clearly the worst part of their setup. For me, I think that lenses might be the worst part. I watch a lot of Netflix and on almost everything I watch I'm seeing bad image quality caused by lenses, not by filters, not the sensor stack, not the sensor, not the digital processing (colour science), not the codec, not editing, not grading, not delivery codecs, not distribution, not the final projection, the weakness is the lens. And when I look at moving from 4K to 6K I think about what lenses people are using and I think "4K isn't your bottleneck". Using a Helios... 4K obviously isn't your bottleneck. Using a Sigma 18-35, Panasonic 12-35 f2.8 or Canon 16-35... I still think that 4K isn't your bottleneck. Using a CP.2... Even then I think 4K isn't your bottleneck. I know it's a taste thing, but when I see lens issues I'm just thinking everyone is spending big dollars to make high-resolution copies of fuzzy 70's polaroids.

I'm planning on getting one of those when OSX drivers are released - the horsepower-per-dollar ratio is pretty good on that. Of course, I'm not sure if getting a better card will help me as my CPU might be the bottleneck. I'm currently talking with one of the BM support providers so am hoping for some wisdom (and performance gains!).

Kraig Adams does it regularly in his videos. His videos are travel films and often go from vlog style to cinematic style. The vlog style is 16:9, hand-held or gorillapod, normal speed, with location sound and is often funny. The cinematic style is 2.35, smooth / often gimbal and sometimes drone footage, often slow-motion (although he's moving away from that now), with music over the top and is beautiful. He has experimented with how to transition between the two and often animates the black bars coming in and going out. It's a very different style to the video @heart0less posted! This is a good example, and he talks about it in the BTS.

I use aperture-priority mode and auto-ISO in combination with zebras. It does the heavy lifting for me, and if I notice that something is pushing into zebras range (I can't recall if I set zebras to 100% or less) then I will manually lower exposure compensation in order to keep the highlights. My GH5 has decent DR and I prefer a less saturated look that shows more DR and doesn't crush the highlights/shadows. To do this I shoot HLG and try to capture the whole DR of the image without clipping and then adjust brightness in post to equalise exposures. As I shoot in 10-bit I can push the image around quite a bit in post and it's not in danger of breaking, so I have that latitude. This fits my "capture everything and make the look in post" approach. If you're trying to get it right in-camera then obviously this isn't how to approach it!

It's a drone lens so the ergonomics are abysmal, but it works, the images aren't bad, and it's significantly cheaper than the Laowa!

Setup a test scene with some shadows/highlights, some bright colours (a test chart is great but not needed), and some skin tones. Film it with your camera at varying exposures, I'd suggest "properly" exposed and +1, +2, +3, -1, -2, and -3 stops. Then in post process the footage so that it's all the same brightness and then look at the images, notice what it does to colour, skin tones, and highlights/shadows. This is what professional cinematographers do to "learn" a new camera or film-stock. They often talk about how a particular camera is a -1 stop camera because they've done the tests and that gives the nicest skin tones or colour or whatever. I'd also suggest that you play with the colour profile to test things like lowering the contrast (most normal profiles have too much contrast for my tastes) and for each change you make you should repeat the above test. You will eventually find that you prefer a profile with a certain contrast, a certain saturation, and a certain exposure level. It's a lot of work, but if it wasn't required then every home video that dad made with a camcorder would be breathtakingly beautiful, and that is obviously not the case.

My opinion is that it would have been great for them to make the mount removable so that people could have the option. Even if it took an hour and specialist tools to do, it would have allowed a level of customisation for those that were interested in it while also not even being noticed by those that would only use the EF mount. Think about how Sigma offers the mount change service on many of its lenses, this might be a similar thing. Maybe a separate version of the P6K will be released, but this makes changing between them a sell/buy exercise instead of a minor change you can do at home. If the P4K/P6K was the right type of camera for me then I would have been willing to sacrifice a bit of resolution when shooting an MFT lens. I use a GH5 as I don't want RAW but I do want IBIS as it suits my shooting style and aesthetic. My lens collection has four major lenses: 8mm MFT, 17.5mm MFT, 40mm FF, and 70-210mm zoom FF. I'm not sure how large the image circles on my MFT lenses are, but I'd be interested in being able to test how large it is and maybe my 8mm lens can actually get a bit wider too, who knows. I see the P4K/P6K cameras as being somewhere in-between a point-and-shoot (where you can't customise anything or plug anything into it) and a cine camera where it is completely modular and you have to attach a dozen different things just to get it to write an image to a storage device. In this sense I think there is room in the market for it to offer some degree of flexibility, and a removable mount with different crop factors would be a win-win where people who want it can use it and people who don't need it wouldn't even know it was there.

Wikipedia is all you should need. https://en.wikipedia.org/wiki/Circle_of_confusion Basically, if I make a scene with three points, the first one is perfectly in focus, the second is a one inch behind the focal plane and is a bit blurry and the third is two inches behind the focal plane. Then I project that image onto two sensors, one is low resolution and the other is high resolution. On the low resolution sensor the point source hits one pixel (one photo site) and the second one is a bit blurry but the photo sites on this sensor are so large that it still only hits one photo site. The third point is so out of focus that it hits more than one photo site. On this sensor the first two points are in focus and the third is out of focus, therefore the DoF is between 2 and 4 inches. On the high resolution sensor the point source hits one pixel (one photo site) and the second one is a bit blurry and because the photo sites on this sensor are so small that it hits multiple photo sites. The third point also hits more than one photo site. On this sensor the first points is in focus and the second and third are out of focus, therefore the DoF is between under 2 inches. This is because the definition of what is in focus and what is out of focus is related to the resolution of the sensor / film-stock. This is what Circle of Confusion is referencing. Also from wikipedia: In digital photography, what is acceptably sharp would be based on the resolution (which is why people pixel-peep by zooming into 100%) and the visual acuity of a digital sensor is its resolution.

Let's cut to the chase with an example. If I set up a lens and adjust focus so that the blur on some writing is just enough to make it unreadable, then it doesn't matter which crop factor or how much resolution I digitise the projection from that lens at, the text will remain unreadable. If I then adjust the lens so that text is just readable, then it will be readable regardless of crop factor and resolution (except if the resolution is too low to resolve the detail in the text). Does that make sense?

That's nice to hear and just reinforces my decision to invest in one. After I played with the Sigma 18-35mm 1.8 for a while I realised that there's a limit to how much weight I want to hand-hold. I know I miss a few shots by having primes in my bag, but if the camera is in my bag because it's too heavy or I end up with RSI in my wrists then I miss all the shots, so I take some over none Everyone can keep their heavy lenses and I'll have a few nice primes!

You know how an avalanche starts with one little rock tumbling down an embankment? Well, embrace it and go all in! Become the first EOSHD member to own an ARRI and a full set of PL cine primes!! Then you can look at people rigging up their P6Ks with V-mount batteries, rods and rails, matt boxes and just say.......

The calculators are correct. The part that is giving you guys a headache is how you measure DoF. Something doesn't just go from being in focus and then you move it a millionth of an inch and it's now blurry. The way that these calculators measure DoF is called the Circle of Confusion, which is how blurry something can get before you can tell it's blurry. As the two different sensors you've chosen are physically different sizes but have roughly the same resolution, the smaller sensor needs something to be better in-focus before it is detectably blurry. https://en.wikipedia.org/wiki/Circle_of_confusion ie, if a tiny detail is being projected onto the sensor and it's one quarter of a pixel wide, then you move the object forwards a bit and now that object is slightly out-of-focus and it's now half-a-pixel-wide it will still appear to be perfectly sharp to that sensor, despite the fact that if you had infinite sensor resolution the DoF of any lens would be zero. Here's the circles of confusion from dofmaster: https://www.dofmaster.com/digital_coc.html

I second what @Geoff CB said. I figure you have a few options: Keep your current gear. It's a pain with crop factors and downscaling, but you get Canon colours, DPAF, and it's the cheapest option. Get a Canon cine camera. You can use your lenses, have DPAF, but crop factor is an issue. Buy something from another brand that will record nice 1080p and you can use your existing lenses (maybe via adapters). None of these will be both FF and have good AF. Buy a completely new system. You'll get something with good AF (Sony is pretty good) and you can get glass to match, but it's a huge cost and a PITA to run two systems. Dealing with crop factors doesn't sound so bad now does it? Seriously, use what you have and use the time to figure out how you work and what you need, and also give the industry time to release more FF cameras with good video AF. The industry is just starting to really go FF so if you can wait then you'll be much better off.

There's a segment of the film world that love exploring alternative processes and chemicals, trying new looks and generally just geeking out about various things. Luckily we have a bunch of things that are meaningless to most people that we all argue passionately about and are saved from having to develop film with coffee to add interest to our days.... I'm pretty sure I read about an photography project about coffee that was also developed with coffee. I can't find it again, but conceptual art is fun sometimes

I agree. It's the only lens still on my wishlist!! Not because it has great bokeh, but because it works in lower light situations (more of blue hour and into the night with city lights and cityscapes). Generally if I'm shooting wide I'm doing it because I want a grand scene, not as a vlog or anything. I do film myself, but more as B-roll and I've found an action camera is better for those shots as they're about me in the place rather than putting the focus on only me. Plus it's easier to hold up an action camera than a GH5 and big MF lens (it's also way less embarrassing too!) This are the shots I'm interested in: @jake all the above are the 8mm SLR Magic.

.....and the slower the lens and the wider it is, the deeper the depth of field will be anyway. Good luck trying to get background blur on the 8mm SLR Magic - even wide open at f4 it's almost the same DoF as the fixed focus of an action camera! (Not quite, but at f16 it sure is!)

ah, good old Barry In terms of sharpness and resolution it's an interesting thing. Our eyes have a huge resolution (in the middle at least) and we don't think of the world as looking 'clinical' or anything, so it's not about resolution. It's about sharpness, which in digital images is typically the alteration of the image to emphasise edges by adding artificial contrast (either deliberately or because of compression) and that is the problem. That's why shooting RAW doesn't look over sharpened or clinical. I don't think I'm searching for the holy grail, I'm just questioning how much resolution is really needed. If you had a 2MP lens and put it on a 4K camera (9MP) then what would be the benefit of upgrading to a 6K (20MP) camera? Nothing I would suggest. Now, if you think about a lens like the Helios, it's probably >9MP in the middle and <9MP around the edges, let's say that half the frame is >9 and half is <9. Now we decide we're going to shoot in golden hour and so we do the right thing and we use the sunlight as a back-light and now we've got CA on the edges. So now there's less than half the frame >9MP. We decide we want a dreamy look to suit the mood of the scene, so we open up the lens to (let's say) f2.8. The Helios is an F2 lens so it's not at its worst, but it doesn't sharpen up completely until you stop down to F4, so now we've got two-thirds or more of the frame under 9MP, and even the parts in the middle that are still relatively sharp have CA intruding into them. At this point would you be saying to yourself "wow, I really wish that I had a 6K sensor"? I really doubt it. The thing is that all lenses are kind of like the Helios, just not quite as bad. I guess the reason for me posting about it now is that I am now aware of how bad most lenses are (I'm seeing lens problems on almost everything I watch now) and I kind of thought that really expensive lenses were much better than they really are. I'm guessing that either people aren't seeing the lens issues like I do now and kind of don't care, or they're aware of them but don't mind. But my question is really that if the lenses are the bottleneck in the image pipeline then why are we optimising something else in that pipeline? It's kind of like taking a super-high-resolution scan of a low quality fax. I just don't see the point. and people seem to be falling over themselves to talk about higher resolutions without any discussion of what it is that we're really trying to achieve.

LOL, yes, that might have given some of it away. I'm interested in the lenses and other bits and pieces too Edit: the more people do something the more that they refine and understand it, so I'm always keen to hear from travel film-makers to get tips and ideas

I'm kind of coming around to this from a different perspective - that of evaluating lenses by their area of worst performance rather than cherry-picking their best. There are absolutely lenses that will resolve 100MP somewhere in the frame under perfect conditions. But what about the corners? and what about CA? and what about halation? If a lens resolves as much as a 100MP sensor in the middle, but edges with contrast have CA still visible when you downscale the image to 1080p (2MP) then is it really still a 100MP lens? and wouldn't adding more resolution in the middle simply make it more obvious that the flaws are so much less? If you check out the images I posted in the other thread, you'll see that the $40,000 Master Anamorphic has CA still visible in a 1080 frame (and it was still visible in the 720 frame I didn't bother to upload) so what are we really talking about? I used to think of the Helios as a charming lens because it was so imperfect, but the truth is that even in 720p some of the best cine lenses available are still a bit like the Helios. Most things in life are hard to see from a huge distance away, and look nice from a sensible distance, but when you get right up close the flaws become visible and the magic kind of fades. I'm beginning to think that with lenses you want to capture the image well enough to see it clearly but not so well it just points out all the flaws. and that happy medium seems to occur around 2-4K.