tupp

Agreed. Raw is great to have, but I've never really needed anything more than 10-bit, 4:4:4, flat and at a nice resolution. A lot of the time, even that is overkill.

Though question since I don't agree with your assessment of the advantages of Medium Format. I still shoot film, both medium format and full frame (small picture). And I definitely don't use medium format to get faster films. If anything I use the opposite. When I was shooting medium format, its primary advantages were: less-grain/higher-resolution; nicer DOF/focus roll-off; and increased color depth. The MF benefits of higher resolution and increased color depth don't necessarily apply to digital medium format cameras, but it would be great to inexpensively have a little of the nice DOF/focus roll-off found in medium format cameras. I used large format for the same reasons (less-grain/higher-resolution, nicer DOF/focus roll-off and increased color depth), which are even further increased with 4"x5", 5"x7" and 8"x10." However, large format view cameras offered the added benefit of Scheimpflug capability when shooting table-top work. LF with view cameras also allowed shifting for architectural work.

Huh? I seem to recall a lot of wide angle lenses ranging from 35mm to 45mm lenses for Mamiya and Bronica, back when I was shooting medium format. Heck, there were Schneider 38mm lenses for 4"x5" cameras. I am not so sure about this. The optical look of medium format lenses has always appeared different to me. Nobody has yet done a conclusive "equivalence" test. I don't know about that. This focal reducer essentially converts full frame into a medium format, so by using MF lenses with this Kipon reducer, one gets all the benefits/properties of MF optics. Judging from @Mattias Burling's test shots, this focal reducer is going to produce amazing images/footage.

Huh? Well, it can be somewhat of an art to pour the top surface of a one-piece mold, but below that surface, everything should be perfectly precise. Also (and, again), this piece might not require the precision you think it needs. You can roughen up the surface of your master with a higher grit to give it a matte finish. However, if you seek a "fancy" texture, then a spray coat might be better, or the finish could just be incorporated into the 3D printing if you are using a hi-res machine. I cannot advise you on which printer is better for your purpose. One must consider printing cost vs. finishing time/effort on the master (and final parts) vs. anticipated production quantity. If you design with enough slop, you should be okay, regardless.

Ha! Yes, such "no-budget" scenarios are unheard of here in the states -- especially on Craigslist.

I 2nd this! @Mattias Burling Also, could you do a DOF "equivalence" test between this set-up and your Digi-Bolex with a prime lens? Thanks for showing these photos. They look spectacular!

Those are nice mold mock-ups, but you might be able to get away with a one-piece silicone mold for each part. To do so, you may have to sacrifice some of the side holes and top recesses and make those as separate drilling/machining operations after casting. To ease such drilling, you can include small pilot/alignment indentations in your master. This could be an easier way to go if you are making small quantities. By the way, for finishing/buffing the surfaces of your master, you can use pro nail buffing blocks or use pro nail buffing sticks. The lowest "grit" surface on theses blocks/sticks should buff your plastic master to a nice glossy shine.

You might also mention something in regards to your ability to bring the project in on time and on/under budget.

In addition to the suggestions above, you might consider a used camera: the Digital Bolex; the Sony F3 (w/4:4:4 capability); the Sony FS700 (w/raw recorder); or a KiniMini refurb. Also, I've heard good things about the DJI X5R.

By the way, PC-7 Epoxy might stick to glass glass better than J-B Weld. Here is a photo of an old PC-7 hardware store display with various objects epoxied to a glass soda bottle: I remember these displays and there was no way to remove the glued objects without breaking the glass bottle.

The easiest solution is a cardellini clamp with a magic arm. If the jaws of the cardellini are too big (visible on the bottle), you could use a thin plumbing pipe clamp with a rubber liner around the bottom of the bottle. Use plumber's epoxy on the head of the clamp's bolt to keep it from rotating and attach the magic arm to the exposed threads at the other end of that bolt. You could likewise make a clamp plumbing strap and a long 1/4-20 bolt. Plumber's epoxy would also work here in keeping the bolt fixed so that you could attach a magic arm to the exposed threads. Rubber scraps from an old bike inner tube could be used as a clamp liner. Or, you could just use J-B Weld to glue two, stacked, 1/4-20 nuts directly to the bottle. J-B Weld sticks to glass (seems to hold up to pot smoke and heat, too!) -- not sure if plumbers epoxy sticks to glass as good as J-B Weld. You might have to score (roughen) the surface of the glass to improve the adherence of the J-B Weld. Screw two 1/4-20 nuts onto the end of a 1/4-20 bolt, then use plenty of J-B Weld around the nuts and around the surface of the bottle. When the J-B Weld has hardened, simply unscrew the bolt from the two nuts, and the nuts will be perfectly aligned within the J-B Weld. Screw the 1/4-20 threads of the magic arm into the epoxied nuts.

I am merely stating fact. There is no absolute correlation between dynamic range and bit depth -- they are two completely independent properties. Yes. That is how binary math works. However, that has no bearing on how many bit depth increments are (or should be) mapped to a signal/dynamic range. The number of "steps" mapped within a signal/dynamic range has nothing to do with the scope of that signal/dynamic range -- that is the point. This is a simple concept/fact. There might be an older machine vision camera with those specs, but it is an irrelevant request -- there are plenty of cameras featuring multiple bit depths yet with only one dynamic range. The dynamic range of these cameras doesn't change because the cameras' bit depth changes. Again, bit depth and dynamic range are two completely independent properties. Also, subject/location contrast has nothing to do with a camera's dynamic range. Standards and their calculation have nothing to do with the basic fact that bit depth and signal/dynamic range are totally different properties. Furthermore, dynamic range is so independent of bit depth that many systems (including numerous cameras) have incredible dynamic range with absolutely zero bit depth.

Raw data can be mapped other ways, too. There is no absolute (nor absolutely "ideal") correlation between dynamic range and bit depth. Dynamic range and bit depth are two completely independent properties. A camera with 14 stops of dynamic range can be mapped with a 6 bit depth, and, likewise, a camera with a dynamic range 6 stops can have a 14 bit depth. Furthermore, there are numerous digital cameras that allow one to choose mapping different bit depths within the cameras' fixed dynamic range (or, more accurately, signal range).

Tell them that you have the expertise to fulfill the director's vision and that you have extensive experience in using the language of the camera to tell the story. Of course, you must be able to follow through from such statements, and it certainly helps to have a reel that demonstrates such ability.

Don't know how/if the camera's noise reduction settings affect the raw stills, but the problem doesn't appear in the raw files. By the way, there are raw files with two different noise settings (NR1 and NR6) available for download, and there is no noise reduction smudging in either. The problem is reduced on the jpegs with noise reduction set to "1." If the noise reduction is working in raw mode, I am not seeing any smudging, so perhaps this problem is just limited to jpeg processing. Wonder how the noise reduction in the GH5's 4:2:2 video will compare with that of the jpegs.

You have a nice idea! Instead of high-end 3D printing or CNC machining, consider silicone molding. You can 3D print a rough master and finish it to your liking, and then make copies of it with the silicone mold. With this method, you have to be mindful that the parts need to be designed to release from the mold, and, of course, you might need plenty of ventilation to safely cure molding compounds. I see a potential problem with your orange part releasing from a mold. Also, don't get too precise with the design/engineering -- in most instances, the precision one thinks is needed is not actually required. Designing with "slop" (sloppy engineering tolerance) is generally a good practice, especially with mating parts and with any manufacturing process that might have shrinkage (injection molding, die casting, extrusion, blow molding, etc.). Design with as much slop as possible, and allow for that slop. It can save you a lot of headaches later on. For your enclosure, it might be good to have about 1/32" (0.8mm) of slop... meaning, for instance, that you should probably design the inner mating surfaces of the orange part to be 1/32" (0.8mm) wider than the outward mating surfaces of the yellow part. Hope this helps.

Of course, you could gain over a stop by using a focal reducer or by using the Metabones BMPCC speed booster.

Not all 80A filters are created equal. You might try one from another manufacturer. Same with CTBs and CTOs (lighting color temp correction gels). Also, filters can fade with age. Nevertheless, your corrections on the test are well within almost anyone's tolerances. Good job matching the two images! I don't see a huge problem in your test (yet I do see a slight difference), but If one finds oneself without correction filters and if one is also concerned about slightly noisy channels, one can often overexpose 1/2 stop (depending on the subject/scene). On the other hand, with well-lit sets, I have never encountered noise problems shooting raw with tungsten and only IR cut filters and ETTR.

Thanks for doing this test! OP is shooting weddings/events, so OP might often encounter incandescent tungsten which starts at less than 3200K (with a higher percentage of IR), and which is likely further dimmed (further increasing the percentage of IR). I don't know if your 500w halogens are 3200K, but using a full CTB correction (80A) probably helps OP color-wise and IR pollution-wise. However, with a full-CTB/80A filter, brightness will suffer a 1 1/2-stop to 2-stop hit, so it might be better to use a 1/2 CTB (80C?) correction.

Blackmagic cameras are sensitive to IR, and tungsten sources emit a high percentage of IR. So, you might want to try an IR cut filter.

Evidently, the TIFF format has metadata designations that include camera make, model, color profile, orientation, exposure, etc. (see the linked exif file below). The highest quality still file that your camera produces will give the best image quality. However, it can be unwieldy or visually detrimental to directly edit raw camera files. Some still photographers first run their raw camera files through a raw image "processor/developing" program (such as Lightroom), before retouching them in an editing program such as Photoshop. I use the open source Darktable as my processing program and GIMP as my photo editor. I just checked the exif info in a tiff that I converted from a raw file using Darktable and the metadata is extensive: tif_exif_info.txt (Note that I was using a manual lens on a Canon EOSM, so there is no lens/f-stop info).

Every raw file "development" program that I have used initially defaults to the camera's settings. Of course, one can usually change from the default settings to adjust white balance, exposure, camera profile, etc., but it probably depends on which program that you are using to convert from raw to .tiff.

Shooting raw stills and comparing them to processed frames can reveal the extent of sharpening or other image processing that is occurring in the camera.

Don't trust auto-focus for anything critical, even if you can see the image momentarily magnified. Again, seeing raw stills could help, as those images should bypass a lot of the in-camera image processing.

Did you use the same 12-35mm f2.8 lens on both cameras, or did you use two different 12-35mm f2.8 lenses? I wonder what the comparison would be like with both cameras set to "natural" and with the sharpening dialed all the way down in both cameras. It would be helpful to also see raw stills. By the way, are you auto-focusing or manually focusing?