sunyata

A problem I've had with youtube is that they have flagged music that I know is not in their database.. music I have personally recorded and that has never been recorded before in studio or live. I have no idea how to get it unflagged, I'm sure it's a slow and laborious process, so I just let it go. If it's a legitimate violation I have no problem not posting. But I think these algos are taking over personally and it's bigger than just music on the internet.. From drone strikes, to front running stocks, to wire taps, or my credit card getting cancelled because I decide to buy diapers for a baby I don't have.. it's getting ridiculous. Maybe I want to make a giant diaper wall... screw you algo.

Awesome.. Love it. Just saw the close-up.. it's really interesting, it looks like film grain up close.

Plot scanline was a way in Shake to visualize what color operations are doing to your blacks (bottom left) and whites (top right). Someone was kind enough to make an expression for this in Nuke. This is a cheat-sheet for the basic functions combining the often redundant names. For example, Photoshop, After Effects and I'm guessing Premiere, have a "levels" tool that is an expand, with a gamma (the midpoint slider) and a min / max at the bottom.. all in one tool. If all you want is an expand, then hopefully this helps show how you can do it with blackpoint and whitepoint, or lift and gain in Resolve. You could then add a gamma right after and have 2/3 of the levels tool.

Don't listen to what people on the internet say.. it's the people that don't say anything that know what they're talking about.

What the hell is going on here? I've totally lost track. Bring back film!

Very interesting debate.. I understand what you are saying. I also think he's making a valid point. Most people are trying to do a conversion in their head when they buy a lens, based on a 35mm film lens, so they think of crop, sharpness and dof conversions as primary importance in price. He's saying if you advertise the crop equivalent, then you should also think about the "dof" equivalent. Unfortunately, he kinda emphasizes this as an empirical aperture number, but he could have said instead "this lens will give you the DOF equivalent of X aperture at 35mm..." and include "assuming you step back and frame the subject the same scale". One thing about the noise issue though.. you have to always consider the limits of our fabrication abilities as they stand today.. in theory the density and size of photosites should have a linear inverse relationship, higher density -- in same format -- the smaller the photosite, but in practice, we are bound by limitations of manufacturing. Leakage and impurities in materials, not to mention impurities in glass and smoothness of lens surface, all become more challenging as you shrink the camera. This is why I still love my film lenses, you couldn't just shrink them down and have exactly the same quality in practice, only in theory... and even that would break at some nano-scale.

Uh oh, heavy topic wrapped in generic subject header.. trouble ahead.

Cmerlin, I'm not sure why you want to transcode with a stand-alone program? The way proxies work is they create a dynamic relationship in the software you're using, so that you see the low resolution file when working (faster), but when you render, it will reference the source hi-res. In After Effects, you can right click on a comp and render a proxy, that will manage the relationship for you.. It's the same in other programs, I don't know about Premiere, but I'm guessing it's in there too. Davinci Resolve makes proxies automatically. You just need to go into the preferences and make sure these programs are using your 1TB RAID.. I'd just use that as dedicated scratch, you should be fine for short format work. For proxy settings, don't use a delivery codec, use uncompressed linear RGB (that is what your display wants to see, it's called display-referred gamut). If you want to do it all in one program, just edit in Resolve. good luck

That's a misleading stat to put out by Sony when it's recording 8bit 4:2:0. This is what the sensor can see before it chooses the right 8 bits. Once it's taken all the data that the sensor is capable of seeing and reducing it to an 8bit integer color space, you will not have many stops to work with in post before you start to see noise/artifacts. But, I don't think the number is technically incorrect, this is a novel sensor (full frame size with low density pixel count). That's why it's kinda ridiculous they are not going to allow at least 10bit HD export. It could be a heat issue, which is really a cost issue, but this camera isn't exactly cheap.

Keep in mind that ProRes is lossy.. if you can bypass compression for your intermediate file, that's the way to go. Michael, He has a level 0 RAIDs so maybe he can work with a 10bit uncompressed linear RGB file instead of ProRes? ffmpeg is great for conversions, it's free and open source. this is a script I wrote to take a folder of mov files and convert them to rgb 10bit dpx sequences for compositing, it also creates subfolders for the sequences: for i in `ls *.mov`; do mkdir `echo $i | sed 's/(.*)..*/1/'`; ffmpeg -i $i -pix_fmt gbrp10le `echo $i | sed 's/(.*)..*/1/'`/$i".%04d.dpx"; done; you could add a resize to the command line with the -s 1920x1080 option and use a mov container codec like "-c:v v410" which is 10bit Uncompressed 4:4:4. Also, I don't know about Premiere, but most programs let you do the reformatting (codecs etc) when you make the proxies from within the application.

Davinci Resolve apparently auto-generates proxies, so if you use one of your RAIDs as the proxy drive, you should be fine. But basically, don't make lower quality 4k files to work with, the standard workflow for film or video is to generate an intermediate at HD (in the old days it was SD) so you can see it at 100% and color grade etc, that can be a proxy or even a dynamic cache file (if the job is small), it depends on the software. Ideal working depth is 32bit float, proxies can be 10bit to 16bit 1/2 float, at HD you should be able to play back uncompressed linear RGB. For backup, if you aren't afraid of tapes, and I know a lot of people that are, you can find deals on LTO-3's on ebay, in the $300 range used and excellent working condition etc. Then if you buy a box of tapes you can get the price down to around $25 or less per TB, this will take up a free PCI slot though and you'll need to find an old SCSI LVD 160 card somewhere, they are also really cheap now. (the SCSI LTO's are much less expensive than the SAS or FC). I gotta do this tape thing because I have clients call me up for something that is over 4 years old, and they need it like "yesterday", so I have to be able to put an entire job back online and I need a catalog. I use bacula, and it will also work with almost any form of storage. Proprietary software like Retrospect has screwed me more times than I can keep track of, I'd stay away, and just manually copying everything onto firewire drives and cramming them randomly onto a shelf is just uncivilized. Good Luck

Way to shoot yourself in the foot again Sony. I think Ebrahim is correct in that they are focused on not cannibalizing an ageing product line, but it needs to be able to export 10bit for that price, even if only at 1080.

ahh, but there is if you work for sony and drink the kool-aid, which i have done a few times (but not electronics). they have this obsessively broad vision for their products, everything must fit tightly into it. i've been considering buying one to take apart to try to hack. probably no easy task.

For anyone interested in color grading I'd highly recommend doing a sticky post or article on color theory. This would be just the vital info for post since it's a broad topic, but like: 1) The CIE 1931 XYZ functions and how we perceive color. 2) Color spaces / gamuts, from film log and digital sensors, to your monitor. 3) Definition of "dynamic range" and "exposure range", also "hdr" vs "ldr" workflows. 4) Color correction examples using a simple plot scanline. Optional: 5) Why dinosaurs and cavemen never lived together.

Going even more off-topic: Anyone interested in the anti-nuclear argument should check out Amory Lovins.

I was waiting for that one fuzzynormal.. boom bang folks. Once when I was a teenager working on a 63 valiant, I got a huge face full of genuine early 60's vintage asbestos brake dust, then I found out it can cause cancer:) Did that stop me from continuing to work on old cars? Uh, no, but now I wear a mask. It's always better to be informed. It's not about what is worse, Fukashima, trans-continental flights, white paint, old brakes, or Thorium lenses, you want to avoid carcinogens as much as possible over your lifespan because it ALL accumulates. I don't think junior is "fear mongering" or trying to create "hysteria" Sean, maybe it's slightly overstated in the thread title, just do your own research. No need to ridicule the guy.

sadly, the place where i'm most exposed to radiation and toxic waste is probably not my camera, it's when i surf in los angeles.

If you're into DIY electronics, there's this too: https://www.sparkfun.com/products/11345

Welcome to the forum junior!

from wikipedia, or course: Thorium dioxide was formerly added to glasses during manufacture to increase their refractive index, producing thoriated glass with up to 40% ThO2 content. These glasses were used in the construction of high-quality photographic lenses. However, the radioactivity of the thorium caused both a safety and pollution hazard and self-degradation of the glass (turning it yellow or brown over time). Lanthanum oxide has replaced thorium dioxide in almost all modern high-index glasses. and this: it was found to be a carcinogen, sometimes causing cholangiocarcinoma (that is with respect to another use for thorium dioxide in the past, as an x-ray contrast agent). I'm not really concerned about 90% of my lenses, but I was just working on an old rangefinder lens, including taking apart the elements and removing fungus, so this is good to know. My main concern wouldn't be beta rays but just doing something stupid like shattering a lens trying to re-glue it or something. Not likely, but I'm going to check if thorium dioxide was used in glass before I work on it now. Thanks for the post junior.

I still want to find a geiger counter now and see what my Zeiss Planar 85mm f1.2 does.. that thing just looks toxic.

That's very interesting.. what do you know about Zeiss C/Y lenses? Apparently it's in the glass itself. some vintage ones on the list: Konica Hexanon 57mm f1.2 Bell & Howell Director Series (Model 1208?) XL Super 8 movie camera; Zoom Lens f: 1.2 F: 9-22.5 mm Carl Zeiss Jena Pancolar 55mm f1.4 (measured at 2360 nSv/h) Carl Zeiss Jena Pancolar 50mm f1.8 "Zebra" Carl Zeiss Jena Biometar 80mm f2.8 "Zebra" "(Only P6 mount version ) Carl Zeiss Jena Flektogon 50mm f4 "Zebra" "(Only P6 mount version ) YouTube Carl Zeiss Jena Prakticar 50mm f1.4 (1st version with engravings around the outer side of barrel)

I wonder if instead of narrow focus, if you could use the scatter info to create longer focus, with more light, i.e. being able to have extreme long focus in low light situations?

I thought those demos were just animatics.. are you sure you can look around the image? I thought that was only possible with waves that penetrate an object and in blade runner.

That is certainly a fast laptop, maybe it can play 10bit 4k, god bless you if it does. I converted part of the 4k video to qHD (960x540) 10bit prores, specifically a shot where the camera pans up the fake eiffel tower (love vegas), and you can see it's really jerky. it looks like maybe a sticky tripod. http://vimeo.com/92897851