-
Posts
1,839 -
Joined
-
Last visited
Content Type
Profiles
Forums
Articles
Everything posted by jcs
-
I find 4K useful for post zoom/crop even for 1080p target content. A7S II 4K looks really nice scaled to 1080p (GH5 has more real detail in 4K). Are you saying the A6500 hasn't freed you from 'worrying about the whole resolution issue"? I'm not sure I understand your point? For medium/close up shots, using a e.g. Black Pro Mist filter on the lens can actually look better for 4K content and display (or adding a little blur in post). For wide shots, the extra detail is typically helpful as long as there are no digital artifacts (aliasing, oversharpening, etc.). Target detail/resolution depends on the needs of the content. More source resolution/detail is always useful and desired if storage costs aren't an issue, right? Would I choose an 8K Red over a 3.4K Alexa? It would depend on the desired final product. For skintones, highlights, DR and overall color the choice is clear For a nature shoot, where lots of post punch-ins and crops are required and final delivery for 4K, the 8K Red might make more sense (would need to test it first, probably with their appropriate 'detail' OLPF). Isn't that obvious? Blurry 1080p does not provide as high of perceived detail as true 1080p? Over-sharpened blurry 1080p starts to look like video and gets worse as sharpening artifacts create 'aliasing pixel crawl'. True 1080p requires oversampling sensor photosites (Nyquist again). The C300 I had killer 1080p from a "4K" sensor (only green was averaged, however that's the primary component for luminance (~.6)) thus the C300 I had very good perceived 1080p (it did alias though). Did you check out the F65 4K example shooting the test chart? The rings were super detailed and alias-free- amazing! ("8k" sensor with 4K target.. Nyquist). Perceived resolution is all about contrast at various spatial resolutions. Starting near the pixel level for convolution sharpening, then going lower frequency with unsharp masking, then getting lower with local contrast enhancement (still using unmask sharp operation, only now we're calling it contrast vs. sharpening), and then with curves and LUTs and finally simple per-pixel multiplication and subtraction (simplest form of contrast). Adding in noise/grain which is higher resolution than the source video can also help to improve perceived resolution: adds higher resolution texture to surfaces. Isn't all of this well known? If this was Yedlin's point these concepts could be shown in a 30s video!
-
I'm using calibrated 4K Dell displays (27" P2715Q and 32" UP3216Q). Part 1 was 1920x1080, watched 1:1, part 2 was 1920x1200, watched 1:1. The difference is even more pronounced watching full screen! It's not that more detail is created when displaying full screen, as the pixels are interpolated (would be same effect with a 1080p display), it's just easier to see the differences, just like blowing up the stills until you can see the pixels. I've done lots of experiments with resolution and perceived detail, including showing that the 5D3 can look decent at 1080p when pos sharpened. And more recently showing that soft 1080p can look OK when scaled to 4K then sharpened and grained, and will easily cut with 4K native material (important sometimes when slomo is only 1080p). You didn't see a difference at 1:1 when he A/B'd, especially in the eyes/eyelashes? It's not dramatic but it is visible. This is for 1080p 1:1. For 4K display, it would be clearly more pronounced when full screen (3840x2160 video on a 4K display), right? So what's Yedlin's point?
-
Upscaling 1080p to 4K and sharpening, or sharpening while upscaling (using a more advanced resampler) to 4K, then adding fine noise/grain, and even using local contrast enhancement (a form of Unsharp Masking), and especially if using Super Resolution (using aliasing information to generate actual real detail over frames of video), will result in a visible perceived detail/quality improvement over just upscaling 1080p to 4K (bilinear/bicubic scaling). I can clearly see detail differences in Yedlin's tests, especially in the eyes and skin detail. I didn't see any in Yedlin's tests that showed anything new, really much ado about nothing! As for super high resolutions not showing more improvement, that makes total sense based on Nyquist sampling theory. Past Nyquist the extra resolution is still providing additional post crop/zoom options, so still useful if one has the storage space. The Sony F65 clearly shows an advantage at "8K" for 4K delivery vs. lower resolution cameras. More info here: "I only later learned that The Revenant (4k mastered and 4k delivered, watched by me sitting in row 6) was 87% shot in 2,5k. A 4k projection that upscales a 2k image or shows a native 4k image will look somewhat more brilliant than anything projected in 2k. So 4k TVs and monitors do have their benefits. It's just that we all grotesquely overestimate this factor." Who is "grossly overestimating this factor" in 2017? It appears most people today agree about story/sound/color/DR/noise etc., and more resolution being useful too. A few years ago EOSHD was all about resolution (still panned the 6D II for no 4K; Canon is protecting the 5D4...), especially around the time of the GH2 hacks (and lots of slamming the 5D3 for being soft), even I made detailed posts regarding various camera resolutions. In 2017 it's more about color, DR, motion, manageable file sizes etc.
-
Agree storytelling (even sound!) is more important- if either of those are off, it really hurts the production regardless of other factors. However those are orthogonal to the image quality debate, right? If everything else is equal- color accuracy, skin tones, noise structure, low light, motion cadence, dynamic range, no sampling artifacts, then more resolution is always useful unless storage space is a major factor. For example, more resolution means you can shoot wider and be able to punch in in post. Super time saver! Image stabilization also benefits from extra resolution. Would I pick a Red or Sony over an Alexa if given the option for the same price? Almost certainly the lower resolution Alexa (ProRes), unless I could first test an F65 to make sure I could easily get great skin tones (and if shooting XAVC was good enough quality. F65 RAW is too data heavy unless someone else is paying for storage ). I shoot mostly "soft" 1080p on the 1DX II, which works fine for medium to close up shots (the aliasing hasn't been a problem so far. If I encounter it will just shoot in 4K and lose a little shallow DOF (and get massively bigger files). Similarly, I shoot mostly 1080p (pretty sharp) on the C300 II for medium and right shot. For full body (green screen) wide, I shoot 4K. On my 4K desktop displays, even on the YouTube, there's a huge difference in quality watching in 4K vs. 1080p: Don't you agree?
-
I am an engineer and Nyquist Sampling Theory applies to all forms of signals, including images (2 dimensional pixel sequences): https://en.wikipedia.org/wiki/Nyquist–Shannon_sampling_theorem Undersampling and Moiré (a symptom of aliasing): Properly sampled (for some reason this image on Wikipedia is larger than the Moiré example): "(well, it applies to anything, but sawtooth waves are effectively of infinite frequency because they contain infinite high odd order harmonics)" - I always thought the simple explanation was sawtooth and square waves are discontinuous functions and require an infinite number of summed sinusoids (Fourier, related to the discrete cosine transform and sampling/reconstruction theory) to exactly reconstruct the discontinuous edges. However in practice it doesn't need to be perfect, as it works fine for graphics and sound. What lead you to believe that Nyquist doesn't apply to image sensor sampling? Do you have link to a math or science paper explaining the mathematics along with example images? We usually agree, especially on audio gear! From https://www.edmundoptics.com/resources/application-notes/imaging/camera-resolution-for-improved-imaging-system-performance/: I do acknowledge that in practice it may be possible to under sample (<= the Nyquist frequency) when using a well-tuned optical low pass filter, such that the aliasing is so minimal in practice it's not worth worrying about. In terms of actual examples I'm aware of, the most alias-free, detailed 4K examples I have seen have all been from well-over 4K sensor resolution, the most detailed seen so far with the Sony F65 ("8K" (kinda)). I would expect 6+ K Red and Alexa 65 to also do very well. I hadn't studied this before, however the so-called Kell Factor is said to reduce the displayable resolution by a factor of .9 for HDTVs: https://en.wikipedia.org/wiki/Kell_factor. A 1080p CCD camera can be displayed at only 1728×972 resolution in order to avoid beat frequency patterns. I'm not sure if I've ever observed this effect on computer monitors (or HDTVs). Didn't find any examples via the google? Have you noticed how detailed, alias-free, and natural 1080p shot at "4k" looks, then downsampled in post to 1080p? The same applies to 4K: "8K" sampled down to 4K looks superior to lower resolution examples. Recall how the F65's "8K" sensor looks way more detailed (and has no apparent aliasing!) vs. all the other cameras in this test: From real-world examples alone I'm calling Shenanigans on statements that Nyquist doesn't apply to images (or 2K doesn't really matter vs 4K, e.g Steve Yedlin's recent videos on resolution). Happy to butter up some popcorn and learn something new if someone has math and example images which show otherwise
-
I can see a noticeable resolution difference between his 4K and 2K examples at 1:1 on Dell 4K desktop monitors. More so, I can see a huge difference in 4K vs. 2K in the theater. For HDTVs, I'm still using 1080p, mostly because there's so little 4K content for viewing. When I do finally upgrade, I'll get larger panels to take advantage of 4K (current HDTVs are 52" and smaller). I did a similar test a while ago to show how the 1DX II's "soft" 1080p can be scaled up to 4K, sharpened in 4K (Lumetri), then fine grain noise added in 4K, to produce an image perceptually similar to a native 4K shot: Watching true 4K (or close to it) on a desktop computer is massively better looking than 2K. A very large 4K HDTV sitting at a similar FOV (which is pretty close for most people) is similarly more impressive than a smaller HDTV in 1080p. I think most people will agree that the native 4K clip looks nicer and more organic than the up-rez'd 2K clip, especially if viewing on a 4K monitor. So while I agree with Steve that we can do tricks with 2K upsampled to 4K and sharpened+grained to come closer to 4K, the truth is 4K is still better than 2K, so I'm not sure what he's debunked? Some day Super Resolution and technology based on Generative Compression will be able to provide massively better upscaling algorithms in NLEs and even future HTDVs (real-time hardware). http://www.infognition.com/articles/what_is_super_resolution.html (still no plugins or NLE support on the commercial market?). Generative Compression algorithms will break an image (or series of frames) into features which can be re-rendered at any desired resolution, similar to rendering vector art / spline curves.
-
Yeah, generative compression applied to moving images (video) will result in a massive reduction in file sizes vs. the current macroblock DCT methods.
-
I remember the Kuro from 2008, it was very nice (ended up getting one of these, and am still using it as a live monitor for the C300 II in the studio): https://www.amazon.com/Sony-Bravia-KDL-52XBR5-52-Inch-1080p/dp/B000WDW6G6. I think the OLEDs have finally caught up (and passed) the top plasmas, but yeah it did take a while!
-
While fractal and wavelet compression made a splash a few years ago (and wavelet is still used today in JPEG2000 and Redcode RAW), there's a lot more to be discovered based on generative mathematics and machine learning. Here's one cool example: http://www.wave.one/icml2017/
-
As embedded computing power increases, the use of RAW will rapidly diminish. There are still breakthroughs ahead based on generative mathematics- fractals, DNA etc., that will encode information without resolution, allowing any desired output to be rendered. Here's one example based on machine learning (not clear if they can generate arbitrary output resolution, however the compression quality is very high):
-
Netflix charges more for 4K content subscription access, and thus wants more "real 4K" content. Perhaps the lawyers mucked this up so people (external lawyers, lol) wouldn't litigate? Everyone at the professional level knows that uprez'd 3.4K Alexa footage spanks most of the competition, and most consumers wouldn't care anyways, even those paying for the 4K sub since it still looks way better than 1080p when viewed on a 4K display sufficiently up close. While a test chart would show truth, I've seen people say crazy things about test chart results, so maybe just sticking with numbers/specs gets them close enough to their 4K goal with the least drama. Lol what does this say about your post?
-
Maybe the FS700 isn't on the list because of aliasing and chroma artifacts, and perhaps skintone color science isn't up to the other cameras? Might also be related to having to use 3rd party devices to get 4K. While we were able to get decent skintones and color from the FS700 (and the slomo was great), it was hit or miss, and so we went back to Canon when we acquired our next professional-level camera: C300 II instead of FS7. While the FS7 has improved color science and overall image quality compared to the FS700, the C300 II still has nicer skin tones with less effort. Even if the FS7 had matching image/color quality, C300 II still wins for native EF lens support and usable AF. With the C300 II set up with ARRI settings, ARRI LUTs can be used and it's like using an Alexa with AF! Is the Alexa still superior in DR and total image quality- absolutely, even though the C300 II is "4K" (which it really isn't- it's less than 4K and aliases!). If Netflix really wanted to make this about 4K quality, they'd base the decision on shooting a 4K test chart. How many professional cameras actually produce a real, measurable, alias-free 4K image? Sony F65 ("8K") Red ("6/7/8K")? Alexa 65 (6.6K)? Pre Nyquist Sampling Theory, we need 2x oversampling to fully capture a signal without aliasing. Along with an optical low-pass filter, it might be possible to get 'close enough' so aliasing is minimal/not visible with less than 2x oversampling, so a test chart is helpful. Here the F65 rules (and maybe Red at 7K; only a 1080p output was provided): So how does the Alexa 65 compare to the F65 and Red shooting a 4K test chart? The only camera I would put in the same league as the Alexa for color and overall look for skin is the F65 based on high-end Hollywood movies. I suspect the Alexa may be much easier to achieve great skintones and that's why the F65 isn't used as often for high-end productions (and the insane data rate of the F65). If Netflix didn't have this 4K requirement, it would probably be dominated by Alexas, which isn't good for competition, so that's a plus in a sense (Alexa 65 is very expensive and rental only). The C200 420 8-bit 4K won't qualify since it's not 10-bit. The 4K raw should qualify if it's log-encode (otherwise must be 16-bit linear to qualify), though it's not true 4K either- not enough photosites and both it and the C700 alias in 4K. What about the GH5 4K 10-bit? Would be interesting to see how it does on the same 4K test chart / conditions Geoff Boyle is shooting.
-
For $299 these work pretty well for 1080p (I added the heat sinks mentioned in the comments (just using rubber bands to hold on): https://www.amazon.com/gp/product/B00I16VQOY/ref=oh_aui_detailpage_o01_s00 For 4K and USB 3.0, this works well with the A7S II and GH4 in 4K: https://www.amazon.com/gp/product/B01H3Y3IL8/.
-
The Zoom F4 is indeed much better than the H4n (I have both and both are for sale! (also have a DR100 MKII for sale)). The noise floor is now usable on the F4, though still not as good as Sound Devices. If that's all that matters, then sure, they are close on noise, even more so with highly compressed online streaming listening. When it comes to sound quality, especially with decent speakers/headphones, any Sound Devices is in another class compared to the Zoom F4/F8! If you can't hear the difference, what headphones or monitors are you using? I'm using the amazing-deal Focal Listens (closed back) and the best-bang-for-buck Stax SRS-3100 w/ Ultra mod (Socas adapter plates and Brainwavz Hybrid Memory Foam pads (also have thick ZMF Lambskin angled pads, though currently prefer the thinner Brainwavz)). Previously: Sony 7506 and Audio Technica ATH-M50 (old model): both are also excellent for the money (and both still used for recording/location monitoring). Headphone amp is Sound Devices USBPre2 (excellent). These are pretty good demos which illustrate the sound quality differences: It's not always about price, sometimes is preference. I always found the RME Fireface 800's mic preamps a bit sterile and digital sounding (harsh). I replaced it with the comparatively cheap Focusrite Scarlett 2i2 (over 10x cheaper) which has nicer preamps- smoother, more natural, and more analog like. While the Scarlett has sufficiently quiet and smooth mic preamps, the Sound Devices USBPre2 mic preamps are much fuller and more detailed across the spectrum. Finally, a professional life-saver feature of the Sound Devices products is the analog limiter- it's basically unclippable. Even if you're very careful and set levels safely (e.g. -12 dB), sh*t happens and you're going to get clipping, especially for live/location/doc work. Using a safety track can help but that's more work in post and still not as good as SD (or higher end) analog limiters. For the price difference between the F4 and MixPre6, there's no way I'd recommend the F4, even if I was a brand ambassador for Zoom. If one needs more channels and doesn't have the budget for equivalent Sound Devices or similar gear, the Zoom F8 makes sense. For indie/hobby work, digital limiters (or no limiter) recorders work fine, as it can be no big deal to go back and re-record once you have reviewed the recordings and find you've got some problems. If time is important or one is paid by someone else, analog limiters are really helpful. Regarding Brand Ambassadoring for Aputure: it's cool to promote products one is compensated for, however as others have noted there are some issues with the current Deity: noise floor and off-axis rejection. While it does sound similar to a 416, it's much noisier, and side & rear rejection is much worse than the 416 and NTG-2. I've only listened on YouTube, however this review says the Deity is noisier than a $249 NTG-2. It appears Aputure tuned the Deity with high gain (hotter, more sensitive) and the budget ran out to keep costs and noise low at the same time. It's a fair compromise, and in noisy environments the noise won't really matter, or noise can be removed in post (FFT/spectral or simple expander/noise gate). Regarding mic placement and usage: there are NO RULES! There are guidelines, however what matters is how the mic and placement actually sound for the conditions- that's it. Mic on camera? Absolutely for run&gun and it can sound great. It's totally fine to use a shotgun indoors in a non-reverberant room. The 416 is also an excellent VO/indoor mic. As long as the room's reflections aren't causing the shotgun to 'phase out' (sh*tty sounding phase effects), they can work great. It only took one time after many years of using the NTG-2 as my main mic to hear nasty reverberant phase effects to finally add a hypercardioid. I picked up both the Audix SCX1-HC ($500) and the Schoeps CMC641 (a lot more). If I 'squint my ears' maybe I can hear a difference between these two mics. I ended up keeping both to do stereo recordings. Even though they aren't matched, they work pretty well together (all the indoor Cosmic Flow shoots use both these mics as a stereo recording on booms above talent. Can you hear the difference?). At that time I also finally upgraded to the CMIT5U, which works well outdoors, indoors, and on camera too! Someone criticized plugging a Schoeps directly into the C300 II- as if it could only be mated with a Sound Devices or better lol. When I'm running both camera and audio, simplicity and reliability are first priority. For critical location shooting for playing clients, the higher quality preamps and analog limiters of a Sound Devices mixer would totally make sense. Here the NTG-2 sounds better indoors than the normally excellent Audio Technica 4053b (one of Curtis Judd's favorites): The 4053b and SCX1-HC tested for dialog here: http://www.4kshooters.net/2016/06/28/five-affordable-boom-microphones-for-capturing-high-quality-indoor-dialogue/ Cables: Neutrik connectors are great and after using cheaper cables really appreciate Mogami quality. Over the years I started out with the best low-cost gear I could find (Rode, Audio Technica, Zoom, Tascam) and slowly upgraded to better gear (Sound Devices, Sennheiser, and Schoeps). You really appreciate the quality of the higher end gear after learning on the budget gear. The Audix SCX1-HC is a special gem- fantastic quality for the price. For internet streaming and listening on a cellphone (where most people consume low-budget/indie/free content), the budget Rode and Audio Technica gear is plenty good!
-
+1 for MixPre6 over MixPre3 and SD over Zoom preamps. For wireless we've been using the Sennheiser G3 with Sennheiser and OscarSoundTech lavs. They have worked very well. We also use the ENG plug with the Rode NTG2 and it works well as a wireless shotgun. For those wanting or needing stereo sound, the Audio Technica BP4029 is a really nice mid-side stereo shotgun mic (5 star rating). You record two channels and can convert from mono to ultra-wide stereo in post and anything in between, using a plugin like MSED (free). Recordings with this mic can have a 3D effect when listening with headphones. For full 3D recording, the Sennheiser Ambeo works really well to capture Ambisonic recordings for 3D manipulation in post (or for real-time VR apps). Ambisonic recordings can be mixed down to stereo with an HRTF, providing excellent 3D spatialization when wearing headphones.
-
I rank the Sound Devices MixPre 3/6 above the Zoom F4/F8 for pure sound quality: smoother, fuller, more natural sounding, more analog like, and of course the amazing analog limiters. I base this on owning a Zoom F4 and a Sound Devices USB Pre2 (which has the same audio topology as the 744T, meaning it sounds as good as the higher end SD recorders) as well as the YouTube/SoundCloud comparison videos: Couldn't find any comparisons of Sound Devices to Zaxcom, Nagra, AETA or other high-end recorders. It seems sound quality doesn't improve after Sound Devices, only features (channels etc.), power system, and size? (that's what I got from a quick peek at Gearslutz.com).
-
"Cheap" spray deodorant (said to work better than expensive brands), hair spray, purpose-made anti-reflection spray, After Affects (and now PP CC): mask, track & blur: http://nofilmschool.com/2015/02/reflective-surfaces-ruin-shots-quick-fixes, https://petapixel.com/2017/07/20/tip-deodorant-spray-can-tone-reflections-product-photos/
-
Lightweight long-sleeve shirt, e.g. Columbia hiking/fishing/sun shirt: http://www.columbia.com/on/demandware.store/Sites-Columbia_US-Site/en_US/Product-Variation?pid=FM7048&dwvar_FM7048_variationColor=010 or https://www.amazon.com/Columbia-Blood-Sleeve-Woven-Shirt/dp/B00QRVNPNU/r, http://www.columbia.com/mens-irico-long-sleeve-shirt-1654422.html?dwvar_1654422_variationColor=492#prefn1=ECsleeve_length&prefn2=technologyInitiative&prefv1=Long+Sleeve&prefv2=Cool&start=0 No black but dark grey, very high quality shirts (I have a few of these): https://www.amazon.com/ExOfficio-Mens-Strip-Sleeve-Medium/dp/B00H9J0PFK/
-
This shot was using a Tiffen variable ND- variable NDs use two polarizers. The reflections look similar to your still: Perhaps wear all black (and especially no white/bright colors).
-
I think this will be possible (and more!) since everything can be computed from the light field capture.
-
Haha, yep. Just a fun short for IG (Shoeps CMC641 and Sound Devices USB Pre 2). Thanks- we'll spend a little more time on the ADR next time
-
It's similar to VR- lots of hype but until certain elements are massively improved, these technologies won't be desired by the mainstream. There will be casualties along the way, though neither will completely disappear. At some point in the future, when VR gets 3D without glasses and computational cameras produce quality HDR 3D video (from high quality 3D depth data), they'll get married and everyone will be happy.
-
Correction- shot with Canon 70-200 F2.8L II at F2.8. First shot is at 70mm, the rest at various focal lengths.