Video Hummus

Hard to tell now but i would be surprised if we dont see a major push by apple to “accelerate all the things”. It’s funny to see the traditional computer hardware and PC YouTube channels, mainly Linus Tech Tips, going all out bashing and doubting and downplaying the Apple silicon performance claims. Why? They don’t even have hardware in their hands to tell. This is the first laptop that will chew through almost any H.265 video file like butter. It is huge deal for people that use resolve or FCPX or just edit and cut video in general.

Based on iPad results. H.265 isn’t a problem.

You forgot to tell me to smash the subscribe button like a psychopath. I gotta say, I’m a bit let down and will not be buying the Ursa 12K.

This process has been happening for 8+ years. iOS and OS X have essentially merged. OS X is no more. It is now MacOS and it shares architecture more in common with iOS. Apple is the master at architecture changes. They’ve done 3 in the past 15 years. PowerPC -> Intel -> ARM. They have had universal binaries for years now starting back to the PowerPC to Intel switch. Universal binaries ship native code for both architecture in one “app bundle”. The result is native performance. For app developers that don’t recompile their code to universal will have to use Rosetta to translate the code on the fly. This sounds bad, but it ties into their “clang” JIT compiler that have been developing for years. It’s world class and one of the fastest JIT compilers in development. Apples move to Apple Silicon is very much been in play for many years and finally a return to Apple developing both hardware and software from the ground up. Which can be good and bad depending on your perspective (increased performance versus lock-in). I am quite impressed with Apple’s stead fast march from acquiring an ARM cpu chipset developer to now leading the world in designing the fastest ARM cpus out there. Fast enough to blow past even Intels best desktop CPUs at a quarter of the power and thermal envelop. Pretty amazing win for Apple and RISC-based ARM CPUs.

Geek bench scores are out. Keep in mind these are synthetic benchmarks but offer a good estimate of performance. This doesn’t take into account the M1 neural engine or Encoder/decoder acceleration chips. MacBook Pro with M1 chip and 16GB RAM posted a single core score of 1714 and a multi-core score of 6802. To put that into some perspective: a 2019 high-end 16-inch MacBook Pro earned a single-core score of 1096 and a multi-core score of 6870. 16 GB is limiting. I wish they had a 32 GB option. However, combined with the extremely fast LDDR memory, the unified SoC and I/O and blazing fast SSD storage, and how MacOS manages and maps memory I suspect 16GB in final cut pro is enough for most projects. Looking good, especially performance per watt!

All of that is up to AMD and NVidia. So far they haven’t put in proper H.265 hardware acceleration for 10, 12, or 14 bit 4:2:2 at all. It is the reason we see H.264/5 4:2:0 flavors in S1H and Fuji cameras. Why Canon didn’t offer a 10-bit 4:2:0 option in the R5 is crazy. Sony did it with the A7SIII. It was just another “roadblock” put into that product along with the insane bitrates and massive file sizes. Oh well.

This is exciting for creators. I’ve been waiting for awhile for a more affordable Mac that can edit more smoothly the 4:2:2 H.264 and especially HEVC. If the 13” MacBook Pro can edit R5 footage wether it be 10-bit 4:2:2 HEVC, 12-bit RAW, or 12-bit Canon RAW Light...I’m going to pull the trigger. 95% of the heavy lifting I do on my Mac is video editing, transcoding, encoding for final delivery. Since this all seems to be natively hardware accelerated with M1, at least in Resolve and FCPX, then I can’t see why I wouldn’t buy it. $1400 for a laptop that can edit 8K RAW on the go. Hopefully we see some exciting things to come for FCPX and Resolve with the neural engine. Like 5 sec warp stabilizer runs on 30 second clips or near instant tracking results.

Why I decided to pass up the A7SIII over the R5; or any forthcoming Canon mirrorless (because the R5 isn’t perfect either). Terrible LCD Only 12MP Compromised S35 crop modes (because of the sensor) Strong video, weak photos for anything but social media (is it even a hybrid, really?) CFExpress Type A (Type B seems to be more widely accepted at the current moment) The aforementioned shitty previews 9M dot EVF is wasted almost completely. A higher res LCD would have made infinitely more sense. E-mount feels more locked in Things I appreciate about the R5 3.2” 2.1M DOT LCD that is bright 5.6M Dot EVF that is more than good Oversampled Crop mode 45MP photos Better IBIS for photos (video IBIS debatable) Can take RF, adapt EF, FD EF lens performance is excellent

Perhaps a FF camera then. Why can’t Sony do this in their A9? 20fps too slow? They can do it, the camera companies are just purists when it comes to taking photos it seems

This looks great on vimeo, imho.

Give it 3 years and people will be saying they prefer the classic dynamic C70 look over the C120.

Isn’t the smartphone just doing auto-magic photo bracketing? Which could be done manually with the GFX 100, albeit with a lot more work. Seems kinda unfair to the GFX 100 merely on a technical standpoint (not on a user experience standpoint). Which means the smartphone took more than one photo, perhaps up to 5, compared to the GFX 100’s 1 RAW. It’s surprising to me camera companies have not embraced this “in camera” processing more, even for special “auto” modes where the user just wants the camera to do everything. Regardless, this is a good example of why camera sales have plummeted in the consumer market. It all comes down to the user experience and the smartphone, in this case, clearly provided the better user experience. If you were to repeat the test and bracket 3 RAW photos from the GFX 100 the results would be superior.

Bummer. AV1 is more of a streaming codec. Amazing quality at low bitrates but the processing times become horrendous as you go up...would be a terrible capture format except for live productions or immediate turn around news or something. I don’t get it. BM doesn’t seem to want to offer BRAW unless through their external recorders. I don’t mind using Apple if I can get hardware acceleration.

I honestly would go Canon with EF glass. Still performs amazing on new EOS R5/R6/R with native adapter. Can use the ND adapter as a bonus. Can adapt the EF to BMPCC4K can use natively on a BMPCC6K, ZCAM, or Kinefinity MAVO. Can use the EF glass + speedbooster on C70 or future CXX camera’s... That or go Sony as they are the only other company with competent AF, tons of lenses, and mirrorless and cinema cameras...you just are more locked into their E-mount.

The AMD 6000 series GPUs may have it. Nvidia seems to be using it to segment and sell it’s incredibly expensive desktop workstation GPU’s. Honestly, if the new ARM MacBooks are anything like iPad Pro performance or better...than it would be cheaper to get a MacBook than an expensive Nvidia Workstation card or the $700 and $999 upcoming AMD cards...if you are in the Mac ecosystem that is. Apple has a huge opportunity with the new ARM based Macs to push heavily into the video editing and “creative” market if even their lower-tier MacBooks can edit 8K and 4K H265 R5 files without breaking a sweat and with excellent battery life even on the go.

Amazing Canon of all companies did it right (within constraints) of offering internal Canon RAW Light in the R5 (with the next firmware update). The landscape, or should I say more accurately, my particular views of what companies are doing the most interesting things...and its Canon and Blackmagic. Hoping BM will embrace RF mount or at least embrace a easily adapted BM mount so you can use whatever lens you want on it. Plus, they are doing internal RAW as well. External recording just sucks. No thank you. Sony is a juggernaut of market control so they aren’t going anywhere either. I just find I don’t much enjoy using their cameras even if the specs are outrageous.

ARRI has this for their cameras but its more of a focus puller cheat device (giving the focus puller a realtime window of the focus plane) than a autofocus method. I’m not sure if it uses LIDAR or ultrasonic.

DJI just released a LiDAR module for their new gimbals. It is more limited than @BTM_Pix device but so for from the video I’ve seen it works well up to about 20 feet.

I actually like the IPhone 12 Mini the most. My favorite iPhone was the boxy 4S. I’m tired of paying $1200 for a huge phone. It’s a tool for me. It should slip in my pocket and disappear until I need to navigate, call, messaged, or lookup public transit times. So glad they went back to the boxy shape and smaller size with the mini. Price is still pretty high.

Confusing, overpriced product. It’s clear L-Mount is the focus for traditional image capture at Panasonic now.

Damn, I think this is another G100 blunder. I don’t know how well that camera has sold but I can’t imagine it’s been very competitive against M50 and other cameras for beginners. Its a GH5S in a different body and a minimal spec boost, but it still falls short compared to Z-Cam and Kinefinity Terra 4K with the so far rumored specs. Is this Panasonic trying to find new opportunities in a low cost rehash of a GH5S that didn’t sell well to begin with? It just seems unusually off-the-mark for Panasonic. We can only speculated how badly Sony has them by the balls when it comes to autofocus AND sensor design and availability.

Really hope Canon continue down this path of small, somewhat affordable EOS Cinema cameras. The R5 is a compromised camera but it makes more sense to me now with the release of the C70. A Canon photo-hybrid R5 with a C70 as dedicated video camera starts to make a lot of sense in a real-world usage for a solo or small crew. A solo operated can lean heavily onto having internal NDs, onboard multi-channel audio, and really great AF, letting us focus on other things like story and getting the shots. The R5 can fill the role as a excellent photo camera with the same excellent AF but still be a useful b-camera. Plus the lenses can be shared. You can still shoot HQ 4K in ASPC mode on the R5 and it looks excellent without being timer-limited. And even though the 4KLQ is not spectacular for YouTube it’s good enough. If you run EF glass you get the useful Nd adapter that is like having built-in ND somewhat.

Despite that topic and all your work you put into investigated it, my eyes do see a difference. It’s shot dependent and wider angle shots with more movement in the scene see the biggest gain. However, almost everything on YouTube is destroyed by heavy compressing codecs to save on bandwidth. There is a threshold for “good enough” that is reach by even budget cameras. As always, content is what you should be worrying about on YouTube not image quality. There are channels out there with multiple 100K subscribers and views that are filmed on GoPro.

DaVinci Resolve has a good upscaler. I doubt it’s available in the free version. There are dedicated programs you can pay money for like Topaz that are pretty amazing. But yes, you are just getting the extra bitrate from YouTube which is what really matters. Sorry, I admit I didn’t read your post closely enough. Maybe upgrading to FF EF would be better in the long run as it can be adapted to many things and works great with the benefit of adapters. that add extra useful features, to Canon RF cameras. Canon is usually always softer looking, sometimes to a fault. I would say the EOR R cameras and onward are soft but not to a fault.

There is some footage in this one.