kye

I updated the FW to 1.6 as soon as it came out, it froze 2 days later on a photo shoot...shutter button stopped working until I restarted it; back screen and everything else kept working, just shutter button wouldn't do anything when I pressed it....not exactly confidence inspiring.
Yes the EVF delay for photos is in all mirrorless cameras but it seemed excessive to me so I opened a ticket with Canon and paid $150 to ship it to them and they returned it saying nothing was wrong. The R7 has the same delay, if a person blinks while taking their picture, in the EVF their eyes are still open, when reviewing the image their eyes are closed, runway shows are the worst for me, the front foot is on the ground in the EVF, in the captured image they have already picked it back up, little things like that are decreasing my keeper count.....definitely making me miss my OVF especially since I typically have a flash setup so I can't just spray and pray; and yes I have tried the higher FPS mode for the EVF.
I had a ballerina jumping and doing twirls for a photo shoot....that was the worst, I was ready to throw the camera out of the window, there was literally no way to do anything except just hope for a few keeper shots; I had multiple flashes setup so again...no spray and pray option due to flash recycle times.
The R7 is quirky too....sometimes the hotshoe doesn't trigger the flash, I know it is not the flash because it is the same flash I have had for years. Also, pressing the record button while it is recording doesn't always stop the recording, requiring multiple presses. After stopping the recording there's a delay of approximately 3-5s where you can't record again as if the buffer is clearing out or something like that; the S5 did not do that.
I really think Canon's cameras have become so complicated that they have all the sophistication of modern computers along with all of their intermittent issues. 
So yes, when I am on a photoshoot I just don't trust the R5 it is that simple; I don't want the photo side of it to lock up or have an issue because I was off shooting some b-roll video since I value reliability above all else.  I got used to a dedicated photo camera with the 5DIV's rock solid reliability (its video was unusable), so after those issues early on I will treat the R5 the same way. Ironically, I trust the R7 more as a hybrid than the R5.
But back to the R6II 🤣 I noticed another feature they brought to the R6II that the S5 had in 2020.....the big red recording box. This is a great feature to verify that it is recording (and another feature that could be brought to the other cameras via a FW update).  I couldn't tell from the CVP video if the box stays on or just shows when it starts recording.
 
I absolutely hate proxy media because of how much time it takes and because of my particular workflow (I only import the exact clips that I need for the project, so I would need to create proxy media each time mid project). Instead, what I do is scrub through and find the piece of the clip I want, then cut it up and add it to the timeline then use optimized media since it is much faster because it is creating a proxy of just that part of the clip. 
It still stutters during the initial review process; mainly on 60FPS media. I am still waiting to see if the NVIDIA 4000 series video cards can accelerate 10 bit 4:2:2 media, for some reason that information is impossible to find and I search about once a week or so. I suppose I could email NVIDIA and ask them. It is very disappointing that in 2022 PC users still have no video card options for 4:2:2.
What you will discover with the EF 70-200 adapted is the same thing that I discovered; AF is the least of your problems. The real problem is the weight of the lens making your camera feel like it is going to tear itself apart. With the EF 70-200 lens on the R5 plus the adapter the whole body was flexing, the adapter and the lens mount was making weird noises, and it really felt like the whole thing was going to fall apart.....vs the 5DIV where the two fit together like a glove.
The flexing and sounds were so bad I ended up getting the RF 70-200 since it is my most used portrait lens; actually, it is my most used photography lens of all time due to runway shows. So yes.....YT reviewers really gloss over a lot and you don't learn details like those until you actually try to use it. I think it is both the length and weight that make the combo so bad, I assume there are RF lenses just as heavy but not as long.

With the benefit of five years of hindsight, what might you have changed if you were to go back in time? 

Try me

The latest batteries for the OG BMPCC get WAY more than that - I got 49 minutes of RAW recording from one 1200mAh battery, and 33 minutes of Prores HQ from a different battery (but it wasn't fully charged).
I unpacked the brand new Wasabi 1200mAh batteries and put them on to charge, so the test was done taking a brand new battery straight out of the charger. I was recording in RAW the card filled up after 23m, with the battery at 54%, so I formatted the card and hit record again, and when it got to 5% battery life it was at 26m, which makes 49m for one the life of a single 1200mAh battery.  That's probably a best-case scenario, but it's pretty darn good if you ask me.  
Second brand new battery recording Prores HQ died around the 20% battery mark, and recorded a 33:22s file.  I'm not sure if it charged properly as the charger kept turning off but it only registered 81% in the camera when I put it in, so maybe not fully charged, but that's also not bad.
This is approaching the battery life of the modern gigapixel monsters that dominate the market now.  The BMPCC battery life complaints are something that used to be bad but are much more 'normal' now, not great, but not unheard of either.

That would be great.
"The pixels are just awful, but it's ok because there's a bazillion of them!" was never an attractive concept.
Of course, photographers have been shooting RAW for way longer than we have in video, and in video they give you more bitrate for higher resolutions, so it was never about the straight number of pixels anyway.  We may get there eventually, but it's not going to be any time soon!

I firmly believe that almost everything can be quantified, but this one falls very far beyond the point of what is practical.
The GH5 tests I saw with the person stepping into and out of frame had the AF recognise a face and change focus across a range of reaction times - sometimes it was fast and other times reluctant, and occasionally the person would just stand there being ignored like a camera nerd at a high-school dance.  There aren't any easy way to quantify this.  GH5 testers couldn't even get the test to replicate, providing a number of hilarious examples were the person was walking around saying how bad it was and it tracking them just fine, and other testers saying it was really good and it doing quite badly.
One YouTuber who got the C70 when it first came out admitted in a follow-up review that he had to stop using the C70 until it had a firmware update or two because when it first arrived it had trouble recognising faces of darker-skinned people.  IIRC he had to hire something to use on commercial shoots because the C70 wasn't ready yet.  Canon can't even test their AF properly and it's one of their key brand differentiators!
If you were to quantify AF performance, not only would you have to have a dozen or so metrics (speed to recognise a face, how out-of-focus the face can be before it recognises a face, maximum tracking speed, how much of the face has to be visible, how far around the side of the face it detects, how bad the lighting has to be for it to recognise a face, etc) but you'd really struggle to quantify the GH5-style lack of reliability except to have an enormous sample size.
Peter McKinnon made a promo video for his new product, and at the 6:12 mark, the camera goes from focusing on the object:

to focusing on his face:

the two frames above are 2 frames apart.  Why did the mighty Canon PDAF randomly choose that moment to change focus to his face from the largest object in the centre of the frame?  Heck knows, but there were even previous frames where more of his face was showing and it didn't choose those times to change focus....
Here's the video linked to that time - judge it for yourself...  
 
That's an AF problem right there.  People tend to think that Canon PDAF or Sony eye-detect PDAF are perfect but in reality they stuff up from time to time, and they tend to think that GH5 is completely useless when it actually gets things right quite a large percentage of the time.  
I've seen shots in vlogs where Canon PDAF cameras just randomly focus on the background when the persons face was visible the whole time.  They're rare, but I've seen at least two that made it to the final edit - we can only guess how many others ended up being cut. 
Any methodology that quantifies AF performance would be useless if it ignored the lack-of-reliability problem (because it would declare the GH5 AF to be great when it's obviously not) and it would be wrong if it gave Canon and Sony a perfect score when they obviously aren't quite there, despite being impressively close.
Sure, you can quantify some aspects of AF performance, but to be even remotely useful, you'd have to test so many variables and some of them would require such incredibly large sample sizes that it just wouldn't be practical.

Yeah, I thought that 12-bit RAW sensor readout (and the corresponding 10bit log) was pretty much the standard on almost all cameras that have a log profile.  I thought that some cameras had a 14-bit readout but it wasn't that common.
Of course, that is just another feather in the cap for 5D with Magic Lantern - 14-bit RAW straight to the card.  Still a standout spec and stand-out image, 14 years later.

The latest batteries for the OG BMPCC get WAY more than that - I got 49 minutes of RAW recording from one 1200mAh battery, and 33 minutes of Prores HQ from a different battery (but it wasn't fully charged).
I unpacked the brand new Wasabi 1200mAh batteries and put them on to charge, so the test was done taking a brand new battery straight out of the charger. I was recording in RAW the card filled up after 23m, with the battery at 54%, so I formatted the card and hit record again, and when it got to 5% battery life it was at 26m, which makes 49m for one the life of a single 1200mAh battery.  That's probably a best-case scenario, but it's pretty darn good if you ask me.  
Second brand new battery recording Prores HQ died around the 20% battery mark, and recorded a 33:22s file.  I'm not sure if it charged properly as the charger kept turning off but it only registered 81% in the camera when I put it in, so maybe not fully charged, but that's also not bad.
This is approaching the battery life of the modern gigapixel monsters that dominate the market now.  The BMPCC battery life complaints are something that used to be bad but are much more 'normal' now, not great, but not unheard of either.

I just got back into film photography. Last month I did an Evil Dead inspired horror photo shoot. This is the first time I've done a conceptual photo shoot with a model. I resurrected my old 35mm Canon Rebel and bought a couple of lenses. This is shot with the Canon 28mm 1.8 (all wide open) on 35mm with Kodak Portra 800. I have some more horror themed photo shoots set up, and some pinup style shoots as well. I also recently got a Yashica Rookie TLR camera and some 120 film.

Wow! I'm going to get some. That's about what I get from a Canon battery on my BMMCC and it's totally acceptable. With the earlier batteries, I had about enough time to set up the date and time, white balance, metadata, etc. and then changed batteries to do the actual shoot.

I finally finished and released this film yesterday. 5 years late is better than never, right? I was dissatisfied with it for a while, and there's a lot I would still change, but I've come to appreciate it. Now. So here it is, at last:
The film has been given a makeover with FilmConvert Nitrate. I'm pretty shocked at how good this a6300 short looks and how much I could recover from overexposed shots. There are only a couple of shots with sorta mushy compression.

Yeah, I thought that 12-bit RAW sensor readout (and the corresponding 10bit log) was pretty much the standard on almost all cameras that have a log profile.  I thought that some cameras had a 14-bit readout but it wasn't that common.
Of course, that is just another feather in the cap for 5D with Magic Lantern - 14-bit RAW straight to the card.  Still a standout spec and stand-out image, 14 years later.

Yes, Canon has nothing that can compare to that lens. I have a friend who shoots with Sony and she raves about that one lens. I definitely agree that a system needs to be as simple as possible and fit your workflow the best so that one lens could definitely decide your future.
I think Sony also still provides 4:2:0 10bit footage so that is a huge timesaver when it comes to editing the source footage if you do not have an M1 Mac. Editing the 60FPS files out of the R5, R6, and R7 is painful on a PC. My system is marginal with the 30FPS files, but I need proxies for 60FPS.
 
 
Not everything comes down to specs, I am very picky with ergonomics when it is my photography camera vs video where it will be held completely differently with a cage and handles. Ergonomics was the reason I picked up a demo Sony camera in the store and put it down and have never considered Sony again.  Also, when a photography camera is my A camera it has to just "feel" right...another thing that no specs can tell you. 
Last but not least, I bought the R6 when I was looking for a hybrid camera; at the time I wasn't looking for a replacement photography camera. Due to the overheating it would not be able to replace my video setup and due to the resolution and way it felt to me when shooting photos with it, it would not be able to replace my 5DIV for reliability, build quality and just the way it felt in the hand.
Even the R5 after the overheating fiasco (still not 100% convinced yet with the latest FW, just because they lifted the timer doesn't mean the reliability won't decrease as it gets hotter), the build quality, the EVF delay, and the freezing just barely puts it in the marginal category for me as a worthy replacement to my 5DIV. 

I have a backlog of 12 client wedding films to edit from scratch if you are that bored? 😬

I’m in the summerhouse and bored.
If you have a shot that you would like a look on then grab a still and post it here. Exr or tiff, and in LOG is nice. Please tell me what camera it’s shot on so I can make an accurate conversion.
 
Wide shots or close ups with skin tones (macro stuff of flowers and objects are less interesting 🫠)
 
And maybe someone wants to challenge me so we can make a nice little competition.
 
Hit me!

Global shutter has been around on a few cameras already like the Blackmagic Production Camera 4K, which people complain about because it can't do ISO 800, and Sony F55 which also took a low light hit. There is a reason why the Alexa isn't global shutter. Because a fast rolling shutter is fine.
Atomos have to compete with Sony if they want to enter the sensor game.
I say good luck to them but get your company ethics sorted first, bye.

I had to earn my pay somehow that day!

Damn!  That's the wrong way to make a day more exciting!

I think the colourists refer to the initial adjustments as "balancing" the image (although it's not a universally used term) and they'll then apply a creative colour grade over the whole scene / film.  
I don't know how technical you like to get with grading, and it's a very very deep and technical topic, but I'll try and give you some examples.  It also depends on what software you're grading in, because Resolve uses Nodes and that means that either some things get done differently to a layer-based package, or it means you can do things that can't get done in a layers-based package.
I had a video in mind that showed it clearly but couldn't find it, but this one describes a similar approach - I've linked to the relevant timestamp:
This is part of the node graph from that grade:

Node 4 takes the balanced image from node 2 and does a serious grade on the image that screws up the skin-tones, but node 6 takes the same balanced image from node 2 and has a key on it to only select the skintones and then he grades in node 6 to push the skintones so they look ok sitting on top of the grade from node 4.  ie, node 6 is at 100% opacity over the top of node 4.
An alternate approach, and the one I was describing, would be to have node 4 do the same grade, and keep node 6 as being a key of the skintones, but instead of node 6 having a grade on it and sit at 100% opacity, I'd keep it having no grade at all and starting from 0% opacity just bring the opacity up until the skintones stop looking wrong.  This technique works well if the grade in node 4 is absolutely full-on, because if that is the case and node 6 was at 100% opacity then you'd have to do a lot of grading on node 6 to get the skintones to sit nicely over the top.
A completely different approach is to grade the whole image with a serious grade, then in a node after that you'd try and get a key of the skintones and then adjust them how you want them.  This is a way to do it when you don't have parallel nodes, but it makes pulling the key much more difficult.  One way to make this easier, if the software allows it, is to pull a key of the skintones from the balanced node before the image has been really messed with but use that key for the skintones node after the serious grade.  This is another awesome feature of Resolve - the blue inputs and outputs on the nodes are for the key, so you can pull a key on any node and then connect that Key Out to the Key In on another node.  
(Side note: you can also connect between the key ins/outs and the image ins/outs which allows you to process the key using the normal controls of a node.  I haven't seen anyone else do this but it works and I've used it in real projects to refine keys etc).
There are other approaches as well, and it's worth setting up a test project and trying to think of all the different approaches and different tools you could use, and just trying each of them to see which ones you like and which ones you find easier to use etc.
Just for fun, here's another video that uses parallel nodes and all kinds of fun stuff to create a strong look:
 

OG BMPCC - costs less to buy outright than renting a C70.....  in 2022!
 

https://www.dpreview.com/news/6357713219/atomos-ceo-reveals-more-details-of-the-company-s-full-frame-8k-global-shutter-image-sensor
 

...then keep on cutting! Until you feel like you've gone too far. Then it is ready. 

LESS THAN thirteen batteries. 
(was constantly running back to the chargers to cycle through the handful of batteries we had, and we were using a big heavy car battery to use to charge up the dead batteries)

Good idea, I'm near the beach so that's definitely possible.
I've found that if you zoom in to 200-400% in post then it's easy to see through the YT compression to what is the original compression, especially if I get lots of movement in the shot.
Also, most content is delivered via some form of streaming (but not all) so if it's not visible on 4K YT then that's a useful conclusion in itself.

I think the colourists refer to the initial adjustments as "balancing" the image (although it's not a universally used term) and they'll then apply a creative colour grade over the whole scene / film.  
I don't know how technical you like to get with grading, and it's a very very deep and technical topic, but I'll try and give you some examples.  It also depends on what software you're grading in, because Resolve uses Nodes and that means that either some things get done differently to a layer-based package, or it means you can do things that can't get done in a layers-based package.
I had a video in mind that showed it clearly but couldn't find it, but this one describes a similar approach - I've linked to the relevant timestamp:
This is part of the node graph from that grade:

Node 4 takes the balanced image from node 2 and does a serious grade on the image that screws up the skin-tones, but node 6 takes the same balanced image from node 2 and has a key on it to only select the skintones and then he grades in node 6 to push the skintones so they look ok sitting on top of the grade from node 4.  ie, node 6 is at 100% opacity over the top of node 4.
An alternate approach, and the one I was describing, would be to have node 4 do the same grade, and keep node 6 as being a key of the skintones, but instead of node 6 having a grade on it and sit at 100% opacity, I'd keep it having no grade at all and starting from 0% opacity just bring the opacity up until the skintones stop looking wrong.  This technique works well if the grade in node 4 is absolutely full-on, because if that is the case and node 6 was at 100% opacity then you'd have to do a lot of grading on node 6 to get the skintones to sit nicely over the top.
A completely different approach is to grade the whole image with a serious grade, then in a node after that you'd try and get a key of the skintones and then adjust them how you want them.  This is a way to do it when you don't have parallel nodes, but it makes pulling the key much more difficult.  One way to make this easier, if the software allows it, is to pull a key of the skintones from the balanced node before the image has been really messed with but use that key for the skintones node after the serious grade.  This is another awesome feature of Resolve - the blue inputs and outputs on the nodes are for the key, so you can pull a key on any node and then connect that Key Out to the Key In on another node.  
(Side note: you can also connect between the key ins/outs and the image ins/outs which allows you to process the key using the normal controls of a node.  I haven't seen anyone else do this but it works and I've used it in real projects to refine keys etc).
There are other approaches as well, and it's worth setting up a test project and trying to think of all the different approaches and different tools you could use, and just trying each of them to see which ones you like and which ones you find easier to use etc.
Just for fun, here's another video that uses parallel nodes and all kinds of fun stuff to create a strong look: