Deadcode

4 hours ago, Slothorp said:

I dit another test with a simple domestic RGB led (hence the flickering) to show how the BMPCC4K reacts in the red channel, both in raw and prores. With other colors, when overexposed, there is a gentle roll-off, but in the red, it clips with a very harsh way, and a strange color shifting. There is also much more noise in the reds. 

 

Could you please upload few DNG from the video above?

10 minutes ago, jonpais said:

I'm guessing ProRes XQ will annihilate ProRes LT.  

We will see it in @mercer 's example.

23 minutes ago, jonpais said:

i have no idea what any of that means, but I'm eager to see what happens.  

It's simpe. He will not see any difference in regular cases. The statement of 700Mbps could be almost twice as good as 400Mbps is only true on paper. I just want him to prove it for himself.

5 hours ago, mercer said:

As I previously noted, I was being hyperbolic with murder and obliterate but that doesn’t negate the fact that ProRes XQ at +700 mbps should look better than 400 mbps of the GH5. And in some ways it does but I don’t know if it looks almost twice as good.

You are shooting with 5D right? Could you please show me what is the difference between ProRes LT (88Mbps for 1080 24p) and ProRes XQ? Just convert Raw to ProRes do a grading and crop the difference please.

"How hard can it be to at least make a small effort to try to match the shots from the two cameras. The one on the bottom is from the video, the one on the top a two-second color correction."

For me this comparison felt like they wanted to make BM look better in most ways. Skintones: bad grade for GH5, Stabilization: "Dont worry, we put it on a gimbal. I mean the BM only," Dynamic range: let's shoot in natural PP with the GH5, and shoot in RAW with BM. 

The dynamic range is impressive. This shot is waaaaay underexposed and still "just" ProRes. But it can be lifted easily

42 minutes ago, IronFilm said:

I had a quick watched (I'll confess, I skimmed a little!). 

And it certainly is not top tier wedding cinematography like I've seen elsewhere. 

However you've got very good solid workmanlike product there, I'd give it a thumbs up pass. 

But whoa...  then I saw your prices! Now, I don't know exactly what is normal rates for the UK (let alone your specific area), but wow, my gut feeling is you must be undercharging! You should lift your rates, as with your current standard of films you're certainly overdelivering for those low prices. (you're starting at only seven hundred???? Crazy)

 

 

Eastern europe wedding cinematography starts from 400 eur. Average price is between 700 - 1200 eur, 1500+ range is very rare.

Weddings costs around between 8000-12000 EUR all together including photographer, videographer, clothes, makeup, hairdresser, meals, and a hotel for 80-100 people.

Average wage is below 1K EUR.

In UK 1500 gbp is low price, 3k is usual. Insane.

I should move to UK ? And you too Jimmy

ohh, and the iPhone XS Max 512GB costs about 1800 eur, actually phones are fckn status symbols.

3 hours ago, IronFilm said:

Just because she is earning $120K/yr doesn't mean each wedding has that budget. Quite the opposite, when she is doing 80 weddings/year that means each wedding is only $1.5K/wedding???? I hope that is not gross income, but net income. As once you subtract expenses.... OUCH!

She is playing the high volume, low price game. 

No wonder she is just rushing through them with minimal equipment. (haven't watched the video, but if she is doing $1.5K weddings then she might be doing it alone or only with one extra shooter)

Here in eastern europe, 1500 usd for a wedding cinematographer means you have a very fancy wedding and your are hireing a pro group with 2 or 3 person ? the wage gap between here and the west is insane.

You should start another thread for each cameras individually.

Plus the 3 new lenses also deserves they own separate threads.

4 hours ago, Mark Romero 2 said:

Thank you for the input.

I can understand your findings. I guess the sadness for me is that the highlight rolloff of Cine4 is... how should we say it... pretty bad... So having the hghlights clip is pretty upsetting to me.

Your testing seems to confirm my (rather haphazard) comparisons between Cine4 and Slog 2. There is not a whole lot more DR in a SLOG2 shot that has been overexposed by two stops. It seems that the main benefit in this case is that SLOG2 gives a less harsh rolloff though.

All other things being equal, in regards to skin tones, would you say that Cine4 at 0 EV gives better skin tones than SLOG2 at around +2 after grading? Or are they about equal? You said with regular grading they both look about equal, so i am assuming that also refers to skin tones being on par with each other when exposed optimally. (I haven't been able to test on skin tones really.)

Yes, that is definitely a concern. Cine4 is much easier to use in the field (on an a6500) but the rolloff is nasty. Even using in-camera LUT (I don't know what sony calls it exactly... gamma assist???) when shooting SLOG2 the issue is that they recommend switching OUT of SLOG when doing your white balance, which is really difficult on a gimbal with the tiny buttons on an a6500.

 

Actually Cine 4 has a very nice highlight roll-off. But it uses 0-255 data level, and even Davinci Resolve cant recognize this in Auto mode, you have to switch it manually.

So technically you have worked with Cine 4 like you throwed away the darkest 0.5 stop and the top 1-1.5 stop DR. (where the roll-off is actually happening :D)

Exposing for Cine 4 properly means 50-60 IRE luminance level for the skin tones, and not 0EV AE

4 hours ago, seanzzxx said:

The reason I stopped shooting Slog was the fact that by overexposing by two stops the skin tones get really pasty and thin most of the times. I much rather have good skintones than some extra dynamic range, however I agree with you on the fact that the Cine curves have terrible highlight roll off.

Actually exposure does not really affect skin tones if it's above the noise level, which is around 40 IRE (ok, this statement is true mostly when recording in 10 bit)

Mark Romero 2

"Well... according to the sony help guide scroll to the bottom of here: http://helpguide.sony.net/ilc/1420/v1/en/contents/TP0000435736.html ), that seems to be what they are saying:

"Setting [ITU709(800%)] or [S-Log2] may cause an error in the white balance custom setup. In that case, set the exposure bright first and then perform custom setup."

Although it is hard for me to interpret that..."

Slog2 is designed to use with S-Gamut. S-Gamut has to be mapped to REC709 Gamut properly (which is not happening usually because the lack of knowledge, this is the reason why everyone is crying about how hard is to grade sony footage) and only 3200K and 5600K correction LUT's are existing right now. So if you are recording for example in 4300K, the mapped colors will be off.

I know everyone is using SLOG2 with S-gamut3.cine, but without gamut mapping the skin tones will not have enough thickness.

Now back to topic: every LOG footage has it's proper exposure levels for skin and 90% reflectance white, and every gamut has to be remapped to REC709 (or wider) colorspace in the grading process.

21 minutes ago, MattH said:

Log is forced under exposure.  Thats how it works.  Thats why it says iso800 when its really iso100.  Its a lower iso pushed to a higher average brightness, but it says iso800 to make you expose lower.  Thats how you get more highlight latitude.  How else could you get it?  It cant be magiced up from nowhere.

So if three stops of extra latitude is too much and you'd rather have just two with cleaner shadows then, sure, over expose a stop in log.

But if you're over exposing by two stops (definitely 3) you may as well just shoot in normal gamma because you are throwing away everything log gives you anyway.

With +2 stop overexposure, you still have around 12 EV dynamic range, and each stop represented about the same amount of luminance levels.

If you are shooting regular gamma, not each stop has the same amount of luma data which can cause limitation in colorgraing.

I tested extensively Cine4 (0EV) against SLOG2 +2EV which represents the same amount of dynamic range, but the low end of the Cine4 is heavily compressed compared to SLOG2. If you lift your shadows Cine4 falls in to pieces. Of course with regular grading both looks good and very close to each other.

My conclusion after tons of testing: expose for the highlights in wide DR scene, using SLOG2 for best looking image, and always expose for the skintones when using Cine4, and let the highlights clip, if necessary to get out of camera nice results.

The idea is great, but now, at the beginning of the affordable 10 bit recording era it is pointless.

From technical point of view, 8 bit LOG recording only lived for a few years, from 2014 (A7s) till nowdays. I think every camera manufacturer will step up to 10 bit in their next camera where throwing out half of the data will not means visible image degradation

17 minutes ago, tellure said:

Thanks for the links!  Some of the 4K hyperlapse in those videos do look better, but the ones with a lot of detail are pretty similar to the Vimeo one, and some are worse.  Of course some of this is no doubt down to YT and Vimeo's crappy bitrates, so maybe when we can get some original files we can see how much of this is the camera.

100% crop from 4K framegrab

 

Want to see good looking image from Hero 7?

Check this handheld shot

And yes, the banding still occurs because of the youtube compression. 

1 hour ago, tellure said:

Damn the compression artifacts in that Vimeo video are pretty terrible.  Sigh.. guess they must be recording their promo videos to an external recorder or something.

100% crop:

You guys are even real? ?

Vimeo is using 12-15 VBR codec to compress 1080p footage. And hyperlapses require much higher bitrate then usual.

Why dont you check normal footages with higher resolution...

35 minutes ago, sanveer said:

I guess after YouTube Reduced revenues for Youtubers/ Vloggers by 10-20 TIMES, they are just happy with whatever freebies they get for new product launches including free flight, stay and a few cameras thrown in. I don't blame them. It's like a millionaire suddenly making $50000-100000 a year from a million just a few months ago.

This is what the GoPro 7 should have ideally focussed on, apart from their stabilization (which wasn't too bad in 30p mode in their previous model, TBO?

1. Make Improvements in low light. With a combination of faster lens and probably larger sensor. 

2. Someone mentioned that the GoPro does 11 stops in Protune. That actually seems like a very high figure for such a tiny camera, especially considering it probably doesn't do 11 stops in stills mode. Maybe it does 11 stops in HDR. But certainly not in video. 

3. Bitrate is still a paltry 78Mb/s in 4K. While it does it in H.265/ HEVC, the video was still very jaggedy and pixelated. Maybe they should target 100mbps with a firmware update.

4. For last 2-3 generations they don't seem to have changed the sensor. It's like they regurgitating stuff now.

 

Because everyone else isn't really pushing them too much, in things apart from price, they've been complacent for the longest. Considering how amazing Sony's tiny sensors are, it strange they (and most other action and small cam makers) aren't doing more. They could have made this with 2 (or 3) cameras, to have zoom options or better low light.

There is a a lot to read here for you

Yes, the bitrate should be higher, but man, you are judgeing the image what you saw on a 8Mbps compressed youtube clip... It actually holds up pretty well, even after colorgrading.

10 minutes ago, sanveer said:

I remember seeing a comparison video between the iPhone X and the GoPro6, and the dynamic range difference (apart from stabilization), was pretty huge. I wish that has been improved. Many Sony Sensor have HDR and 30fps+ full sensor readout. It's only a matter of faster processing and better algorithms and small sensor cameras can be vastly improved.

I saw videos of the GP7 and they seem yo have fixed that muddy sound of the in-built mics too.  So I guess they're improving.  

GoPro's have around 11 stop dynamic range in Protune Native Flat mode. Actually the dynamic range is very good for the sensor size

12 minutes ago, sanveer said:

Is stabilization the only major improvement to the GoPro7 Black (apart from live streaming at 720p)? 

What else do you need?

2 hours ago, kaylee said:

is this like... on the table? is that possible??? ?

No it's not. The max resolution in full sensor mode is 2240x1260. 

4K raw is barely possible, the SD card interface max writing speed is 95MB/s just like the CF card interface in 5D3. Which means with 10 bit compressed raw around 3.5K will be the max possible for continuous shooting. And because of the 30MP sensor  it's gonna be over 2x crop.

But let's hope for the best.

Introducing Blackmagic RAW in 27 minutes long video

The built in EVF is dimming too in 4K just like the back LCD. It is almost impossible to see anything on the screen or on the EVF in bright sunlight when you are recording 4K. At 1080p both are perfectly bright.

With field monitor it will be fine. The image is identical in 4K to the FS5. The rolling shutter is much worse on the A6500, but the low light performance is much better.

1 hour ago, Yurolov said:

I'm unimpressed by this footage.

The minimal amount of jello while shooting handheld is good, it means the rolling shutter is minimal.

Also the low light section is nice.

I agree with @Anaconda_ "look" is decided in post with this camera, however that Contax Zeiss 50mm 1.4 is smoooth.

7 minutes ago, TheRenaissanceMan said:

" So I'd look at the website and check the guy out. 16 pages of op eds, think pieces, comparisons, and advertisements for his products later, I'd find Andrew Reid's last actual review: part 1 of the GH5 review, with the promise of a part 2 that never materialized. No finished work shot with it, just some ungraded clips available for download. Posted August of last year. And many more pages until his next latest review. I would say "eh, this guy doesn't do much in the way of reviews, just camera news and coverage of the market," and pursue other media outlets. 

I'm still waiting the Part 2 of the following reviews: 

1Dx Mark II, A7S II, A7R II, KineMax, Atomos Shogun, A7s, Digital Bolex, BMPCC

Long ago, cameras like 5D3 deserved a lot of tests and comparison on eosHD. Nowdays barely any footage.

Fortunetly the skill is still there (M50 sample, Pro Color V4, 3.5K ML RAW) but the passion is fading..

Just now, Andrew Reid said:

No gratitude whatsoever. You're as bad as the people on the Facebook thread. My family is probably reading that. Who wants to put up with this shit over an f-ing camera?

You are an influencer, just like vloggers.

Your articles reach a lot of people. The tone of your voice in your last article is not appropiate. What did you expect?

1 minute ago, Andrew Reid said:

How's this for some nice PR John?

https://www.eoshd.com/2018/09/why-i-wont-be-buying-the-upcoming-blackmagic-pocket-cinema-camera-4k/

This is the post what clearly explains why you dont get invites to any brand anymore. Childish. 

John's post are perfectly reasonable, Blackmagic dont want comparisons against other cameras, they want to provide enjoyable images and emotions with the camera. 

Have you see the parachute footage from Eos R? It's all about smiling people enjoying the moment in the sunset. You want to capture the feeling, and we can provide you a tool for that, this is the message. Not low light noise comparisons.

Use Colorizer. It was created by @Attila Bakos and this is exaclty what it was made for.

Nice image!