Blackmagic Forum

Note Suwanchote wrote:

Wayne Steven wrote:

Note Suwanchote wrote:Agreed completely. There's a lot of CA on that lens. I'm curious to know what his aperture was. Also his settings too.

F8 didn't somebody say? Yeah, the aberrations were surprising.

I've shot with the 18-35mm on the 12K, not at f8 though, and it does not look like that. There's something off there.

He got his Sigma 18-35mm modded to cinema. So possibly something could have happened there that the 12K is showing that a lower resolution camera wouldn't have show. Can't say for certain however. But I just compared the Sigma 18-35mm Art against the Angēnieux Optimo DP 16-42mm & DZOFilm Pictor 20-55mm and found that it didn't hold up against the Angēnieux as well as the DZOFilm did. We're releasing that video next week.

And, your footage with the 18-35mm Art looked fantastic. So clearly the lens does produce great images. And, I would say mine still does too. So as much as I'll knock the lens compared to higher priced glass, it's still a fine lens. Your footage proves that.

Wayne Steven wrote:I'm thinking it's a site and YouTube channel to avoid. Plus now hotrod video might have so much interest it blows out the delivery times mentioned?

I hate YouTube. I mean it might be fun to do something but to act like this fur a serious living. What next, chuck a Red 8k weapon in and see if it floats on water?

Well if it floats doesn't that make it a witch? Wait...

Blackmagic is being used to make the cameras float... it is witchcraft!

jallen0 wrote:Just watched a video of the 12K in the hands of a customer and they talked about an issue when zooming in at 400%.

youtu.be/kSCD5rk6QCs

Interested in seeing how this works out.

Oh, and the amount of anxiety I had with the unboxing was a bit off the charts.

this guy is an idiot and only YT kicks out

.

Ahhg, can't we get a descent micro 8k. I'm sick of reading 12k stuff.

I take this video less seriously than some. I see it as an entertaining, amusing send-up of high tech camera unboxing videos. The bit at the end is perfect - a little "drama" to get the punters to tune in next week.

Got a kick out of Armando Ferreira's response to the "drama" (pinned): "I’m coming over" [too bad this forum won’t print his runner icon ]

Apparently it worked. Their subscribers are up, in 2 1/2 days they’re at 19,000 views (surely over 20,000 in the two or three hours), and they've got people talking about them, including, so far, 54 posts here.

Lol. Some people think inceased GDP is great, good for the economy. People on their death bed is great for GDP.

John Brawley wrote:500-5000% blowups

https://www.flickr.com/gp/johnbrawley/o2D689

JB

Thanks for sharing these pictures. They speak for them self and put all “worries” to the side IMO.

I don’t believe Epic Light Media is bashing the camera that hard. They clearly state that it seem like a great camera, but “they have worries” regarding some issues. Issues that they clearly haven’t spent much time investing into before releasing the video. They probably rushed the video to get some attention, spiced with some attempt of humor, not that surprising today. Maybe they have already realized that it got something to do with their lenses and the fact they used 18:1 compression? Lets see if they are man enough to admit their fault and correct it. That’s how you know if you want to continue follow their Youtube channel or not. Honesty or tech babble bs. Well, some bs can be entertaining but that need to be some epic heavy bs.

Oyvind Fiksdal wrote:

John Brawley wrote:500-5000% blowups

https://www.flickr.com/gp/johnbrawley/o2D689

JB

Thanks for sharing these pictures. They speak for them self and put all “worries” to the side IMO.

I don’t believe Epic Light Media is bashing the camera that hard. They clearly state that it seem like a great camera, but “they have worries” regarding some issues. Issues that they clearly haven’t spent much time investing into before releasing the video. They probably rushed the video to get some attention, spiced with some attempt of humor, not that surprising today. Maybe they have already realized that it got something to do with their lenses and the fact they used 18:1 compression? Lets see if they are man enough to admit their fault and correct it. That’s how you know if you want to continue follow their Youtube channel or not. Honesty or tech babble bs. Well, some bs can be entertaining but that need to be some epic heavy bs.

Ive seen this being referenced on a few sites though (that there’s something wrong with the blow ups and noise)

It’s bad information in my view.

The shot I posted the blow ups from by the way was something like 60:1 (Q5) compression.

JN

John Brawley wrote:
The shot I posted the blow ups from by the way was something like 60:1 (Q5) compression.

JN

Ok thanks for the info. Q5 is of course harder to analyse since it will shift compression based on the content in the recorded image. It can go much lower than 18:1 or higher, intellectually so. Based on that, I can’t say that it’s a fair comparison between Q5 and 18:1. You know this, but the guys who was “worried” maybe didn’t. I also believe, by pure speculation, that 18:1 is a fairly beta stage compression compared to let’s say 12:1. Anyhow. It’s the sum of all parts.

My appologies, the gy-HD10 had a ccd sensor not CMOS.

I've had an e-mail correspondence going with the owner of Epic Light Media the last few days. I sent him a focus chart and he tested the lenses. The Sigma Art 18-35 is sharper than their Cinema Modified 18-35 Sigma done by Cinematics from the focus charts test. So the lens they shot the 12K footage with is definitely flawed.

I conveyed all input from the Blackmagic Camera Team regarding shimming to the owner. So, Tim Schumann's information about leaving the 0.1 shim in with the other two shims has been provided. I sent a screenshot of the post Tim Schumann wrote to the owner.

I was informed they are shooting more test footage today. They will be using different lenses for those tests. So, we'll have to see what their future tests yield.

timbutt2 wrote:I've had an e-mail correspondence going with the owner of Epic Light Media the last few days. I sent him a focus chart and he tested the lenses. The Sigma Art 18-35 is sharper than their Cinema Modified 18-35 Sigma done by Cinematics from the focus charts test. So the lens they shot the 12K footage with is definitely flawed.

But, than again. Soft lenses are not particularly flawed looking back at history. It’s just an artistic choice. Right? Nikon never competed agains canon on who got the sharpest lens. Nikon focused on another aspect.. smooth skin. Ironically we now have the chance to change that harsh sharpens, digitally, with modern computers. So maybe you are right.
Cinema lenses, particularly, are flawed.. and expensive. It’s interesting because we start to agree, in masses, that film is flawed compare to digital sensors. And we see lenses popping, cheap and sharp.

I agree with another take on the 12k sensor. It’s not about sharpness or 400% cropping. It’s the quality of the pixels. Sharp or not. why are we suddenly possessed about cropping anyways ?

Oyvind Fiksdal wrote:

timbutt2 wrote:I've had an e-mail correspondence going with the owner of Epic Light Media the last few days. I sent him a focus chart and he tested the lenses. The Sigma Art 18-35 is sharper than their Cinema Modified 18-35 Sigma done by Cinematics from the focus charts test. So the lens they shot the 12K footage with is definitely flawed.

But, than again. Soft lenses are not particularly flawed looking back at history. It’s just an artistic choice. Right? Nikon never competed agains canon on who got the sharpest lens. Nikon focused on another aspect.. smooth skin. Ironically we now have the chance to change that harsh sharpens, digitally, with modern computers. So maybe you are right.
Cinema lenses, particularly, are flawed.. and expensive. It’s interesting because we start to agree, in masses, that film is flawed compare to digital sensors. And we see lenses popping, cheap and sharp.

I agree with another take on the 12k sensor. It’s not about sharpness or 400% cropping. It’s the quality of the pixels. Sharp or not. why are we suddenly possessed about cropping anyways ?

I personally didn’t order it for extreme cropping, I imagine a lot of reviewers will focus on exploiting the flaws at extreme breaking points even with higher resolving lenses like Zeiss Otus for clicks.


Sent from my iPhone using Tapatalk

timbutt2 wrote:I've had an e-mail correspondence going with the owner of Epic Light Media the last few days. I sent him a focus chart and he tested the lenses. The Sigma Art 18-35 is sharper than their Cinema Modified 18-35 Sigma done by Cinematics from the focus charts test. So the lens they shot the 12K footage with is definitely flawed.

I conveyed all input from the Blackmagic Camera Team regarding shimming to the owner. So, Tim Schumann's information about leaving the 0.1 shim in with the other two shims has been provided. I sent a screenshot of the post Tim Schumann wrote to the owner.

I was informed they are shooting more test footage today. They will be using different lenses for those tests. So, we'll have to see what their future tests yield.

SuperPro Bro? You had a resolution chart and lens handling options and they didn't.

Wayne Steven wrote:

timbutt2 wrote:I've had an e-mail correspondence going with the owner of Epic Light Media the last few days. I sent him a focus chart and he tested the lenses. The Sigma Art 18-35 is sharper than their Cinema Modified 18-35 Sigma done by Cinematics from the focus charts test. So the lens they shot the 12K footage with is definitely flawed.

I conveyed all input from the Blackmagic Camera Team regarding shimming to the owner. So, Tim Schumann's information about leaving the 0.1 shim in with the other two shims has been provided. I sent a screenshot of the post Tim Schumann wrote to the owner.

I was informed they are shooting more test footage today. They will be using different lenses for those tests. So, we'll have to see what their future tests yield.

SuperPro Bro? You had a resolution chart and lens handling options and they didn't.

Only $200. Worth the investment: https://www.ducloslenses.com/collections/test-equipment

This is grab still from 12K timeline (original) and zoom in photoshop 100%. Look like vintage Lens with miss focus or something? I think Sigma 18-35 f1.8 using f 4 -5.6 will get better resolve.

There is no excuse for not having a test chart: https://stephen-westin.com/misc/res-chart.html

timbutt2 wrote:I've had an e-mail correspondence going with the owner of Epic Light Media the last few days. I sent him a focus chart and he tested the lenses. The Sigma Art 18-35 is sharper than their Cinema Modified 18-35 Sigma done by Cinematics from the focus charts test. So the lens they shot the 12K footage with is definitely flawed.
.

That is some useful information if further testing holds up. I have both the 18-35 and 50-100 Art that I was thinking about sending out for the conversion as Canon is no longer in the crop stills camera business. One thing the conversion does do is make the lens parfocal I wonder if that is a factor in the IQ loss? If one of the costs is to lose the sharpness they have in the non-converted form then it is a non-starter for me.

I am surprised to hear you say the Pictor holds up better as the shots I have seen of the Pictor all felt pretty soft but I haven't seen it directly compared to the 18-35. I look forward to seeing your video. Please post the link when it is ready.

Some new test shots



and

Misha, thanks very much for the 4k samples.

Thanks for sharing these new tests!

This is the Guy to thank

https://www.guymolin.com/

It's all in Dutch, so you might want to use google translate or something similar.

Uli Plank wrote:There is no excuse for not having a test chart: https://stephen-westin.com/misc/res-chart.html

Another rather unique resolution testing lens chart from a camera operator: P.A.T. Accessories.

https://www.fdtimes.com/2020/09/28/p-a-t-lens-charts/

That's a great chart, better than ISO 12233, but I wrote 'no excuse' – not even money

The advantage of that more primitive chart is automatic analysis to avoid subjective judgement. But then, aren't the imperfections of some lenses also their beauty in the eye of the beholder?

From Ricks link:


There seems to be a mirror in the middle of the chart. That's a clever way to make sure the camera is centered and aligned!

My Issue with test charts of all kinds is that the focus distance is proportional to the focal length so wide lenses have to be tested at closer distances than telephoto lenses to fill the frame. All lenses behave differently in their across the frame performance with distance. Some lenses are optimised for infinity and others for closer distances and add to this field curvature which may or may not benefit your subject it adds up to unreliable results IME. Oh and a mirror won't help as it if there is any decentring ( and there will be as no lens is perfect) then this will result in the focal plane being out of parallel with the target and thus change edge resolution. OK these test charts probably do have a use if you are doing flat artwork copy jobs but in the real world forget it.....

I dare to contradict here, as much as I also like to judge lenses by their 'feeling' too.

A mirror helps normally, you just keep it flat on the chart and center the camera in the mirror image. A lens must be extremely de-centered if that doesn't fit. Trust me, I have checked such situations by measuring the distance to the corners for reproduction purposes. A chart will then actually show you if there is de-centering by showing differences of focus in the corners.

Sure, many lenses, wides in particular, are not flat-field. But you can still check them by adjusting the focus and you'll even see how strong the curvature is. BTW, there are lenses which can adjust the field curvature.

Finally, I print my test charts in different sizes, a large one for the wides and a smaller one for longer lenses.
Any lens with acceptable quality should be good enough when you can keep some distance.

Of course, quality at infinity is another subject.

And then, a camera with 12K photocells will challenge any lens. Heck, a Red One with 4.5K was clearly showing the difference between a Zeiss compact prime and a master prime – and they were considered soft by many because of their strong OLPF.

Measuring to the corners only establishes the camera position in relation to the test chart and not the orientation. All lenses have in addition to decentring optical centres slightly off the center of the sensor so neither a mirror nor framing will ensure the focal plane is parallel with the chart. OK it's not much but enough to throw off the focus plane on the target to confuse decentring with alignment. Testing at infinity eliminates these variables.

jallen0 wrote:Just watched a video of the 12K in the hands of a customer and they talked about an issue when zooming in at 400%.

Interested in seeing how this works out.

Oh, and the amount of anxiety I had with the unboxing was a bit off the charts.

Pointless video.

Well Captivate Media are stoked! We've just placed our order!!
We'd love to share our thoughts of the new camera as soon as we get our hands on it.

Can you do a comparison of dynamic range compared to the Xiaomi mi 10 ultra in highest bit photo mode.

Has extra stops.

New video posted and I cannot tell if he is honestly trying to "help" or is looking to expose flaws to "hurt". In any event it would be interesting to see if others who have this camera can recreate his findings.

jallen0 wrote:New video posted and I cannot tell if he is honestly trying to "help" or is looking to expose flaws to "hurt". In any event it would be interesting to see if others who have this camera can recreate his findings.

Don't know is its helping or hindering but that is bad Moire. Without seeing the original footage though hard to tell whats at fault there.

I've shot on Reds and had to get the interviewee to change shirts in the past. So I'm not ruling out user error there but until we see more we won't know more about moire.

If I’m not mistaken, he shot everything in 12k, right?
I wonder if moire is lessened when you shoot 8k, 6k, 4k full sensor.

Also, punching in that much is something I’d never do, so tests of resolution don’t really excite or deter me from choosing this camera. Still, I’d love to see a comparison with a UMP G2 on Dynamic Range and Low Light.

Haven’t seen the video but on moire...

It can happen on any camera. Even with an OLPF.

With no OLPF it’s far easier. But the 12k resolution raises the threshold so in theory there’s less likelihood.

JB

I'd assumed at 12k the sensor would out resolve most lenses (esp cinema lenses that were never designed for this kind of resolution) and moiré would be an non issue. I still get it on stills with my A7rIV (9.5k) in the center of most lenses at mid apertures under F11 but it's not as obvious as on my previous 42 and 36mp Sony's. If you are paying 10k you can probably afford an aftermarket OLPF ( Rawlite?). If you are going to be punching in you had better be sure you are using 100% focus mag and if anything's moving forget it. I really can't see a real world situation where punching in in post would be a better option that getting it right in camera. It is however a great way of exaggerating focus errors, lens imperfections, noise and moiré for test videos.....

You can provoke moiré on any camera with a single sensor, even those with an OLPF. Nobody is filtering them down to a point where color moiré is completely suppressed, since that would mean an 8K camera is turned into a 2K version to get below the Shannon/Nyquist-limit of the color-filter pattern.

I wonder, though, which lens was used at which aperture for that test. And then, I'd rather like to see the same shot in 4K out of the camera.

Yes Uli, at 4k, are you below the limit in this odd sensor pattern? I think the way Braw blur might have done a good job of that, but along the lines of what you said, nit great detail.

I wish they would at least give a technical outline of the technology in Braw. Like the true compression engine currently, and in the past, and filtering. Which is hardly anything to ask, moderators?

Been a bit out of it. Just watched the video. Guess the RawLite OLPF is still needed. Not an issue in my book.

However, I would love if Blackmagic could make their own OLPF to sell with their cameras. Definitely something they should look into doing for the next generation of cameras.

Still, doesn't change how impressed I am with this new sensor. It's all about the colors and dynamic range for me and not the resolution. Although I'm sure 12K would be really nice for VFX. Can't wait to see the next Michael Bay explosion in 12K. Haha.

It's beyond understanding that blackmagic still doesn't want to put an OLPF in their cameras. Especially the ones designed for the pros. I've said it many many times during my reviews and on forums. Extreme sharpness is definitely not what a top image is about, especially at the cost of severe moire and aliasing (they should go for improved dynamic range and noise performance for instance). And you never know when it will show. I made comparisons with the bmpcc4K and the UMP 4.6K. Sometimes the 4K had less, there's no rule. Even a very high pixel count can't prevent from moiré.

It can also happen with things you'd consider harmless. Like a simple cotton t-shirt of plain color. If you have a packshot to do, with writings or fabrics, you're so nervous something's gonna show up.

If you rent a Sony or a Canon camera, you know you won't have all these reliability issues. It's a shame because otherwise Blackmagic have awesome products. The codecs are vastly superior in terms of image fidelity and workflow. The UI is the best. Honestly, I can tolerate some shortcomings with a camera I paid 1300€ but as soon as we're talking about the UMP and so on, it's just not acceptable. The noise pattern of the UMP was also a no go (and other problems I mentionned in other threads) and ended up returning the camera.

Right now I shoot a LOT with my BMPCC4K and the touchscreen randomly stops responding.

Now Blackmagic is well installed in the camera landscape and has more money, a customer base. Why don't they go with an ultra reliable product. Once again especially at 10K, when there are great sony and canon offerings. There's a point where a professionnal is more interested in the reliability of his tool more than anything.

I don't wanna bash blackmagic just to be negative or always complain, I love the brand. I just give my genuine feedback of the experience I had with many of their cameras. Today, the only reasons I stick to the bmpcc4K is the easy to edit codecs (braw, prores) and image quality (when nothing goes wrong). But ergonomically besides the UI, there's still this battery problem with their pocket cameras, their different bugs with the touch screen, lack of IR and OLPF filter and so on.

You can say whatever you want, but there's NO good reason not to put an OLPF in a video (not photo) camera.

Wayne Steven wrote:I wish they would at least give a technical outline of the technology in Braw. Like the true compression engine currently, and in the past, and filtering. Which is hardly anything to ask, moderators?

I am pretty sure they open sourced BRAW. If so you will likely find some technical docs where ever you would find the source code.

I think that video is worth a watch provided there's no user error. That moire is pretty noticeable, and not sure i've ever seen moire on Chino pants before.

lost_soul wrote:

Wayne Steven wrote:I wish they would at least give a technical outline of the technology in Braw. Like the true compression engine currently, and in the past, and filtering. Which is hardly anything to ask, moderators?

I am pretty sure they open sourced BRAW. If so you will likely find some technical docs where ever you would find the source code.

It's not. Some people have hacked into the old version providing little detail, but this move to 5:1 in the 12k, might indicate a significant new change to the codec.

lost_soul wrote:I am pretty sure they open sourced BRAW. If so you will likely find some technical docs where ever you would find the source code.

They did not open sourced BRAW but they would like everyone to believe they did it, it makes look them great and open. Some people still think they did it, for an unknown reason. They provided a decoding compiled SDK, big difference

Olivier Chira wrote:Extreme sharpness is definitely not what a top image is about, especially at the cost of severe moire and aliasing (they should go for improved dynamic range and noise performance for instance). And you never know when it will show. I made comparisons with the bmpcc4K and the UMP 4.6K. Sometimes the 4K had less, there's no rule. Even a very high pixel count can't prevent from moiré.

We'll need to make some tests with the URSA 12K but maybe there exists some downscaling algorithms that are optimized to reduce moire patterns. Maybe that could help here ! Shoot 12K, downscale to an awesome final 8K or 4K image

Exactly! Just like Red, making R3D or BRAW is their secret 'coke recipe'.
They both offer an SDK for decoding and working with it, so NLEs can integrate these formats.

It shouldn't be really complicated to encode / decode BRAW. Maybe there's some "secret sauce" as you say in the encoding and the handling of the sensor noise though

I'm not in their head but open sourcing only has drawbacks : potential camera competitors could bring a BRAW competitor, they would have to deal with more software developers, the "BRAW brand" would deteriorate and be diluted etc.

Anyway back to the topic, it would be great if some of you send us your URSA 12K test files for our own tests for Autokroma BRAW Studio !
Cheers

jallen0 wrote:New video posted and I cannot tell if he is honestly trying to "help" or is looking to expose flaws to "hurt". In any event it would be interesting to see if others who have this camera can recreate his findings.

Am doing post production on a Commercial at the moment that we shot on the 12K for technical reasons. We had no moire problems but we weren't that close to anything that could have exhibited it. The other comments regarding noise and detail were very consistent with what I have been seeing in post. We were using an Optimo zoom but some of the footage felt every so slightly soft, I put some of that down to BRAW as I have noticed this when we moved from CDNG. In some situations I was able to sharpen the footage which it seemed to take pretty well, which I always take as a good sign (i.e. no internal sharpening). Really like the colours straight out of the box (ACES), they feel so true.

This is a cross post from the FB group from me:

Timothy Naylor Having used the camera for months now, I would say If you're shooting 12K and 8K, you'll likely not need an OLPF. Anything below that regularly shooting, there's no real free lunch. It is a 12K / 80MPx camera as advertised with some astonishing extras.

I've only seen such moire pop up when shooting 4K in camera, and I've seen an example of it at 6K cropped. And I imagine that the footage of him talking is shot at 4K, as there wouldn't be much need (or ability for them as it sounds like they were having trouble even dealing with the 12K data to begin with) for them to shoot it at 12K.

Here is a perfect example of where I would expect moire to be intense/unavoidable on my other BMD cameras. This guys shirt should and would be dancing, and it's clean.

https://drive.google.com/file/d/1yLMZkH ... sp=sharing

I appreciate BMD for making the grand effort of packing the camera with as many features as possible. 12K/8K/4K full sensor, High frame rates in all three of those, and even higher frame rates in crop modes is a lot for a one camera body to do. There has got to be some kind of trade off moving away from native 12K resolution.

Personally, I'm not interested in having an OLPF stock and would rather go to RAWLITE for that.

What I am now interested in, thanks to the conversation here on the boards, is deciding on Sigma, Milvus or Otus lenses to get closer to having the full potential of the new sensor at disposal.

Kholi Hicks wrote:This is a cross post from the FB group from me:

Thanks for posting that here, that is the explanation that makes the most sense to me.