or Connect
AVS › AVS Forum › Other Areas of Interest › Camcorders › Haters gonna hate: RAW videos in your pocket!!!!
New Posts  All Forums:Forum Nav:

Haters gonna hate: RAW videos in your pocket!!!! - Page 3

post #61 of 170
Quote:
Originally Posted by thedest View Post



Those were shot in the same place. Can you see the bottom frames? Cant you see the difference? Please.. and im the one closing my eyes?

Shot in the same place and that's it? That's all we need to know? Did you look at the two Vimeo videos you posted showing them as some kind of 'proof' and asking me if I couldn't see the difference? They were shot at entirely different vantage points with entirely different conditions at entirely different focal lengths. Good God man, get a grip. Seriously!!!!!!!!!!!!!!!!!!!!!!!!!!!!! You honestly think that was a comparison? It's obvious that you have no idea as to what constitutes a valid A/B between two pieces of equipment.

As for the pix that are also supposed to be convincing since they were shot at the same spot, two of the three were not shot at the same spot. As for the third, again:

Time of day makes no difference?

Weather conditions makes no difference?

Season makes no difference?

And I'm the one that needs to learn about imagery? Read what I wrote about being at the Grand Canyon at different times of the day, different seasons or different weather conditions. Do you really think because two cameras shot from the same spot on different days, different seasons, different lighting, that it somehow constitutes a valid comparison???

You and I are on such different wavelengths, there really is no point in continuing this. I'm done with our back & forth.
post #62 of 170
Quote:
Originally Posted by thedest View Post

.



.

Sometimes words are not needed biggrin.gif

I believe we have discussed that particular comparison before. Both clips were taken from the same camera, but with the 8 bit clip they truncated the upper and lower ends of the original 12 bit clip to "mimic" 8 bits. So the detail in the highlights and lowlights were deliberately thrown away. After that they enhanced the missing bits to accentuate them in the 12 bit clip, so it shows a bigger difference than you would see in real footage (in real footage the visual impact of detail at the high and low end is not as great as in the midrange). It is an artificial comparison, not one between separate cameras.

That assumes of course that all cameras have the same dynamic range and extra bits just extends it. Unfortunately that is not true. A camera shooting in 8 bits may have a higher dynamic range than one shooting in 12 bits. 12 bits would give you finer gradations but it is NOT the same thing as dynamic range.

In one of the threads on the BM forum some guy showed something similar, with an indoor foreground subject and a window in the backdrop. He then proudly proceeded to "grade" his footage so that the subject matter through the window was in high detail. But in the process of doing that he washed out a lot of the midrange detail in the foreground (the part the viewer would be focussing on). Basically he muted his primary subject matter so he could accentuating background detail. This is the sort of thing that is jarring to a viewer. It is also the sort of thing that will create compression artifacts in wrong areas of the image on delivery, since compression engines need something to compress (and will choose the less detailed areas to do that in). This is the problem with your approach, you are so fixated on dynamic range that you neglect other aspects of the image that are more important.

Another thing, as far as image quality is concerned, at the end of the day you are going to be delivering 8 bit compressed video. If you have well shot footage you probably don't need to alter it much or at all (beyond cutting/slicing), in which case it isn't going to matter if your footage is 8 or 12 bits, RAW or compressed.
post #63 of 170
Thread Starter 
Quote:
Originally Posted by Ken Ross View Post

Shot in the same place and that's it? That's all we need to know? Did you look at the two Vimeo videos you posted showing them as some kind of 'proof' and asking me if I couldn't see the difference? They were shot at entirely different vantage points with entirely different conditions at entirely different focal lengths.

Even if one of them were shot during the night in Venus and the other one during the day in Pluto I would still see the differences in DR, color depth and detail! If you cant, im sorry for demanding so much of you!
Quote:
Originally Posted by Ken Ross View Post

As for the pix that are also supposed to be convincing since they were shot at the same spot, two of the three were not shot at the same spot. As for the third, again:

Time of day makes no difference?

Weather conditions makes no difference?

Season makes no difference?

Im not sure, but I think that they were shot in the same day, at almost the same time.

And LOL, im waiting for you to say that the air humidity can make a big difference? Maybe Bill's humor can also increase the image quality? LOL.. Please..

.



.

Again, cant you see the differences here, or do you think thats the result of a weather condition? Damn, thats why our discussion is not getting anywhere.
Quote:
Originally Posted by Tugela View Post

I believe we have discussed that particular comparison before. Both clips were taken from the same camera, but with the 8 bit clip they truncated the upper and lower ends of the original 12 bit clip to "mimic" 8 bits. So the detail in the highlights and lowlights were deliberately thrown away. After that they enhanced the missing bits to accentuate them in the 12 bit clip, so it shows a bigger difference than you would see in real footage (in real footage the visual impact of detail at the high and low end is not as great as in the midrange). It is an artificial comparison, not one between separate cameras.

Nope, thats actually the opposite. Please, watch that video again.

When you convert a 12-bit source into an 8-bit file (with the workflow he used) you are actually creating an 8-bit file with more info than what any other 8-bit camera can create. So that "fake" 8-bit file has actually more data than any other 8-bit native file.

They have accentuated the 12-bit file, because when you shoot 12-bit RAW, you always do that. Simple as that.

Its not an artificial comparison, its actually very intelligent and it was made by one of the best pros in the market, one that has had the chance to work with all the best cameras available. He actually was one the first ones to use DSLRs in professional stuff, and he also owns a GH3 and many other great 8-bit cameras. Now watch again his video and listen to him talking about the RAW video from the Blackmagic against ALL 8-bit cameras.

I know that those 8-bit cameras have some tricks to "increase" the DR, and they are pretty ridiculous. People think that dynamic range is simply being able to see shadow detail and cloud detail, but THATS WRONG. Dynamic Range is all about smooth gradations between the highlights and the shadows. To try to mimic a great dynamic range, those 8-bit cameras have an algorithm to privilege highlights. What happens? When you are in a high contrast scene, it CRUSHES the shadows, transforming them into a black mud, and they also crush some midtones - and that creates horrible stuff like orange skintones.

So with them you have two options: you can have your highlights and destroy the shadows and crush the midtones OR your can blow out your highlights.
Quote:
Originally Posted by Tugela View Post

That assumes of course that all cameras have the same dynamic range and extra bits just extends it. Unfortunately that is not true. A camera shooting in 8 bits may have a higher dynamic range than one shooting in 12 bits. 12 bits would give you finer gradations but it is NOT the same thing as dynamic range.

Nope. Wrong again.

Mathematically speaking, 12-bits will ALWAYS have more dynamic range than 8-bits.

Even if a digital camera can capture a large dynamic range, the precision at which light measurements are translated into digital values may limit usable dynamic range. We have an alanog to digital converter that makes that, and its totally dependent of bits of precision.

So, an 8-bit file can only have a contrast ratio of 256 levels, and that means that an 8-bit file can only have 8 stops of dynamic range. And im not even talking about GOOD dynamic range, because part of that dynamic range can be garbage because of the noise levels.

You can use some picture profile vodoos or create better algorithms for the cameras, and that can create a better "visible" dynamic range. And thats what good cameras (like the GH3) can do. They are smart enough to privilege the important levels, so it LOOKS like it has a great dynamic range.

But what is actually happening is that the camera is crushing a little bit more a shadow or a midtone to give you a better look. That may fool most people, but numbers dont lie, and I can clearly see the differences.

When my eyes hit an 8-bit video I may think: WOW! That looks great!!

But when my eyes hit a real high dynamic range video I think: WOW! Thats so freaking cinematic!
Quote:
Originally Posted by Tugela View Post

In one of the threads on the BM forum some guy showed something similar, with an indoor foreground subject and a window in the backdrop. He then proudly proceeded to "grade" his footage so that the subject matter through the window was in high detail. But in the process of doing that he washed out a lot of the midrange detail in the foreground (the part the viewer would be focussing on). Basically he muted his primary subject matter so he could accentuating background detail.

Well, thats his fault, not mine nor the camera's. The BM forum has a lot of garbage. Most users posting there have no idea what they are doing. They are migrating from iphones to a cinema camera. When they realize they cant get decent results they go to the official forum to complain.
Quote:
Originally Posted by Tugela View Post

This is the problem with your approach, you are so fixated on dynamic range that you neglect other aspects of the image that are more important.

Nope. You guys should read my posts more carefully.

I will say again. With RAW you have better resolution/details, better color reproduction, more dynamic range, you can make it sharper, more saturated etc.

In ALL aspects of IMAGE quality it is better.
Quote:
Originally Posted by Tugela View Post

Another thing, as far as image quality is concerned, at the end of the day you are going to be delivering 8 bit compressed video. If you have well shot footage you probably don't need to alter it much or at all (beyond cutting/slicing), in which case it isn't going to matter if your footage is 8 or 12 bits, RAW or compressed.

Yeah.. thats a RAW Blackmagic video graded by me and exported as an 8-bit video with a bad codec and recompressed by youtube (and I have less than 6 months of experience with post processing)

It looks like any other 8-bit video, right? rolleyes.gif

.

.
Edited by thedest - 11/14/13 at 4:54am
post #64 of 170
Quote:
Originally Posted by Tugela View Post

This is the problem with your approach, you are so fixated on dynamic range that you neglect other aspects of the image that are more important.

I've tried to get through to him on that point, but to no avail. I've given up. Very tough to communicate with people that firmly believe they know all there is to know.

I found it amusing that he's trying to convince me that the two Grand Canyon shots were shot on the same day at about the same time. Of course it's clear from the third set of 'comparisons' (I use that word while wincing) pix that the vegetation is very different in the two shots.

As I said, I'm done, this has just gotten so over the top.
Edited by Ken Ross - 11/14/13 at 5:36am
post #65 of 170
Thread Starter 
Ken, its really funny how you nit pick about small things and you simply ignore me when I talk about facts. If they were shot in two different decades, the difference is still pretty clear. Do you want to talk about vegetation? biggrin.gif I may learn some stuff with you

And here I am, giving you guys FACTS AND EXAMPLES and all that you guys are giving me is "I THINK THIS. I THINK THAT."

Please, where are your facts? In what aspects of image quality a 12-bit RAW file loses for a compressed one?

Im really curious to know HOW an original file like RAW can be worse than a compressed file based on a RAW. That kinda defies the laws of physics. You are taking away information of a RAW file when you compress it and the result is: a better file! biggrin.gif Its like magic!

We can all talk about how good those 8-bit cameras are in another topic - and I do that all the time, because I like some 8-bit cameras - but please, we are in another league here.

A mercedes is a great car (and it has airbags, YAY!), but when we talk about performance, its no match to an F1 car. You guys are not getting the feeling of that discussion!
post #66 of 170
Thread Starter 
So now we have 2 more examples. One of them is dedicated to Ken, and the other one to Tugela.

First we have a test to show Ken's point of view. He thinks that the difference is not that big between average 8-bit cameras and a BMPCC RAW. So here it is, the GH1 vs the BMPCC in the same place.
Quote:
My GH1 can make very nice images, but it cannot shoot this room and creat any useable footage.
The BMPCC in RAW mode, 800 ISO, w/ a Voigtlander 25mm @f.1 actually sees far more than my eye does in this light...

It's very interesting to know that in a pinch I can manage to get a shot where lighting cannot be controlled - documentary style.
Quote:
The scene is white balanced to around 3000k and it looks terrible with any camera NOT shooting RAW.

LINK FOR THE COMPARISON: https://vimeo.com/79365687

Now, is it only better than an 8-bit average camera? Lets see how it compares to the Magic Lantern RAW of the 50D

Quote:
Right off the bat, even in preview mode, I could tell the BMPCC had more dynamic range than the 50D.

And working the DNG's in Adobe Camera RAW only confirmed this. The BMPCC's 13 stops of dynamic range is much more flexible compared to the RAW from the 50D.

LINK FOR THE COMPARISON: https://vimeo.com/79359239

And look, Tugela, the interior and the exterior are well exposed! And there are no crushed midtones!


NEXT!
post #67 of 170
I can see the difference. There's more flexibility in post with the BMPCC. That zoom-in comparison shot with all the noise on the Canon really illustrates the advantages of the BMPCC.

but... biggrin.gif

Being that two people, who are into video, can't agree what image looks better over the other...just think how non-video people see it - they don't. Unless you will be the only one ever seeing your footage, the average person will never be able to tell either way.
If you type in a model number of a $150 HD point and shoot into YouTube, you will find videos with comments such as "amazing quality". (In a sense, they are right because we are spoiled by the quick advances of technological roll-outs/yet the basic technology is still amazing.)

There is a basic conflict here in that many on this forum are looking more for done-in-camera process/ready to upload. This forum is not populated by people into fimmaking/post-processing.

thedest, that basic difference in approach here may always frustrate you.

A forum like dvxuser get into the geeky discussions about bits and DR:
http://www.dvxuser.com/V6/forum.php?s=12fbbed0d04b150b25cc601aaa3d71b9

They even have their own, separate Black Magic forum, so you can bask in all its bitty glory biggrin.gif
http://www.bmcuser.com/
post #68 of 170
Thread Starter 
Now more stuff for you Ken.

Here you can see the HACKED GH3 against RAW video in the same place, at the same time.

.



.

And here you can see the same camera with the same lens in the same spot at the same time of the day shooting 8-bit vs RAW.

.

post #69 of 170
Thread Starter 
Quote:
Originally Posted by xfws View Post

Being that two people, who are into video, can't agree what image looks better over the other...

Even my girlfriend can see the differences between RAW and h264 when I show them side by side. I dont know whats happening to Ken. Maybe he is doing a prank or something.
Quote:
Originally Posted by xfws View Post

Unless you will be the only one ever seeing your footage, the average person will never be able to tell either way.
If you type in a model number of a $150 HD point and shoot into YouTube, you will find videos with comments such as "amazing quality". (In a sense, they are right because we are spoiled by the quick advances of technological roll-outs/yet the basic technology is still amazing.)

Thats true. People cant see with technical eyes the differences between them, but they can feel it. One year ago I knew nothing about cameras (I had never heard about ISO, sensor, framerates etc back then), but even at that time, when going to the cinemas I could feel that the image in that screen had something different than most TV shows. Even if they had the same actors from TV shows, they looked different.

Everyone I know shoot JPEG pictures with point and shoot cameras. When I show up and show them my RAW pictures taken with the NEX or the 60D they get all excited. They dont know whats better, they just know it is.

If that were totally true, why is no one shooting Hollywood movies with 4k cellphones? And thats the answer. Ordinary people can feel the difference.

And RAW video is not a technological advance. Film cameras were awesome in resolution, color reproduction and dynamic range. When we entered the digital era, we actually gave a step back, because those digital images were far worse than the film images. Thats why a lot of movie directors still shoot movies with film cameras.

What we are doing with RAW now, is getting back to a similar image quality (and process, because you have to develop the image).
Quote:
Originally Posted by xfws View Post

There is a basic conflict here in that many on this forum are looking more for done-in-camera process/ready to upload. This forum is not populated by people into fimmaking/post-processing.

thedest, that basic difference in approach here may always frustrate you.

I know that, but im not frustrated. They are learning by force biggrin.gif

Im pretty sure that some guys are considering or are waiting for a new camera with higher depths. They may not like the BMPCC, but they are definitely waiting for a 4:2:2, 4:4:4 high dynamic range camera!
Edited by thedest - 11/14/13 at 7:46am
post #70 of 170
RX10 videos are bad because no one remaps 0-255 input into 16-235 output so the clip highlights, crush blacks, and lower DR. BMPCC videos are bad because people do not apply simple LUTs, like Captain Hook's, and instead half grade the stream.
post #71 of 170
The one thing I find amusing is how the samples from the GH3 look so bad. It really isn't too difficult to get those shots without totally blowing out the sky as depicted in those grabs. I'm sure anyone here that owns a GH3 can attest to the fact that they can get far better imagery out of their cameras than that A/B. Sometimes I honestly think the guys that do this are deliberately overexposing, underexposing or just generally whacking the exposure to show how good sample B is. I've taken many videos where I've gotten beautifully exposed landscape AND sky at the same time.

Granted I don't own the Canon, but once again I see nothing so challenging in a few of those grabs that shadow detail should have been crushed like that. Hell, in the Canon shots, not only is the sky blown but the shadow detail is crushed. Really? C'mon. IMO this is not so much a question of the beauty of RAW, this is a question of either poorly exposed original footage or someone 'tampering' to make a point. As someone here likes to say repeatedly when one lousy BMPCC example after another is presented, 'it's not the fault of the camera'. biggrin.gif
post #72 of 170
Thread Starter 
Quote:
Originally Posted by Rohela1122 View Post

RX10 videos are bad because no one remaps 0-255 input into 16-235 output so the clip highlights, crush blacks, and lower DR.

Welcome to the forum! Here we fight a lot, but we are all good guys! biggrin.gif Dont get the wrong idea. Ken loves me! wink.gif

You are talking about the way those 8-bit cameras organize the DR data. I have talked about that many times here. And I "remap" all my 8-bit videos. That "remaping" is only needed when you have overexposed shots. When you dont have a high contrast scene, there is nothing you can do.

Here you can see an example that I have posted not a long time ago. I work with the full range on my NLE and its really easy to see the full DR of those cameras.

.



.

But even doing that, the footage is very limited compared to higher end formats. The dynamic range will never come close. That information looks like a new thing for the dvxusers, but thats really old news. Some guys call that "super white", and its present on most 8-bit cameras. You can also recover some shadow, but it has so many noise that its useless.
Quote:
Originally Posted by Rohela1122 View Post

BMPCC videos are bad because people do not apply simple LUTs, like Captain Hook's, and instead half grade the stream.

That is not true. Ken is talking a lot about having the best possible native colors. When you apply a LUT, you are changing color info, and doing that, what Ken says become totally true.

Its becoming a joke to see those BMPCC video samples. When I see someone saying that he used a LUT or Colorista or Magic Bullet the first thing that comes to my mind is: HE KNOWS NOTHING ABOUT COLOR CORRECTION CONCEPTS.

You dont need LUTs or Colorista or Copy of films or whatever to grade a BMPCC video. You just have to apply basic concepts to get a real life look.

Here is RAW Blackmagic video. One was graded by me using only basic concepts of color correction on Premiere Pro. The other one was graded by a pro, using Resolve and some LUT.

.



.

LUTs are usefull if you want to create a look. If you want real life images, its not the way to go. People are using them a lot because they are lazy and dont want to learn how to work.
Quote:
Originally Posted by Ken Ross View Post

The one thing I find amusing is how the samples from the GH3 look so bad. It really isn't too difficult to get those shots without totally blowing out the sky as depicted in those grabs.

You need to look better sir.

The exposure is not wrong. Both cameras were exposed for the trees. As you can see, in both of them the exposure for the trees are correct.

So when you expose the GH3 for your subject (trees) you overexpose the sky. That doesnt happen when you have more dynamic range.

Of course you can expose the GH3 for the skyes. But if you do that you will throw the trees to the shadows. You can create decent looks with the GH3 that way, but thats a trick to handle with the low DR.

One more important thing that you are not seeing. The GH3 is using a flat profile, so its DR is being pushed to the limits, and even that way it cant match the RAW. The standard picture profile has even less DR.

That comment is from an experienced shooter, that likes to do short-movies with the GH3

"I consider myself more as a m4/3 shooter than a GH3 film-maker, so I am happy to have alternative m4/3 compatible cameras that can fully utilize my m4/3 glass.

The Blackmagic Pocket Cine camera fills a very big hole in the current GHx line-up...

- Wide Dynamic range ( 13 stops, compared to the GH3's 9 stops )
- Low compressed high bit-rate 4:2:2 recording
- RAW recording ( coming very soon )

Having a BMPCC body in your camera kit ( along with your GH3 ) means that you can handle challenging lighting situations where the GH3 does poorly. "


Quote:
Originally Posted by Ken Ross View Post

this is a question of either poorly exposed original footage or someone 'tampering' to make a point.

On those cameras, in a high contrast scene, if you expose your video for the skyes, you will crush midtones and shadows. Again, its math. The DR is translated into numbers. And the numbers in 8-bit are lower.

And thats the comment of a guy that is shooting a movie with a RED SCARLET, a 5DMK3 RAW and a Blackmagic Pocket

"Pocket smoked the Scarlet @ ISO 1600. It also had cleaner blacks than 5D3-raw @1600, and more dynamic range.

I'm very, very happy with the Pocket."

Edited by thedest - 11/14/13 at 10:08am
post #73 of 170
Not every LUT is a simple lookup designed to apply a look.. some just restore the image to a neutral state., Away from the cmoressed state inherent in RAW. That is what the Captian Hook LUT does specifically for BM camera RAW. I also hate Colorista and the like.

I am not attacking anyone here but I look at the reviews of both the BMPCC, which I owned, and the RX10, and I see simple mistakes in almost all review videos, and then see people nitpicking over flawed samples.
post #74 of 170
Properly exposed trees, bright, properly exposed skies, all at the same time and just done about a minute ago with the GX7. How is this even possible without the BMPCC? Sorry, I think some of those shots are utter nonsense. Anyone with good equipment can properly expose both. If you're pointing toward the sun, that's something else, but my God, our cameras have enough dynamic range to do better than some of the garbage we're being shown here.

I'm just so tired of some of the exaggerations here, really.

https://vimeo.com/79420915

https://vimeo.com/79420916
post #75 of 170
Thread Starter 
Quote:
Originally Posted by Rohela1122 View Post

Not every LUT is a simple lookup designed to apply a look.. some just restore the image to a neutral state., Away from the cmoressed state inherent in RAW. That is what the Captian Hook LUT does specifically for BM camera RAW. I also hate Colorista and the like.

Yup, that LUT tries to achieve a natural look. The problem is that when you make that process automatic, you can miss some shots, and most new users do that. They just apply the LUT and have no idea how to refine it.

And recreating a natural look is faster than loading a LUT and refining it. People are using too much LUT without the need to do it. Its the same as processing a RAW picture, and no one uses a LUT for that.
Quote:
Originally Posted by Rohela1122 View Post

I am not attacking anyone here but I look at the reviews of both the BMPCC, which I owned, and the RX10, and I see simple mistakes in almost all review videos, and then see people nitpicking over flawed samples.

The only one doing that is Ken. He just cant accept that RAW is better than a highly compressed h264, as funny as this may sound.
Quote:
Originally Posted by Ken Ross View Post

Properly exposed trees, bright, properly exposed skies, all at the same time and just done about a minute ago with the GX7. How is this even possible without the BMPCC?

You have just criticized me for comparing 2 different videos, and now you are doing the same? rolleyes.gif

Please, the comparison I posted was made in the same spot at the same time. You cant argue with that. Its obvious that in that comparison the scene had more dynamic range than in yours. You cant argue that the scene is well exposed for the trees and you cant argue that the sky is missing data in the GH3. Period. In that situation, to get a decent image on the Gh3 you would have to underexpose the trees. And poof! Now you can see the RAW shining!

And your video has a lot of blown out stuff! (just to nit pick biggrin.gif )

Exposing well trees and skies is not a challenge. The challenge is dealing with high dynamic situations.

And surprise! Even a cellphone can do that in the right situations!

.



.

I dont understand why you are trying so hard. You are the only person in the planet that cant see the advantages of RAW over h264. There are plenty of specialists showing examples, we have all the numbers to prove the superiority, we have lots of comparisons.

Can you please watch that again?

https://vimeo.com/79365687

You are getting me bored Ken. You just dont have more arguments. Why dont you go to 8-bit camera threads?
Edited by thedest - 11/14/13 at 11:42am
post #76 of 170
Are there advantages of RAW over AVCHD, for IQ yes. But a film is not just IQ potential within the camera, many things go into suing some of that potential, lighting, rigs, set design, etc. The average person could careless about doing these things. In the same way the average person will never exploit the full potential of a RAW camera. So the real question is if another type of camera can actually produce good enough images. All you need is good enough images to tell a story. A story does not depend upon maximum image quality, it depends upon the totality of the effort. Because most people don't make the effort that a RAW camera requires, they will actually produce better pictures using a camera with a baked in look. The Zacuto test showed that a low res low DR camera, the GH2, could actually be preferred in a blind test by Francis Ford Coppola. I think he know a few things about filmmaking.
post #77 of 170
Thread Starter 
Yep. All that you say is true.

But please, thats a topic about IMAGE QUALITY. We have lots of forums that discuss scripts, audio, acting, lighting, rigs, decoration etc

And about an average person not being able to exploit the full potential of RAW, its not right. Just go to flickr and look for amateur pictures that were shot in RAW. Some of them are simply amazing. RAW video is the same. People just have to learn that.

And I agree when you say that most people will produce a better image with cooked videos (if they dont want to learn how to work with higher end stuff)

The Zacuto Shootout also showed that those famous film makers dont know a lot about their gear. In one of the episodes they were impressed to realize the differences between ProRes and h264. I remember that a famous guy kept asking for them to flip back and forth the h264 X ProRes footage. All the guys said at the end that they were going to change their workflow after that.

Being a pro doesnt mean that much. There are lots of good hobbyists out there. And there are differences between directors, producers, shooters, actors etc.

In another episode they said that once we get RAW videos on digital cameras, films will be dead (that video was published before Magic Lantern RAW and the Blackmagic cameras).
post #78 of 170
Thedest, it is you that is boring me with your never-ending pushing and selling. You'd think you were the BM sales manager.

As for the video i posted, I see no areas that are blown, help me out. And i stand by my assertion that the A/Bs do not look right. There is little there that is so challenging for a Canon or GH3.

You also totally misssed the poiint in the videos I posted. You give people the impressionb that only a BM can properly expose a darker and brighter area at the same time. That was the point.
post #79 of 170
Thread Starter 
Quote:
Originally Posted by Ken Ross View Post

Thedest, it is you that is boring me with your never-ending pushing and selling. You'd think you were the BM sales manager.

Im not boring you, because you keep coming back to my topics, over and over again. If you cant see the advantages of RAW and dont want to learn what they are, please, feel free to leave. There are great topics about 8-bit cameras here!

The only person that I said that sould try out this camera was mark. I think that most of the other guys here are good to go with an EOS M. And im definitely not trying to sell it to you.

Im showing how awesome this new format is, and YOU are the only one that cant understand that. There is no other format in the market that holds more data than a RAW file. And we have that at a consumer price. If you cant get excited with that, you have to understand that you are another kind of consumer! Just like for some people a GH3 is an overkill and an iphone is more than enough.

The BMPCC brings the same revolution that the firsts DSLRs made when they started to offer RAW pictures. Now, put on the market a DSLR that cant shoot RAW pictures and try to sell it!

What im saying can be applied to lots of other cameras. The problem is that they cost the price of a car. Soon we will see the release of the Digital Bolex, wich will have a similar performance to the BMPCC. I'll definitely rave about it either.
Quote:
Originally Posted by Ken Ross View Post

As for the video i posted, I see no areas that are blown, help me out.

You said you have been working with video editing for years. Im tired of showing you all the facts, and you keep saying wrong things. Load your video in your NLE and look at the YC Waveform. I can count a lot of blown out spots, and thats why those videos look cheap and NOT cinematic. Its really amazing that you cant see them, because they are right at your face! Im sure that they can be corrected if you expose the camera correctly, which you didnt.
Quote:
Originally Posted by Ken Ross View Post

There is little there that is so challenging for a Canon or GH3.

Words cannot express how much you dont know what you are talking about. You should stop disagreeing and start learning. They cant handle the dynamic range, they cant render real colors and they cant handle the amount of detail.
Quote:
Originally Posted by Ken Ross View Post

You also totally misssed the poiint in the videos I posted. You give people the impressionb that only a BM can properly expose a darker and brighter area at the same time. That was the point.

Im not missing the point! Im a great fan of the GX7/GH3, the NEX series, the Canon APS cameras etc, and they are more than enough for MOST people.

I've never said that the BM is the only one that can expose correctly. I only said that technically it will ALWAYS perform better, in ANY SITUATION, SPECIALLY IN HIGH CONTRAST SCENES. It will ALWAYS have more detail, better colors, better gradation, better EVERYTHING.

That doesnt mean you cant do great videos with those other cameras.

How many times do I have to say that im talking about performance???
Edited by thedest - 11/14/13 at 1:13pm
post #80 of 170
Quote:
Originally Posted by thedest View Post

Im not boring you, because you keep coming back to my topics, over and over again.

Guess I could say the same. You are on a evangelistic mission or so it seems.

Here is my problem with you from Day One...and I promise this is the last time I will respond to you in this thread:

* You have extolled the virtues of the BMPCC and that's fine. I never claimed the greater dynamic range or shooting RAW was a bad thing. It can obviously be a good thing in the right hands.

* It is the last part of the above point that has always bugged me about your pitches. You continually dismiss out of hand the horrific end-results that almost every owner whose work you have posted seems to get. Likewise you continually dismiss this as 'OE'. At some point you must come to the conclusion this is an art and not a science. Most people either don't take the time to master the art or simply find it too difficult to do.

* You show examples of how in '10 seconds' you can produce better results than the other guys. I've found virtually all your results to be off too. Perhaps not as bad as some others, but still off. And let's be fair, you are working with one frame, not an entire video.

* You have never address the very real issue of not achieving a proper color balance at the shoot location. Attempting to do that in post, hours, days or weeks later is nothing more than guesswork. There is absolutely no better substitute for achieving this accuracy than when on site and while shooting. AND, your point above that our 8 bit cameras 'can't render real colors' so underscores my issues with you, I couldn't have done a better job. That statement is so far off-base, so inaccurate, so over the top, so naive, that I just don't know what to say.

* You continually dismiss the issues with the BMPCC camera itself, no VF, lousy audio, lousy ergonomics, poor battery life and other technical issues that I've read on the forums that escape me for the moment. It is very unfair to promote the BMPCC (and let's face it guy, you ARE promoting this camera) and dismiss, minimize or even chastise those that point out the many weaknesses of the unit. I find a fair & balanced discussion far more convincing and believable than the tact you have taken throughout.

* I will again stand by my perception that some of the samples you provided (not created by you) are suspect IMO. Why? Because I've worked with many cameras that produce far better results than that under similar conditions. Really, no joking. In looking at some of those pictures produced by Canons, GH3s and others, you'd think they came from a cheap cellphone. That's why I am somewhat suspect that the 'author' of those A/Bs was trying to make a point and, shall we say, stretched the point a bit.

* Somewhat related to the above point is that you yourself have denigrated '8 bit' cameras as if they could do no better than a cellphone from 5 years ago.

Again, I am not arguing the point of better dynamic range or RAW output, but I am saying that is not nearly as easy as you make it sound and the results prove it time and time again. Aside from that are the very distinct issues with the BMPCC itself. You have continually 'poo pooed' those deficiencies with the very dismissive 'oh, you can add a VF....oh, you can add an outboard mike....oh, you can add multiple batteries'. So the point here is that many of us are not looking to carry a production studio around with us, we are shooting for fun. If you're an Indie movie-maker, that's something else. Some of us are simply looking for excellent results, in a very portable package, from those 'horribly handicapped, hideously AVCHD equipped' 8 bit trashy cameras that have produced some extraordinarily good results for many many people.

Regarding the video I posted, the point was to show a properly exposed sky and the trees in the background, nothing else. That was achieved. That was done in 5 seconds.

I promise (I really do) that I will no longer respond to your posts in this thread. I think we've beaten this to death so many times there's no body to recover. smile.gif
Edited by Ken Ross - 11/14/13 at 1:32pm
post #81 of 170
I continue to be bothered by the pervasive posting in all forums of some of the worst-looking videos I have ever seen coming from the Black Magic cameras, particularly because people just seem to ignore how bad they are. The "explanation" seems to be that these folks have no idea what they are doing when they "color grade", with most using some pre-sets (LUTs) that pretend to be "cinematic." This term from the evidence appears to mean that colors are totally unrealistic, or there is hardly any color at all. Avoidance of hot spots and lack of detail in shadows seems to be the only objective.

Is this just incompetence? some aesthetic that has escaped us regular guys? Or is it next to impossible to get good colors from the camera without twenty years experience as a grader?

To check this out, without having to get any of these cameras, I downloaded a RAW dng "clip" (folder of RAW frames) that was shot by a BMPCC. Here is a converted jpeg of one original frame
right from the camera (no editing whatsoever):



This is awful, but it is, well, raw. So, into Lightroom to treat the raw picture (actually the 200 RAW pictures) just like I would a raw still (which is all it actually is). Using standard photography tools - WB temperature, saturation, contrast settings I tuned to taste (now here we get Ken's point - how do I know what the scene really looked like?, but it's a scene in somebody's kitchen, so I can guess). Here is what I came up with (same jpeg compression):



This seems more like a regular video shot, and now we can admire the detail, lack of noise, etc. (although there does not appear to be any challenging dynamics). Indeed, it is hard to believe this is just a video frame grab (1920x1080). The point is, at least with the new RAW files, it should not be an insurmountable challenge to create videos that might in fact be better than anything any small camera could, with a somewhat unfriendly camera and a lot more editing work.

For people who shoot RAW stills, shooting RAW video should be appealing. The dream of not worrying about recompression and compression artifacts in video is now realized, whether one does especially creative editing or just wants a realistic depiction of what one sees, in a tiny camera. Too bad the best lens for this camera costs $1,200 (an f2.8 IS zoom).
Edited by markr041 - 11/14/13 at 3:34pm
post #82 of 170
Quote:
Originally Posted by markr041 View Post

For people who shoot RAW stills, shooting RAW video should be appealing. The dream of not worrying about recompression and compression artifacts in video is now realized, whether one does especially creative editing or just wants a realistic depiction of what one sees, in a tiny camera. Too bad the best lens for this camera costs $1,200 (an f2.8 IS zoom).

So for me Mark the question is, forgetting the work that would be involved, how do we achieve a 'realistic depiction' of what one sees days or weeks later? Do we really remember what that scene looked like? Do we really remember the highlights as our eyes saw it? Do we really remember the subtle shades of green, red or what have we?

So here's the point, if by a 'realistic depiction' we mean a 'believable' depiction, I can buy into that. "Believable" can go in many different directions before someone says "hey, that just doesn't look right". Your kitchen 'touch up' hits the nail on the head. It looks 'believable' and probably wouldn't raise any eyebrows except with the person whose kitchen that is, but it probably isn't 'accurate'. So if that's the case, we have quite a bit of latitude. But 'realistic' and 'accurate' have two very different meanings IMO. Therein, at least for me, lies the problem.

Now compound this issue with the fact that when we go out shooting, we're usually at multiple sites, in multiple lighting conditions with a resulting multitude of different color renditions. If we want to achieve accuracy for this entire shoot, as opposed to a 'realistic depiction' of these multiple scenes at multiple sites, we've got quite a headache in store for us.
post #83 of 170
I think you are correct - you really do have to recreate the scene in the editor. The BMPCC lcd screen shows focus, but not really the color. There is no auto WB, you just set the relevant temperature, use zebras to get the right exposure, and use peaking to get the right focus. Hard to know what you got until you see it in an editor and then can fine tune.

Here's an idea - take a still photo for each different scene/site, and then use those as a guide later! Match the video frames to the pictures (which have even more resolution and dynamic range). That's what painters do. Of course, the BMPCC does not take stills, so this requires two cameras (two pockets, maybe two photographers).
post #84 of 170
Mark, it's funny you said that, because I was thinking exactly that when I wrote my last post. I thought here we go again, more equipment to make up for the deficiencies of not getting it right with the acquisition camera.

So now all we need is an auxiliary VF, outboard mikes or recorders, extra camera for stills to recreate the scene for later color benchmarking and extra batteries...lots of them. Oh, forgot the extra photographer! Here I thought I was out for just a fun shoot. wink.gif
post #85 of 170
You all might consider reading the book "How to Shoot Video That Doesn't Suck" by Steve Stockman. You can play with or argue about cameras forever. But somehow this equipment exists to tell a story. This argument may be interesting, but consider making a video!
post #86 of 170
Quote:
Originally Posted by bsprague View Post

You all might consider reading the book "How to Shoot Video That Doesn't Suck" by Steve Stockman. You can play with or argue about cameras forever. But somehow this equipment exists to tell a story. This argument may be interesting, but consider making a video!

Well, I have 155 videos on Vimeo, but one needs to master now some editing tools to work with RAW and check that one can actually get a good video, before shooting.
post #87 of 170
Quote:
Originally Posted by markr041 View Post

Well, I have 155 videos on Vimeo, but one needs to master now some editing tools to work with RAW and check that one can actually get a good video, before shooting.
You are well ahead of me on Vimeo. I agree. I'm trying to suggest using the equipment to make videos worth watching rather that endlessly arguing about equipment.
post #88 of 170
Quote:
Originally Posted by bsprague View Post

You are well ahead of me on Vimeo. I agree. I'm trying to suggest using the equipment to make videos worth watching rather that endlessly arguing about equipment.

Well, evidently the only videos of mine worth watching are my silly clips using the EOS M RAW capability - over 1500 plays, and hundreds of downloads. If I put "Test Video" in a video name, I can up the plays by a lot... another big winner is focus pulls on the 55-250mm Canon lens. Many more plays than dancing in Shanghai, or strolling in Nanjing, or roaming Hong Kong.
post #89 of 170
Quote:
Originally Posted by thedest View Post


Nope. Wrong again.

Mathematically speaking, 12-bits will ALWAYS have more dynamic range than 8-bits.

Even if a digital camera can capture a large dynamic range, the precision at which light measurements are translated into digital values may limit usable dynamic range. We have an alanog to digital converter that makes that, and its totally dependent of bits of precision.

So, an 8-bit file can only have a contrast ratio of 256 levels, and that means that an 8-bit file can only have 8 stops of dynamic range. And im not even talking about GOOD dynamic range, because part of that dynamic range can be garbage because of the noise levels.

You can use some picture profile vodoos or create better algorithms for the cameras, and that can create a better "visible" dynamic range. And thats what good cameras (like the GH3) can do. They are smart enough to privilege the important levels, so it LOOKS like it has a great dynamic range.

But what is actually happening is that the camera is crushing a little bit more a shadow or a midtone to give you a better look. That may fool most people, but numbers dont lie, and I can clearly see the differences.

.

I don't think you know what dynamic range is. It has nothing to do with the bit depth.

Sensors work by responding to light and generating a signal. At the high end of exposure the response becomes saturated and that represents the upper limit of your detection. At the low end eventually the signal will be lost in noise, so you get no usable signal. These represent the white and black points of the sensor respectively, and that is the dynamic range of the sensor. The data coming off the individual sites on the sensor will be an analog signal. This is passed through a DAC to generate your digital response. The resolution of the signal can be any number of bits depending on the how the hardware of the camera has been set up, but the dynamic range will be the same no matter if you have a resolution of 4, 8, 12 or 100 (or whatever) bits. It is like taking a meter stick and dividing it up into centimeters or millimeters. The stick stays the same length, all you are doing is changing the divisions within the length.

In a real camera digital filters will be used at the high and low ends to eliminate poor quality data. At the low end you will have mostly noise, and at the high end the response curve for the sensor will become non linear. Those are things to avoid, so they will be cut off. These will be the black and white points for your camera, it's effective dynamic range. Within that dynamic range the actual data will be resolved into X bits, depending the depth the manufacturer sets. They might use a linear response curve for that resolution, or they might use something like an S curve if they want a non-linear response.

Now, suppose (for example) you have sensor A. Sensor A loses signal in noise at 100 light units, and saturates at 1000 light units. The camera manufacturer uses filters to set the black point at 150 units so that your blacks are really black and not noise. Then they use a filter to set the white point at 900 units to eliminate the non-linear response near saturation. That gives them a nice linear response between their black and white points to work with. Your effective dynamic range for the sensor is 900 units, and for the camera 750 units.

But, someone else produces sensor B with superior technology, that has less noise and saturates at a higher level. Let us say that sensor B loses signal at 50 units, and saturates at 1500 units. The camera manufacturer sets the black point at 80 units, and the white point at 1350 units. So, For sensor B the effective dynamic range of the sensor is 1450 units, and for the camera 1270 units.

Camera A might describe that 750 unit dynamic range using 12 bits, while camera B might describe its 1270 unit dynamic range in 8 bits. Note that Camera B still has a larger dynamic range, even though it is using fewer bits to describe it. So, number of bits is NOT dynamic range.

Here is another scenario. Lets assume some new company decides to build an enthusiast camera. Lets call the company BlueSmoke. The folks at BlueSmoke take the everyday sensor A (note, not the state of the art sensor B) and they decide to shift the black and white point filters outside the normal specification. The black point filter is dropped down to 50 units, while the white point filter is raised to 1000 units. So now they have a camera with a dynamic range of 950 units. You might think this is great, you have more dynamic range than other cameras, BUT, this camera will have noise near the black point, and distorted color near the white point (since the non-linearity of the response curves differs for different color channels as saturation is approached). They will say, this is not ideal, but, since we are selling to enthusiasts, they will fix these things (which the camera normally does for you) in post. They will say, now we can see detail in blacks (even thought there is noise) and whiter than whites (even though this screws with color balance), but you will have to do (a lot) of fiddling and modification.

And that is fine for people who are skilled at doing this sort of thing. Unfortunately most of the people who buy BlueSmoke's camera are not in this group, and worse, don't have a clue. And consequently you will end up with a lot of really bad footage from BlueSmoke's camera.

As I see it, this is basically the situation we are facing with respect to older and newer cameras on the market under discussion today.
post #90 of 170
Tugela, great explanation! smile.gif
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Camcorders
AVS › AVS Forum › Other Areas of Interest › Camcorders › Haters gonna hate: RAW videos in your pocket!!!!