O.M.G. Are you serious?!?

[Roberto Lequeux]

Again I am no guru in any sort of way... but if I understand correctly, then this couldn't be too far from the truth:

~78%

So if I understand correctly, 5K x 78% = 4K?

Then with 2k, which is still tiny bit bigger than 1080p, you are making your image 4 times smaller. You will have four debayered 12-bit pixels of information, 16-bit if you go with FF35 windowed to S35 5K, that get turned into just one of your pixels... so you have 1/4th the possibilities of the amount of noise in any given frame to make it to your 2k output. Or go bigger, use a FF35 or even the 645 and stop down further if you are trying to avoid shallow DoF.

That doesn't sound like Fan-boy ethers to me.

 

[GlennChan]

LOL.. but are you seriously going to consider that everyone sits at that distance? And do you only consider that we are talking about TV broadcast? There are other things out there, like Blu-Ray, 1080p on-demand type services are popping up last time I heard.

Regarding Fox's test results, I wouldn't be surprised. Try this test:

Take an image into Photoshop. Especially Red images... try the milk girls stills on red.com
Duplicate the image onto a new layer.
Size the image to 50% using Bicubic sharper. (This method seems to implicitly add edge sharpening, which makes things sharper.)
Blow it back up to the original size.
Turn the layer on/off to compare between the two images.

What I see is that the top layer looks sharper, even though its resolution is roughly half that of the original.

And there's a way to tell the two apart: try to pick out small detail like the little hairs we have on our skin, try to read small text, pores on people's faces, etc. In the lower resolution image you can't really make out that stuff.
I'm guessing that most people don't know how to look for this... which would explain why people sometimes debate the resolution of a camera (e.g. is Red 4K? Is the HVX a 1920x1080 camera? etc.). I think a lot of people are fooled by sharpness/edge enhancing tricks, that's why there are these debates. And part of it might be dogma, e.g. film=good, bayer=bad, etc.

1b- It also depends on how the test was performed. I am guessing they are displaying images on a broadcast CRT, which tends to be somewhat 'blurry'. With fixed-pixel display technologies like LCD, plasma, etc. then any rescaling of the image will introduce scaling artifacts. This can make it easier to spot the difference.
http://www.glennchan.info/broadcast-monitors/scaling-artifacts/scaling-artifacts.htm

(The example in my article is a zone plate test pattern, which is extreme and rarely crops up in real productions. But you can see aliasing as crawling on edges and it sometimes shows up in fabrics and other fine detail.)

2- There is a phenomenon called hyperacuity...
http://www.michaelbach.de/ot/lum_hyperacuity/index.html

Technically resolution needs to be much higher than what we have now if we want to make the resolution part completely seamless. (Or, project film.)

 

[M Most]

LOL.. but are you seriously going to consider that everyone sits at that distance?

Not everyone. But lots of people do.

I'd like to know how those tests that Fox performed were carried out... was it a test to see if they should opt for 1080i or 720p out of their cameras? Which cameras? Newsroom cameras? Sitcom cameras? What cameras? What sensor? Cause unless they were native 1920x1080 cameras with detail filled frames for comparison... something that didn't ruin the shot too much with bad compression...

The tests had nothing whatsoever to do with cameras, it had to do with formats and displays. Most of the material they used was from Fox prime time series, sourced from both film and video. As far as compression goes, one of their reasons for going with 720p vs. 1080i was that 720p at its broadcast frame rate of 60fps requires less compression than 1080i/60. But that's the only compression issue they were looking at. These tests weren't conducted yesterday, they were done a number of years ago. But the results are still valid.

And what about DNR?! What about DoF? What about the ability to do a TON more in post with RAW images?????

You keep talking about cameras. I'm talking about displays, which is what the original statement that was quoted was also talking about.

Again... sorry if I sound upset, but I am just sick of people trying to knock us down.

I don't know where you get that idea. All I said was that from a viewers' standpoint, what was said in the original quote had validity. I didn't "knock down" anybody or anything.

 

[Dj Joofa]

And there's a way to tell the two apart: try to pick out small detail like the little hairs we have on our skin, try to read small text, pores on people's faces, etc. In the lower resolution image you can't really make out that stuff.

It is inherently dangerous to base conclusions on a software result. Several well-known software implementations have made mistakes in the implementations. (For. e.g. it appears a very popular image processing software, VirtualDub, has made a few mistakes in its implementation of resampling. Similarly I have reason to believe that the very popular software, ImageMagick, has made some similar mistakes in understanding the problem at least.)

You can't just assume if a piece of software is from a well-known vendor, it must always be free of errors -- not errors in programming, but errors in the fundamental understanding of many algorithms.

 

[Dylan Reeve]

FWIW the EBU concluded after much study that 720p is preferable to 1080i (however 1080p is better than both).

There's a few whitepapers about it.

 

[GlennChan]

It is inherently dangerous to base conclusions on a software result.

Hi Joofa,
I use the example of Photoshop because I am familiar with it and somewhat/fairly common. Does it have the best resampling algorithms? No.

But I don't think you would need the best resampling algorithms to demonstrate my point. If Photoshop's algorithms aren't that good and they can make things look sharper than the original, then better algorithms can only do better.

If somebody has Photoshop on their system, then they can quickly illustrate/check the concepts for themselves.

2- I'm not so sure that there is a right answer to resampling. Our models of human vision aren't really there and the textbook answer for resampling looks far inferior to stuff like this:
http://www.alienskin.com/blowup/blowup_examples.aspx
(I've sent you a link to that product before.)

 

[Roberto Lequeux]

Not everyone. But lots of people do.



The tests had nothing whatsoever to do with cameras, it had to do with formats and displays. Most of the material they used was from Fox prime time series, sourced from both film and video. As far as compression goes, one of their reasons for going with 720p vs. 1080i was that 720p at its broadcast frame rate of 60fps requires less compression than 1080i/60. But that's the only compression issue they were looking at. These tests weren't conducted yesterday, they were done a number of years ago. But the results are still valid.



You keep talking about cameras. I'm talking about displays, which is what the original statement that was quoted was also talking about.



I don't know where you get that idea. All I said was that from a viewers' standpoint, what was said in the original quote had validity. I didn't "knock down" anybody or anything.

Mmost... the e-mail was about the camera, and I assumed you were on topic.

DNR matters, period. Regardless of what you are going to view this on... even if it were half 480... being able to capture your environment matters on any screen because you either got it on the can or not.

I have a winter project I am working on, if I was going to only release it on YouTube I would still require the most DNR I can get or I will have immense holes of clipped info... and all I would be able to do to it in post is to bring that white pixel down, but it moves in a block together with all the other white pixels next to it unless you start drawing in there... I want to do sky replacement, especially on specific scenes to augment the mood... but ground replacement? The snow is certainly an extreme example but it illustrates a point.

A bigger poit to make about DNR is that with a pile of DNR you can save boatloads of money... say that you show up at a location and notice that your camera has enough sensitivity to work without having to raise the exposure... you could have chosen the location specifically because of this... having enough DNR makes this sort of situation more likely... there may be already some big sources in there... you would normally have to bring up the rest to within the stops that you can capture... but if your camera captures the dark areas AND the bright areas, you can have your talent walk from one to the other without fear, and also get to shoot the location the way you intended or however you think it should be... you may end up needing half the crew and lighting package.

Then you go to the editing booth knowing that you can tweak the bejesus out of it cause you have RAW... but continue to be amazed at just how much you can do...

You may only have gotten enough sensitivity cause you went with the 645, or FF35, or because Scarlet S35 was affordable enough for you to go with a sensor that is four times as big as the 2/3" camera you would have otherwise gone with... or the 1/3" + adapter rig that only had 4-5 stops of DNR and needs heaps of light...

More IS better... especially when more SAVES you MORE $$$... or when more saves your ass... or when more means that someone WILL buy your finished project because you can tell him that you can give him the 2k RAW footage so that he can make it look better...

Glenn, I see your point and all I can add to that is that having RAW instead of 4:2:2 stuff only makes your point that much more valid.

FWIW the EBU concluded after much study that 720p is preferable to 1080i (however 1080p is better than both).

There's a few whitepapers about it.

Yea, I remember now.. that's how it works. And that guy said 1080, not 1080i. I assume 1080 to be "full" 1080, and 1080i to be 1080i but that's still an assumption on my part...I guess.. either that or I assume that he considers broadcast TV as the only medium.

Again, don't mean to pick on anything that isn't there... sorry if I did Mmost, but the language on that guy's e-mail was certainly clear otherwise and I couldn't stay quiet about it. IMO a company doing the things that Red is doing deserves clarification and respect that that guy didn't afford. Maybe he just doesn't know better. :innocent:

 

[Dj Joofa]

Hi Joofa,
I use the example of Photoshop because I am familiar with it and somewhat/fairly common ...

Hi Glenn,

Okay, I buy your point.

2- I'm not so sure that there is a right answer to resampling. Our models of human vision aren't really there and the textbook answer for resampling looks far inferior to stuff like this:
http://www.alienskin.com/blowup/blowup_examples.aspx
(I've sent you a link to that product before.)

Thanks for the link again. The results on the link are impressive. However, the website mainly compares to biCubic (perhaps Photoshop has biCubic and that is why). However, there exist better techniques than biCubic. Maybe that website you point is using (a variant) of such techniques. Of course, you won't find them in many textbooks as many textbooks are already dated.

 

[M Most]

DNR matters, period. ...require the most DNR I can get ...make about DNR is that with a pile of DNR you can save boatloads of money... having enough DNR makes this sort of situation more likely...

Warning: I'm going to play terminology police (again).

DNR is a common abbreviation for Digital (or, in the past, Dynamic) Noise Reduction. If you want to talk about dynamic range, call it dynamic range, or maybe DR, but DNR is already taken and means something very specific to myself and anyone else here with a modicum of experience in the video and film industry. It's very confusing and annoying to have to translate it due to people here wanting to appropriate terminology for their own personal use. Just because Jim misuses the term doesn't mean everyone else should as well.

BTW, if the original post you're talking about was where I think it was, it was likely written by Walter Graff. Walter is quite opinionated, but he's also very, very experienced as a shooter, an editor, and as a director/producer. Never assume people don't know what they're talking about just because they happen to say something you disagree with.

 

[Roberto Lequeux]

Warning: I'm going to play terminology police (again).

DNR is a common abbreviation for Digital (or, in the past, Dynamic) Noise Reduction. If you want to talk about dynamic range, call it dynamic range, or maybe DR, but DNR is already taken and means something very specific to myself and anyone else here with a modicum of experience in the video and film industry. It's very confusing and annoying to have to translate it due to people here wanting to appropriate terminology for their own personal use. Just because Jim misuses the term doesn't mean everyone else should as well.

Thanks! I must do that more often that you imagine. It helps to have someone point out where I might be confusing people by using conflicting terminology or acronyms. I certainly was referring to dynamic range and not dynamic noise reduction, as you noticed.

BTW, if the original post you're talking about was where I think it was, it was likely written by Walter Graff. Walter is quite opinionated, but he's also very, very experienced as a shooter, an editor, and as a director/producer. Never assume people don't know what they're talking about just because they happen to say something you disagree with.

I actually never though he didn't know better... that's what got me so upset and got me to post here to blow off some steam.

In fact I think what he said was horse-@#%# BECAUSE he knows better...

It sounded to me like a BS rant trying to knock down a camera that would increase the competition around him. That's just a guess, cause there is little else that could explain suck loads of crap.

If that is right, then again: it should be about the work, the talent, not the gear...

I am sure we'll see him talking in a completely different light about Red's cameras in the near future... but his initial response sound to me of resistance to change. Completely understandable, but inappropriate to say the least.

He does know better, and that is what makes it so bad. I mean WTF... I an a fanboy in most peoples eyes, because I am so exited about Red, not because I am some sort of idiot that buys cameras based on what's in vogue. If people go around talking like he did then clients will automatically ignore me, never look at me... all because nobody called him out. It is the whole Mythbusters theme...

Now don't get me wrong... R1 post is not easy... BUT same goes for film! The problems were never with Red, but with how NEW Red's technology is. There are thousands of people that could do post for R1 out of their houses for you, that know everything about it and how to get it to look best. There are builds that have no issues... at all...

I am so sorry to continue ranting about this, on and on... but if we are going to get clients and production houses to take us seriously then we need to make sure that Red is not bashed down anymore... this BS needs to stop.

The good thing is that soon enough R1 owners won't be alone fighting this fight... reinforcements are on their way boys. :innocent:

 

[Charles Angus]

I am often impressed and often unimpressed by the level of discussion at this unnamed other forum.

I also read the OP's quoted message, and it would fall into unimpressed. Why do they shoot SDTV on 35mm then? Why not shoot it on tube cams from the 70's, they have sufficient resolution for highly compressed broadcast.

Overkill always seems like a silly argument. Overkill as much as you can afford to (in time and money).

 

[Roberto Lequeux]

I am often impressed and often unimpressed by the level of discussion at this unnamed other forum.

Yes, to the point that it starts to sound like a burocratic bull-@#$! forum at times... all about deciding what is in Voge and what will be cool... unfortunately that works at a certain level. If enough "big" duded say this is in and that is out then clients that don't know better follow those trends... of course this is a bit exaggerated or speaks only of a portion of the industry... the hollywood or brainless part.

Again, I don't mean to offend anyone... just blowing off steam.

I want to shoot my own shit and just try to sell it after the fact, always... that would be my perfect life.

 

[David Birdy]

Yes, to the point that it starts to sound like a burocratic bull-@#$! forum at times...


__________________________________________________________________
Sounds like Hollyweird to me!

No offence.. you must know what I mean!



Great points here! I guess one point to be made is that viewers of succesful
Shows will watch them on many formats, many different distances and many different screen sizes.

Start with the most resolution possible, then for distribution send what ever, format your client requests!

I was in charge of a movie release to HD digital theaters. Much of The original
footage was Betacam and 3/4" mixed in with film, 720P & 1080i of Racing event, (lots of motion etc).

I made to call to use the "best" upconverter available in 2006,) Alchemist
Dave
The movie premier in Charlotte from the SRW machine was fantastic.
A new movie is coming out and they decided to forgo the cost of the (Alchemist) and up convert in the Avid....will there be a noticeable resolution diffence in the final products if you compare them???? Time will tell.....


BTW
I can see the difference between 1080P & 720P on a 1080P monitor...



Dave

 

[M Most]

Yes, to the point that it starts to sound like a burocratic bull-@#$! forum at times... all about deciding what is in Voge and what will be cool... unfortunately that works at a certain level. If enough "big" duded say this is in and that is out then clients that don't know better follow those trends... of course this is a bit exaggerated or speaks only of a portion of the industry... the hollywood or brainless part.

Again, I don't mean to offend anyone...

Well, you have offended just about anyone who has spent their life in this industry, especially those of us who have spent most of it in "Hollywood" - aka, Los Angeles, aka the center of the industry, whether one likes it or not - and who do not consider ourselves either bureaucratic, trendy, or brainless.

It is both the blessing and the curse of those who are less experienced in something to question the abilities, knowledge, and validity of those who are more experienced, particularly if they are successful. It is a blessing because naivety allows one to ask questions that are not limited by knowledge, to think outside the box. It is a curse because it also invokes a degree of jealousy and misunderstanding where there can just as easily be education. Experienced professionals of the type that you seem to be referring to are unconcerned about the things you're accusing them of being overly concerned with. Trendiness, what is "cool," holding back younger talent - an experienced pro craves not these things (sorry, George). It's almost always about the ability of the tools to do the job under given circumstances, both foreseen and unforeseen. When you've done something for a long time, you tend to know a lot more about what works and what doesn't, particularly when you're dealing with physical things. I think if you read the posts to which you're referring a bit closer, you'll find that the questions raised are primarily about the practicalities of physical designs and the practicality of post solutions that allow one to meet the rather brutal turnarounds and quality demands that are common in the high end of the industry, but almost unknown in the indie community. Instead of dismissing those who have been there before you, and been there for quite some time, as dinosaurs, you might want to consider that eventually - if you're talented, hard working, and to some degree lucky - you'll be one of them. You'll see things much clearer then than you likely do now. That's not a knock, that's a promise.

 

[Roberto Lequeux]

I completely agree that there are TONS of things that will remain unforseen to me for a long time... I do not have experience on my side and wisdom only comes from experience. As harsh as I am on others I can be on myself, ten fold. Trust me when I say that if I had to shoot something now, with a tight deadline and on a tight budget... I'll be the first to cry for help. The person I cry to will be someone with experience. I'll gladly work with the guy that posted this even... I can take it, and that is the only reason why I allow myself to dish it.

I didn't mean to "offend" wise experienced or otherwise accomplished people... I look up to people like this...I want to be one one day.

Allow my to try to be a bit more clear in my whirlwind of thoughts and developing knowledge... perhaps the issue I have is not only the technical atrocity of his statement... now I DO understand that he may be speaking of other aspects that come through clearly to other experienced people... when you've worked for ever you speak of other intangibles using the same words that people like me still associate with more specific elements.

However... we are already at a point where all that you require to properly edit 2k footage is a dual quad Mac Pro, FC, and the main thing being enough fast hard drive space... if you want to grade in 2k you might need more but for the concerns of the initial statement you might only need to render out to ProRes after having done a first light on something like Red Alert. In turn you get a killer image, all the benefits of Misterium and the original RAW files...

I don't know... maybe you are 100% right and nobody at that level of experience misunderstood that he was talking about being able to shoot, edit and forget... maybe all of the people at that level understood that he was speaking about getting a decent quality for broadcast or DVD, perhaps even blue ray...

...what I do know is that the benefits of shooting Red are immense, that they are here, proven by the R1 latest builds, and that often they will make a big difference in your production budget by raising your post just a bit.

But... maybe there are some gross unforeseen aspects that will hit me over the head... what am I saying... there WILL be... I don't have enough experience to say "maybe" .

However... and this might be the biggest point... perhaps I am not taking into account how much more willing and ready I am to dive into digital... perhaps large RAID enclosures, bit depth, processing power and the sort don't have any effect on my. Maybe that plays a role... now... don't be getting upset again... cause the art goes beyond the technical so nobody should be insulted here... ok? I only trying to get to the bottom of my communicative divide cause if I can't work with accomplished people I won't learn as fast.

Also... perhaps I know the Red crowd better... perhaps I assumed that he knew why we are exited about Red. There is a LARGE group of people that are exited to be able to buy into tools that can deliver a certain potential quality... and these people are more likely to want to work on personal projects. Projects that all of the wise accomplished people already did... ages ago... projects that proved their capabilities... projects we wouldn't get to do without owning... perhaps this is the biggest misscommunication between me and "they".

Finally... I must admit that I too am weary and defensive... defensive of my tools and their perceived validity... perhaps I was a bit jumpy... but there is no way that you could argue that the wording on the e-mail, which I got sent to my personal e-mail for no good reason was already worded in such a defensive way...

...I wish I could speak the short hand dialect that long time pros have... I get it most of the time...

...again, do NOT take offense... nobody should other than the guy that sent me an e-mail saying things such as fanboy and other things that basically meant that I was not being logical.


Just as I have to and DO respect "them" I feel the same should be done for "us"... if I have to understand you you should, in a perfect world, try to understand "us".