Paul Vicente
Well-known member
Of course, but in order to silence the critics. I for one think the footage looked really, really good.
The Footage from Leo's shoot...
- Thread starter Ticheli
- Start date
- Status
Of course, but in order to silence the critics. I for one think the footage looked really, really good.
hscully
Veteran
This is a great discussion. I have read every post in this thread and it confirms a nagging concern of mine. One of our projects, in the spring is a web series. Regarding this test, I don't see egregious levels of noise beyond the ISO 800 shot and I trust the experience of the folks doing this test. I need to because I do not have your years of experience. Professionals are disagreeing as to why this test footage looks noisy on various monitors, viewed through either computer video cards or other output devices. We will have no control over how our series will present to our viewers because their rigs will be much more poorly controlled. I also wonder about what's best for streaming on devices such as Roku. What I see of this test footage on that with a 1080p LCD Panasonic Viera looks fine, much like the the better of the two stills of the girl in the black dress. I see no noise in the dress at all.
So I'm following closely for best practices regarding camera settings, compression settings and formats for streaming on the web. Also I'd like to learn more about how best to calibrate a desktop monitor. Thanks, Barry for the short tutorial on PLUGE. I can follow that. I use Vimeo to run trailers, etc. and I have seen lots of variation in quality from my desktop to what's running there so I would really appreciate any guidance on what's best for mitigating the issues raised here.
Thanks for the hard work on the test and thanks for the hugely useful conversation.
So I'm following closely for best practices regarding camera settings, compression settings and formats for streaming on the web. Also I'd like to learn more about how best to calibrate a desktop monitor. Thanks, Barry for the short tutorial on PLUGE. I can follow that. I use Vimeo to run trailers, etc. and I have seen lots of variation in quality from my desktop to what's running there so I would really appreciate any guidance on what's best for mitigating the issues raised here.
Thanks for the hard work on the test and thanks for the hugely useful conversation.
Last edited:
Justyn
Veteran
It's good that things are becoming more civil.... but when comparing this camera to the 7D is kind of like pigs to silk purses. I"ve stopped shooting with the 7D simply because it's s super soft camera that doesn't resolve as much detail it seems as even my HMC150. When it was 7D footage alone it was fine cause it didn't have anything to compare itself to, but when it's on the timeline with the 150 and even worse the 40, it just looks soft and slightly out of focus even when the focus was spot on. Now with tests coming back saying that the GH2 is resolving twice as much detail as the 7D, I would expect the 100 to be on par with that...
But the biggest reason this camera shouldn't be even considered is the PROCESS to which these two are worked. The 7D is a PITA to work with and compromises all around the game. Even when I had it at my disposal in my kit bag the 150 was grabbed 100 times more often cause it's a joy to shoot with and I know what I'm going to get with it. The 7D is a craps shoot and sometimes it's stunning but othertimes it's just garbage...
This of course is my subjective judgement but also the judgement of a lot of professionals that I know.. Especially the audio guys. They are happy to be moving back to a proper video camera and I for one will be too. Sure I'm getting the GH2 but again that's a B camera and if I'm going to grab one thing to shoot with it's going to be the AF100...
One final point.. I was in on all the noise stuff with the HVX and the rants and pissing and moaning about it.. but once I got the camera and worked it's quirks I produced with it for 3 solid years both locally, nationally and for broadcast. When a doc I shot showed at Cannes no one said bubkiss about the footage other than it looked good and that the content was compelling... And not to be an elitistis here, but I think we all should have a certificate to indicate we have an escrow account or savings or whatever that we are intended to purchase this camera and that our opinions are because we are vested. I'ts kind of like me ranting about the epic but there's no way in hell the wife is going to let me buy two of them... That's as much as I paid for my house.... Same for the Alexa which I'd love to own, but in reality I'm trying to keep things within my grasp and Panasonic has created a loyal buyer her not only because they deliver but because they put a huge amount of trust and faith in Barry... and also because Jan is a huge part of this website... I came on here cause I was going to buy the HVX and this was the site with the most info and creating the most buzz, just as it is now...
So pony up or don't but who really cares if you do or dont'. Who cares if I do? Nobody other than those that know me or those that will benefit from the footage I post here and my opinions... After watching public speaking on HBO, which I highly respect, I like the idea that not everyone should be a filmmaker.. She was talking about writing but it applies here. Or more appropriately, not that everyone's opinnion should matter on the same level. Being an elitist or not, but that's the way of the world and the way of artist. We all have a pedigree of some sort but of course we are all crying out in the wilderness and it's just opinions....
It's like the funny idea that other people's eyeballs don't work as well as yours or that somehow it's tained because it does. Well keep it up Barry and all and I for one will be grateful for your efforts.
But the biggest reason this camera shouldn't be even considered is the PROCESS to which these two are worked. The 7D is a PITA to work with and compromises all around the game. Even when I had it at my disposal in my kit bag the 150 was grabbed 100 times more often cause it's a joy to shoot with and I know what I'm going to get with it. The 7D is a craps shoot and sometimes it's stunning but othertimes it's just garbage...
This of course is my subjective judgement but also the judgement of a lot of professionals that I know.. Especially the audio guys. They are happy to be moving back to a proper video camera and I for one will be too. Sure I'm getting the GH2 but again that's a B camera and if I'm going to grab one thing to shoot with it's going to be the AF100...
One final point.. I was in on all the noise stuff with the HVX and the rants and pissing and moaning about it.. but once I got the camera and worked it's quirks I produced with it for 3 solid years both locally, nationally and for broadcast. When a doc I shot showed at Cannes no one said bubkiss about the footage other than it looked good and that the content was compelling... And not to be an elitistis here, but I think we all should have a certificate to indicate we have an escrow account or savings or whatever that we are intended to purchase this camera and that our opinions are because we are vested. I'ts kind of like me ranting about the epic but there's no way in hell the wife is going to let me buy two of them... That's as much as I paid for my house.... Same for the Alexa which I'd love to own, but in reality I'm trying to keep things within my grasp and Panasonic has created a loyal buyer her not only because they deliver but because they put a huge amount of trust and faith in Barry... and also because Jan is a huge part of this website... I came on here cause I was going to buy the HVX and this was the site with the most info and creating the most buzz, just as it is now...
So pony up or don't but who really cares if you do or dont'. Who cares if I do? Nobody other than those that know me or those that will benefit from the footage I post here and my opinions... After watching public speaking on HBO, which I highly respect, I like the idea that not everyone should be a filmmaker.. She was talking about writing but it applies here. Or more appropriately, not that everyone's opinnion should matter on the same level. Being an elitist or not, but that's the way of the world and the way of artist. We all have a pedigree of some sort but of course we are all crying out in the wilderness and it's just opinions....
It's like the funny idea that other people's eyeballs don't work as well as yours or that somehow it's tained because it does. Well keep it up Barry and all and I for one will be grateful for your efforts.
Brandon Rice
Distributed Producer
I just viewed the vimeo clip on my normal Dell monitor here, and it looks FANTASTIC! Anyone who says there is "horrible noise" in this footage is CRAZY. 
Looks brilliant, thanks for posting it!
Looks brilliant, thanks for posting it!
I just viewed the vimeo clip on my normal Dell monitor here, and it looks FANTASTIC! Anyone who says there is "horrible noise" in this footage is CRAZY.
Looks brilliant, thanks for posting it!
OK, guess that settles it!. Jesus.
Melvin Harris
Black Flag Member
Well, I did the opposite... I watched it on shyte computer monitors in a middle school and a shyte tv after further compressing it and you know what, I didn't care about the noise.
Maybe that counts for something.
I didn't care.
Harry Potter 7.1 had a number of focus and resolution issues among other things, and not to mention the mask that was not dealt with over a characters face... yes, a circular mask around one of the death eaters face in the Malfoy house during the liberation scene. A mask. Yes. Both me and my wife saw it. Yes. A mask.
But...
I didn't ask for my money back, nor did I write angry letters to anyone saying that it is my right to bring the errors to their attention because my opinion matters. I observed it because I am cursed with knowledge, knowledge that keeps me in a "crystal tower," or more aptly a dungeon made of pixels and glass, but others do not share my affliction. Other people can enjoy content and look past certain shortcomings.
Maybe to commercial clients it matters, and that absolutely should be addressed by none other than the intrepid DVXuser membership, but I guarantee to people who watch content for content, and not noise levels and moire and aliasing and all of that other stuff... they don't care.
The Signal
Collateral
Irreversible (Good Lord, that movie- whew)
28 Days Later
Open Water
Manic
Paranormal Activity
Crank 2
And for good measure here is Roger Ebert's review of Paranormal 1.
http://rogerebert.suntimes.com/apps/pbcs.dll/article?AID=/20091007/REVIEWS/910089996
He in no way mentions noise, low light performance, or that another camera would have been better to reduce artifacting or all of that dribble.
They don't care.
It's the experience.
Please do not forget the experience that some of us are trying to create.
So by all means, figure it out.
However, I think it looks great. My people thing it looks great. They buy movies. I make movies. Ergo- sign me up for the AF100 train (noise and all).
Maybe that counts for something.
I didn't care.
Harry Potter 7.1 had a number of focus and resolution issues among other things, and not to mention the mask that was not dealt with over a characters face... yes, a circular mask around one of the death eaters face in the Malfoy house during the liberation scene. A mask. Yes. Both me and my wife saw it. Yes. A mask.
But...
I didn't ask for my money back, nor did I write angry letters to anyone saying that it is my right to bring the errors to their attention because my opinion matters. I observed it because I am cursed with knowledge, knowledge that keeps me in a "crystal tower," or more aptly a dungeon made of pixels and glass, but others do not share my affliction. Other people can enjoy content and look past certain shortcomings.
Maybe to commercial clients it matters, and that absolutely should be addressed by none other than the intrepid DVXuser membership, but I guarantee to people who watch content for content, and not noise levels and moire and aliasing and all of that other stuff... they don't care.
The Signal
Collateral
Irreversible (Good Lord, that movie- whew)
28 Days Later
Open Water
Manic
Paranormal Activity
Crank 2
And for good measure here is Roger Ebert's review of Paranormal 1.
http://rogerebert.suntimes.com/apps/pbcs.dll/article?AID=/20091007/REVIEWS/910089996
He in no way mentions noise, low light performance, or that another camera would have been better to reduce artifacting or all of that dribble.
They don't care.
It's the experience.
Please do not forget the experience that some of us are trying to create.
So by all means, figure it out.
However, I think it looks great. My people thing it looks great. They buy movies. I make movies. Ergo- sign me up for the AF100 train (noise and all).
Justyn
Veteran
Jesus does indeed save, but I don't think he's complaining too much about the noise. Now he might be a bit pissed at Mel Gibson or maybe he's thinking of 5K for 5k, but somehow I don't think so...
Brandon Rice
Distributed Producer
Great post, content is king. I see "noise" in nearly every film I watch, at home or in the theater, but as you said, it doesn't bother me or matter. The content is always king.
I've only viewed this on my MacBook Pro, but I can see noise in the blacks on the first model and in the candle scene, but I've got to say that I was looking for it and it was at an entirely acceptable level. I would go as far as to say it was minuscule. What excited me was how good the variable frame footage looked. So many people have been begging for 1080P 60 VFR and now they get it and it looks great. I can't wait to be in a position to order this camera.
Just to add to the fire I saw a good bit of noise as well. Watched the mov file with QT on a Dell monitor that probably isnt calibrated. I figured it was just my monitor on all but two shots. The noise on the candle shots I figured was just what it was because of the low lighting. The one "unusable" shot for me is the first shot of the model in the black dress. The noise on her dress was dancing around so much i thought it was distracting. I wonder, was her dress a flat black or did it have a pattern?
Also I wasnt looking for noise, in fact i saw the stream before anyone commented and later downloaded it before seeing people talk about the noise.
Also I wasnt looking for noise, in fact i saw the stream before anyone commented and later downloaded it before seeing people talk about the noise.
derbrocks09
Well-known member
I saw the same thing. But here is the point. IT'S NOT THE CAMERA EVERYONE!
There is something weird going on with the compression work flow or something. I later watched the same clip on my laptop and it looked 100 times better.
So yes this is a very interesting thread because there is some pitfall in the work flow at work here. I find it very interesting and I can't wait until the mystery is solved.
But it IS NOT the camera. Clearly those who have used it and viewed the images live know what they are talking about. Also for those who are not seeing wicked noise quit complaining about those who have been reporting it. It's a serious work flow problem that we have just discovered IMO. It's NOT the camera. But still we should try and figure out what the hell is going on here.
Just my 2 cents.
nyvz
Veteran
Good to see a well-shot af100 test video for once. For those of you seeing excessive noise and comparing to 7D footage, keep in mind:
1. Quicktime h.264 decoding has always had gamma shift issues, and they have never definitively been fixed
2. Quicktime X, Quicktime 7, and FCP always seem displaye and decoded different gammas, at least I find this to be true for h.264 movs from a 7D, and none of them quite display it correctly like CS5 does
3. I've had many problems with incorrect decoding of P2 DVCProHD HVX footage in windows, producing visibly noisier blacks and excessive blocking which look similar to the left image in the side-by-side wmpc-quicktime picture posted in this forum, so this may somehow be connected
4. The Canon 7D footage posted here as comparison shows clear FCP workflow issues that are very common, significantly increasing the contrast and saturation of the footage, so its black level and noise content is not to be trusted
To anyone using this footage to judge the noise characteristics of the AF100, unless you are positive your workflow and monitoring path is not susceptible to these or any other issues, I think it would be best that you work on overcoming in your system before making any judgements about this or any other camera.
While I am worried about noise in the AF100 and I expect it to be just a bit noisier than a 5D/7D but with more finely detailed noise, I think the much bigger problem is latitude and gamma curves. Why are there no test videos showing cinegammas? Are they not implemented yet, not to be seen until production models are released? Who in their right mind uses anything but cine-gamma? This test footage along with just about every other uses other gamma curves that have way too much contrast and saturation and hard white clipping for any kind of consideration for digital cinema applications in my opinion. Just look at the over-exposure portion of this test, non-cine gammas are pretty much unuseable unless you happen to be shooting high-key material and skintone exposure is always nailed right at 50-70ire.
1. Quicktime h.264 decoding has always had gamma shift issues, and they have never definitively been fixed
2. Quicktime X, Quicktime 7, and FCP always seem displaye and decoded different gammas, at least I find this to be true for h.264 movs from a 7D, and none of them quite display it correctly like CS5 does
3. I've had many problems with incorrect decoding of P2 DVCProHD HVX footage in windows, producing visibly noisier blacks and excessive blocking which look similar to the left image in the side-by-side wmpc-quicktime picture posted in this forum, so this may somehow be connected
4. The Canon 7D footage posted here as comparison shows clear FCP workflow issues that are very common, significantly increasing the contrast and saturation of the footage, so its black level and noise content is not to be trusted
To anyone using this footage to judge the noise characteristics of the AF100, unless you are positive your workflow and monitoring path is not susceptible to these or any other issues, I think it would be best that you work on overcoming in your system before making any judgements about this or any other camera.
While I am worried about noise in the AF100 and I expect it to be just a bit noisier than a 5D/7D but with more finely detailed noise, I think the much bigger problem is latitude and gamma curves. Why are there no test videos showing cinegammas? Are they not implemented yet, not to be seen until production models are released? Who in their right mind uses anything but cine-gamma? This test footage along with just about every other uses other gamma curves that have way too much contrast and saturation and hard white clipping for any kind of consideration for digital cinema applications in my opinion. Just look at the over-exposure portion of this test, non-cine gammas are pretty much unuseable unless you happen to be shooting high-key material and skintone exposure is always nailed right at 50-70ire.
Barry_Green
Moderator
Leo's quote, while we were there: "Oh, I totally believe 10 and a half stops. Easily."
Barry_Green
Moderator
This is the experience I've observed every time I've seen anyone who's looked at the footage on a proper display. Hence the exasperation when folks go complaining like this -- it's like there's totally two different worlds out there. It is, truly, enough to drive someone mad.
D10S
Well-known member
Barry I understand the frustration. I think a lot of people on here A) like to find fault B) don't make a living from their camera. I am getting the AF100. I will make well over 6 figures from it. I did so with the HVX200 despite its faults. The camera is what you make it. My clients were very happy with my images before this camera, they will be happy with my images after I have it. I can rent a RED or something else when I need that quality.
Michael Carter
Veteran
Barry, for months now I've been agape at the forum response to the AF100 - no matter how many times it's been said, we're still getting "why pay $4k for a Gh2" and "AVCHD aliasing sucks, I know this from my 7D" and "no way this will be a RED killer" (as if that's the market for the AF100) and on & on. Seemed clear to me from the first released specs that this could be a no-brainer for people in many markets, after hearing from initial users like yourself I feel it indeed is a no-brainer, and I expect it to have an impact that's something like the original DVX did.
Watching DVX user grow to thousands of users explains some of this - the forums are beginning to feel like walking into a room with a bunch of kids who could finally afford a T2i or GH2 (hey, I do a lot of work with a T2i) and are not liking the "DSLR Killer" comments regarding the AF100. But if creative directors are asking for 7D/5D now, I'll bet you a good bottle of scotch many will be asking for the Af100 and I won't be surprised if there's a "5D was sooooo last week" attitude before long among more tech-savvy clients.
My hat is really off to you for your patience in responding these and other topics, seemingly endlessly. I dunno why you're so nice to us.
Watching DVX user grow to thousands of users explains some of this - the forums are beginning to feel like walking into a room with a bunch of kids who could finally afford a T2i or GH2 (hey, I do a lot of work with a T2i) and are not liking the "DSLR Killer" comments regarding the AF100. But if creative directors are asking for 7D/5D now, I'll bet you a good bottle of scotch many will be asking for the Af100 and I won't be surprised if there's a "5D was sooooo last week" attitude before long among more tech-savvy clients.
My hat is really off to you for your patience in responding these and other topics, seemingly endlessly. I dunno why you're so nice to us.
Barry_Green
Moderator
Regarding monitors, variations in monitor setups, lousy monitors, and web streaming:
It is simply not possible to go chasing after every conceivable misalignment a consumer might do. Furthermore, I would argue that it's not the content creator's responsibility to do that. It is the content creator's responsibility to deliver content that looks excellent on a properly-calibrated, properly-configured monitor. If the footage looks different on someone else's un-calibrated monitor, that's their fault. You cannot possibly anticipate, judge, grade for, and re-grade for, a moving target like that. The reasons that color bars and calibrations exist, is because there can only be one possible interpretation of "good" -- and anyone's monitor that varies from that, will degrade the presentation of your image. We can't educate everyone in the world, nor can we accomodate every possible way that they screw up.
There are people out there with 6-bit LCD screens. On those screens, 8-bit or 10-bit or 12-bit footage is going to show horrible banding. Should we light for a 4-stop contrast ratio so that our footage won't show banding on their screens? No! If the footage looks lousy on their lousy screens, the onus is on them to upgrade their screens! And yes, I said it, and I mean it. The only way to accomodate the lousy screen, is to degrade the footage on everyone's screens. Is that really what you want?
I could probably go in post and crush this footage so that it looks absolutely noiseless on every computer monitor in existence. Would that make you happy? It shouldn't, because it would mean taking 10.5 stops of dynamic range and scrunching it down to maybe 4 or 5 stops. It would mean taking 256 shades of color per channel, and reducing it to maybe 64. That isn't the solution.
Disney/Pixar faced this same problem, because consumers have horribly misadjusted televisions out there. Their solution was to include the THX Optimizer program on their DVDs. As in, if the consumer's television is lousy, it's their responsibility to fix it.
If you recorded a symphony on a 96KHz 24-bit top-of-the-line audio recorder, it's probably not going to sound that brilliant on $2.99 Wal-Mart earbud headphones. Should we compress the dynamic range down to the point where the earbuds can reproduce it? If you do, you'll probably destroy at least 75% of the dynamic range of that concert. Is that a wise course of action? I rebel at the notion.
If your edited, finished production looks excellent on a high-quality, properly-calibrated display, then you've done your job. If someone has an NTSC TV with the tint so far off that The Incredible Hulk looks blue instead of green -- are you supposed to color correct the Hulk? Of course not. You shot him green, you edited him green, he's green on your properly-set-up display. The video recorded him as green. He's green. If the TV makes him look blue, then fix the TV, not the footage.
It is simply not possible to go chasing after every conceivable misalignment a consumer might do. Furthermore, I would argue that it's not the content creator's responsibility to do that. It is the content creator's responsibility to deliver content that looks excellent on a properly-calibrated, properly-configured monitor. If the footage looks different on someone else's un-calibrated monitor, that's their fault. You cannot possibly anticipate, judge, grade for, and re-grade for, a moving target like that. The reasons that color bars and calibrations exist, is because there can only be one possible interpretation of "good" -- and anyone's monitor that varies from that, will degrade the presentation of your image. We can't educate everyone in the world, nor can we accomodate every possible way that they screw up.
There are people out there with 6-bit LCD screens. On those screens, 8-bit or 10-bit or 12-bit footage is going to show horrible banding. Should we light for a 4-stop contrast ratio so that our footage won't show banding on their screens? No! If the footage looks lousy on their lousy screens, the onus is on them to upgrade their screens! And yes, I said it, and I mean it. The only way to accomodate the lousy screen, is to degrade the footage on everyone's screens. Is that really what you want?
I could probably go in post and crush this footage so that it looks absolutely noiseless on every computer monitor in existence. Would that make you happy? It shouldn't, because it would mean taking 10.5 stops of dynamic range and scrunching it down to maybe 4 or 5 stops. It would mean taking 256 shades of color per channel, and reducing it to maybe 64. That isn't the solution.
Disney/Pixar faced this same problem, because consumers have horribly misadjusted televisions out there. Their solution was to include the THX Optimizer program on their DVDs. As in, if the consumer's television is lousy, it's their responsibility to fix it.
If you recorded a symphony on a 96KHz 24-bit top-of-the-line audio recorder, it's probably not going to sound that brilliant on $2.99 Wal-Mart earbud headphones. Should we compress the dynamic range down to the point where the earbuds can reproduce it? If you do, you'll probably destroy at least 75% of the dynamic range of that concert. Is that a wise course of action? I rebel at the notion.
If your edited, finished production looks excellent on a high-quality, properly-calibrated display, then you've done your job. If someone has an NTSC TV with the tint so far off that The Incredible Hulk looks blue instead of green -- are you supposed to color correct the Hulk? Of course not. You shot him green, you edited him green, he's green on your properly-set-up display. The video recorded him as green. He's green. If the TV makes him look blue, then fix the TV, not the footage.
vicrattlehead
Well-known member
There are two different worlds: Professionals and hobbyists. And we're getting quite an earful from the hobbyists these days.
Barry and D10S you are both wrong. The HVX is too noisy for professional use. It will be the end of Panasonic and major networks are only buying and using JVCs. Let's see, it's 2005 now you will both be proven wrong by 2010. No one will be using the HVX then. Stop telling us to calibrate our monitors as if you have professional experience and the problem could be that simple, or that we're just not aware of the normal amount of noise in all digital video.
Oh wait. Sorry. Wrong thread and I accidentally time travelled again. I just get confused with such similar arguments / comments / reactions to people looking at this footage on computer monitors to what they said about the HVX in 2005
Too bad Barry doesn't have a track record being proven 100% right over this exact same issue on Panny's HVX with 5 years of data / hindsight. Oh wait, never mind, he does:
http://www.dvxuser.com/V6/showthrea...o-s-shoot...&p=2173961&viewfull=1#post2173961
Oh wait. Sorry. Wrong thread and I accidentally time travelled again. I just get confused with such similar arguments / comments / reactions to people looking at this footage on computer monitors to what they said about the HVX in 2005
Too bad Barry doesn't have a track record being proven 100% right over this exact same issue on Panny's HVX with 5 years of data / hindsight. Oh wait, never mind, he does:
http://www.dvxuser.com/V6/showthrea...o-s-shoot...&p=2173961&viewfull=1#post2173961
- Status