Misc. Discussion cut out of M-X Reel thread

[Rick Burnett]

It's not on the original 4k for sure.

Graeme

What would cause the coloring in the Moire pattern? I always thought with my 7D that they were skipping lines before the demosaicing which caused the red/green and blue/green coloring in whites due to it throwing out whole lines of red or blue.

To me it would seem the reduction in resolution would have to occur to the raw data before debayering to cause color in the Moire.

 

[David Rasberry]

The low light performance looks absolutely fantastic, that said, what *really* concerns me is the colorful moire at 00:00:38. This is an issue I have been fighting on the 7D and is just as colorful as the results I get with the 7D.

Thoughts?

OK. So long as he didn't replace anything I wouldn't mind some chroma key tweaking to the color of the sky. What I am looking at there is the detail captured on the sun-lit hill and the detail in the shadow. I am only concerned with dynamic range and I feel I got a good grasp of the lighting there.

Amazing Low-light BUT at 00:19 ( White singlet ) and at 00:20 ( Audi logo) i see a lot of Moirè.

I sold my Canon DSLR for Moirè issues... and even if with M-X this moirè appears more natural is really annoyng .

This moirè is present in raw files too?

Thank you

If you look closely you will see a very fine crosshatch texture on the wheel surface around the logo. There will be moire near the Nyquist limit anytime two fixed high freqeuncy patterns are overlaid, the pixel field and the surface texture in this case. You are seeing the Nyquist limit of 2k, not the camera original.

 

[Rob Ruffo]

I certainly agree, I think someone just needs to fix that render! I only bring it up due to the point there are 3 characteristic sets people seems to fixate on with CMOS sensors (for good reason)

1. Moire / Aliasing
2. Sensitivity (DR, noise floor, etc)
3. Skew / Jello (Rolling shutter)

There are obvious other characteristics as well that are important, I am just mentioning what items seem to be the pain in most peoples sides. The Canon and Nikon DLSRs with video fail 1 and 3 pretty bad. They are very good for 2. Red, theoretically (and I say this because its not in the hands of the consumers just yet) excels at 1, 2 and 3. That said, I still want to see it

Moire is not a CMOS thing, at all. Red has never had particular problems with moire. Moire has to do with a badly tuned low pass filter. DSLRs have lwo pass filters tuned for taking 21MP or whatever stills, so there is nothing to smooth out 2MP video. That and line-skipping create the worst moire of any video camera ever sold to consumers. Can anyone think of any cam in history that gave worse moire than the 5D/7D?

Many CMSO cameras show no moire problems: Red, EX1, any modern DSLR when taking STILLS, etc.

Jello has also been hugely overstated. Most pro CMOS VIDEO CAMS (not still cams with video afterthoughts) do not really have much of a serious problem when used with professional care to proper motion. In real life it's only a PITA when you are using strobe lights or flashes of any kind in your scene. Otherwise you don't see it.

Sensitivity on CMOS can be excellent. CCD is not really better.

5D and 7D are not really video cameras. They have all kinds of problems no video camera engineer would ever let through the door.

The problems you see in the post are most likely due to bad downsizing algorythm choices in post. They should fix this.

CMOS or not means nothing. Its' like saying a French restaurant is just like Arby's because both serve meat and vegetables.

 

[Rick Burnett]

I wasn't trying to imply the Moire was a CMOS issue, just that people are focusing on it since the 1D/5D/7D have such a problem with it and are so very popular at the moment. I understand the physics behind the LPF and the issue that the LPF is designed for the max resolution of a still frame on these DSLR cameras.

If the LPF was designed for 2MP video, as has been explained elsewhere, the image would be much worse due to the line skipping on the 1D/5D/7D (in terms of sharpness).

The real issue with the 1D/5D/7D is that if you could get the raw out, or have downsampling from the full image down to your target resolution without just skipping lines, you'd get a much better result. Unfortunately, most people feel the processors of these cameras just cannot handle the resizing of the full frame, or can the whole frame be unloaded from the sensor fast enough, or third, can it get the raw out through the CF port (doubtful). Hence the reason this camera is so much cheaper than the Red, it is NOT as powerful in the least.

While I absolutely agree the #D are not video cameras per say, I use them as many people do as such and get GREAT results when working within their limitations. I prefer working around the limitations of the 7D instead of using our HVX200 because the results look so much better. Naturally, as I am, many people will make the step from the #D to a Red camera as the costs make it not so prohibitive. That said, when someone shows this footage to someone else, who's used these cheaper cameras, they, like me, will notice that Moire first and get concerned (which I am not now given Graeme's response), but a lot of people might not be as technical as me and instantly get turned off to something that has plagued them relentlessly! First impressions do make a difference.

 

[Rob Ruffo]

I wasn't trying to imply the Moire was a CMOS issue, just that people are focusing on it since the 1D/5D/7D have such a problem with it and are so very popular at the moment. I understand the physics behind the LPF and the issue that the LPF is designed for the max resolution of a still frame on these DSLR cameras.

If the LPF was designed for 2MP video, as has been explained elsewhere, the image would be much worse due to the line skipping on the 1D/5D/7D (in terms of sharpness).

The real issue with the 1D/5D/7D is that if you could get the raw out, or have downsampling from the full image down to your target resolution without just skipping lines, you'd get a much better result. Unfortunately, most people feel the processors of these cameras just cannot handle the resizing of the full frame, or can the whole frame be unloaded from the sensor fast enough, or third, can it get the raw out through the CF port (doubtful). Hence the reason this camera is so much cheaper than the Red, it is NOT as powerful in the least.

While I absolutely agree the #D are not video cameras per say, I use them as many people do as such and get GREAT results when working within their limitations. I prefer working around the limitations of the 7D instead of using our HVX200 because the results look so much better. Naturally, as I am, many people will make the step from the #D to a Red camera as the costs make it not so prohibitive. That said, when someone shows this footage to someone else, who's used these cheaper cameras, they, like me, will notice that Moire first and get concerned (which I am not now given Graeme's response), but a lot of people might not be as technical as me and instantly get turned off to something that has plagued them relentlessly! First impressions do make a difference.

Yes, they need to redo the posts on those clips, but 5D/7D shooters do not get "great results" for 1080p output. They may be proving their ability, they may be learning something but 5D footage does not really look great, even in the extremely limited circumstances where it supposedly shines, like a close-up with blurry background. These cams do weird things with skin pores and eyes brows and other facial features that I for one find highly disturbing.

People also talk about how prohibitive Reds are, but making an indy movie costs a fortune, no matter what. The camera is such a small part of that cost that it's a bit silly to save that "last $15 000" by cutting corners on the least expensive and most important element.

If you can't afford a Red, fine, but at least buy a 35mm adapter. Paired with a good camera , the results are much more acceptable to HD broadcast than 5D/7D's. If you don't believe me, ask a broadcaster.

 

[Rick Burnett]

Yes, they need to redo the posts on those clips, but 5D/7D shooters do not get "great results" for 1080p output. They may be proving their ability, they may be learning something but 5D footage does not really look great, even in the extremely limited circumstances where it supposedly shines, like a close-up with blurry background. These cams do weird things with skin pores and eyes brows and other facial features that I for one find highly disturbing.

People also talk about how prohibitive Reds are, but making an indy movie costs a fortune, no matter what. The camera is such a small part of that cost that it's a bit silly to save that "last $15 000" by cutting corners on the least expensive and most important element.

If you can't afford a Red, fine, but at least buy a 35mm adapter. Paired with a good camera , the results are much more acceptable to HD broadcast than 5D/7D's. If you don't believe me, ask a broadcaster.

I think that is definitely a matter of opinion with regards to the 5D/7D, as I personally find the footage I get from it to look great. Clearly a lot of other people feel this way as well, and yes, we are using the 1080p output (I don't even bother with 720p, it's horrible). I just finished shooting a project that will in fact be broadcast.

I have used other setups with a 35mm adapter and I wasn't happy with the workflow on set or with the camera itself with regards to the focusing screen and everything. It just made more variables to go wrong.

What really changed my mind was looking at the use of the 5D for the intro shots on Saturday Night Live and seeing the quality. Not to mention the plethora of other 5D/7D commercials and videos I've looked at that the quality has been incredible.

Maybe you have a more discriminating eye that I do, but the people who pay me for my work seem to me ecstatic about the results. (Not to mention I am too).

I agree, Indy film making has a million other costs in it, but I never indicated the cost of the Red setup made it prohibitive to purchase. That said, the difference in quality between say $3000 in my 7D and lenses definitely does not have an equivalent in straight video cameras I have seen in terms of cinematic appeal. Granted, I traded one set of headaches for another. And not to mention light sensitivity. There is NO OTHER camera even close in this price point that I can shoot at night with little light and get so little noise in the images.

The point being though is that Red has put together a package (Scarlet Fixed) that to me is absolutely a game changer, because it's at a price point with features that NOTHING I have seen compares too (existing or announced). In that, which I am sure, is that Red is putting these at a price point where a higher volume in sales will recoup their R&D and get them integrated into more locations everywhere.

That said, there are a different set of discriminating eyes and a different set of requirements that are going to drive some of those sales in different market segments. One of those segments is people owning the 5D/7D. Whether you like these cameras or not, there are A LOT of people that do like them. I am sure Red would like to capture that user-base. The fastest way to do that is show them how all their woe's will be gone with their cameras. That's all. Others probably freaked out at first with the Moire because they too are tired of dealing with it elsewhere.

 

[Chris Kenny]

I think that is definitely a matter of opinion with regards to the 5D/7D, as I personally find the footage I get from it to look great. Clearly a lot of other people feel this way as well, and yes, we are using the 1080p output (I don't even bother with 720p, it's horrible). I just finished shooting a project that will in fact be broadcast.

Well, the thing about the vDSLRs is that usually -- the vast majority oft the time, even -- the footage looks great. But every now and then you'll find that you pointed one at some surface with the wrong short of high-frequency detail and it ruined the shot. So you have to be working on a project that's tolerant of this sort of thing -- where the client won't totally freak out if one or two shots are blown, or where reshoots aren't too much of a hassle. Checking shots where you suspect this sort of problem might arise immediately on-set helps, of course.

 

[Rick Burnett]

Well, the thing about the vDSLRs is that usually -- the vast majority oft the time, even -- the footage looks great. But every now and then you'll find that you pointed one at some surface with the wrong short of high-frequency detail and it ruined the shot. So you have to be working on a project that's tolerant of this sort of thing -- where the client won't totally freak out if one or two shots are blown, or where reshoots aren't too much of a hassle. Checking shots where you suspect this sort of problem might arise immediately on-set helps, of course.

Totally agreed, and I shoot a lot more coverage because of this if I see anything that I think might be a problem (especially in urban settings).

It's worse too, because on the camera, there is NO WAY to detect Moire/Aliasing until you dump it off the camera or review it with the HDMI port AFTER recording since during recording the port outputs 480p from what I am told. I'm thinking of getting the DP1 from smallhd.com which is a 9inch true 720p display to use with the 7D and then the SFixed eventually to see if I can at least check the footage before transferring to another viewing source. There is no way to zoom into 1:1 on playing back video on the 5D or 7D.

This is also why my upgrade path will be from the SFixed to the S35 eventually as I like the results with the 7D sized sensor(APS-C) and they aren't very far off from one another.

 

[Roberto Lequeux]

Yea, I keep hoping the European guys will chime in. I didn't mean for you to think I was asking you directly. I am just very curious to hear the T-stop used on the 18mm, hear which where shot on the 18, hopefully get a 4k grab or two! This is the first we've seen of the 18mm I think.

We should all clarify however that all of this microscope analysis comes not from us finding nasty issues with the reel or the footage, but because we are all so :willy_nilly: about what we are seeing that we can't stop asking questions... well, that's me at least.

We might not be commenting on the reel itself because we have learned to expect so much from all things associated with Red, but the production value put into this Reel is very significant. I enjoyed the reel in a big way! Yet here we are, whining about moire and stair stepping even though we know this is a weirdness prone delivery codec. So much so that several people brought up pro-sumer VDSLRs, even though that is like comparing Argentine steak with BBQ flavored Fritos.

It seems, as Jim already said, the only true digital competition will be Arri. I don't see them shooting reels yet either. I am also very aware that for the cost of the rental of one Arri I will be likely to get two R1-X's! That means oodles of awesomeness to me.

Let's not forget the cost:IQ of Red's Primes! Sure they are a little heavy... but they are a very small fraction of the cost of MPs and deliver something that can be compared, which says a lot since MPs are the undisputed king of IQ. Also, it is thanks to Red gear that we can afford to rent MPs for a week, that we can afford to rent a specialty rig for an action sequence, or buy post gear!

Thanks to Red we are also working with negatives! And "unlimited lab budgets"! (if you know how to do your own stuff and made a small investment) Just make sure to finish with a pro cause the only thing that Red can't give us is talent... I heard they are working on AI to be implemented into RedCine-X by 2020 though.

OK, I'll go stfu in my corner now. I got some work to do, and some prayers to say for the LA incentive to bring some work back to the city.

Thank you to all that made this reel!

And thank you to everyone at Red for rocking the digital scene!

 

[Rick Burnett]

Hey, I represent those remarks! First, thank you for posting the 4K stills, very much appreciated and they look absolutely awesome. That is what I wanted to see. I cannot wait to get my hands on a Red, money is actually sitting and waiting.

And like I said, while the SFixed might be functionality in a league of its own, price wise it is not, say, compared to the 1D/5D. For that reason, people from the prosumer market (and admittedly that is me at the moment) are over hear looking at these videos saying "Hmmmm, is it worth upgrading". Of course, I know that answer already myself, but I'm not ready to send this video to all my other prosumer buddies yet until the quality of the final result is as good as it can be.

It's like posting an MP3 of something I recorded with my Apogee Rosetta ADC audio converters I want to send them something that blows them away on ALL levels and leaves nothing to criticism except maybe the artistic relevance which you know there will ALWAYS be complainers

And in all excited honesty, remember when the 5D and 7D came out and all those cool videos started showing up on Vimeo and the like? I CANNOT wait to see all the Scarlet stuff that pops up with people just running the camera through ever imaginable test. You know, like; water falls, carnivals, motorcross, coke commercials, someone's girlfriend Of course, I jest, I went out and shot some of the same things when I got my 7D to see the results first hand. Not to mention how many slow motion videos are gonna be out there!

So again, thanks for all the hard work on pushing the sensor. You're just teasing us more now

I think it goes probably without saying that everyone is just hella excited. May/June just cannot get here fast enough for me

 

[Todd M.]

Yea, I keep hoping the European guys will chime in. I didn't mean for you to think I was asking you directly. I am just very curious to hear the T-stop used on the 18mm, hear which where shot on the 18, hopefully get a 4k grab or two! This is the first we've seen of the 18mm I think.

We should all clarify however that all of this microscope analysis comes not from us finding nasty issues with the reel or the footage, but because we are all so :willy_nilly: about what we are seeing that we can't stop asking questions... well, that's me at least.

We might not be commenting on the reel itself because we have learned to expect so much from all things associated with Red, but the production value put into this Reel is very significant. I enjoyed the reel in a big way! Yet here we are, whining about moire and stair stepping even though we know this is a weirdness prone delivery codec. So much so that several people brought up pro-sumer VDSLRs, even though that is like comparing Argentine steak with BBQ flavored Fritos.

It seems, as Jim already said, the only true digital competition will be Arri. I don't see them shooting reels yet either. I am also very aware that for the cost of the rental of one Arri I will be likely to get two R1-X's! That means oodles of awesomeness to me.

Let's not forget the cost:IQ of Red's Primes! Sure they are a little heavy... but they are a very small fraction of the cost of MPs and deliver something that can be compared, which says a lot since MPs are the undisputed king of IQ. Also, it is thanks to Red gear that we can afford to rent MPs for a week, that we can afford to rent a specialty rig for an action sequence, or buy post gear!

Thanks to Red we are also working with negatives! And "unlimited lab budgets"! (if you know how to do your own stuff and made a small investment) Just make sure to finish with a pro cause the only thing that Red can't give us is talent... I heard they are working on AI to be implemented into RedCine-X by 2020 though.

OK, I'll go stfu in my corner now. I got some work to do, and some prayers to say for the LA incentive to bring some work back to the city.

Thank you to all that made this reel!

And thank you to everyone at Red for rocking the digital scene!

Watch what you say about BBQ Fritos Roberto! :yikes:

Good post. Seconded.

 

[Rob Ruffo]

I think that is definitely a matter of opinion with regards to the 5D/7D, as I personally find the footage I get from it to look great. Clearly a lot of other people feel this way as well, and yes, we are using the 1080p output (I don't even bother with 720p, it's horrible). I just finished shooting a project that will in fact be broadcast.

I have used other setups with a 35mm adapter and I wasn't happy with the workflow on set or with the camera itself with regards to the focusing screen and everything. It just made more variables to go wrong.

What really changed my mind was looking at the use of the 5D for the intro shots on Saturday Night Live and seeing the quality. Not to mention the plethora of other 5D/7D commercials and videos I've looked at that the quality has been incredible.

Maybe you have a more discriminating eye that I do, but the people who pay me for my work seem to me ecstatic about the results. (Not to mention I am too).

I agree, Indy film making has a million other costs in it, but I never indicated the cost of the Red setup made it prohibitive to purchase. That said, the difference in quality between say $3000 in my 7D and lenses definitely does not have an equivalent in straight video cameras I have seen in terms of cinematic appeal. Granted, I traded one set of headaches for another. And not to mention light sensitivity. There is NO OTHER camera even close in this price point that I can shoot at night with little light and get so little noise in the images.

The point being though is that Red has put together a package (Scarlet Fixed) that to me is absolutely a game changer, because it's at a price point with features that NOTHING I have seen compares too (existing or announced). In that, which I am sure, is that Red is putting these at a price point where a higher volume in sales will recoup their R&D and get them integrated into more locations everywhere.

That said, there are a different set of discriminating eyes and a different set of requirements that are going to drive some of those sales in different market segments. One of those segments is people owning the 5D/7D. Whether you like these cameras or not, there are A LOT of people that do like them. I am sure Red would like to capture that user-base. The fastest way to do that is show them how all their woe's will be gone with their cameras. That's all. Others probably freaked out at first with the Moire because they too are tired of dealing with it elsewhere.

Broadcast standards are not really opinion. They are obejctive technical standards. You say your footage looks great, but most likely you have not seen this footage recompressed for broadcast over an over-crowded cable stream. Take a good look at SNL's opening sequence shot on 5D - it looks like a 3 pixel gaussian blur was applied on it. Problems like aliasing become far worse issues when recompressed at low bitrates for cable, as all that false information does not play nice with the way Mpeg compression works, which assumes a smooth, non aliased image (the kind a Red, a Varicam, an Ex1 or any pro or semi-pro video camera produces) to perform its perceptual tricks.

For a few seconds of fast editing, this is not great but OK, but a whole TV show looking like this would be hard to watch. (BTW - People keep talking about commercials, but which commercials? Local spots for bars?)

Most 5D owners know nothing of this - nor do they need to because their work goes to internet only, or as some form of demo reel. That's fine. However, people who rely on the false promises these tools and their generally not so informed fan-base make are in for a big surprise when they try to sell their finished work for broadcast. Remember there are many levels of broadcast standards: Major Network = high. Music video = low. Community television = very low. Local advert = none "Hey, you paid for your time spot so..."

False promises are a dangerous thing, because they lead people to make decisions based on false information. A friend here in Montreal, for example, impressed by 5D footage he saw ON THE INTERNET at 480p or so, and teh SNL spot he saw on his 42 inch, 720p TV, began shooting a feature on a 5D. After wasting roughly 100K on production days that did not result in much useable footage the destination network deemed passable, (30% or so was destroyed by moire that could not be fixed, another 25% had other issues) he is now shooting on Red, with an Ex1/35mm adapter as a B-cam.

The Ex1/35mm adapter footage in some cases looks less good than the 5D stuff on a small editing monitor (for example the bokeh is less complex) , but never when viewed on a large screen, and to a much larger degree the 35 adapter footage looks way, way nicer when recompressed for TV transmission. Sharper, cleaner, far less prone to macroblocking. the Red of course blows both away.

Now, how many 5D fans have done informed side by side comparisons of their camera vs. other options as seen on air? I'd say virtually none.

The low resolution fo the Internet, coupled with a lack of knowledge about how cable compression works, coupled with a lot of wishful thinking has created very dangerous illusions around DSLRs, illusions that risk hurting a lot of people's careers.

I wish it were true that I could buy a $2500 camera and shoot a whole movie with it. This would save me a bit of money I could put into other things like slightly longer shooting days and slightly more elabarorate sets.

The Red and other really pro tools are worth the money, and I write this as a caution not only to other filmamkers, but to suits who think they can pinch a 5D penny and get away with it (this has been hurting Red owner-operators, very unfairly). There is affordable lunch, good value lunch (Red, EX1 high quality new 35mm adapters which are far better than the old ones who gave the tech a bad rep.) but free lunch, while it would be nice, does not exist.

I've been reading about information MEMEs lately, and the 5D is a myth that persists because so many people desperately wish it to be true - I know people will flame for this post, and their passion will reflect their wish for this - which is also accompanied by a similarly fanciful wish for there not to be a need for lighting, not to be a need for pro, trained actors, for a pro script written over many months rather than the typical few weeks most no budget projects allocate for this - all these are wishes connected to the dream of making a movie without training, without Hollywood connections, of having their dream of being directors and DOPs come true essentially for free - without the need for the huge sacrifices and persistence that are needed, in the real world, for this.

You say you didn't like the 35mm adapter workflow. Forgive me for being harsh, but how lazy are you? It's really no big deal. Yes there are variables, and you wake up 30 minutes earlier to make sure everything is perfect. If you don't like working hard and having a few things little things to worry about, maybe you chose the wrong business.

People love the idea of easy shortcuts. In this business, sooner or later you accept that there are none, and sit down for the real hard - and long - work of getting there. The good news is the effort is worth it.

 

[Rick Burnett]

You say you didn't like the 35mm adapter workflow. Forgive me for being harsh, but how lazy are you? It's really no big deal. Yes there are variables, and you wake up 30 minutes earlier to make sure everything is perfect. If you don't like working hard and having a few things little things to worry about, maybe you chose the wrong business.

I will certainly give you that my broadcast experience is very limited, but the response we've gotten back so far has been positive. I'll have to respond when the entirety of this project is over as to the experience and results.

Regarding the 35mm workflow, it is not about laziness at all, the point is, adding in more areas for issues to crop up just adds more wasted time that could be used doing something else. Just because you do something the hard way doesn't make you necessarily a hard worker. I prefer an optimized workflow so I am not wasting time when trying to change out lens, deal with extra loss of light, deal with focusing from two aspects instead of one. Alignment issues. Not to mention the absolutely ridiculous amount of added gear to my rig. You might be okay with this, I am not. I've used two different 35mm setups and even built one for a HV30 for fun and while the results were nice, the effort of getting them seemed like a lot of extra work to me.

Would I shoot a feature length film with the 5D? No, but I would use it as a B camera. People jump into using all kinds of technology without understanding all the downfalls. I didn't do this. I have studied my 7D both from a research level and by my own tests. I know where the camera shines, and I know where it fails horribly. And I also know that the Red One blows it out of the water for usability.

That said, the LAST THING I am gonna do is go buy a camera that doesn't work like I want and costs a whole lot more. This is why I am here, this is why I am waiting for the Scarlet's to come out.

I am getting paid for projects now with my 7D and the results have made the customers ecstatic with the results. I'm not shooting Hollywood blockbuster material here, but I am shooting broadcast music videos at the moment.

All I can say is people need to know their equipment. I'm not having a problem and neither are my customers. Its their opinion that matters. And if the customers we have come to us because we cost less, because we don't have a $60k camera we've got to pay for, so be it. They have a need, and a quality level they want to obtain, and we provide that. I don't see how I could be faulted for that logic.

If someone spent $100k in production and didn't realize the problems on DAY ONE of shooting, I think they should have been picking a different field. I go over all my footage in very fine detail as I am a perfectionist. A day of test shooting with the 5D will reveal ALL problems if you really know how to test a camera. I don't mean to be harsh, but this isn't rocket science.

All that said, believe me when I tell you, if the Red One cost $5k I'd have one tomorrow I'm not wasting money on a video camera I can't upgrade cause when these new cameras come out, people will probably be giving away things like the HVX200 to try and get one.

 

[Rob Ruffo]

Totally agreed, and I shoot a lot more coverage because of this if I see anything that I think might be a problem (especially in urban settings).

It's worse too, because on the camera, there is NO WAY to detect Moire/Aliasing until you dump it off the camera or review it with the HDMI port AFTER recording since during recording the port outputs 480p from what I am told. I'm thinking of getting the DP1 from smallhd.com which is a 9inch true 720p display to use with the 7D and then the SFixed eventually to see if I can at least check the footage before transferring to another viewing source. There is no way to zoom into 1:1 on playing back video on the 5D or 7D.

This is also why my upgrade path will be from the SFixed to the S35 eventually as I like the results with the 7D sized sensor(APS-C) and they aren't very far off from one another.

This is why on a pro shoot that matters, only a fool would use a DLSR if he/she had any choice at all- sorry but true.

 

[Kholi Hicks]

Broadcast standards are not really opinion. They are obejctive technical standards. You say your footage looks great, but most likely you have not seen this footage recompressed for broadcast over an over-crowded cable stream. Take a good look at SNL's opening sequence shot on 5D - it looks like a 3 pixel gaussian blur was applied on it. Problems like aliasing become far worse issues when recompressed at low bitrates for cable, as all that false information does not play nice with the way Mpeg compression works, which assumes a smooth, non aliased image (the kind a Red, a Varicam, an Ex1 or any pro or semi-pro video camera produces) to perform its perceptual tricks.

For a few seconds of fast editing, this is not great but OK, but a whole TV show looking like this would be hard to watch. (BTW - People keep talking about commercials, but which commercials? Local spots for bars?)

Most 5D owners know nothing of this - nor do they need to because their work goes to internet only, or as some form of demo reel. That's fine. However, people who rely on the false promises these tools and their generally not so informed fan-base make are in for a big surprise when they try to sell their finished work for broadcast. Remember there are many levels of broadcast standards: Major Network = high. Music video = low. Community television = very low. Local advert = none "Hey, you paid for your time spot so..."

False promises are a dangerous thing, because they lead people to make decisions based on false information. A friend here in Montreal, for example, impressed by 5D footage he saw ON THE INTERNET at 480p or so, and teh SNL spot he saw on his 42 inch, 720p TV, began shooting a feature on a 5D. After wasting roughly 100K on production days that did not result in much useable footage the destination network deemed passable, (30% or so was destroyed by moire that could not be fixed, another 25% had other issues) he is now shooting on Red, with an Ex1/35mm adapter as a B-cam.

The Ex1/35mm adapter footage in some cases looks less good than the 5D stuff on a small editing monitor (for example the bokeh is less complex) , but never when viewed on a large screen, and to a much larger degree the 35 adapter footage looks way, way nicer when recompressed for TV transmission. Sharper, cleaner, far less prone to macroblocking. the Red of course blows both away.

Now, how many 5D fans have done informed side by side comparisons of their camera vs. other options as seen on air? I'd say virtually none.

The low resolution fo the Internet, coupled with a lack of knowledge about how cable compression works, coupled with a lot of wishful thinking has created very dangerous illusions around DSLRs, illusions that risk hurting a lot of people's careers.

I wish it were true that I could buy a $2500 camera and shoot a whole movie with it. This would save me a bit of money I could put into other things like slightly longer shooting days and slightly more elabarorate sets.

The Red and other really pro tools are worth the money, and I write this as a caution not only to other filmamkers, but to suits who think they can pinch a 5D penny and get away with it (this has been hurting Red owner-operators, very unfairly). There is affordable lunch, good value lunch (Red, EX1 high quality new 35mm adapters which are far better than the old ones who gave the tech a bad rep.) but free lunch, while it would be nice, does not exist.

I've been reading about information MEMEs lately, and the 5D is a myth that persists because so many people desperately wish it to be true - I know people will flame for this post, and their passion will reflect their wish for this - which is also accompanied by a similarly fanciful wish for there not to be a need for lighting, not to be a need for pro, trained actors, for a pro script written over many months rather than the typical few weeks most no budget projects allocate for this - all these are wishes connected to the dream of making a movie without training, without Hollywood connections, of having their dream of being directors and DOPs come true essentially for free - without the need for the huge sacrifices and persistence that are needed, in the real world, for this.

You say you didn't like the 35mm adapter workflow. Forgive me for being harsh, but how lazy are you? It's really no big deal. Yes there are variables, and you wake up 30 minutes earlier to make sure everything is perfect. If you don't like working hard and having a few things little things to worry about, maybe you chose the wrong business.

People love the idea of easy shortcuts. In this business, sooner or later you accept that there are none, and sit down for the real hard - and long - work of getting there. The good news is the effort is worth it.

** golfclap.mp3 **

 

[Rob Ruffo]

I will certainly give you that my broadcast experience is very limited, but the response we've gotten back so far has been positive. I'll have to respond when the entirety of this project is over as to the experience and results.

Regarding the 35mm workflow, it is not about laziness at all, the point is, adding in more areas for issues to crop up just adds more wasted time that could be used doing something else. Just because you do something the hard way doesn't make you necessarily a hard worker. I prefer an optimized workflow so I am not wasting time when trying to change out lens, deal with extra loss of light, deal with focusing from two aspects instead of one. Alignment issues. Not to mention the absolutely ridiculous amount of added gear to my rig. You might be okay with this, I am not. I've used two different 35mm setups and even built one for a HV30 for fun and while the results were nice, the effort of getting them seemed like a lot of extra work to me.

Would I shoot a feature length film with the 5D? No, but I would use it as a B camera. People jump into using all kinds of technology without understanding all the downfalls. I didn't do this. I have studied my 7D both from a research level and by my own tests. I know where the camera shines, and I know where it fails horribly. And I also know that the Red One blows it out of the water for usability.

That said, the LAST THING I am gonna do is go buy a camera that doesn't work like I want and costs a whole lot more. This is why I am here, this is why I am waiting for the Scarlet's to come out.

I am getting paid for projects now with my 7D and the results have made the customers ecstatic with the results. I'm not shooting Hollywood blockbuster material here, but I am shooting broadcast music videos at the moment.

All I can say is people need to know their equipment. I'm not having a problem and neither are my customers. Its their opinion that matters. And if the customers we have come to us because we cost less, because we don't have a $60k camera we've got to pay for, so be it. They have a need, and a quality level they want to obtain, and we provide that. I don't see how I could be faulted for that logic.

If someone spent $100k in production and didn't realize the problems on DAY ONE of shooting, I think they should have been picking a different field. I go over all my footage in very fine detail as I am a perfectionist. A day of test shooting with the 5D will reveal ALL problems if you really know how to test a camera. I don't mean to be harsh, but this isn't rocket science.

All that said, believe me when I tell you, if the Red One cost $5k I'd have one tomorrow I'm not wasting money on a video camera I can't upgrade cause when these new cameras come out, people will probably be giving away things like the HVX200 to try and get one.

The 100K was lost because of a number of things related to scheduling repercussion, but mostly my friend's unwillingness to admit a serious error. He did do some tests - he's just one of those guys who "doesn't believe in focus charts". He also did not foresee just how harsh the network would be. Fact is, you yourself refuse to believe that 5D footage is not really broadcastable in HD - in SD yes, as long as you never shoot any buildings in focus and spend a lot of time checking wardrobe for moire, but it's not really an HD camera, and here on my 1080p projector, it really shows that it is not.

False detail can look OK on a first generation of compression, but not when broadcast. I learned the hard way using an old AVID system, back in the day. It had 6:1 compression. Looked fine in the suite. On air it looked - well, I lost 2 months work because of how it looked. Compression and aliasing problems are exponentially compounded through the broadcasting process.

Another thing many people fail to see is that by using cheap tools, you are limiting yourslef to a cheap market, building a reel of cheap work, and condeming yourself to working the low end. Why not rent the best stuff? Why not aim for higher level clients?

 

[Rick Burnett]

In many ways, because I am using the 7D, it's harder because I do have a lot of limitations to work around. I can't just point and shoot and expect everything to come out great. Believe me, I'd rather not deal with these problems. But the work I do is on the side, and not as frequent enough to warrant buying a $50k camera. Renting around here is very prohibitive with the Red One unfortunately.

I thought long and hard before buying the 7D after getting burned buying the Nikon D90 which is one of the worst cameras I have EVER purchased that did video. Yes, as most point out, it's a stills camera. I wasn't looking for perfection, but I was looking at something even slightly usable, which it was not!

I like to shoot a lot of night stuff which was the draw of the 7D. No other camera I have used has such great response at night with a F1.4 lens. Well, nothing I can afford.

The 7D has to hold me over for maybe 4 or 5 months. To me, that is worth it, then I can pick up a Scarlet Fixed to start with and then eventually the S35 brain. Luckily, the larger full length projects I am involved with do not have to start soon and the testing we are doing is more just location scouting.

For the paid projects, like I said, so far they've looked good. When the client gets back to us on the final result, I'll be sure to comment back one way or another as to how it went.

But like I said, this is a stop gap. I have NO interest in investing in older technology when the new ways are right around the corner and I needed a decent stills anyways.

Regardless, thanks for the advice and information. I am like a sponge.

 

[Emanuel A.]

Broadcast standards are not really opinion. They are obejctive technical standards. You say your footage looks great, but most likely you have not seen this footage recompressed for broadcast over an over-crowded cable stream. Take a good look at SNL's opening sequence shot on 5D - it looks like a 3 pixel gaussian blur was applied on it. Problems like aliasing become far worse issues when recompressed at low bitrates for cable, as all that false information does not play nice with the way Mpeg compression works, which assumes a smooth, non aliased image (the kind a Red, a Varicam, an Ex1 or any pro or semi-pro video camera produces) to perform its perceptual tricks.

For a few seconds of fast editing, this is not great but OK, but a whole TV show looking like this would be hard to watch. (BTW - People keep talking about commercials, but which commercials? Local spots for bars?)

Most 5D owners know nothing of this - nor do they need to because their work goes to internet only, or as some form of demo reel. That's fine. However, people who rely on the false promises these tools and their generally not so informed fan-base make are in for a big surprise when they try to sell their finished work for broadcast. Remember there are many levels of broadcast standards: Major Network = high. Music video = low. Community television = very low. Local advert = none "Hey, you paid for your time spot so..."

False promises are a dangerous thing, because they lead people to make decisions based on false information. A friend here in Montreal, for example, impressed by 5D footage he saw ON THE INTERNET at 480p or so, and teh SNL spot he saw on his 42 inch, 720p TV, began shooting a feature on a 5D. After wasting roughly 100K on production days that did not result in much useable footage the destination network deemed passable, (30% or so was destroyed by moire that could not be fixed, another 25% had other issues) he is now shooting on Red, with an Ex1/35mm adapter as a B-cam.

The Ex1/35mm adapter footage in some cases looks less good than the 5D stuff on a small editing monitor (for example the bokeh is less complex) , but never when viewed on a large screen, and to a much larger degree the 35 adapter footage looks way, way nicer when recompressed for TV transmission. Sharper, cleaner, far less prone to macroblocking. the Red of course blows both away.

Now, how many 5D fans have done informed side by side comparisons of their camera vs. other options as seen on air? I'd say virtually none.

The low resolution fo the Internet, coupled with a lack of knowledge about how cable compression works, coupled with a lot of wishful thinking has created very dangerous illusions around DSLRs, illusions that risk hurting a lot of people's careers.

I wish it were true that I could buy a $2500 camera and shoot a whole movie with it. This would save me a bit of money I could put into other things like slightly longer shooting days and slightly more elabarorate sets.

The Red and other really pro tools are worth the money, and I write this as a caution not only to other filmamkers, but to suits who think they can pinch a 5D penny and get away with it (this has been hurting Red owner-operators, very unfairly). There is affordable lunch, good value lunch (Red, EX1 high quality new 35mm adapters which are far better than the old ones who gave the tech a bad rep.) but free lunch, while it would be nice, does not exist.

I've been reading about information MEMEs lately, and the 5D is a myth that persists because so many people desperately wish it to be true - I know people will flame for this post, and their passion will reflect their wish for this - which is also accompanied by a similarly fanciful wish for there not to be a need for lighting, not to be a need for pro, trained actors, for a pro script written over many months rather than the typical few weeks most no budget projects allocate for this - all these are wishes connected to the dream of making a movie without training, without Hollywood connections, of having their dream of being directors and DOPs come true essentially for free - without the need for the huge sacrifices and persistence that are needed, in the real world, for this.

You say you didn't like the 35mm adapter workflow. Forgive me for being harsh, but how lazy are you? It's really no big deal. Yes there are variables, and you wake up 30 minutes earlier to make sure everything is perfect. If you don't like working hard and having a few things little things to worry about, maybe you chose the wrong business.

People love the idea of easy shortcuts. In this business, sooner or later you accept that there are none, and sit down for the real hard - and long - work of getting there. The good news is the effort is worth it.

Great post and advice.

Rick, nice blog out there:
http://www.asylumstudioproductions.com/

E.

 

[Rick Burnett]

Thanks! I try to keep track of all that I learn and my experiments with video/audio. Can't wait to start posting some tests with the Scarlet!

 

[Emanuel A.]

5D Mark II/7D (as same as Scarlet will do as well too) has come to serve the art and the artists, not the industry standards or the usual establishment of a few