arrestthisman
Well-known member
Okay, the following is a rant with no real point, and it touches on stuff that may not be flame retardant; but it's on my mind. I just hope we can discuss this in a civil manner, assuming anyone is interested. I'm not very sure of any of it, and that's my qualification. I feel almost selfish posting here, but nobody is being forced to read it unless they find it interesting or relevant, so I guess that's kind of an insecurity/non-sense.
So here goes:
The question turning over in my mind was, for me, if I had 50,000$ to blow, what would make the best set up specifically for me. An Ikonoskop dII, a Red One (or maybe scarlet, assuming scarlet is still a fixed lens that comes close in quality with redcode 36), or an HPX170. I shoot almost exclusively in available light. I also like a deep depth of field, so for me 35mm adapters make no sense and larger sensors for DOF are of no use to me. It's usually a very small crew, pretty much run and gun. For delivery it would most likely be on the internet, although I'm sure there would be a few screenings (not necessarily distribution, but local screenings with some groups I have arrangements with) in a small theater setting using digital projection. (Which BTW, looks awesome at the Cinerama in Seattle.)
My goal would be the best quality/weight/size/workflow ratio, in addition to the "look" that I want. I used to own a DVX100 (original). Then an HVX200, which I'm selling.
But in theory, I'm going over an ideal world where I have millions of dollars to spend.
First there's the Ikonoskop. I have to admit, I'm pretty smitten over this camera, however, it is a single sensor 1920 x 1080, and I haven't seen any footage. 3.3lbs sounds awesome. I like the cartridge system, but the uncompressed workflow seems a bit unwieldy at 80gbs/15 minutes. That said, I could always pick a good codec (I'm partial to DNxHD 220 even though I hate Avid with a serious passion), and transcode.
What I don't like is that being a single sensor, it will likely use a bayer pattern, and resolve something like 70 to 78% of the actual resolution. So in terms of plain old resolution it doesn't seem likely to resolve much more than the HPX170. It also doesn't have a flip out LCD, and there's only 1 method of focus assist announced. I won't get into a color debate, but just using a loose analogy. If the Red one needs to be recorded @ 4k, and downsampled to 2k (based on a post from Graeme Nattress I just read five minutes ago) to achieve 4:4:4 equivalent color, then the color is NOT going to be full 4:4:4 on the Ikonoskop at 1920x1080. It's stupid to compare the chroma 3-chip to 1-chip, I know, but I'm just trying to get an idea of the advantages, and it doesn't seem like the Ikonoskop would have an Advantage here. The advantage would be uncompressed, RAW (although some would argue this isn't really an advantage if you get the picture correct in-cam), and 12bit color. I've never worked in 12-bit color, so I cannot speak to how much better in the real world it would look over 8-bit.
I would love to get a chance to play with a dII, and I love the tiny formfactor, but at least in theory, the much more cumbersome workflow and lack of streamlined conveniences (also no Hd-SDI out) doesn't necessarily make it a total net-gain. Or if it is then just not night and day, or AS night and day as it would seem. Interchangeable lenses aren't a big deal to me.
Then there's the Red One. (f23, and the whole lot of the giant digital cinema cameras are out of the running in this little hypothetical odyssey, because I'd never want to lug around something that heavy, or need a crew that big to operate them.)
At 10 pounds it kind of tips the scales for me, but I suppose it's close to doable, plus it's a precedent for scarlet, so for the sake of discussion it may be useful. But I have a copy of some Raw footage (the infamous race car stuff), plus Redcine on my computer. The latest build is still pretty buggy and sluggish. I did notice at full rez/quality, previewing it, I could see a lot of compression artifacts. They were fine in texture, and not ugly, but they were there. I was not expecting that. So I exported it in 1080p uncompressed, in case it was the viewer. Still there. I know the new builds have improved picture quality, but I did I little math, and for a 4k image to have the same compression as 1080p DVCPRO HD (yes I know one is wavelet and one is DCT), it would require 66mb/sec or more. This is almost double the redcode's 36mb, and 12-bit color at that. DVCPRO HD is only 8-bit. But leaving the bit comparison out that still is a 12:1 versus 6.7:1 compression ratio at the highest quality the camera offers.
I know it's wavelet, but if it's twice as efficient, it's because it HAS to be. I just found that interesting. I'm not saying Red is a bad camera. I think it's a badass camera. I would love to play around with one in the flesh. The footage is spectacular and has a neat look to it, it's just not--I think--the divine holy grail I was led to believe. (It did look MUCH better than some stills I got from the Dalsa, which was very disappointing.)
I took a still of a well exposed HVX200 clip, and compared it to some still of the Red One that was in daylight, and there was a noticeable difference no doubt. But it wasn't as big of a difference as what I was expecting. What was neat was that the both stills had a hand present in the picture, so I could compare finger to finger, nail to nail. The color tones were similar. I just found that very interesting. And maybe this is all just my eye, but for my content, it's my eye that counts. Furthermore, under heavy magnification, the edges of Red footage reminded me more of EX1 footage (4:2:0) than HVX200 footage (4:2:2). It seemed to be right in between, and again, I'm aware that chroma subsampling numbers are not useful for RAW footage, but I don't know any other way to quantify it at all, so it's better than no number, especially because I'm not trying to convince anyone of anything. It's just an observation of mine, so please don't bite my head off (of course feel free to point out error). And keep in mind this is all before downsampling, and at 4k you have a lot to work with.
The crux of all this is that with an endless budget, I'm struggling to find a camera (in my head) that would beat the HPX170 for the size/weight/workflow/quality ratio. If it were just the size and weight, then I think unless the DR was really bad or it was noisy, or some other unexpected bad thing, the Ikonoskop would beat the HPX170 definitively. Even then with a 640x480 simple color viewfinder, unless you had a monitor around, critical focus would be a bit harder. But that's not a deal-breaker, so I still think it's probably better footage.
We still have to wait until tomorrow to see what Scarlet 2.0 has in-store for us. But if it uses the same compression scheme as the Red One, it will still be better, but not by much (for my eye/preference), and as of right now it doesn't exist--again if you take workflow into account, it's the same problem as the Ikonoskop. The Red One is too big for the stuff I do anyway. Not feasible, so I think (soon to be) Scarlet vs dII vs HPX170 is more meaningful for my purposes.
I keep thinking of Inland Empire and Lynch's comments about working with the PD-150. He wasn't happy with the quality, but taking into account the total ease of use, and speed of online editing, it was the right balance for the way he wanted and still wants to work.
I think the little Panasonic cams seem to be the right balance for a lot of people, and definitely for me. At least right now. We should all be grateful we have so many wonderful tools to pick from these days. And it's only going to get better.
I do think, and this is where I hope anybody still reading would comment on--that for my purposes the Ikonoskop would be the most logical "step up" from the HPX170 if I had the money. Agree?
I think that would be the dream team. dII + an HPX170 on the shelf, and pick which one is best for the job.
Happy Scarlet day -24 hours!
So here goes:
The question turning over in my mind was, for me, if I had 50,000$ to blow, what would make the best set up specifically for me. An Ikonoskop dII, a Red One (or maybe scarlet, assuming scarlet is still a fixed lens that comes close in quality with redcode 36), or an HPX170. I shoot almost exclusively in available light. I also like a deep depth of field, so for me 35mm adapters make no sense and larger sensors for DOF are of no use to me. It's usually a very small crew, pretty much run and gun. For delivery it would most likely be on the internet, although I'm sure there would be a few screenings (not necessarily distribution, but local screenings with some groups I have arrangements with) in a small theater setting using digital projection. (Which BTW, looks awesome at the Cinerama in Seattle.)
My goal would be the best quality/weight/size/workflow ratio, in addition to the "look" that I want. I used to own a DVX100 (original). Then an HVX200, which I'm selling.
But in theory, I'm going over an ideal world where I have millions of dollars to spend.
First there's the Ikonoskop. I have to admit, I'm pretty smitten over this camera, however, it is a single sensor 1920 x 1080, and I haven't seen any footage. 3.3lbs sounds awesome. I like the cartridge system, but the uncompressed workflow seems a bit unwieldy at 80gbs/15 minutes. That said, I could always pick a good codec (I'm partial to DNxHD 220 even though I hate Avid with a serious passion), and transcode.
What I don't like is that being a single sensor, it will likely use a bayer pattern, and resolve something like 70 to 78% of the actual resolution. So in terms of plain old resolution it doesn't seem likely to resolve much more than the HPX170. It also doesn't have a flip out LCD, and there's only 1 method of focus assist announced. I won't get into a color debate, but just using a loose analogy. If the Red one needs to be recorded @ 4k, and downsampled to 2k (based on a post from Graeme Nattress I just read five minutes ago) to achieve 4:4:4 equivalent color, then the color is NOT going to be full 4:4:4 on the Ikonoskop at 1920x1080. It's stupid to compare the chroma 3-chip to 1-chip, I know, but I'm just trying to get an idea of the advantages, and it doesn't seem like the Ikonoskop would have an Advantage here. The advantage would be uncompressed, RAW (although some would argue this isn't really an advantage if you get the picture correct in-cam), and 12bit color. I've never worked in 12-bit color, so I cannot speak to how much better in the real world it would look over 8-bit.
I would love to get a chance to play with a dII, and I love the tiny formfactor, but at least in theory, the much more cumbersome workflow and lack of streamlined conveniences (also no Hd-SDI out) doesn't necessarily make it a total net-gain. Or if it is then just not night and day, or AS night and day as it would seem. Interchangeable lenses aren't a big deal to me.
Then there's the Red One. (f23, and the whole lot of the giant digital cinema cameras are out of the running in this little hypothetical odyssey, because I'd never want to lug around something that heavy, or need a crew that big to operate them.)
At 10 pounds it kind of tips the scales for me, but I suppose it's close to doable, plus it's a precedent for scarlet, so for the sake of discussion it may be useful. But I have a copy of some Raw footage (the infamous race car stuff), plus Redcine on my computer. The latest build is still pretty buggy and sluggish. I did notice at full rez/quality, previewing it, I could see a lot of compression artifacts. They were fine in texture, and not ugly, but they were there. I was not expecting that. So I exported it in 1080p uncompressed, in case it was the viewer. Still there. I know the new builds have improved picture quality, but I did I little math, and for a 4k image to have the same compression as 1080p DVCPRO HD (yes I know one is wavelet and one is DCT), it would require 66mb/sec or more. This is almost double the redcode's 36mb, and 12-bit color at that. DVCPRO HD is only 8-bit. But leaving the bit comparison out that still is a 12:1 versus 6.7:1 compression ratio at the highest quality the camera offers.
I know it's wavelet, but if it's twice as efficient, it's because it HAS to be. I just found that interesting. I'm not saying Red is a bad camera. I think it's a badass camera. I would love to play around with one in the flesh. The footage is spectacular and has a neat look to it, it's just not--I think--the divine holy grail I was led to believe. (It did look MUCH better than some stills I got from the Dalsa, which was very disappointing.)
I took a still of a well exposed HVX200 clip, and compared it to some still of the Red One that was in daylight, and there was a noticeable difference no doubt. But it wasn't as big of a difference as what I was expecting. What was neat was that the both stills had a hand present in the picture, so I could compare finger to finger, nail to nail. The color tones were similar. I just found that very interesting. And maybe this is all just my eye, but for my content, it's my eye that counts. Furthermore, under heavy magnification, the edges of Red footage reminded me more of EX1 footage (4:2:0) than HVX200 footage (4:2:2). It seemed to be right in between, and again, I'm aware that chroma subsampling numbers are not useful for RAW footage, but I don't know any other way to quantify it at all, so it's better than no number, especially because I'm not trying to convince anyone of anything. It's just an observation of mine, so please don't bite my head off (of course feel free to point out error). And keep in mind this is all before downsampling, and at 4k you have a lot to work with.
The crux of all this is that with an endless budget, I'm struggling to find a camera (in my head) that would beat the HPX170 for the size/weight/workflow/quality ratio. If it were just the size and weight, then I think unless the DR was really bad or it was noisy, or some other unexpected bad thing, the Ikonoskop would beat the HPX170 definitively. Even then with a 640x480 simple color viewfinder, unless you had a monitor around, critical focus would be a bit harder. But that's not a deal-breaker, so I still think it's probably better footage.
We still have to wait until tomorrow to see what Scarlet 2.0 has in-store for us. But if it uses the same compression scheme as the Red One, it will still be better, but not by much (for my eye/preference), and as of right now it doesn't exist--again if you take workflow into account, it's the same problem as the Ikonoskop. The Red One is too big for the stuff I do anyway. Not feasible, so I think (soon to be) Scarlet vs dII vs HPX170 is more meaningful for my purposes.
I keep thinking of Inland Empire and Lynch's comments about working with the PD-150. He wasn't happy with the quality, but taking into account the total ease of use, and speed of online editing, it was the right balance for the way he wanted and still wants to work.
I think the little Panasonic cams seem to be the right balance for a lot of people, and definitely for me. At least right now. We should all be grateful we have so many wonderful tools to pick from these days. And it's only going to get better.
I do think, and this is where I hope anybody still reading would comment on--that for my purposes the Ikonoskop would be the most logical "step up" from the HPX170 if I had the money. Agree?
I think that would be the dream team. dII + an HPX170 on the shelf, and pick which one is best for the job.
Happy Scarlet day -24 hours!
