Duke M.
Veteran
I've been doing a lot of thinking about this issue because too many people get caught up in the hype rather than facts.
I'd like to lay out my thoughts and welcome people to point out where I'm wrong or add their own thoughts.
RESOLUTION: As soon as people start discussing this someone always throws out the assertion that the Scarlet image will be vastly superior to the EX1 (or any other pro-prosumer camera) because the Scarlet will be 3072 wide and the EX1 is 1920 wide. That's not as certain as you might think.
The Scarlet is a single CMOS chip camera with a debayer system to separate the colors. The EX1 is a 3 CMOS chip camera and from what I've been needs no debayer system because it separates colors with prisms and separate sensors.
The thing is a debayer system reduces resolution by ~30%. For this reason people have speculated that the 3K Scarlet will really have resolution of about 2.1k. The Scarlet images will need to be down rezed to 2k or 1920.
Looking at the specs the EX1 has 'effective' resolution of 1920x1080, but the sensors have a bit more pixels and down rezed a bit.
Net result (I believe but we don't actually know) is that the actual resolution of the two cameras will be very, very similar. There's not a whole lot of difference between 1920 and 2072 (2.1k) anyway.
COMPRESSION:
It is my understanding that the RAW files are wavelet compressed. Wavelet compression is nearly (not completely) lossless. The EX1 has compression and 4.2.0 color but the codec is very hard to break. In addition a nanoflash can improve the file format and improve color to 4.2.2. In ordinary shooting the regular EX1 files look fabulous.
JELLO AND SKEW:
Another battle cry of the Scarletteers is that the Scarlet will have a much faster reading of the sensor, maybe 50% faster rate, so less jello and skew. The problem with that is the Scarlet will have 50% more pixels. We can't test yet, but the Scarlet may have similar CMOS skew characteristics as the EX1. To be fair its only an issue in a limited number of high movement situations and even the Red One has been demonstrated (at reduser) to have some skew issues in the right/wrong circumstances.
COLORS AND WORKFLOW:
One of the issues with one CMOS chips is the colors tend to be flat. The Red cameras work around this by recording RAW to allow post color correction. There are both advantages (you can push things more in post since the meta data isn't baked in) and disadvantages (your client on the set thinks it looks like crap and more post production work.)
Whether this is an advantage or disadvantage seems that is will depend on the circumstances of the project. Certainly if you can get the look nearly perfect in camera the work flow is faster. If you can't get it right in camera the end result may be better if you need to tweak a lot in post.
DOF CONTROL:
A 2/3" sensor gives better depth of field (DOF) control, right? Not automatically.
There are three factors at work. Aperture, lens length and sensor size, and two of those are interconnected. (Some people will argue the sensor size doesn't matter, but with a larger sensor the effect of the lens length changes, so it is a consideration.)
I loved Barry Green's article on DOF. One of the pertinent points is that to get a 1/2" chip to get the same DOF as a 2/3" chip you only need to open the aperture by 2/3 of a stop.
Again, we don't know for sure because the exact specs on the Scarlet lens haven't been published, but the ones that have been mentioned are a T3 8x zoom, which would be ~f2.8.
Now the EX1 goes to f1.9, which is more than 2/3 of a stop on the wide end. In short, the EX1 will have a potential for narrower DOF due to the wider aperture.
The other aspect of DOF is the length of the lens. Since the Scarlet fixed will be limited to 8x magification, the 14x of the EX1 will give many more options to control the DOF. (Sure you can throw a 2x adapter on the Scarlet, but you'll lose some resolution and there goes that resolution advantage.) Or if you go with the Scarlet 2/3" with removable lens it appears that all the little accessories start to get pricey.
CONCLUSION:
So is there really an advantage in the 2/3" Scarlet over the EX1? Resolution, minor advantage. Skew, can't tell yet, but its still a CMOS camera. Workflow, disadvange for me. Compression, slight EX1 disadvantage. DOF, scarlet disadvantage. Something you can actually use right now? EX1 huge advantage. In balance it seems like a wash to me.
What do you think?
I'd like to lay out my thoughts and welcome people to point out where I'm wrong or add their own thoughts.
RESOLUTION: As soon as people start discussing this someone always throws out the assertion that the Scarlet image will be vastly superior to the EX1 (or any other pro-prosumer camera) because the Scarlet will be 3072 wide and the EX1 is 1920 wide. That's not as certain as you might think.
The Scarlet is a single CMOS chip camera with a debayer system to separate the colors. The EX1 is a 3 CMOS chip camera and from what I've been needs no debayer system because it separates colors with prisms and separate sensors.
The thing is a debayer system reduces resolution by ~30%. For this reason people have speculated that the 3K Scarlet will really have resolution of about 2.1k. The Scarlet images will need to be down rezed to 2k or 1920.
Looking at the specs the EX1 has 'effective' resolution of 1920x1080, but the sensors have a bit more pixels and down rezed a bit.
Net result (I believe but we don't actually know) is that the actual resolution of the two cameras will be very, very similar. There's not a whole lot of difference between 1920 and 2072 (2.1k) anyway.
COMPRESSION:
It is my understanding that the RAW files are wavelet compressed. Wavelet compression is nearly (not completely) lossless. The EX1 has compression and 4.2.0 color but the codec is very hard to break. In addition a nanoflash can improve the file format and improve color to 4.2.2. In ordinary shooting the regular EX1 files look fabulous.
JELLO AND SKEW:
Another battle cry of the Scarletteers is that the Scarlet will have a much faster reading of the sensor, maybe 50% faster rate, so less jello and skew. The problem with that is the Scarlet will have 50% more pixels. We can't test yet, but the Scarlet may have similar CMOS skew characteristics as the EX1. To be fair its only an issue in a limited number of high movement situations and even the Red One has been demonstrated (at reduser) to have some skew issues in the right/wrong circumstances.
COLORS AND WORKFLOW:
One of the issues with one CMOS chips is the colors tend to be flat. The Red cameras work around this by recording RAW to allow post color correction. There are both advantages (you can push things more in post since the meta data isn't baked in) and disadvantages (your client on the set thinks it looks like crap and more post production work.)
Whether this is an advantage or disadvantage seems that is will depend on the circumstances of the project. Certainly if you can get the look nearly perfect in camera the work flow is faster. If you can't get it right in camera the end result may be better if you need to tweak a lot in post.
DOF CONTROL:
A 2/3" sensor gives better depth of field (DOF) control, right? Not automatically.
There are three factors at work. Aperture, lens length and sensor size, and two of those are interconnected. (Some people will argue the sensor size doesn't matter, but with a larger sensor the effect of the lens length changes, so it is a consideration.)
I loved Barry Green's article on DOF. One of the pertinent points is that to get a 1/2" chip to get the same DOF as a 2/3" chip you only need to open the aperture by 2/3 of a stop.
Again, we don't know for sure because the exact specs on the Scarlet lens haven't been published, but the ones that have been mentioned are a T3 8x zoom, which would be ~f2.8.
Now the EX1 goes to f1.9, which is more than 2/3 of a stop on the wide end. In short, the EX1 will have a potential for narrower DOF due to the wider aperture.
The other aspect of DOF is the length of the lens. Since the Scarlet fixed will be limited to 8x magification, the 14x of the EX1 will give many more options to control the DOF. (Sure you can throw a 2x adapter on the Scarlet, but you'll lose some resolution and there goes that resolution advantage.) Or if you go with the Scarlet 2/3" with removable lens it appears that all the little accessories start to get pricey.
CONCLUSION:
So is there really an advantage in the 2/3" Scarlet over the EX1? Resolution, minor advantage. Skew, can't tell yet, but its still a CMOS camera. Workflow, disadvange for me. Compression, slight EX1 disadvantage. DOF, scarlet disadvantage. Something you can actually use right now? EX1 huge advantage. In balance it seems like a wash to me.
What do you think?
Last edited: