Digital Noise and camera sensor

O

offbeatbryce

Well-known member
Does anyone here know what kind of sensors are in cameras that film for Saturday Night Live? I'm asking because some perrromances of artists on that show sometimes don't have bright lights on the face etc but there is no digital video noise at all. Is that cause if the sensor or do the cameras used for SNL and other big productions have a Noise filter? I can't even film a scene with one light on in a dark room withoutit getting really noisy.
 
DLD said:
They're probably using the 2/3" Sony or the 3 1/3" Ikegami cameras.

Something like this - https://www.youtube.com/watch?v=EH5W7VvBw3Y

And there's no noise because there's plenty of light.
Click to expand...


I have a canon xa25. If I film someone on a stage with not much lighting the band behind them has big blocks of digital noise. Here is a screengrab from the SNL performance. the singers in the back aren't lit that well and you hardly see any digital noise but when I try this there is big blocks everywhere over the video. Even more strange is I can't even get an image as bright as the SNL screengrab without turning my gain up to like 15 which is odd.

Even more strange is with my f stop all the way open the highest luma value on the waveform in premiere for skin tone is at 35 IRE which makes the whole picture dark so I have to crank up the exposure in the Lumetri panel.




SNLScreenGrab.jpg
 
Last edited:
DLD is right, you are comparing your $2,000.00 consumer camera with a tiny 1/2.84" imager to professional $80k - $120k cameras with significantly larger imagers (I tried researching which stage cameras are used on SNL but couldn't find anything other than they were evaluating 4k cameras back in 2013 so they probably went with some kind of nice 4k camera, Hitachi, Sony, Panasonic or Ikegami). I interviewed SNL DP Alex Buono a few years ago but it was just about what he uses for the open and all of the insert segments when they do commercial spoofs and other inserts so it wasn't the studio stuff, he has nothing to do with that.

What you are seeing on your camera is a tiny, realtively noisy imager with a low signal to noise ratio and what you are seeing on the NBC cameras is a much, much higher end camera with a bigger sensor and a much better s/n ratio and DSP. Keep in mind also that you are seeing the SNL output on your consumer television set or computer screen that has been through multiple compression generations, if you were in the SNL master control looking at the image on one of their 4k monitors, it would look MUCH cleaner than what you are seeing at home. it's simple, you are comparing horsepower and 0-60 times between a Kia Elantra and a Ferrari.

Another factor is that exposure isn't always what you think it is. That Lorde performance above apparently is low lit, but if you were there in person with a light meter on that stage, you would find that what looks like low light is often pretty brightly lit, they just iris down the cameras or otherwise treat the signal to look low lit. I once shot a dance film for a talented director and I could not get him to understand that we needed to light and expose the dance sequences correctly, then he could drop the blacks and levels in post to look dark and mysterious. He inisisted on me radically underexposing the signal as we shot the performances and the end product ended up being noisy mud as I told him it would be. This has been one of my most challenging tasks as a DP, trying to convince producers and directors that what you see with your eyes and on production monitors isn't what you get. Try to shoot Log on a not LUT monitor and see what happens with most directors, they freak out and think it looks like hell, when we all know that correctly exposed log is usually going to end up looking better than any baked in look because post software and LUTs are more sophisticated and capable than most in-camera processing. Most DPs and gaffers have ran into this. There are reasons to underexpose occasionally but you typically end up with a much better result exposing "your negative" (your camera's sensor) correctly and doing special effects in post, especially on low end cameras.
 
Last edited:
puredrifting said:
.... Another factor is that exposure isn't always what you think it is. That Lorde performance above apparently is low lit, but if you were there in person with a light meter on that stage, you would find that what looks like low light is often pretty brightly lit, they just iris down the cameras or otherwise treat the signal to look low lit...
Click to expand...
Exactly. Plus, on a multi-cam production like the SNL, one can easily mix multiple exposures. (+ multiple apertures + multiple everything). So, one camera is at T2 (as an example) wide and close and another is T5.6 and further away and each camera aims at a different performer on stage, each of whom may stand under different lights.

As a bowl of wax, this couldn't be any different than a one man band.
 
puredrifting said:
DLD is right, you are comparing your $2,000.00 consumer camera with a tiny 1/2.84" imager to professional $80k - $120k cameras with significantly larger imagers (I tried researching which stage cameras are used on SNL but couldn't find anything other than they were evaluating 4k cameras back in 2013 so they probably went with some kind of nice 4k camera, Hitachi, Sony, Panasonic or Ikegami). I interviewed SNL DP Alex Buono a few years ago but it was just about what he uses for the open and all of the insert segments when they do commercial spoofs and other inserts so it wasn't the studio stuff, he has nothing to do with that.

What you are seeing on your camera is a tiny, realtively noisy imager with a low signal to noise ratio and what you are seeing on the NBC cameras is a much, much higher end camera with a bigger sensor and a much better s/n ratio and DSP. Keep in mind also that you are seeing the SNL output on your consumer television set or computer screen that has been through multiple compression generations, if you were in the SNL master control looking at the image on one of their 4k monitors, it would look MUCH cleaner than what you are seeing at home. it's simple, you are comparing horsepower and 0-60 times between a Kia Elantra and a Ferrari.

Another factor is that exposure isn't always what you think it is. That Lorde performance above apparently is low lit, but if you were there in person with a light meter on that stage, you would find that what looks like low light is often pretty brightly lit, they just iris down the cameras or otherwise treat the signal to look low lit. I once shot a dance film for a talented director and I could not get him to understand that we needed to light and expose the dance sequences correctly, then he could drop the blacks and levels in post to look dark and mysterious. He inisisted on me radically underexposing the signal as we shot the performances and the end product ended up being noisy mud as I told him it would be. This has been one of my most challenging tasks as a DP, trying to convince producers and directors that what you see with your eyes and on production monitors isn't what you get. Try to shoot Log on a not LUT monitor and see what happens with most directors, they freak out and think it looks like hell, when we all know that correctly exposed log is usually going to end up looking better than any baked in look because post software and LUTs are more sophisticated and capable than most in-camera processing. Most DPs and gaffers have ran into this. There are reasons to underexpose occasionally but you typically end up with a much better result exposing "your negative" (your camera's sensor) correctly and doing special effects in post, especially on low end cameras.
Click to expand...


Thanks for that info. If I get a 4k camera from BlackMagic will I have similar results to SNL or will I still be a what my Canon xa25 is? My only recording format options are AVCHD and MP4. I just don't want to rush out and buy any 4K camera dn not be happy with it. Also (newb question) what exactly is Slog? I've heard the term many times but can't quite figure out what it is. Is it a setting n cameras or field monitors?
 
offbeatbryce said:
Thanks for that info. If I get a 4k camera from BlackMagic will I have similar results to SNL or will I still be a what my Canon xa25 is?
Click to expand...

NO! If you want to get SNL results you'll need to get a camera and post processing equipment like SNL, along with their cameramen and post production engineers. What part of "you are comparing horsepower and 0-60 times between a Kia Elantra and a Ferrari" didn't you understand? I'm not trying to be a dick, but there's no way you can ask the same question in a different way and have someone tell you "yes, you can buy a piece of s*%t camera and get the same quality as a camera that's not a piece of s*%t". If you could get the same results with cheaper cameras don't you think SNL would save their money and not buy those $80,000+ cameras?
 
Reggie Brown said:
NO! If you want to get SNL results you'll need to get a camera and post processing equipment like SNL, along with their cameramen and post production engineers. What part of "you are comparing horsepower and 0-60 times between a Kia Elantra and a Ferrari" didn't you understand? I'm not trying to be a dick, but there's no way you can ask the same question in a different way and have someone tell you "yes, you can buy a piece of s*%t camera and get the same quality as a camera that's not a piece of s*%t". If you could get the same results with cheaper cameras don't you think SNL would save their money and not buy those $80,000+ cameras?
Click to expand...

right I understand that but I've heard how any 4K camera is like hollywood
 
If any 4k camera was like that... Arri 4k cams, Red 4k cams, Canon 4k cinema cams, along with Sony 4k cine cams would tank because the producers would definitely cut cost and wouldn't allow the dp/cinematographer to use the budget money on them. Besides, it's more than about the camera being 4k. You can have a 4k camera and someone else can have a 720p, if your 4k camera is always out of focus and the person with the 720p camera nail focus, guess which footage is going to look better. 4k is just a resolution and just because a manufacturer says there camera is 4k doesn't mean that their sensor is going to be sharp and/or top notch. There's a lot that goes into camera technology and specs. Please at least somewhat study that before you needlessly spend your cash on camera equipment.
 
Reggie Brown said:
If any 4k camera was like that... Arri 4k cams, Red 4k cams, Canon 4k cinema cams, along with Sony 4k cine cams would tank because the producers would definitely cut cost and wouldn't allow the dp/cinematographer to use the budget money on them. Besides, it's more than about the camera being 4k. You can have a 4k camera and someone else can have a 720p, if your 4k camera is always out of focus and the person with the 720p camera nail focus, guess which footage is going to look better. 4k is just a resolution and just because a manufacturer says there camera is 4k doesn't mean that their sensor is going to be sharp and/or top notch. There's a lot that goes into camera technology and specs. Please at least somewhat study that before you needlessly spend your cash on camera equipment.
Click to expand...

So what would be a good camera that has a better sensor than my Canon Cameras?
 
It comes down to personal taste. The gh4 is good but suck in low light and it's a micro four third sensor, Sony is good but ppl complain about the colors and how they render skin tones, Canon is good but ppl complain about their slim offerings of 4k, blackmagic is good but suck in low light and you pretty much have to use the native iso to avoid fpn, Arri is good and I've never heard a complaint, Red is good and the complaints I have heard were nitpicking complaints. I wouldn't say any sensor is "better" than the other, but a particular sensor may be "better" for a particular need. If you really needed to get a candid shot of something and all you had at the time was your phone, at that time that's the best sensor for that shot. Dps do camera and lens test before making a movie to test to see which technology would help achieve the look they're going for, sometimes it's a Canon camera, sometimes it's a Sony...etc.
 
Last edited:
Just saw a GH5 test, where a live band was shot at 6400 ISO, with very little remaining noise after the required work in post (Neat, I presume). Then, for the money, A7S/II is great at low light but to get reach with lenses might turn expensive, with the manual Tamron or Samyang probably being the best buy. The next step up - and that's a big step - is Panasonic Varicam LT, which has a dual analog gain and is capable of shooting at 5000 ISO with very little noise. Of course, the LT shooting package is in the mid-$20,000.
 
You must log in or register to reply here.
Back
Top