You guys should add a third score for the DR tests. You have the SNR 1 and 2. But I would LOVE to have a separate doctored version of SNR 2 where you apply noise reduction in post, highlight recovery, etc. That way, people can actually see what these cameras are capable of with solid post-treatment. Right now, the tests favor cameras with solid in-camera noise reduction, which is misleading. The over-under exposure test helps to show the real dynamic range, but it would be cool to have it reflected in a score.
Good idea - though maybe including highlight recovery complicates it a bit. I imagine for instance that changing the hue on a sunset shot where highlight recovery is doing a lot of heavy lifting, would be noticeably worse than shifting the hue on a shot where the last stop is full RGB... I just mean that it's not merely a technicality.
@@DaCarnival that is true to an extent, but at least in the case of the blackmagic cams, I’ve found that last stop to be more than usable and a primary reason I shoot blackmagic. While it may not be perfect it gives you an entire stop of really nice highlight rolloff if nothing else
I understand where you are heading Kyle, but there is no way to make this a standardized test. The latitude test exactly reflects your request: to show you whats possible in post.
If you could put your subtitles above the youtube's subtitles, for those who doesn't understand english 100%, it would be nice! thank you for this excellent test CineD, you're like the Opta data for cameras
What was the ISO setting on the footage for the DR test? Yes, I know ISO is not a physical setting but it is a remapping of middle grey. I would suspect, if you set the iso to 6400, you would find about 5 more stops….. Middle grey would be at usable stop 11. This has already been tested with the Xyla 21.
What do you mean 5 more stops? Isn't moving middle grey just about DR allocation, so what you gain in the highlights you lose in the shadow? Yes, I know there are optimal (native) ISO values, but 5 stops will not magically appear...
@@DaCarnival Well, when you consider they are using IMATEST, it isn’t reporting accurately. And then when you watch the under/over tests, they show how there are many more stops of usable DR than their 13 stops says.
they did the test at 800 iso, the raptor vv's native iso. if you went to 6400 iso sure, you'd gain stops in the highlights but you'd lose stops in the shadows. the camera has 13 usable stops of dynamic range no matter where you're allocating your iso/middle grey. i'm not entirely sure what you're suggesting as it seems out of the realm of actual camera performance and disregards the entire spectrum of dynamic range. I'm confused where you're seeing more usable stops than 13. Check out 15:35 where they run through the 13 usable stops.
I saw the under/over exposure test from cvp, and in the under v-raptor was incredible. I know is controversial but Iwould be great if you could do a test for the stitch artifact, this is the only reason I didnt and woudnt buy this camera....
That was a super interesting watch! Happy to hear Gunther's direct impressions and how he puts things in perspective. Regarding the clipped channels and highlight recovery, does the test pattern generates light in D65?
Weird request: Could you do Tests with older Cinema Cameras, like the original C300, C100 or Sony F3 or fs100s? Some of them are gaining popularity again and back when they were released, no one did proper dr tests. Would love to see that. But that’s probably just me😂
Which lenses do you use for the tests? Always the same, right? It seems a bit low-con. But maybe is only my monitor :) And especially, why don't you use NDs or IRNDs to compensate the exposure and rather use iris and also shutter? Thanks for what you do! This is just friendly curiosity, not some sort of critique. I'm a big fan of ya'll!
It would be really interesting to see how much actual special resolution a camera like this records and how that compares to a "lower resolution" camera like the Alexa LF or 35. Like a simple lines per mm chart. I bet the V-Raptor will have resolution failure because of sensor noise and will not significantly out resolve the Alexa 35 or even LF. I think it's so interesting that this test shows once again that the only sensible camera RED makes is the Komodo. 8K is stupid for narrative work and the reason RED is pushing their marketing for the motion picture industry is because there is a lot of clout and money in it. I remember in a marketing call for the Monstro they actually talked about how a fish farmer was using an 8K RED camera on a drone to fly over his ponds and then use the 8K to crop in to see which fish were pregnant. And this was not a "Here's a fun anecdote" kinda thing. This was a serious and proper use case for them why these cameras were the right choice for Hollywood films. The sacrifice of color, noice performance, and latitude for resolution is not what people for narrative work need. And you can also see in the work of most Hollywood filmmakers that use RED that they don't have a full grip on the tools they use and how they work. Listing to John Schwartzman an absolute legend in cinematography talking about resolution, sensor size, and lenses in just painful because he just doesn't know how they actually and how to use them to his advantage. He knows how to shoot and light but he doesn't know how photo sites size and sensor size really influence an image and he just buys what RED tells him without questioning it.
I mean, didn't the tests show that the noise is great and that the red suffers in terms of the highlights as it always has? I mean, it probably won't out significantly resolve the Alexa 35 in terms of spatial resolution just because they're similar pixel pitch anyways, that falls in line with the line pairs per mm part if you do the calculations. And Isn't the komodo pretty noisy? I'd argue the better contender in the sensible red camera bit would be the gemini.
@@zebius4157 It used to be, but by now there are consumer cameras with better high ISO performance than the Gemini and the Alexa 35 seems to be pretty great at that too.
@@zebius4157 Well if you look at interviews of people like John Schwarzman and Henry Braham and make it sound like resolution is the key to color and DR as well. Also some analog people switched to RED simply believing inaccurate scan resolution equivalents and thinking the high resolution will be closest to film. While an Alexa is much closer in terms of DR and noise characteristics. Just look interviews up by people who’re really loyal to red. You gonna pull your hair out at some stuff they think. Like how far they can push in because of the resolution and then forgetting that they’d hit the noise much quicker than on a 4K camera for the small crops they do. Or their absolute non understanding of large format.
@@VariTimo well, of course they're gonna say that if the interview is through Red, they have to, it's marketing more than anything else. Like, are you going to expect people interviewed by Kodak to point out you can get a similar result with digital? I saw a Kodak interview article with Linus sandgren for don't look up, and he said that film is more environmentally friendly than digital because it's not on hard drives, which lmao. Also, are you saying any crop is unusable on higher resolution cameras in general? Theoretically I could understand that but in practice, with the cameras available not just from red but others, you'll be just fine, maybe in the early days of digital cinema cameras but definitely not now, hell, David fincher able to frame in 6k from the dragon and extract to 4k, and that's the dragon. Also, what do you mean "non understanding of large format"? Do they say that that focal lengths literally change when they don't in actuality? Do they say there's a different geometric perspective when the big determining factor is camera placement? Also, you act like this is special to red interviews and is just not a thing that happens with other camera manufacturers, I saw a dude in an interview for Arri claim there was more color with the Alexa 65 because it was a higher resolution, so again, what's so special?
So if I somehow ever get my hands on a v raptor, I should underexpose. Noted. Also does any1 know why red doesn't allow users to turn of highlight recovery?
I would not recommend arbitrarliy underexposing all footage shot on REDCODE RAW. It’s best to keep an eye on the histogram and traffic lights. The traffic lights read the raw data directly off the sensor. When 2% or more of pixels are overexposed, the traffic lights will turn on. When the highlight traffic lights are off and your histogram isn’t all bunched up to the right, you’re okay. You can even rate the camera at ISO 400 instead, effectively overexposing, for cleaner footage. Then, in bright scenes with lots of important highlight detail, raise the ISO to 800. Arri lens specialist Art Adams halved the recommended ISO rating of all his cinema cameras, including the Alexa. This is especially useful for HDR, where noise is much more visible. The CineD tests are only looking at the noise in SDR.
@@JonPais thankyou for your input. But I was half joking. I'm probably never gonna see the raptor in my life lol. For context I shoot on a 100 dollar camera.
If you use the traffic lights and histogram, exposure is crazy simple. The traffic lights read the RAW data directly off the sensor. The highlight traffic lights will turn on when 2% of pixels are overexposed. For best exposure, except in situations where lights or strong reflections are directly hitting the sensor, the highlight traffic lights should never be on.
Does anyone, can anyone recommend a rucksack for the V-Raptor? Body, 2 Lenses med size, Monitor, Top and side handle and cage (broken down ), two batteries and charger. For plane carry on.
The problem I have with your tests is that they are only in full stops. You need a better setup to measure half stops because its a safe bet that Red have better testing gear and they also know the best ISO to test at, even though Reds dont have a fixed ISO, the ISO setting affects dynamic range. I see half a stop at both ends on your charts even after you threw out one stop.
No idea where you get the idea that we are only measuring in half stops - we measure in 0.1 stops. Just check our Camera Databases that allow you to view and compare all results: www.cined.com/camera-database/
You are arbitrarily picking start stops and end point stops. Example the Raptor youve dismissed the first stop because you claim lack of colour info, fair enough but you are starting the count from the next full mark on the chart. Your chart isnt showing where the colour kicks in between stops. At the other end you've ignored another stop where it touches bottom in the log yet its still slightly up and being on the log curve thats a stop in my book. @@CineD
It's damn good, especially for the price and form factor, but I wish Red wouldn't try to overcompensate for it's shortcomings via software like Tesla and instead be more forthcoming about it's 13 stops of dynamic range. Also, you have the ability to turn off highlight rolloff in ipp2 settings, why couldn't they have the option to turn off highlight recovery?
![How is the RED V-RAPTOR [X] SO GOOD?!](/img/n.gif)
These follow up conversations are great
I would like to see how these results are adjusted for the new V-Raptor X with Global Shutter?
You guys should add a third score for the DR tests. You have the SNR 1 and 2. But I would LOVE to have a separate doctored version of SNR 2 where you apply noise reduction in post, highlight recovery, etc. That way, people can actually see what these cameras are capable of with solid post-treatment. Right now, the tests favor cameras with solid in-camera noise reduction, which is misleading. The over-under exposure test helps to show the real dynamic range, but it would be cool to have it reflected in a score.
I also think this is a great idea, it gives you an idea for how much you can really pull from the sensor after some work
I think this is a decent idea
Good idea - though maybe including highlight recovery complicates it a bit. I imagine for instance that changing the hue on a sunset shot where highlight recovery is doing a lot of heavy lifting, would be noticeably worse than shifting the hue on a shot where the last stop is full RGB... I just mean that it's not merely a technicality.
@@DaCarnival that is true to an extent, but at least in the case of the blackmagic cams, I’ve found that last stop to be more than usable and a primary reason I shoot blackmagic. While it may not be perfect it gives you an entire stop of really nice highlight rolloff if nothing else
I understand where you are heading Kyle, but there is no way to make this a standardized test. The latitude test exactly reflects your request: to show you whats possible in post.
Thank You very much for this, really love this format. Can't wait to se more cameras.
If you could put your subtitles above the youtube's subtitles, for those who doesn't understand english 100%, it would be nice! thank you for this excellent test CineD, you're like the Opta data for cameras
You can just move the subtitles, if youre talking about youtube generated. Click and drag anywhere
Ursa Cine 12K FF ?
Fascinating content guys, really enjoyed this!
Thank you guys
That guy knows what he is talking about :)
quite reasonable result, thanks for the explanation
Very impressive for the size of the camera.
Slightly strange question - Which Lapel mic are you using with the DJI Mic? Looks like it has a locking 3.5mm jack and I can't get mine to work!
What was the ISO setting on the footage for the DR test? Yes, I know ISO is not a physical setting but it is a remapping of middle grey. I would suspect, if you set the iso to 6400, you would find about 5 more stops….. Middle grey would be at usable stop 11. This has already been tested with the Xyla 21.
Very true
What do you mean 5 more stops? Isn't moving middle grey just about DR allocation, so what you gain in the highlights you lose in the shadow? Yes, I know there are optimal (native) ISO values, but 5 stops will not magically appear...
@@DaCarnival Well, when you consider they are using IMATEST, it isn’t reporting accurately. And then when you watch the under/over tests, they show how there are many more stops of usable DR than their 13 stops says.
@@DaCarnival this is correct
they did the test at 800 iso, the raptor vv's native iso. if you went to 6400 iso sure, you'd gain stops in the highlights but you'd lose stops in the shadows. the camera has 13 usable stops of dynamic range no matter where you're allocating your iso/middle grey. i'm not entirely sure what you're suggesting as it seems out of the realm of actual camera performance and disregards the entire spectrum of dynamic range. I'm confused where you're seeing more usable stops than 13. Check out 15:35 where they run through the 13 usable stops.
I saw the under/over exposure test from cvp, and in the under v-raptor was incredible. I know is controversial but Iwould be great if you could do a test for the stitch artifact, this is the only reason I didnt and woudnt buy this camera....
That's also what 'scares me' to use it in projects. Would be great to have some updates.
Can u please compare the low light capabilities with low light of the glorious Gemini?
Would love if you could get your hands on the MAVO EDGE 8K
Me too
That was a super interesting watch! Happy to hear Gunther's direct impressions and how he puts things in perspective.
Regarding the clipped channels and highlight recovery, does the test pattern generates light in D65?
Weird request: Could you do Tests with older Cinema Cameras, like the original C300, C100 or Sony F3 or fs100s? Some of them are gaining popularity again and back when they were released, no one did proper dr tests. Would love to see that. But that’s probably just me😂
They did some older cameras pre-2018. I think it's fair to assume they were almost all SNR=2
Which lenses do you use for the tests? Always the same, right? It seems a bit low-con. But maybe is only my monitor :) And especially, why don't you use NDs or IRNDs to compensate the exposure and rather use iris and also shutter? Thanks for what you do! This is just friendly curiosity, not some sort of critique. I'm a big fan of ya'll!
Latitude tests should be performed under relatively flat lighting. ND filters will introduce color shifts, so not an option.
@@JonPais I mean, really good ND shouldn't. I. E. Tiffen. That's why they're called, neutral, right? 😊
@@gabrielebuffa2520 You’re kidding, right?
It would be really interesting to see how much actual special resolution a camera like this records and how that compares to a "lower resolution" camera like the Alexa LF or 35. Like a simple lines per mm chart. I bet the V-Raptor will have resolution failure because of sensor noise and will not significantly out resolve the Alexa 35 or even LF.
I think it's so interesting that this test shows once again that the only sensible camera RED makes is the Komodo. 8K is stupid for narrative work and the reason RED is pushing their marketing for the motion picture industry is because there is a lot of clout and money in it. I remember in a marketing call for the Monstro they actually talked about how a fish farmer was using an 8K RED camera on a drone to fly over his ponds and then use the 8K to crop in to see which fish were pregnant. And this was not a "Here's a fun anecdote" kinda thing. This was a serious and proper use case for them why these cameras were the right choice for Hollywood films. The sacrifice of color, noice performance, and latitude for resolution is not what people for narrative work need. And you can also see in the work of most Hollywood filmmakers that use RED that they don't have a full grip on the tools they use and how they work. Listing to John Schwartzman an absolute legend in cinematography talking about resolution, sensor size, and lenses in just painful because he just doesn't know how they actually and how to use them to his advantage. He knows how to shoot and light but he doesn't know how photo sites size and sensor size really influence an image and he just buys what RED tells him without questioning it.
I mean, didn't the tests show that the noise is great and that the red suffers in terms of the highlights as it always has? I mean, it probably won't out significantly resolve the Alexa 35 in terms of spatial resolution just because they're similar pixel pitch anyways, that falls in line with the line pairs per mm part if you do the calculations. And Isn't the komodo pretty noisy? I'd argue the better contender in the sensible red camera bit would be the gemini.
Also, how would you say that most Hollywood red users "don't have a full grip on the tools they use and how they work"?
@@zebius4157 It used to be, but by now there are consumer cameras with better high ISO performance than the Gemini and the Alexa 35 seems to be pretty great at that too.
@@zebius4157 Well if you look at interviews of people like John Schwarzman and Henry Braham and make it sound like resolution is the key to color and DR as well. Also some analog people switched to RED simply believing inaccurate scan resolution equivalents and thinking the high resolution will be closest to film. While an Alexa is much closer in terms of DR and noise characteristics. Just look interviews up by people who’re really loyal to red. You gonna pull your hair out at some stuff they think. Like how far they can push in because of the resolution and then forgetting that they’d hit the noise much quicker than on a 4K camera for the small crops they do. Or their absolute non understanding of large format.
@@VariTimo well, of course they're gonna say that if the interview is through Red, they have to, it's marketing more than anything else. Like, are you going to expect people interviewed by Kodak to point out you can get a similar result with digital? I saw a Kodak interview article with Linus sandgren for don't look up, and he said that film is more environmentally friendly than digital because it's not on hard drives, which lmao. Also, are you saying any crop is unusable on higher resolution cameras in general? Theoretically I could understand that but in practice, with the cameras available not just from red but others, you'll be just fine, maybe in the early days of digital cinema cameras but definitely not now, hell, David fincher able to frame in 6k from the dragon and extract to 4k, and that's the dragon. Also, what do you mean "non understanding of large format"? Do they say that that focal lengths literally change when they don't in actuality? Do they say there's a different geometric perspective when the big determining factor is camera placement? Also, you act like this is special to red interviews and is just not a thing that happens with other camera manufacturers, I saw a dude in an interview for Arri claim there was more color with the Alexa 65 because it was a higher resolution, so again, what's so special?
Do for mavo edge 8k
Will you make a sortable table with all of your results combined? Sort of how DXOmark has. So we can compare every camera together
Their website was recently overhauled and you can compare different cameras and sort them by their different properties if I'm not mistaken
Here we go: www.cined.com/databases/
Thank you!
So if I somehow ever get my hands on a v raptor, I should underexpose. Noted.
Also does any1 know why red doesn't allow users to turn of highlight recovery?
I would not recommend arbitrarliy underexposing all footage shot on REDCODE RAW. It’s best to keep an eye on the histogram and traffic lights. The traffic lights read the raw data directly off the sensor. When 2% or more of pixels are overexposed, the traffic lights will turn on. When the highlight traffic lights are off and your histogram isn’t all bunched up to the right, you’re okay. You can even rate the camera at ISO 400 instead, effectively overexposing, for cleaner footage. Then, in bright scenes with lots of important highlight detail, raise the ISO to 800. Arri lens specialist Art Adams halved the recommended ISO rating of all his cinema cameras, including the Alexa. This is especially useful for HDR, where noise is much more visible. The CineD tests are only looking at the noise in SDR.
@@JonPais thankyou for your input. But I was half joking. I'm probably never gonna see the raptor in my life lol. For context I shoot on a 100 dollar camera.
the moire shirt :D:D:D
So basically you can only overexpose highlights by 2 stops, and let the shadow detail fall in place.
If you use the traffic lights and histogram, exposure is crazy simple. The traffic lights read the RAW data directly off the sensor. The highlight traffic lights will turn on when 2% of pixels are overexposed. For best exposure, except in situations where lights or strong reflections are directly hitting the sensor, the highlight traffic lights should never be on.
Does anyone, can anyone recommend a rucksack for the V-Raptor? Body, 2 Lenses med size, Monitor, Top and side handle and cage (broken down ), two batteries and charger. For plane carry on.
Highlights suck but much better shadow performance than the Alexa.
Arri alexa better in terms of color depth compared red camera !!!.
The problem I have with your tests is that they are only in full stops. You need a better setup to measure half stops because its a safe bet that Red have better testing gear and they also know the best ISO to test at, even though Reds dont have a fixed ISO, the ISO setting affects dynamic range.
I see half a stop at both ends on your charts even after you threw out one stop.
No idea where you get the idea that we are only measuring in half stops - we measure in 0.1 stops. Just check our Camera Databases that allow you to view and compare all results: www.cined.com/camera-database/
You are arbitrarily picking start stops and end point stops. Example the Raptor youve dismissed the first stop because you claim lack of colour info, fair enough but you are starting the count from the next full mark on the chart. Your chart isnt showing where the colour kicks in between stops. At the other end you've ignored another stop where it touches bottom in the log yet its still slightly up and being on the log curve thats a stop in my book. @@CineD
It's damn good, especially for the price and form factor, but I wish Red wouldn't try to overcompensate for it's shortcomings via software like Tesla and instead be more forthcoming about it's 13 stops of dynamic range. Also, you have the ability to turn off highlight rolloff in ipp2 settings, why couldn't they have the option to turn off highlight recovery?
Tesla? 🤣🤣🤣
@@JonPais alright, just so I know, who are you arguing in favor of? Red or Tesla?
It's the color information in the highlights bit that makes me wish red would stop it's bullshit and make a monochrome vraptor already.
jezz .. stop comparing a 24,000 camera with a 90,000 camera…
Arri alexa better in terms of color depth compared red camera !!!.