I can't tell you how much I appreciate comparisons like this. Sometimes there is nothing like seeing the footage right next to each other. I make my living shooting stills but video is becoming a larger component of that living. With the excitement of the D5 and 1DX Mark II, my next camera for video is going to be the Sony FS7. I am tired of wonky workflows, low fps, motion jpeg, lack of log and always a disregard for professional video results. I have been studying the cameras and this one is it. I appreciate your post. Thanks!
The Blackmagic Camera looks like it was in video mode, to achieve the full dynamic range you have to shoot at the stated native iso and shoot in film mode. i shoot with blackmagic cameras & the footage doesn't look right to me...
+Derron Campbell Everything I've seen suggests of Blackmagic cameras suggests that it was shot in video mode. I know the Ursa has less dynamic range (12 stops I think) due to global shutter but this doesn't look right to me at all. By the way, there is a similar setup done with an Alexa and a BMCC on youtube and the results are much more what I would expect.
A better way to see the dynamic range performance of the cameras would be to shoot raw and color grade the under and over exposures to match the proper exposure. It is really hard to understand this way.
+Matt Black no they're not...the "new" sensor (4.6K) that will be featured in the new URSA-mini is faaaaar superior to the one tested here, with a latitude of 15-stops.
Not being a fan boy by any stretch. I own two of the cameras in this video. Despite what the comments say, this IS a test of dynamic range. But it's testing for absolute worst case scenario i.e. your camera operator somehow forgets to adjust something so badly that you over or underexpose by FIVE stops. That being said, it is not a good test of dynamic range by any means. A good test is a properly exposed scene with very dark parts and very bright spots. Seeing which camera can resolve better under those situations AFTER color correction is the really the only thing that matters.
Ilford HP5 S pushed to 1000 ASA. (13-14 stops). No job after the take. Yes, B&W film, Breathless 1960. ASC was Raoul Coutard . The technique was derived from the colonial wars coverage (Indochina) were Coutard learned how to work with minimum gear. He used some light bulbs for inside takes... like 400W. That's what u get inside with indirect light : around 800 ASA 2.8 to 4.0 @1/48s (depends on ur lens). So basic ISO for Alexa and Amira today is 800 ASA.
I'm absolutely in love with the Arri Alexa/Amira 'look' --The way they deal with highlight roloff is incredibly filmic. Would you say that the FS7 comes pretty close in dynamic range to the Arris? I've yet to see any 'real world' 'direct sunlight' comparisons of this. According to your test here, the FS7 and the Arri are on-par when it comes to dynamic range. But, all of the other FS7 tests I'm seeing on TH-cam and Vimeo all just feel so 'video' - I think it's because people aren't grading the footage right. Think you could ever do an Amira/Alexa/Fs7 outdoor comparison?
Octopus Films I was also wondering about this, I own an FS7 and regularly shoot on an URSA so was very surprised by these results. What settings were the different cameras on? It would be interesting to know that information for comparison.
Something that was pointed out in the comments, that I'd be most interested in seeing... is a test where you adjust the exposure levels by way of physical light output, rather than maintaining uniform lighting for all tests. It would appear that you adjusted exposure by modifying Aperture and/or possibly by way of ISO manipulation/or even shutter speed... i.e. camera settings. In my real world use, I'm going to have a target aperture, fixed shutter speed, and ideal native ISO. My exposure adjustments for the most part, will be done by way of light output (dimmer), manipulation (flags & diffusion), and light filtration (ND). So while I find this test intriguing, it doesn't do much for me as far as real-world situations and how these perform in an underexposed environment (due to lack of light) and or an Overexposed environment (excess levels of light). My assumption is that the sensors behave quite differently in these scenarios and this test was more of a precursor or variant to a more accurate way of demonstrating. I do thank you for this test, it was very well put together.
What did you expose for? the guy, the wall, the chart, the bulb, what was your point of reference for "proper exposure"?? appreciate the test but this is very much a light by monitor typical comparison, would be great to see this metered. You could set a light up 10ft away from the subject pumping out a controlled 100-300 foot candles, pick the control camera (Amira) and then rate the other cameras in context with what the Amira found middle grey at, then from there we can see 1, 2, 3, 4 stop over/under. Thanks again for the comparison.
+Anthony Brown We used a light meter and exposed for the guys face. Thanks for your comments, we will certainly take it in to consideration if we shoot another test like this.
+Ross Leeson you may always do that, I certainly wouldn't in every single case but I would in this instance, regardless we have no idea what "proper exposure" is, was it a waveform middle grey reading of the face (which would be under exposed) was it a spot or waveform reading of a grey card that was angled towards the key, was it an incident reading pointed at the key or straight back to the camera, what is the lighting setup, what foot candle readings are you getting and at what distance to the subject are they and then what exposure are you getting? You see, one of these approaches is how it would be approached on a narrative film set or within a sit down corporate interview but for testing you would first establish a ratio of light via distance (eg two 300w fresnels, one 6 ft away and the other 9ft), you would take a foot candle reading of the light and note that down lastly you would take an incident reading with the meter facing the camera dude
Oh c'mon, the C300 was worse than the Ursa...look at 5 stops over and you can still see the stand on the Ursa shot but it's pretty much completely gone in the C300 shot. The biggest take away for me is: all of them look horrible at 5 stops over and under, so save yourself some money, get the cheapest camera and light and expose correctly. In the end it's the story that counts, people don't go to movies to look at dynamic range.
One of the most revealing things about this is how it shows how Sony rate their dynamic range, when Arri say the Amira is 14 stops, and Sony say the FS7 is 14 stops, and the Amira clearly has more dynamic range than the FS7, there is a lot of marketing bollocks going around!
nope the new one (4.6k / Pro) has 15 stops and it rly looks how it should. And even the big 4k ursa should have done far better. The profile looks not flat at all. I think they shot in video not film mode with a less flat image and dynamic range. With tilm mode and native iso it compares to mid range red cameras and the new one comes close to the alexa. Watch the comparison by The Slanted Lense showing they compare very well unless you don't over oder under expose for more than 4 stops +/- there the ursa reaches it's limit but that is also the reason why it costs only 5500$ and not 100,000$
There is no comparison of DR between the C300/C-Log and C300 MII/C-Log 2. The C-300 is not relevant to a DR test conducted in 2015. Considering the other cameras in this test, the C300 MII would have been the correct camera for comparison.
I may not know a lot about the way the electronics working inside the cameras but if there are differences, this variable can be eliminated by getting a LED light and changing the source light instead of changing the F stop (sorry is that what you did?). Anyway, I would not go up too much in light intensity since in most cases if that's done is deliberate and you can choose not do it, although I can see the importance of showing let's say a bright day window with outside content showing well and the person's face standing in front of the window in a relatively dark room would have more information on their face. That kind of stuff is only beautiful on the film stock as far as I know. But there are lots of low light scenarios and it's good to compare the strength of the cameras (AND THE FACT THAT CRANKING UP THE ISO IS NOT GOOD ENOUGH SINCE AS YOU DO THAT YOU ARE REDUCING THE DYNAMIC RANGE). However the base line of each camera is a bit different so they need to be dealt with differently for what gives the proper balance of mid tone and not loosing extreme low or high light areas and then it should be brought to the post and see what the end result is after adjusting the curve. Once in that level a camera is holding up well, then it can set a standard practice.
Hey man, I dont think your methodology is very sound. The passport color checker is a terrible tool for checking dyamic range. You do not have the in between values. I have done 3 dynamic range tests against the fs7, the alexa. the ursa consistently gets 6 stops over at 800 and 7 stops over at 1600 iso. I just it tested against the fs7. the highlights are EXACTLY the same for both cameras at or about 6 stops plus a 1/6 stop. The shadows both go down to 5 stops. Then it is just a qualitative question of how much noise you can take for each camera and which noise pattern is more acceptable to you. I also noticed that at the exact same settings the fs7 exposes a stop brighter at 800EI. don't know what that is about but it probably affected your test results. Also the ursa might be in video mode.
+Gary Alfaro Hay que ser muy idiota para entrar a una conversación en un idioma y hacer preguntas en otro, esperando que realmente te contesten. Si no hablás inglés, buscá una conversación en español o, quedáte aquí y no preguntés nada.
Well, as others have probably pointed out (JamieA242), underexposing or overexposing shots IS NOT how you test dynamic range. By its own definition, testing a camera's dynamic range would involve shooting a scene that contains a wide dynamic range of exposures (e.g., the inside of a dark room with a brightly lit window). That would test how well the camera handles the darkly lit to brightly lit areas in the shot (its dynamic rage). This tests, are therefore, useless for determining dynamic range.
The Ursa is still better than C300. Don't know how you hot it to perform that badly. Here is a link comparing the BMCC with F5 in what seems to be a very easy to judge test: th-cam.com/video/uWssco9kDHg/w-d-xo.html I know the BMCC is rated 13 stops instead of 12 of the URSA, but making it look that bad just wasn't convincing for me. We've matched the Epic even with the pocket with hardly any visible difference. The thing is, if you had RED footage in your comparison, it would also look bad without proper post. That's the difference between the BMs and REDs to Sony and Canon, which has no true LOG btw. The amount of highlights you can pull from the BM cameras is actually slightly more than what you can get out of Epic. That's powerful, unlike this test.
People who have access to that kind of money, would spend 20 times more to achieve 10% better results. The cheapest products always win in the price to performance ratio test.
yeah but the fs7 has many functions missing like internal raw recording. the arri cameras are pure machines serving the best result. they maybe have specs or comparison footage that is not so awesome compared to other cameras but even if it is so in practical work they just serv better results the fs7 needs more add ons and is more usable for documentary.
dafuq? xD i love canon and they are sure great for documentary etc. but amira and ursa will do a much better job on shooting narrative. Feeling and dynamic range are nothing compared to the amira
I can't tell you how much I appreciate comparisons like this. Sometimes there is nothing like seeing the footage right next to each other. I make my living shooting stills but video is becoming a larger component of that living. With the excitement of the D5 and 1DX Mark II, my next camera for video is going to be the Sony FS7. I am tired of wonky workflows, low fps, motion jpeg, lack of log and always a disregard for professional video results. I have been studying the cameras and this one is it. I appreciate your post. Thanks!
Excellent work Josh! Thank you for taking the time to make this video.
The Blackmagic Camera looks like it was in video mode, to achieve the full dynamic range you have to shoot at the stated native iso and shoot in film mode. i shoot with blackmagic cameras & the footage doesn't look right to me...
+Derron Campbell Everything I've seen suggests of Blackmagic cameras suggests that it was shot in video mode. I know the Ursa has less dynamic range (12 stops I think) due to global shutter but this doesn't look right to me at all. By the way, there is a similar setup done with an Alexa and a BMCC on youtube and the results are much more what I would expect.
FS7 is fantastic... Can't get enough of it. :-)
Good job! We are very pleased with the fs7
ParaPix It is a great camera! We've only shot a little bit with it, but we love it so far.
A better way to see the dynamic range performance of the cameras would be to shoot raw and color grade the under and over exposures to match the proper exposure. It is really hard to understand this way.
Is there the same lens for each camera?
just ordered ursa mini, and i saw this. what a day
The URSA Mini will better than all these cameras combined. Feel good mate - that is one hell of a camera.
+Marko H the ursa and ursa mini are the exact same sensor, ie the sensor in this video.
+Matt Black no they're not...the "new" sensor (4.6K) that will be featured in the new URSA-mini is faaaaar superior to the one tested here, with a latitude of 15-stops.
+habsrus Sure but that's presumably not the Ursa-mini he ordered. Think before you speak.
Not being a fan boy by any stretch. I own two of the cameras in this video. Despite what the comments say, this IS a test of dynamic range. But it's testing for absolute worst case scenario i.e. your camera operator somehow forgets to adjust something so badly that you over or underexpose by FIVE stops.
That being said, it is not a good test of dynamic range by any means. A good test is a properly exposed scene with very dark parts and very bright spots. Seeing which camera can resolve better under those situations AFTER color correction is the really the only thing that matters.
Exactly.
Preach
Ilford HP5 S pushed to 1000 ASA. (13-14 stops).
No job after the take. Yes, B&W film, Breathless 1960. ASC was Raoul Coutard . The technique was derived from the colonial wars coverage (Indochina) were Coutard learned how to work with minimum gear. He used some light bulbs for inside takes... like 400W. That's what u get inside with indirect light : around 800 ASA 2.8 to 4.0 @1/48s (depends on ur lens). So basic ISO for Alexa and Amira today is 800 ASA.
No, you're referring to a camera's latitude, not its dynamic range I believe. They are related but not the same thing.
I'm absolutely in love with the Arri Alexa/Amira 'look' --The way they deal with highlight roloff is incredibly filmic. Would you say that the FS7 comes pretty close in dynamic range to the Arris? I've yet to see any 'real world' 'direct sunlight' comparisons of this. According to your test here, the FS7 and the Arri are on-par when it comes to dynamic range. But, all of the other FS7 tests I'm seeing on TH-cam and Vimeo all just feel so 'video' - I think it's because people aren't grading the footage right. Think you could ever do an Amira/Alexa/Fs7 outdoor comparison?
Nice! Keep up the good work guys.
What you guys would suggest in terms of dynamic range .... THE fs700 or the ursa mini ?
4.6k is dope at dynamic range but the 4k is rly far away from the 4.6k so if you want 4.6k/pro go for it.
Are the c300 & URSA shooting log? They don't look now where near as flat
Octopus Films I was also wondering about this, I own an FS7 and regularly shoot on an URSA so was very surprised by these results. What settings were the different cameras on? It would be interesting to know that information for comparison.
Octopus Films The Ursa doesn't have a log setting, only what they call "Film" which is the flattest color space. C300 is shooting C-Log.
+Stray Angel Films ok so what mode were you shooting in? I have the URSA 4K and my footage doesn't look this bad. That doesn't look like film mode.
Something that was pointed out in the comments, that I'd be most interested in seeing... is a test where you adjust the exposure levels by way of physical light output, rather than maintaining uniform lighting for all tests.
It would appear that you adjusted exposure by modifying Aperture and/or possibly by way of ISO manipulation/or even shutter speed... i.e. camera settings.
In my real world use, I'm going to have a target aperture, fixed shutter speed, and ideal native ISO. My exposure adjustments for the most part, will be done by way of light output (dimmer), manipulation (flags & diffusion), and light filtration (ND).
So while I find this test intriguing, it doesn't do much for me as far as real-world situations and how these perform in an underexposed environment (due to lack of light) and or an Overexposed environment (excess levels of light). My assumption is that the sensors behave quite differently in these scenarios and this test was more of a precursor or variant to a more accurate way of demonstrating.
I do thank you for this test, it was very well put together.
Thanks for the test! :) Am pretty impressed by the amira. What lens was used on the amira btw?
Would be nice to see the footage graded. Or at least a DL link so we could grade it ourselves for comparison.
What did you expose for? the guy, the wall, the chart, the bulb, what was your point of reference for "proper exposure"?? appreciate the test but this is very much a light by monitor typical comparison, would be great to see this metered. You could set a light up 10ft away from the subject pumping out a controlled 100-300 foot candles, pick the control camera (Amira) and then rate the other cameras in context with what the Amira found middle grey at, then from there we can see 1, 2, 3, 4 stop over/under. Thanks again for the comparison.
+Anthony Brown We used a light meter and exposed for the guys face. Thanks for your comments, we will certainly take it in to consideration if we shoot another test like this.
You always expose the face in a headshot dude
You always expose the face in a headshot dude
+Ross Leeson you may always do that, I certainly wouldn't in every single case but I would in this instance, regardless we have no idea what "proper exposure" is, was it a waveform middle grey reading of the face (which would be under exposed) was it a spot or waveform reading of a grey card that was angled towards the key, was it an incident reading pointed at the key or straight back to the camera, what is the lighting setup, what foot candle readings are you getting and at what distance to the subject are they and then what exposure are you getting? You see, one of these approaches is how it would be approached on a narrative film set or within a sit down corporate interview but for testing you would first establish a ratio of light via distance (eg two 300w fresnels, one 6 ft away and the other 9ft), you would take a foot candle reading of the light and note that down lastly you would take an incident reading with the meter facing the camera dude
Oh c'mon, the C300 was worse than the Ursa...look at 5 stops over and you can still see the stand on the Ursa shot but it's pretty much completely gone in the C300 shot.
The biggest take away for me is: all of them look horrible at 5 stops over and under, so save yourself some money, get the cheapest camera and light and expose correctly. In the end it's the story that counts, people don't go to movies to look at dynamic range.
dynamic range is important. Without it the audience wont see whats going on in certain scenes the camera cant capture.
One of the most revealing things about this is how it shows how Sony rate their dynamic range, when Arri say the Amira is 14 stops, and Sony say the FS7 is 14 stops, and the Amira clearly has more dynamic range than the FS7, there is a lot of marketing bollocks going around!
what's even more funny is that they claim the Ursa has 15 stops of dynamic range LMAO. it was the worse
nope the new one (4.6k / Pro) has 15 stops and it rly looks how it should. And even the big 4k ursa should have done far better. The profile looks not flat at all. I think they shot in video not film mode with a less flat image and dynamic range. With tilm mode and native iso it compares to mid range red cameras and the new one comes close to the alexa. Watch the comparison by The Slanted Lense showing they compare very well unless you don't over oder under expose for more than 4 stops +/- there the ursa reaches it's limit but that is also the reason why it costs only 5500$ and not 100,000$
How does the C300mk2 (and C-Log2 and 3) compare with the previous C300?
There is no comparison of DR between the C300/C-Log and C300 MII/C-Log 2. The C-300 is not relevant to a DR test conducted in 2015. Considering the other cameras in this test, the C300 MII would have been the correct camera for comparison.
A transmission step wedge is the way to go to measure dynamic range.
I may not know a lot about the way the electronics working inside the cameras but if there are differences, this variable can be eliminated by getting a LED light and changing the source light instead of changing the F stop (sorry is that what you did?). Anyway, I would not go up too much in light intensity since in most cases if that's done is deliberate and you can choose not do it, although I can see the importance of showing let's say a bright day window with outside content showing well and the person's face standing in front of the window in a relatively dark room would have more information on their face. That kind of stuff is only beautiful on the film stock as far as I know. But there are lots of low light scenarios and it's good to compare the strength of the cameras (AND THE FACT THAT CRANKING UP THE ISO IS NOT GOOD ENOUGH SINCE AS YOU DO THAT YOU ARE REDUCING THE DYNAMIC RANGE). However the base line of each camera is a bit different so they need to be dealt with differently for what gives the proper balance of mid tone and not loosing extreme low or high light areas and then it should be brought to the post and see what the end result is after adjusting the curve. Once in that level a camera is holding up well, then it can set a standard practice.
Hey man, I dont think your methodology is very sound. The passport color checker is a terrible tool for checking dyamic range. You do not have the in between values. I have done 3 dynamic range tests against the fs7, the alexa. the ursa consistently gets 6 stops over at 800 and 7 stops over at 1600 iso. I just it tested against the fs7. the highlights are EXACTLY the same for both cameras at or about 6 stops plus a 1/6 stop. The shadows both go down to 5 stops. Then it is just a qualitative question of how much noise you can take for each camera and which noise pattern is more acceptable to you. I also noticed that at the exact same settings the fs7 exposes a stop brighter at 800EI. don't know what that is about but it probably affected your test results. Also the ursa might be in video mode.
hola me comparten un email para hacerles unas preguntas sobre cámaras para circuito cerrado y cine ?? gracias
+Gary Alfaro Hay que ser muy idiota para entrar a una conversación en un idioma y hacer preguntas en otro, esperando que realmente te contesten. Si no hablás inglés, buscá una conversación en español o, quedáte aquí y no preguntés nada.
which best camera
fs7 and canon for documentary etc,
ursa for indie/budget narrative
amira big production
that isnt how you test dynamic range.
Well, as others have probably pointed out (JamieA242), underexposing or overexposing shots IS NOT how you test dynamic range. By its own definition, testing a camera's dynamic range would involve shooting a scene that contains a wide dynamic range of exposures (e.g., the inside of a dark room with a brightly lit window). That would test how well the camera handles the darkly lit to brightly lit areas in the shot (its dynamic rage). This tests, are therefore, useless for determining dynamic range.
Doo bad that the VariCam LT wasn’t in there.
I know that Sony sensors dominate the world of photography... But also in film... Thats awesome
The Ursa is still better than C300. Don't know how you hot it to perform that badly. Here is a link comparing the BMCC with F5 in what seems to be a very easy to judge test: th-cam.com/video/uWssco9kDHg/w-d-xo.html
I know the BMCC is rated 13 stops instead of 12 of the URSA, but making it look that bad just wasn't convincing for me. We've matched the Epic even with the pocket with hardly any visible difference. The thing is, if you had RED footage in your comparison, it would also look bad without proper post. That's the difference between the BMs and REDs to Sony and Canon, which has no true LOG btw. The amount of highlights you can pull from the BM cameras is actually slightly more than what you can get out of Epic. That's powerful, unlike this test.
+Marko H +1 for that these words would have been my, if i hadn't found your comment.
What is it that the manufacturers are doing in their tests to be able to achieve such different (misleading) results compared to real world tests?
Why did you guys cut the talking head when the guy is talking about the camera test to a profile shot? It is extremely distracting
How much longer can the giant hipster beard last? How long has it been now? Five years?
i own an ursa.. and i gotta say.. its true.. Dynamic Range is quite an issue
yeah it's only 2000-3000 bucks for ursa or ursa mini but the dynamic range compared to the 4.6k made me stop thinking the 4k would be the best choice
Yeah, but at "proper exposure", the Ursa looked just as nice as the Amira, for thousands less. Moral: hire a great DP.
ursa looked like a digital camera, amira looked like film.
c300 20,000 iso
So crual :/ ! Good job.
Come on...you could buy 10 FS-7s with one Amira??? Arri please keep your camera. Thank you
People who have access to that kind of money, would spend 20 times more to achieve 10% better results. The cheapest products always win in the price to performance ratio test.
And how about we think this way.... Can sony make a camera of Amira quality within 60k price range? :D
yeah but the fs7 has many functions missing like internal raw recording. the arri cameras are pure machines serving the best result. they maybe have specs or comparison footage that is not so awesome compared to other cameras but even if it is so in practical work they just serv better results
the fs7 needs more add ons and is more usable for documentary.
bullshit c300 way better than 60k camera ,what you see is not what we see on youtube
lol you wish
dafuq? xD
i love canon and they are sure great for documentary etc. but amira and ursa will do a much better job on shooting narrative.
Feeling and dynamic range are nothing compared to the amira