Tom Pollock Posted January 6, 2011 Share Posted January 6, 2011 Does Alexa high dynamic range process work and how so? Thanks for any tips. Link to comment Share on other sites More sharing options...
Premium Member Adrian Sierkowski Posted January 6, 2011 Premium Member Share Posted January 6, 2011 Turn on and hit record ;) Seriously though, no. The Alexa has a very wide range out of the box and as such does not incorporate an HDR mode when shooting as this time nor does it really need one in my own opinion. Link to comment Share on other sites More sharing options...
Mitch Gross Posted January 6, 2011 Share Posted January 6, 2011 The ALEXA double-pulls from each photosite using dual drivers that operate simultaneously, not sequentially. The result is an imager with more than 14 stops of usable Dynamic Range (I can see more than 15 on a scope above the noise floor) and no temporal artifacting. Link to comment Share on other sites More sharing options...
Premium Member Adrian Sierkowski Posted January 6, 2011 Premium Member Share Posted January 6, 2011 Hey Mitch, didn't know that it had dual driver.. that's interesting.. would it be possible, do you think, to tweak those drivers even more beyond factory spec, should Arri wish to, or allow users to do it so as to make a form of HDR, recording etc? Just theoretically, of course. Link to comment Share on other sites More sharing options...
Mitch Gross Posted January 6, 2011 Share Posted January 6, 2011 The camera as designed is set to deliver an optimized image. Push them further and one would likely introduce noise or other issues. ARRI's engineers are continually improving and updating the camera with new firmware releases. If it can be improved they will do it for everyone. Coming next month is 3200ISO on the camera. Link to comment Share on other sites More sharing options...
Tom Pollock Posted January 6, 2011 Author Share Posted January 6, 2011 thanks Mitch, yes, I am refering to the double pull system, can you set a different exposure for the same frame, for instance, one frame underexposed and one frame overexposed. Is there special drives required for the double pull system? Link to comment Share on other sites More sharing options...
Mathew Rudenberg Posted January 6, 2011 Share Posted January 6, 2011 The ALEXA double-pulls from each photosite using dual drivers that operate simultaneously, not sequentially. The result is an imager with more than 14 stops of usable Dynamic Range (I can see more than 15 on a scope above the noise floor) and no temporal artifacting. That is rather interesting - if the drivers are pulling from the same photosite over the same period of time how do they capture different exposures? I can understand the Epic approach of varying the shutter speed but this seems to be something altogether different... Link to comment Share on other sites More sharing options...
Premium Member Vincent Sweeney Posted January 6, 2011 Premium Member Share Posted January 6, 2011 Is a DR range beyond 15-16 stops really that usable anymore? Obviously I can see where it might be useful in some certain shot that a film may encounter but after a while, things look too unnatural and thus, likely distracting. The HDR race seems less useful from here on out compared to things like ergonomics, color depth, highlight roll-off appeal, focus assisting improvements, etc. Link to comment Share on other sites More sharing options...
Premium Member David Mullen ASC Posted January 6, 2011 Premium Member Share Posted January 6, 2011 If you are going to replace color negative film with a digital equivalent, it has to be somewhat HDR , "semi" HDR. Film has around 15-stops of DR with less contrast in the shoulder and toe of the characteristic curve. To emulate that, you need a pretty wide DR. To get highlights to "roll off" rather than get clipped off, you need some excess information at the high end that you can compress into a flatter gamma curve. And I believe that color depth and dynamic range are somewhat interconnected, more exposure information means more color information as well. Sure, a 20-stop HDR would be sort of special use situations, most of the time people won't need to use it. But how often have we even been on a FILM shoot where we wished we could shoot just by window light in some restaurant or diner and yet still hold detail outside the windows, the way your eye sees the room, rather than light the room to simulate that effect? Just today I was in an old desert diner on a scout and thinking how I wished I didn't have to add light through the windows to balance the interior and exterior because it looked so nice to my eyes. Link to comment Share on other sites More sharing options...
Premium Member Vincent Sweeney Posted January 6, 2011 Premium Member Share Posted January 6, 2011 I guess with 20 stops available, the colorist's job will become that much more important, esp. for rushed shoots. I have yet to see an HDR still with a human face in it that looked acceptable, which makes me wonder what trends will come along soon. Link to comment Share on other sites More sharing options...
georg lamshöft Posted January 6, 2011 Share Posted January 6, 2011 @Mathew Rudenberg The analogue sensor signal is amplified before the AD-conversion. These amplifiers are also used to change ASA in most cameras, that's why it differs from pushing the image in post. AD-converters are defined by many different parameters (bit depth, noise, speed...) that are often not easily understandable - just putting a high-bit-converter behind the amplifier will create high-bit files (14bit, 16bit) but not necessarily a better (more dynamic range, better tonal transitions, lower noise) image. ALEXA reads out two amplifier signals per pixel, one is a high-gain (strong amplification), the other one is a low-gain signal. They use 14bit-ADCs to convert the two signals and merge them into one true 16bit-file. What I'm tryiing to say: The exact reason why they use this specific technique remains unclear, theoretically, a low-noise amplifier and a high-quality 16bit-ADC could extract the very same low-noise, high-DR signal from the photosite. But maybe those amplifiers or ADCs are not available or they use the redundant information within the two 14bit-conversions to reduce random noise? We don't know. Since the sensor-readout is fast enough to process more than two full-readouts within 1/24s, HDR "RED-style" would be an easy thing to do - just shooting 2 times the frame rate and altering exposure every second readout - basically every CMOS-sensor can do that. Then you just need a software to merge two images into one - no magic here. But how do you get rid of the temporal artifacts? You don't - at least never all of it. Link to comment Share on other sites More sharing options...
Premium Member John Sprung Posted January 6, 2011 Premium Member Share Posted January 6, 2011 ..., theoretically, a low-noise amplifier and a high-quality 16bit-ADC could extract the very same low-noise, high-DR signal from the photosite. No doubt they tried a single amplifier, which is what the other camera makers use, and found that the results from the bi-amp setup were significantly better. It's not easy to get all three, low noise, high dynamic range, and high gain, all from one analog amplifier. It's sort of analogous to the reason we have woofers and tweeters in audio. -- J.S. Link to comment Share on other sites More sharing options...
Premium Member Phil Rhodes Posted January 6, 2011 Premium Member Share Posted January 6, 2011 Further to John's comments above, it really is a case of "why don't they just..." Well, maybe they don't because it's actually quite hard to make a video amplifier to run at HD bandwidths that does everything at once... Link to comment Share on other sites More sharing options...
georg lamshöft Posted January 7, 2011 Share Posted January 7, 2011 They ended up with the lowest-noise, highest DR and best tonal range digital camera system around. So why should we blame them for this unusual technical approach? We should rather ask why other manufacturers don't incorporate this technology... Link to comment Share on other sites More sharing options...
Tom Pollock Posted January 7, 2011 Author Share Posted January 7, 2011 Thank you everyone for your input, what I was trying to inquire about is related to Mitch's comment about how the Alexa double pulls from each photosite simultaneously. I am trying to achieve a look known in the still world as High Dynamic Range, where two exposures are taken, one three stops over, and one three stops under, the two shots are put together creating an interesting look. I had heard that this might be possible in the Alexa rather than utilizing a beam splitter and two cameras. Link to comment Share on other sites More sharing options...
Premium Member John Sprung Posted January 7, 2011 Premium Member Share Posted January 7, 2011 I am trying to achieve a look known in the still world as High Dynamic Range, where two exposures are taken, one three stops over, and one three stops under, the two shots are put together creating an interesting look. The overview is that things go from photons to electrons to bits. What you're thinking about is going two ways all the way from the photon part of it, and putting it together as bits. That's not quite what Arri does, they split it up at the electron stage, and put it back together as bits. -- J.S. Link to comment Share on other sites More sharing options...
Premium Member Peter Moretti Posted January 7, 2011 Premium Member Share Posted January 7, 2011 No doubt they tried a single amplifier, which is what the other camera makers use, and found that the results from the bi-amp setup were significantly better. It's not easy to get all three, low noise, high dynamic range, and high gain, all from one analog amplifier. It's sort of analogous to the reason we have woofers and tweeters in audio. -- J.S. Maybe you're on to something there John, not surprising ;). As many stereophile setups use two amplifiers, one for the tweeters and one for the woofers, a.k.a. bi-amping. Link to comment Share on other sites More sharing options...
Mitch Gross Posted January 7, 2011 Share Posted January 7, 2011 I think the term "bi-amping" is a great way to describe what ARRI does on the ALEXA. Tom, sorry but it is not set up to do what you are talking about doing. Link to comment Share on other sites More sharing options...
Premium Member John Sprung Posted January 7, 2011 Premium Member Share Posted January 7, 2011 Maybe you're on to something there John, not surprising ;). As many stereophile setups use two amplifiers, one for the tweeters and one for the woofers, a.k.a. bi-amping. Thanks -- It's not quite a perfect analogy, since in audio you're dividing up high and low frequencies, not the loud and soft parts.... -- J.S. Link to comment Share on other sites More sharing options...
Premium Member Peter Moretti Posted January 8, 2011 Premium Member Share Posted January 8, 2011 Thanks -- It's not quite a perfect analogy, since in audio you're dividing up high and low frequencies, not the loud and soft parts.... -- J.S. True, but what analogies are perfect ;). BTW, I believe stereophiles bi-amp speakers b/c the low frequencies take a lot more energy to create than do the high frequencies. Makes sense when you compare a woofer's movement to a tweeter's vibration. Apparently the energy drain caused by powering the woofer causes some type of distortion that bleeds into the high frequencies. Also the quality of the amplification need apparently differs w/ speaker driver type. Woofers need fast, clean, copious power. Solid state amps can provide just that. Tweeters need more warm smooth amplification (hope I'm not anthropomorphizing too much) and tube amps provide just that. How this all relates to the Alexa, I'm not sure. Link to comment Share on other sites More sharing options...
Jamie Metzger Posted January 11, 2011 Share Posted January 11, 2011 The camera as designed is set to deliver an optimized image. Push them further and one would likely introduce noise or other issues. ARRI's engineers are continually improving and updating the camera with new firmware releases. If it can be improved they will do it for everyone. Coming next month is 3200ISO on the camera. 3200 seems unnecessary to me. I've heard requests for slower ISO's ;) Link to comment Share on other sites More sharing options...
Mitch Gross Posted January 11, 2011 Share Posted January 11, 2011 I'll mention it to them. Of course one can't ND to a higher ISO ... Link to comment Share on other sites More sharing options...
Chris Millar Posted January 11, 2011 Share Posted January 11, 2011 Help me understand it... you have a photosite value - its goes through two independent amplifiers - resulting in two values, one low and one high so they intrinsically connected by the value of the high amplification factor divided by the low ? you chop out the high values from the low signal and the low from the high - squish it about so they meet nicely in the middle and also have nice heads, shoulders, knees and toes Why wasn't this done ages ago ? Link to comment Share on other sites More sharing options...
Stuart Brereton Posted January 11, 2011 Share Posted January 11, 2011 3200 seems unnecessary to me. I've heard requests for slower ISO's ;) Indeed, I'd like to see a 50asa setting for digital cinema cameras. It would be great to be able to shoot in bright sunshine without 5 or 6 stops of ND Link to comment Share on other sites More sharing options...
Premium Member John Sprung Posted January 11, 2011 Premium Member Share Posted January 11, 2011 Why wasn't this done ages ago ? Really good ideas often seem obvious in hindsight. But this one required two things to become a reality: An engineer smart enough to think it up, and a company willing to spend the money to develop it in silicon. -- J.S. Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now