Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Dynamic Range

Now here the thing.
11 1/3 sounds and is great but i've no relative measure to judge it.
Of course I haven't even got my camera yet. I know Graeme did this test well, even where he needed to get photoshop involved to be able to see the entire chart on one take.
But my point is what is a relative comparison..what would be the DR of say a Nikon D200 or 300 using the same technique and rigor as what Graeme did?
I guess someone could counter what's the point of comparing it with a D200 as well and I'd have no answer except it's one more tool in modelling the camera. So how would/does the camera DR compare to another known 'still or moving' camera using the same Stouffer technigue?

Dave,

ps: sounds a shame they don't make a Stouffer test strip with the black section like they do for the Chroma du monde where the black is simply a hole opening onto a black felt lined box to reduce reflections.
 
The Stouffer doesn't need the black-hole because it's a transmissive chart. The tranmission of the black card it's mounted on (I'm handy with a craft knife) is essentially zero. We block out all stray light, so the whole chart is effectively sitting in a black hole.

As for a DSLR - I'd expect somewhere between 10 to 12, but it's hard to get linear pixel data out of their RAW software, so that makes it hard to plot the linearity graphs that I do. You could do the same kind of pixel peeing analysis though, but again, as you can see on the DPReview DR test charts, it's very hard to get the RAW software to show all the range the sensor is capable of.

Graeme
 
david farland said:
But my point is what is a relative comparison..what would be the DR of say a Nikon D200 or 300 using the same technique and rigor as what Graeme did?
Click to expand...

Same wedge - Nikon D200.
http://www.dpreview.com/reviews/NikonD200/page22.asp
Canon EOS 40D
http://www.dpreview.com/reviews/canoneos40d/page20.asp

BTW, I'd hazard their estimates seem more conservative than Graeme's.

david farland said:
I guess someone could counter what's the point of comparing it with a D200 as well and I'd have no answer except it's one more tool in modelling the camera.
Click to expand...

david farland said:
So how would/does the camera DR compare to another known 'still or moving' camera using the same Stouffer technigue?
Click to expand...

As said above, my HV20 gets around 9 stops using the similar criteria to Graeme's.

Also, you've seen the extensive stuff on Cinematography.net, right?
http://www.cinematography.net/compare.htm

Note their Red tests were with done with an pre-production sensor, sub-optimal Red Cine settings, etc. Those need to be updated.

Bruce Allen
www.boacinema.com
 
I don't see the CML style test of DR philosophically sound. Geoff does them that way because he's always done them that way, but to me, there's too many variable, too many images to compare, too much that can go wrong. I much prefer the DP review method, which for DSLRs would work better if RAW converters allows you access to the linear sensor data.

If you look at the DPreview links, to "ACR Best" settings for raw, the Canon goes from 5 to 37 (or 32 wedges) and the Nikon from 6 to 40, (or 34 wedges) which is bang in line with the numbers I quoted above :-), 32 being 10.6 stops, and 34 being 11.3 stops. I don't think better RAW software would make the numbers better, but would make measuring them and plotting them a lot easier :-)

Graeme
 
Thanks Graeme. I'll read more on your or others techniques for how to run the Stouffer test but right now I'm guessing the light source is behind the strip it it radiates or tranmits as you say and the box stops stray light. It's a beautiful 25 degree sat morning so I'm trying to supress the inner geek in reading up in Stouffer testing...but until I can see what the Red can do in showing the latitude in shadows and roll off it's highlights I must be content with figures. Talking about figures I always thought it was totally unfair that the eyes could only view 6.5 stops in a single moment but will vary it's iris/chemical response? as it scans a scene to get 20 stops. I'm guessing there's a bucket load of different things that come into play when viewing a moving image as opposed to a still image. Things like action/characters on screen, racking focus etc. Someone was telling me how left and right eyed people will scan a scene differently.

Dave

stoppress: Thanks Bruce!!!
 
Hi Graeme

How does the red compare to the D200 in terms of signal to noise ratio?

Emmanuel
 
Well, DR = SNR. I've not measured a D200, so I don't know precisely.

All the sensors out there, that are used for cinematography are totally linear, which means that when they clip, they hard clip. Therefore all their dynamic range is below that, ie shadow detail if you look at it that way. What you do when you expose is you set the base ISO so as to bias the exposure to give you some stops above a middle point and some stops below. Then you use a tonal curve to map that exposure range down to a view able range that looks nice and contrasty.

Graeme
 
So the lower the ISO, the more headroom below and less above, and vice versa ?

Any chance of being able to create a LUT with those tonal curves in RedCine and then to upload it in the camera for monitoring ?

Emmanuel
 
Hi Graeme

I'd be interested in hearing more about the previously mentioned tool that interpolated clipped channel info from any unclipped channels.. Having some low color accurate luma detail in blown highlights (even if it amount to not much better than clever noise) would be great..

Michael
 
Policar said:
This makes me kind of wary. Looking at the chart, you can see that almost all the latitude is in the shadows and the anecdotal evidence supports this: 3 stops more in the shadows, 3 stops less in the highlights. You could push three stops (ISO 1200?) and have the same exposure latitude in both directions, but then it would be so grainy.

Three stops less in the highlights means you need 8 times less light before blow outs. There are situations (mostly exteriors) where this would pose a real problem. I don't understand why digital has so much trouble holding on to highlight detail. The camera still looks amazing, but this really seems like a limitation. A lot of DP's complain about too much shadow detail in Vision 2 500T, and I suppose this means that the 3 extra stops beyond this aren't too useful under conventional circumstances.

Not a deal-breaker, but this doesn't seem like the best news I've ever heard.
Click to expand...

Policar,

I had some time with a RED at Rivetal Rentals' open house about a week and a half ago. 1200 ISO looked really good! Yes it was grainier than 320 to 500 ISO, but no granier than Vision 2 500 T would have been under the same studio lights. The amount of grain was barely perceptible, and completely acceptable.

-Thor
 
Michael, that control is in RedCine (Highlight) and RedAlert (DRX) and works very well. Nothing much to add other than "try it".

Emmanuel, the in-camera controls give you good LUT control, or in the future you'll be able to do it in our apps and upload to camera.

Graeme
 
kosmos3d said:
The one thing that resonates from all of them is RED is better than film by 3 stops in the shadows and worse than film for 3 stops in the highlights. so the range is basically the same, it is just skewed toward the dark, so a great camera for low light and night shots, much better than film in that case. Not bad for bright stuff, but you must protect for the highlights.
Click to expand...

So if you had lit the camera 3 stops under it would have had exactly the same DR and Sensitivity as the film you compared it to.
You wouldnt have said 'the camera is great for night shots and 'not bad' for bright stuff'. When you would treat it as if it was 640 ASA or higher stock this same camera would suddenly perform better in the highlights and worse in the shadows.

In general I think the image of the Red would be much more film-like if it was treated like 640 ASA or even higher. In post you would need to make a nice knee rolloff for the highlights, make sure nothing hits the 100% when exporting from Redcine or Red Alert.

I think in this RAW processing a lot of people make a lot of mistakes. They probably slide the exposure slider to the left [EDIT: I meant to the right] without realizing ths clips a lot of highlights that were there in the original footage. I've witnessed this happening more than once, and wrong conclusions were drawn from the clipped image.
 
Sliding exposure to the left lowers the exposure, and would reduce clipping, not increase it.

There's a histogram and zebras so you can quickly and easily see clipping.

As for RAW processing, that's an art people will have to learn.....

Graeme
 
That's why there's a "brightness" control to brighten up the image without clipping highlights so much, just crushing them.

Graeme
 
Thor Wixom said:
Policar,

I had some time with a RED at Rivetal Rentals' open house about a week and a half ago. 1200 ISO looked really good! Yes it was grainier than 320 to 500 ISO, but no granier than Vision 2 500 T would have been under the same studio lights. The amount of grain was barely perceptible, and completely acceptable.

-Thor
Click to expand...

Hey Thor. What were you viewing the RED footage on to accurately compare its graininess to that of V2 500 T?
 
Blue said:
Hey Thor. What were you viewing the RED footage on to accurately compare its graininess to that of V2 500 T?
Click to expand...

The crew at Rivetal were capturing 4k files and loading them into Smoke. They were zooming in, color correcting, etc.

It wasn't a side by side comparison between Red and Vision 2 500T. Just an impression from having worked with that film stock and knowing what it looks like in various lighting conditions. There were several there who agreed with me that 1200 ISO looked very similar.

Before my HDV days, that was my film of choice.

I plan to do some actual testing after I get my camera. That should be in about 30 days or less.

I have a client that wants me to show comparisons between Red and several of the popular film stocks, including Vision 2 500T.

I plan to post the results. I think they will be close to my impressions, but we shall see.

-Thor
 
to clarify, my comments come directly from one of the directors that shot with my camera as a test against 35mm, not sure what stock he used, i guess i can find out, but he also is very good friends with the owner of scratch and has used the program for many years and was able to manipulate the footage in a very advanced way in redcine... i am certain he knows what he is talking about, but i guess when you get your cameras and have 15 dinos (lights, not me) working on your job, you will be able to draw your own conclusions.
 
Stephen Webb said:
In the interests of being the first to post in this forum, I thought I'd ask cinematographers out there who've actually tested the camera what they think the dynamic range actually is.

I've yet to see a full and controlled exposure and latitude test, but what I've been hearing recently hasn't sounded too good. In fact one guy whose been testing recently has insisted that there's only 6 stops, something I find hard to believe (that'd be worse than DigiBeta) - but then I'm not a cinematographer.

Has anyone done any thorough tests?
Click to expand...

I recently had a conversation with a director friend regarding DR on Red vs film based on my limited experience with the camera. My conclusion was that because RED's response curve is more linear than film, it all depends on where you put the noise. If you expose so that the noise is present in the image you will have tremendous range in your highlights. If you expose so that there is no noise, you will feel your highlights like on HD. That's not based on any science that's just my feeling having tried to shoot and light a little bit for a Red. Filmic, when exposed low, beautiful video when exposed higher. Almost seems like common sense.

I understand its important to try to quantify this, especially for your light meter. But I think there is a limitation to the usefulness of this quantification as well.

Rest assurred it's much more DR than conventional HD. Somewhere around film. If you expose to have similar amount of grain as film, you will have filmic highlights. That's my first impression.

IBloom
 
You must log in or register to reply here.
Back
Top