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 :)

Question about Dynamic Range

S

sbcooler

Well-known member
Joined
Apr 5, 2008
Messages
153
Reaction score
0
Points
0
Location
New York
Can someone(s) give me the approximate dynamic range of these cameras. I want to understand the difference and I can't seem to find this anywhere. I understand the Red Camera has a theoretical dynamic range of over 11 stops...but in actuality it is more like 8-9...is that about right?

Ok, here's my list..best guesses is fine..just trying to figure this out and how much room for improvement I have over my current setups.

Red Camera, Scarlet (best guess), Sony Z1u, Canon HV20, Sony HC1, Sony Ex-1, Panasonics HVX 3000, Sony F900, Sony F950, Sony F23.

Anyone?
 
To know that, you'd have to test them using the same methodology and factor in that some of these cameras apply gamma curves to the signal, so there are limits to how much you can make all the testing parameters the same.

You'd also have to decide where the cut-off point is in the shadow end where the noise makes the information unusable.
 
Thx, but!

Thx, but!

Hey thanks for saying how to do it...but really what I'm curious about is how the RedOne and Scarlet will compare to my current cameras...better I know..but how much better in Dynamic Range...again I own a Sony Z1u, Canon HV20, Sony HC1 and Sony HC7...I'm curious to how many stops approximately each cam would be rated at?

I see people on here talk about the Sony F23 or the Red this way..I'm curious to how my cameras compare...I know not nearly as good..trying to quantify it.

Thx if anyone can guess at this for my cameras and soon Scarlet-
 
You're going to have to wait until Scarlet is released, or until some kind of test is done with it. It's really not a feasible question to ask at this point unfortunately. All there is right now is pure speculation.
 
What about?

What about?

What about the RedOne's dynamic range compared to the Sony Z1u...how many f-stops is the Sony Z1u would you estimate?

What about the single chip Canon HV20?

Thanks-
 
Thanks, read the site and some articles on the cams...really didn't understand how to compare these cameras and the images to each other.

I am trying to decide what cameras to buy and want to study and thorougly understand all the differences over the next few months before I purchase.

If I could get real world tests from camera people or articles rating how many stops they get under real world circumstance that would be ideal. I did read some of that on the F23 vs RedOne initial test.

I want to know how my Sony Z1u stacks up as far as how many stops, etc.
 
You cannot compare Sony Z1 to RED.

Would you compare BMW/Mercedes, etc... to KIA?

RED's DR is not a miracle, however the rolloff in the whites is great if developed correctly. I find 9 stops in a real world scenario for myself working. Others might see it differently. As David Mullen says, it all depends how much noise you are willing to accept.

RED's strengh is the sensors size and the ability to use 35mm lenses.

Sonys Z1 and the likes are another class.

Hans
 
Thanks Hans,

I know the Sony Z1u is not in the same class as Red. I just want to make sure if I get a RED that there are no down sides to the much less pro cam Sony Z1u. And if there is a downside or any I want to know every one of them before purchasing.

And I am just real curious to how many stops the Sony Z1u and lesser yet Canon HV20 can handle compared to the Red of say 9 stops all other things being equal.

Just a guess would be interesting to me? Thx-
 
Well i hope you also look into other factors because dynamic range is just one of the issues. I would say red gives you a vastly noticeable increase in dynamic range compared to the z1u but haven't tested it. Saying that with the better data via red raw, you are going to push, pull and crush a lot more in post and that will get you better images as a result. There is no doubt in my mind. Add depth of field and the pristine sensor and there is really no comparison. You will be getting 10 times better image with red for sure at 5 times the price. Usually in this industry you get 2 x better image for 5 times the price. Have a look at some of the footage shot with red and then compare it with the z1. In terms of what can be done it is probably better to compare a jeep against a vespa.
 
Hi Sbcooler,

The DR of the Sony Z1 is considerably smaller. I got both and I can say this from my own experience. I never tested the DR of the Sony Z1 since we only use it for casting or layouts.

The Sony's HDV codec makes so much noise and artefacts that no one would say that this camera is a 1080p thing. The REDs oversampling capabilities when downconverted from 4K to 1080p are second to none, excluding high end cameras like F23, D20, etc... (I haven't shot with them, so only hearsay...)

There are may downsides of the RED compared to the Sony Z1:

1. Weight.
2. DOF, focus is much harder to handle.
3. You don't get what you see.
4. Complicated workflow.

I don't have to list the pros, do I?...

Hans
 
I shoot rock concerts and you are SOOOO right about the HDV codec breaking down and pixellating in very large blocks especially when there is changing lighting effects and camera movement which is pretty much all the time! It's almost unusable and looks really unprofessional. If I can get the band to keep the lights on and not flash them/strobe them it's much better, but often they don't want that. And I don't blame them!

I use an Apple Computer so I don't know that the workflow will be hard. The tape based system I use today sucks, so some things might get speeded up...DOF and focus could take some getting use to.

Does the Red have an auto focus for run and gun? I might have to brush up and really practice focusing on the fly in rock concert settings? Also I wonder how fast the lenses are too adjust to really dark scenes and then a spotlight with a really bright comes on. So many lighting challenges.

Is the Red the camera of choice for what I'm am doing....or is it some other make.

The Scarlet might be even better for me because of the smaller sensor (2/3 inch) and easier focus...after all U2 was shot with 18 Sony F950's which uses a 2/3 inch sensor and I'd say that came out pretty good!

I suppose you are saying the detail in highlights and the blacks will be far greater than the Sony Z1u as well? I guess that's a given.

Someone told me the Sony Z1u does 6 or 7 stops or so compared to maybe 9 for the Red...does that sound about right?
 
Sbcooler, you have to evaluate the whole chain, sensor-camera-post.
I have noticed that maybe theoretical RED sensor-camera limit is 11 stops but the final post results is in my case is more than 14 stops.
I made a clip of black granite wall on the bright blue sky and green grass background.
See picture in this post:
http://www.reduser.net/forum/showthread.php?p=203601#post203601
(This sky out there is not blown out, it shows full texture, just my Dell 8 bit screen can’t take the full DR of the 14 bit input, plus jpg compression here.)

What came out of RED was a full 12 bit (12 stops) histogram with both blacks not crashed and whites not blown, thus acceptable. Not without the help of little helper in the release 15 (traffic light and mini histogram that you see real time while shooting) Coming from DLSR background shooting RAW since long time I got used to having such real time indicators on the camera.

Now I converted this clip from 12 bit RED RAW to 16 bit TIFF linear and processed it further.
Well I could stretch these 12 bits of REDRAW all over 16 bit of TIFF but from artistic point of view I ended up with 14 bit of light stretched from end to end just by adding some post processing ,making this clip looking the way I like it.
Make sure that your grading tools works with a minimum 16 bit (steps) resolution, otherwise you just stretch 12 stops of DR across 10 or 8 stops limit in bands all over the 16 bit DR space.
Sounds complicated, yes it does, though people with experience in still photography shooting RAW know what I am talking about.
Now , I could make intentional attempt to stretch from blacks to whites (histogram) across all 16 bits of the available DR of 16 bit TIFF but this will not be necessary desired effect.

So when we are talking film, you are locked in the film DR. When we are talking cameras that do not shoot RAW then you are locked in this camera DR but when we are talking RED capable of shooting RAW, you have to evaluate the whole sensor-camera-post chain and it is getting better not the worse at the end.
 
The RED does have the capability to record longer takes with the REDraid but for handheld work I would NEVER use the REDraid but the CF cards. The REDraid has two HD raid 0 striped in its enclosure and hence is prone to failure due to its concept.
On the contary to the CF cards; they are solid state and will take a fair amount of shaking. CF cards can only record 4 minutes of 4K.

The worklfow for a multi camera concert shoot with REDOne will generate a huge amont of footage close to that one will expect on a feature shoot. The current state of the RED workflow is clumsy even a fabulous app like Crimson does help a lot.

1. Offline edit (needs converting all proxise to ProRes, a day or two - at least)

2. Reconforming with Crimson and REDCine (in case you have done all right and did not mess things up, another 2 or 4 days)

3. Colour grading (for a 60 minute show at least 1, better 2 days)

So, you spend a week or so (in case you did all right...) just on RED's workflow peculiarities and shortcomings.

If you like the look of the U2 concert and a convinient postpro stick to Sony HDCAM or Panasonics counter parts.

RED strength is cinematoraphy in its different flavours (feature, commercials, high end docs etc...) but not multi camera recordings.

Hans
 
Andrew said:
Sbcooler, you have to evaluate the whole chain, sensor-camera-post.
I have noticed that maybe theoretical RED sensor-camera limit is 11 stops but the final post results is in my case more than 14 stops.
Click to expand...

IMHO, one cannot invent more DR in post than the sensor delivers.

Hans
 
Good posts...as far as getting more dynamic range, I do cheat a little bit in post and darken the blacks and up the whites some of the time...again when it looks better....so it's not really more dynamic range but the end result looks a little better and makes it seem that way.

As far as shooting rock shows...the Sony HDCam is not the best solution although I love the footage from those cams...I hate to work on tape and no less spend a ton of money on these kinds of cams...that will be worth half as much as soon as I buy it.

I wonder if the Scarlet will be my way to go...much more budgeted for what I do. And maybe just one main bigger cam for a center stage shot...either a Red or Sony EX-1 or possibly a Sony F950...and then 5 or 6 Scarlets floating around for close ups and a mid range shot.

Hmmm....food for thought.

So is it your guess that Red is not the ideal cam for shooting live rock shows I guess? It still should look MUCH better than shooting shows now with my Sony Z1u I'd guess. And I'm wondering if the Red or Scarlet would look much better than the Sony EX-1 for rock shows. I think the EX-1 might have rolling shutter issues worse than the Red though. I saw something to that effect in a still frame.

There is something about the Red Codec that I trust more than even HD Cam for shooting this fast changing frames...I don't know what to do. I guess I'll just keep talking to people and get more input. Maybe it will be more clear in a few months.

Thanks again all...any other feedback I am listening-
 
Hans, actually you can.... It's very possible to trade resolution for DR in post.

Graeme
 
Hans, it is like saying you can’t transmit 20Mbits/sec HD video in 3.5Mhz bandwidth of SD channels.

Well with Quadrature modulation we actually transmit 30Mbits/sec in standard TV channels.

DR on the sensor is kind of analog thing, more light you put more current or voltage you get out of the sensor cell.

Well, Mysterium has 12 electrical stops of light recognition with the last (dark one) bordering with the noise level. Yes, I know it is not very scientific explanation but very logical one.

Saying that in the post we can’t push DR because where it comes from, or it is artificial way of adding stops, I will just ask you to watch histogram when you are adjusting colors in post.

Right away you will see that it widen from 12 bits to more as soon as you touch color control, like a spunch that was very dry and got water and expands a bit.
Original picture (data) is very dry and mathematically squeezed in to the 12 bits.

You have to be lucky if you do not have to touch it (correct it) in post.
If you limit yourself to the 12 bits all the way down the post pipe you will have trouble to adjust anything and still keep it looking good.
Actually I see one stop of say, 12 bit blue or green, getting stretched in to the even 12 bit level with much finer micro steps of 16 or even 32 bit control. It is kind of replacing the 12 steps slope with 32 bit slope with approximation directed by your request, not by computer cold blooded average calculation.

So I will disagree with Sbcooler that it is just artificial adding of DR. It is your eye evaluating in artistic way intensity needed here and you have choice of 32/16 bit resolution and it will take you from necessity of 12 stops in the picture to bit more, usually two stops, in order to finish the job.
Also at the end you scale it down from 4K to 2K and here you get extra stop (theoretically two stops) if you do it right.
Nothing artificial here.

Hey, you don’t use dry dry spunch to clean your car, it has to be soft and smooth.

It is how I think about the RAW, it is just dry picture that have to be smoothed out by 16 or 32 bit fine tuning and soaked by 16 bit TIFF stretch.

Do you like metaphors?
 
Hans von Sonntag said:
Where does the DR come from... Sounds a bit like a Perpetuum Mobile to me...
Click to expand...

If you scale your footage down, you're decreasing noise and therefore adding to the usable dynamic range. You could always add some noise reduction in post as well.
 
You must log in or register to reply here.
Back
Top