Jump to content

RED BUILD 14 FULL TEST


Zac Halberd

Recommended Posts

  • Premium Member
There isnt´t the usual O with a slash through it, but there is a divide in the body at the approximate location of the sensor (maybe 2mm behind it, it has been discussed in on reduser.net - I haven´t yet opened my camera to check, but it seems to be accurate.)

You really want that spot to be clearly and unmistakably labeled though, to avoid any confusion in the heat of the battle.

Link to comment
Share on other sites

  • Replies 112
  • Created
  • Last Reply

Top Posters In This Topic

You really want that spot to be clearly and unmistakably labeled though, to avoid any confusion in the heat of the battle.

 

I agree, I had the two of the most experienced focus pullers in Norway in my house yesterday (prepping for a shoot with RED), they were both full of questions - and I am sure they will manage.

The less you leave to chance the better results you will get, but anyone pofessional will insist on the brief that we did - and get it right.

 

Hopefully RED will put the markings on there soon, as a camera owner I am free to mark my own camera - I might.

Edited by Eirik Tyrihjel
Link to comment
Share on other sites

  • Site Sponsor
The Origin does not have an on-chip ADC.

 

The reason for its high performance is that it uses a very old but very low-noise CCD technology that dates back to the early 1970s. While this is by far the most sensitive of all the electronic imaging technologies, the drawback is that it needs a mechancial shutter to work properly.

 

Manufacturers like RED have been able to improve the performance of their cameras using microlenses although there is some debate about the desirability of this, as it can cause "portholing" effects with some lenses.

 

 

Hmmm, Dalsa actually has their own foundry to produce sensors, such as the ones fitted to the Spirit and Opportunity rovers on Mars, one small example of the company's work. I do not believe the sensor in the Origin is a "old" "1970's" device, in fact the CCD is probably the most optimized for D-Cine work out there , yet.

 

CCD's regularly work without mechanical shutters in hi end HD video cameras, the mechanical shutter on the Origin is to facilitate the optical viewfinder a hi value added benefit IMO.

 

The design of the Origin sensor also avoids the use of a rolling shutter by having two CCD's stacked such that a complete frame is dumped back to the underlying ccd frame to be read avoiding any temporal issues with reading the contents of the sensor, also I believe the A/D's on the Origin are the only 16bit ones being used as most are 10-12bits.

 

I have had Origin 4K footage to work with on a Baselight4 and it looks to me to be a better picture in terms of dynamic range and color fidelity when compared to red footage I have seen. The counter to that is that after a few weeks of Origin rental one could buy a Red...

 

As I said befor i think the RED is an excellent example of a motivated company building a camera from off the shelf parts and at that it is probably 78% of the quality of the Dalsa for 1/10th the price, pretty good.

 

 

-Rob-

Link to comment
Share on other sites

  • Premium Member

It's a bit confusing as to why, but the Dalsa has a 14-bit A/D that writes data to 16-bit, I guess the other two bits are for something else than what the camera sensor captured. It has something to do with creating headroom for color-correction somehow.

Link to comment
Share on other sites

  • Premium Member
Hmmm, Dalsa actually has their own foundry to produce sensors, such as the ones fitted to the Spirit and Opportunity rovers on Mars, one small example of the company's work. I do not believe the sensor in the Origin is a "old" "1970's" device, in fact the CCD is probably the most optimized for D-Cine work out there , yet.

 

CCD's regularly work without mechanical shutters in hi end HD video cameras, the mechanical shutter on the Origin is to facilitate the optical viewfinder a hi value added benefit IMO.

 

The design of the Origin sensor also avoids the use of a rolling shutter by having two CCD's stacked such that a complete frame is dumped back to the underlying ccd frame to be read avoiding any temporal issues with reading the contents of the sensor, also I believe the A/D's on the Origin are the only 16bit ones being used as most are 10-12bits.

 

 

As I said befor i think the RED is an excellent example of a motivated company building a camera from off the shelf parts and at that it is probably 78% of the quality of the Dalsa for 1/10th the price, pretty good.

-Rob-

 

"I do not believe the sensor in the Origin is a "old" "1970's" device,"

Please read my post again. I did not say it is an "old 1970s device" I said it "uses an old 1970s technology".

 

Your car uses 1880s technology, but I doubt it looks or performs much like the 1880s product.

 

"The design of the Origin sensor also avoids the use of a rolling shutter by having two CCD's stacked such that a complete frame is dumped back to the underlying ccd frame to be read avoiding any temporal issues with reading the contents of the sensor, also I believe the A/D's on the Origin are the only 16bit ones being used as most are 10-12bits."

 

Exactly so, and that is precisely the architecture of the first commercial CCD camera sensors produced in the 1970s. For example a 640 x 480 NTSC type sensor would have 640 x 960 photocells arranged as 640 960-element CCD delay lines. The "bottom" 640 x 480 array is covered with an opaque material so it is unaffected by light.

 

During the frame blanking period, the accumulated charges from the "top" 480 photosites would be shunted down into the "bottom" 480, where they would then be shunted out "sideways" to form the lines of a video signal. While this is being done the "top" 640 x 480 array would then accumulate a new set of charges for the next frame of video.

 

The enormous advantage of this architecture is that the photosensors also serve as the CCD delay line elements, which means that most of the silicon surface is available for gathering photons, rather than signal processing.

 

The major disadvantage is that the photocells are still exposed to light while the "down-shunting" process is taking place, producing vertical smearing. This can be alleviated by shunting the charges as quickly as possible, but there is a limit to how quickly this can be done.

 

The most effective way to eliminate problem this is to incorporate a mechanical shutter, and several Philips/BTS SD broadcast cameras did use precisely this technique.

 

To prevent this problem without resorting to mechanical means, the photocells have to be fabricated separately from the CCD delay lines, which takes silicon "real estate" away for the light sensing areas which reduces the signal-to-noise ratio. Also, despite manufacturers' best efforts, some light will always leak into the CCD elements,. producing the characteristic vertical smear on highlights.

 

By reverting to the simple Frame Transfer architecture and a mechanical shutter, Dalsa have managed to squeeze every last ounce of perfromance out of the CCD sensor.

 

By no means is this meant to imply that Dalsa's 2008 Frame Transfer chip has a performance like a 1970s chip, any more than I would suggest modern movie film is no better than 70s movie film, even though modern films and 70s films still work on exactly the same principles.

Link to comment
Share on other sites

There isnt´t the usual O with a slash through it, but there is a divide in the body at the approximate location of the sensor (maybe 2mm behind it, it has been discussed in on reduser.net - I haven´t yet opened my camera to check, but it seems to be accurate.)

 

I also don´t recall any "slow delivery of focus hook", I believe they are widely availalbe.

 

There seems to be comments on some non availability at times. Certainly, this item seems not to ordered with the camera by some purchasers.

 

I gather the RED are now putting the location mark on the camera.

Link to comment
Share on other sites

  • Premium Member
First of Brad,

 

Don't let the door hit your ass on the way out.. I want to see posts from guys like Zac. That was a fantastic post and very accurate. He was not either for the Red nor was he against it. It was a perfect unbiased opinion. Then you go and look at our pal Phil and what he posts. Give me a break. I expect to see guys not liking the camera but at least post a detailed reason as to why and what equipment was being used. So- to answer your question..... that is why I come here. I am not a Fan Boy ( as so many of you have called us "reservation holders" )

I want accurate information- and I don't want it from a guy who has a personal vendetta against guys at RED.

You said you were leaving for good, but now you're telling me to leave...OK, whatever.

If you want to SEE posts from guys like Zac and not from Phil, then I have a suggestion....READ the posts by people like Zac, and IGNORE the posts by people like Phil. It's very simple.

I don't know if I'd call Zac's post "a perfect unbiased opinion" (really, how can you declare an opinion perfect anyway?). And how do you know it's accurate? I thought you were still waiting on your camera? How would you know?

I'd say that Zac did his best to describe his experience with the camera. It was certainly informative, and a good post. And I'm sure we'll all hear hundreds of other opinions about the camera in the coming weeks, months, and years...some of which will coincide with Zac's post and some that won't. I doubt any of those opinions will be perfect though. By definition, an opinion can't be perfect. Phil's opinion isn't perfect either. But he has valid points, just as Zac does. If you have a problem with the way Phil, or anyone else, addresses those problems then it's probably in your best interest to ignore his posts, since it appears that you don't want to hear anything that you don't agree with.

If you really want accurate information, like you say you do, then you'll read as much as you can and then do your own tests. Because only YOU can know how the camera will work for YOU. To act as if you can get all the information you need from one source is folly. Do that, and you're sure to be surprised and disappointed.

 

I have no experience with 35mm. I have been using prosumer cameras over the last 20 years. I have won two Emmys and several regional Addys for my work - using prosumer cameras. So is the Red an upgrade? Hell yeah... Will it help me and my production company? I certainly hope so.

I'm glad you've had success. I really am. But to bring up awards as a preemptive defense is a bit much. Awards are often not worth the plastic they're molded from, so absolutely everyone is going to take that with a grain of salt, whether the awards were deserved or not. And to tout your awards and then say that you've been shooting on prosumer cameras for 20 years proves that the camera and medium are not the most important part of the equation. So your argument about Red helping your 20 year old, award winning production company holds no water. What help do you need when it's clear that you're quite proud of yourself and your accomplishments, and making award winning spots already? And you've been doing it for 20 years without even shooting film? Sounds quite impressive! I'd love to see some of your spots from 20 years ago. Which prosumer camera were you shooting on then? Where can I see some of your stuff?

 

 

You guys need to stop and adjust your hearing aids. Listen to the crap that is said from guys like him. My mission is to dis-credit him.

 

Thats it

Luckily, I don't have to wear my hearing aid while reading.

What particular "crap" should I be listening for that is "said from guys like him"? Please instruct me...I'm waiting with my hearing aid turned on full blast. I thought you disagreed with him....why would you tell people to listen to him?

If Phil's posts are so worth dis-crediting, then they will be dis-credited on their own. They won't need your help. It would actually be interesting if you would address actual technical issues that he brought up for a change and talk about why he's wrong about those things. Who cares whether Phil has a personal vendetta or not. You have a personal vendetta against Phil, so how are you any different?

Link to comment
Share on other sites

  • Site Sponsor
"I do not believe the sensor in the Origin is a "old" "1970's" device,"

Please read my post again. I did not say it is an "old 1970s device" I said it "uses an old 1970s technology".

 

Your car uses 1880s technology, but I doubt it looks or performs much like the 1880s product.

 

 

The enormous advantage of this architecture is that the photosensors also serve as the CCD delay line elements, which means that most of the silicon surface is available for gathering photons, rather than signal processing.

 

The major disadvantage is that the photocells are still exposed to light while the "down-shunting" process is taking place, producing vertical smearing. This can be alleviated by shunting the charges as quickly as possible, but there is a limit to how quickly this can be done.

 

The most effective way to eliminate problem this is to incorporate a mechanical shutter, and several Philips/BTS SD broadcast cameras did use precisely this technique.

 

By reverting to the simple Frame Transfer architecture and a mechanical shutter, Dalsa have managed to squeeze every last ounce of perfromance out of the CCD sensor.

 

By no means is this meant to imply that Dalsa's 2008 Frame Transfer chip has a performance like a 1970s chip, any more than I would suggest modern movie film is no better than 70s movie film, even though modern films and 70s films still work on exactly the same principles.

 

I did not know that the dalsa's sensor relied on the mechanical shutter to avoid smearing effects, I suppose building any complex camera system is a series of design decisions and compromises..

 

Here is an interesting page I found on dalsa's site with a bunch of info about cmos and ccd http://www.dalsa.com/markets/ccd_vs_cmos.asp

 

I suppose any technology can be made to perform if the engineering can be worked out, audi fielded a very successful diesel powered gtp car the r10-tdi that won at sebring.

 

I was impressed with the origin footage that I had to look at for a few months, it seemed to have very good dynamic range and good skin tone rendering and color fidelity but I did not think it was as good as ecn scans from the northlight in this regard.

 

Go 1970's !!

 

-Rob-

Link to comment
Share on other sites

  • Premium Member
Somebody said in one of these forums (can't remember where, sorry) that you can easily scratch the sensor when mounting a lense??? No way. The sensor is behind a coated piece of high quality glass, and is far enough back into the camera to be perfectly safe. Interestingly enough though, the glass that protects the sensor has a red coating or something over it, so I was wondering if that's why they call it the RED??? Or were they just being clever, and puting a red tinge to the sensor for fun? I'm guessing it's some sort of UHT coating? Any confirmation Jim?

So far I've only seen Reds at trade shows, and haven't looked in the hole with the lens off. (I have looked in the Genesis, its chip looks kinda dark green.)

 

Most likely the piece of glass you see isn't there for protection. It's probably the OLPF -- Optical Low Pass Filter. An OLPF is absolutely necessary on every chip camera. It rolls off fine detail that would otherwise alias uncorrectably against the sampling structure of the chip. Look up Nyquist's theorem, Harry Nyquist and Claude Shannon if you want to know the details on why you have to filter out the details....

 

The red stuff is probably an anti-reflection coating, such as is used on lenses.

 

 

 

 

-- J.S.

Link to comment
Share on other sites

Hello guys

 

Just to add some engineering point of view here.

 

I was the guy that used the Imatest in this forum first and I pointed out it as a DR measurement tool. So I have a responsibility to clear up few things?

 

If someone reads what the Imatest numbers meaning, he will easily understand that the maximum USABLE latitude could only be ?mid-high? in which the noise level is NOT more than ¼ of an F-stop.

 

Above that level everything is almost a crap, so ?mid? is half a stop noise and "low" is ONE FULL STOP, I don?t won?t to mention the "total" which is 5 STOPS of noise level!!!

 

So what Jim is actually saying is that the camera has an EFFECTIVE DR of 9 and ¼ F-stops!!! In the mid high range. NOT to mention the excessive noise reduction which is evident to his screen grab, the curve is almost vertical in the noise spectrum?

 

So where in the hell did he see the 11,5 stops?

 

Except if we consider the 5 stops of noise an acceptable level?

 

An old Varicam F can have the same numbers with NO noise reduction!!!

If I use noise reduction, I can get up to 10 ½ in the mid high level but with a noise spectrum as vertical as RED has? but this will create plastic faces? and David will not be very happy.

 

If someone is in interest, see my objective latitude test that I did back in September and notice how much steeper is the noise spectrum curve (that Jim is pointing) than from all the cameras I have tested?

 

The noise curve is the bottom right curve on the Imatest report picture.

 

The link is http://www.motionfx.gr/files/Latitude%20Ob...son%20Sep07.pdf

 

David, what an engineer is trying to do, is, to maximize the performance of the tools an artist is using. In order to do that, it needs an accurate method to measure the results of his efforts. There for an artist should put the minimum of his time to really understand what an engineer is trying to tell him with his numbers. So please read slowly what am writing and try to understand what am saying, probably there is some helpful information hidden in our efforts.

 

A humble engineer?

Link to comment
Share on other sites

Dude, I'm no engineer so your charts are a bit baffling without careful study. However I was suprised to see that, according to your tests, the HVX200 has almost a 2 stop wider dynamic range than the RED!? Is that correct??? If so, that's very interesting. Would like to see more actually. I'm just making an observation here.

Link to comment
Share on other sites

  • Site Sponsor
It's a bit confusing as to why, but the Dalsa has a 14-bit A/D that writes data to 16-bit, I guess the other two bits are for something else than what the camera sensor captured. It has something to do with creating headroom for color-correction somehow.

 

 

Some sort of dithering or oversampling? I assume that the voltages in the "bucket" on any type of sensor would be somewhat small and the precision that the bucket is quantized at I assume makes up a large portion of the "subtlety" and DR of the picture?? So even if a photo-bucket can read 16 stops of dynamic range the problem of actually reading that electrical signal with enough precision becomes a barrier to the DR of the camera?

 

-Rob-

Link to comment
Share on other sites

Some sort of dithering or oversampling? I assume that the voltages in the "bucket" on any type of sensor would be somewhat small and the precision that the bucket is quantized at I assume makes up a large portion of the "subtlety" and DR of the picture?? So even if a photo-bucket can read 16 stops of dynamic range the problem of actually reading that electrical signal with enough precision becomes a barrier to the DR of the camera?

 

The data is probably just padded to 16 bits because digital systems like dealing with byte-sized (e.g. 8-bit) chunks.

Link to comment
Share on other sites

Zac the essence of my message was:

 

Look the performance that?s above ?mid-high??

I?m posting a more digested version of the summary?

 

Latitude_Comparison.jpg

 

So now I thing its better...

 

Green is better red is worst.

 

Now see what Jim showed to us in his post:

 

StoufferImatest.jpg

 

Its easily understood that the camera doesn?t have 11,5 stops of usable latitude...

 

That?s my two eurocents, dude's

Link to comment
Share on other sites

  • Premium Member

Hi Evangelos,

 

9.26 stops, thats about right with what I have seen & others have claimed. Expecting some improvement with Build 15.

 

Stephen

 

 

Zac the essence of my message was:

 

Look the performance that?s above ?mid-high??

I?m posting a more digested version of the summary?

 

Latitude_Comparison.jpg

 

So now I thing its better...

 

Green is better red is worst.

 

Now see what Jim showed to us in his post:

 

StoufferImatest.jpg

 

Its easily understood that the camera doesn?t have 11,5 stops of usable latitude...

 

That?s my two eurocents, dude's

Link to comment
Share on other sites

1) Adam's test, which was interpreted above, from the image Adam showed of him conducting the test, allowed stray light to hit the chart. That means it never got dark enough. It was a poorly conducted "quick" test and not representative of the proper tests that Adam can perform, and not representative of the camera either.

 

2) Imatest can give wildly different results depending on the curve applied to the data, even if that curve DOES NOT effect dynamic range.

 

3) Lattitude is the amount you can over expose or under expose an image and still pull it back and have it look ok. It's related to dynamic range, but IS NOT Dynamic Range. If you mean Lattitude, say it, or if you mean Dynamic Range, say that, but don't mix the two - it's confusing.

 

4) We have always quoted Dynamic Range, as measured on a Stouffer test chart, with the measurement being the counting of the wedges from clipping through to the point where you can no longer see the difference between one wedge and the next. This is a useful measurement as it's repeatable and easy to measure. It doesn't requite complex software, and ignores issues like what curve you use on the data.

 

5) Every digital camera has it's dynamic range noise limited. It is up to you, the user, to explore the available dynamic range and see what your tolerance is.

 

Graeme

Link to comment
Share on other sites

I have updated the table in order to include Graeme chart.

 

I can't accept that chart since its not independently captured and processed.

 

Regardless to my personal objection I have included it in order to be easier the comparisons.

 

Graeme I'm aware of your objections but please don't claim 11,5 stops since there aren't visible even with Imatest.

 

Stephen there is more to it, the DINSTICT steps are just 28,2 so 28,2 x 0,333 = 9,39 F-stops and the rest are INDINSTICT steps (not clearly visible), but the software, due to the advanced algorithms Imatest is using, can see deeply into noise and can report up to 11,4 stops. (the 0,333 F-stop is for 0,1D step for Stouffer in contrast of 0,15D or 0,5 F-stop Danes chart)

 

If you compare that to the Varicam results all the steps are visible clearly up to 12 stops.

 

In addition to that, there is a clear saturation in the latitude behavior of the camera around 9 stops. It seems that the sensor (or the post processing on the DSP) is so pushed to produce latitude that it has no headroom above 9,5 stops and that is evident also in the Adam Wilt test.

 

 

Common_patern.jpg

 

 

The only difference between Adam and Graeme is the blue channel noise reduction; all the rest is almost similar. Its also evident the fatigue of the latitude from a point on, that the noise curves is climbing very aggressively.

 

Varicam F goes up to -4 Log exposure with no more than 0,9 noise and RED cant reach more than -3 Log exposure with almost double noise levels, above 1,5 to 2,0 ?

 

And still I haven't test Viper or Varicam H or HPX3000 or even F23.

 

The -4 Log exposure is equivalent to 4,0D Visual density in the blacks.

Link to comment
Share on other sites

Graeme, I have no comment on the color of your face? but the image that?s being published and used in this thread is from your web site and it was your BEST effort in terms of curves and noise manipulation, even so they simple can?t be compared with a stupid old camera like Varicam F? don?t mention HPX3000 or Viper?

 

Don?t go to your B plan immediately and try to depreciate my argument when there is no other way to fight it. You are a very good engineer, and you know good guys sometimes are making mistakes to?

 

So stick to your A plan and stop trying to convince me that the sky is RED, just try to fix proper latitude to your camera, starting with a new better performing sensor because resolution is not everything, color and latitude IS everything.

 

Don?t be another marketing resolution victim?

 

All the Oscar?s awards in our resent history of cinema went to movies that we see in our local theaters with no more than 750 lines resolution (or much less)? Didn?t we laugh with them? Didn?t we cry with them? Didn?t we respect them?

 

The scenario, the acting, the directing, the art of photography, the colors and the light are magic, NOT resolution.-

 

(Except if you make cameras for military espionage satellites?)

Link to comment
Share on other sites

  • Premium Member

Hi Evangelos,

 

I would be very interested for you or Graeme to do a Red v Viper test.

My original fall out with Jim was based on the fact I did not believe Red would have a greater DR than a Viper.

 

I hope to be proved wrong.

 

Stephen

 

 

[

Link to comment
Share on other sites

  • Premium Member

Claudio Miranda's tests clearly showed that the F23 has more latitude than the Red.

 

I agree with Evangelos, resolution is just one part of the equation, color and dynamic range are much more important.

Link to comment
Share on other sites

Claudio Miranda's tests clearly showed that the F23 has more latitude than the Red.

 

I agree with Evangelos, resolution is just one part of the equation, color and dynamic range are much more important.

 

And Claudio's test was with an old build with issues, which you forgot to mention. You seem to be selective in the facts you choose to post.

 

Jim

Link to comment
Share on other sites

  • Premium Member
And Claudio's test was with an old build with issues, which you forgot to mention. You seem to be selective in the facts you choose to post.

 

Jim

 

Hi Jim,

 

I understand that DR is changed little since Claudia's test. He had to repeat his first test with another camera due to 'issues'

 

Are you saying that a Red with Build 14 beta will now beat an F23? I understand that Build 15 beta is not yet available.

 

Stephen

Link to comment
Share on other sites

Evangelos, my only real comment is that I find the measuring DR hard. It's not an easy test to perform. There are many parameters that can ruin the results, especially comparing across cameras, and across test done in different ways by different people. Not least, once your shooting is done, measuring the results in a meaningful way is also tricky.

 

Then there's the fact that measuring noise is not the same as viewing noise. The eye can see that two different noises, with the same PSNR can look very different.

 

Stephen, I'd love to have the time to develop better ways of measuring DR, based on linear light values from calibrated charts, and have it automated, but I have much more pressing tasks. Such a scheme would be very useful for internal R&D and I'm sure I'll get to it at some point.

 

Graeme

Link to comment
Share on other sites

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 account

Sign in

Already have an account? Sign in here.

Sign In Now

Forum Sponsors

Metropolis Post

New Pro Video - New and Used Equipment

Gamma Ray Digital Inc

Broadcast Solutions Inc

Visual Products

Film Gears

CINELEASE

BOKEH RENTALS

CineLab

Cinematography Books and Gear



×
×
  • Create New...