D-20 v. Red Filmout Test

[M Most]

There were some circuit board changes in the D20 recently that reduced noise and allowed them to rate it a little faster, plus make their "LOG C" mode more usable. Although if you are recording RAW data, you wouldn't use a LOG mode.

Actually, you might if compression is involved. In fact, Cineform does just that, which represents a significant philosophical difference between Cineform and Redcode. Of course, the log curve is added in the compression pass, not by the camera itself, which only passes on the raw Bayer data.

 

[StephenMurphy]

The s.two is an HD video recorder. This is considerably different than recording RAW bayer data, even if the word "raw" is often misused these days when referring to the dual link video output.

Is there a new version of the S.two that is a straight data recorder?

Mike,
No idea re. the new S.two - best to check with Steve R.
I would have used the term "RAW" to describe any unprocessed information from a digital film camera - but I stand corrected
Can the Codex record raw bayer data from the D20?

 

[David Mullen ASC]

Thanks Mike -- you reminded me of something I forgot about the specs for the new Phantom CineMag, that it stores linear RAW data from the Phantom (usually 14-bit) in 10-bit LOG form but exports it as linear RAW again. Using LOG as a form of data compression of a RAW file was something I hadn't heard of before.

 

[M Most]

Mike,

Can the Codex record raw bayer data from the D20?

Once Arri permits it. Codex currently records (and debayers) Dalsa bayer data, so I would imagine they would do the same for other formats, assuming the camera manufacturers allow them access to information about their particular format.

 

[mradeck]

You're incorrect about effects in FCP being limited to 8bit. Most allow for 32 float YCbCr, which, is transparent to the 12bit RGB we deliver to the floating point engine.

I have other informations. Only some thirdparty effects are able to render in high precision yuv!

Quicktime reference movies have applied to them whatever settings you chose in camera. If you want different ones, you can make new references in RedAlert.

That's incorrect! We tried to chance the colormatrix with redalert. Made new qt refs but colormatrix stays to camera rgb. So you have no correct color reproduction when you edit with them. Changing colormatrix needs rendertime, so qtref are not able to change that, is my explanation.

 

[mradeck]

Can you post some links, FTP, etc to the sample outputs so we can digest and make up our own opinions ... (rather than just read your verbal one ?)

Cheers

i understand your point but think about that:

don't believe any test you haven't faked by yourself! Many words that are posted in this forum are blablabla. And a picture posted here is not a proof. Could be from any digital camera.

show me any picture that shows a better picture then arri D20 recorded on HDCAM SR or SII!?

Make your own tests and then believe or not.

at the moment i'm too busy to provide material. Sorry But its on my timeline.

 

[mradeck]

there's nothing suspect, i was the postprotrainer on a red workshop

there's nothing suspect, i was the postprotrainer on a red workshop

Just curious... where did you get 5 RED's in Germany? How many were "x" cameras (new ones)?

Sorry to ask... just seems a bit odd from a 1st time poster. The more I read your post, the more suspect it sounds...

Jim

look here. translate the part with the red testing team! 40 students on red workshop. rental house owns 5 reds all delivered in jan 2008 to munich germany
http://ludwigkameraverleih.de/

my websites:

www.michael-radeck.de
www.avidcutter.de
www.hdforum.de

I'm Postproductions Supervisor - Colorgrader- Productspecialist for FCP and AVID at www.dve.de

 

[Graeme Nattress]

I have other informations. Only some thirdparty effects are able to render in high precision yuv!

That's incorrect! We tried to chance the colormatrix with redalert. Made new qt refs but colormatrix stays to camera rgb. So you have no correct color reproduction when you edit with them. Changing colormatrix needs rendertime, so qtref are not able to change that, is my explanation.

Well, perhaps you'd try it yourself, and see that indeed many inbuilt and 3rd part effects can render successfully in 32bit float YCbCr!

The lack of matrix choice in the Quicktime codec has nothing to do with render time as there's a full matrix in there for white balance now. The explanation is very much more simple - that's the first beta of the codec, and the colour matrix was hard wired, rather than being allowed to be parameterized via metadata. That will be fixed when the codec is updated.

Graeme

 

[mradeck]

no drugs, real research & investigation

no drugs, real research & investigation

MR mradeck, did you try new mushrooms recently ?
Some infos are not properly going to your brain, i am afraid for your health !!

what do you mean? what's your knowledge?

here's my:


look here. translate the part with the red testing team! 40 students on red workshop. rental house owns 5 reds all delivered in jan 2008 to munich germany
http://ludwigkameraverleih.de/
we had 1 FCP, 1 DVS Clipster 4k, 1 Iridas Speedgrade DI, 1 AVID DS Nitris, 1 Assimilate Scratch, 1 4k Sony Projector (i'm productspecialist for that also), 1 2k christie dig projector, 2k arrilaser printout and projected in compare to the digital projection. 5 days of testing, played out to RGB HDCAM SR. About 25 minutes, all testmaterial is about 1 Terrabyte diskspace.

my websites:

www.michael-radeck.de
www.avidcutter.de
www.hdforum.de

I'm Postproductions Supervisor - Colorgrader- Productspecialist for FCP and AVID at www.dve.de

 

[mradeck]

Well, perhaps you'd try it yourself, and see that indeed many inbuilt and 3rd part effects can render successfully in 32bit float YCbCr!

The lack of matrix choice in the Quicktime codec has nothing to do with render time as there's a full matrix in there for white balance now. The explanation is very much more simple - that's the first beta of the codec, and the colour matrix was hard wired, rather than being allowed to be parameterized via metadata. That will be fixed when the codec is updated.

Graeme

So tell when beta beta beta will finish and we get a final codec.

FinalCuts YUV - Rendering with some 32 bit but most 8 bit rendering with RGB material degraded to tv-levels not preserving superblack, stay to a "no go" for me! Thats amateurs finishing level for enough for standard def tv not for cinema or good HD TV!!

 

[Graeme Nattress]

It's being worked on now.

The issue you have with RGB in FCP is the codec, not FCP. A RGB codec must be written so that it presents a 32 float YCbCr option to FCP, and then it will use it. If not, then it will go down the 8bit RGB pathway.

Graeme

 

[Bruce Allen]

Is FCP still limited to only 8bit support for After Effects plugins?

In that case my beloved GenArts Sapphire won't be friendly with the Red / FCP combo, right?

You'd have to finish on another system? Avid / discreet / Scratch / etc? Not too big of a problem for me - Sapphire always translated nicely for finishing.

Bruce Allen
www.boacinema.com

 

[mradeck]

Btw: You can measure the much higher resolution. We did so not only in the tests we shot in nov/oct - but with any decent chart. But you should be in focus.

the problem is, that many people think, ah there's a new cheap camera we could use it and produce the same quality as other productions with much higher budgets, most of them are tv-video-cameramen's who has no 35 mm knowledge, and film cameramen's believe they could use there own old 16mm lenses with this new camera and so on... the chance that those people produce 4k pictures which have lower res then 2k is very high. And most of the pictures posted in many websites from red are lower then 2k res. not sharper then good HD material. That's reality. And many of them haven't the chance to compare to other good cameras. May be the know HDV or HDCAM. I work daily with all good best stuff of D20 Thompson Viper F23, F950, 1500er, any camera what you want. I have supported many cameratest, student workshops, HD-Workshops over the last 7 years. I'm Trainer for HD-Postproduction. On Our Workshops we provide all post systems: Quantel IQ 4k, EQ, discreet lustre, AVID DS Nitris, Sympony Nitris, Final Cut with color what you want.

Tell me, how many RED-Owner has worked with all that stuff? 1 percent? Well i have.
On the last handsonhd - workshop in hannover i have worked with michael ballhaus german DOP from Hollywood, he was trainer of the camera masterclass, they shot on D20, Viper, F750 a shortfilm, i graded this material, i was head of postproduction, and main trainer of the post classes where we integrated the cameraclasses.

With 35 mm you get only 1k res in cinema at the end focus pulling / sharpness isn't a big issue as it will be with HD or 2k or more. If you have a good highres digital projection you will see the problem and also the differnce in quality. Also with many hdcam-tv-productions we have this problem, we had not before with standard definition.

Noise was and is an issue with the D20 in lowlight, reds cameras behaved much more silent in the same light. How high did you set the ASAs in redcine?

i have spent many many days of grading D20 material, especially dark material. Noisefree as nothing else. If you try to push reds material its horrible noise from compression artefacts from codec and noise from chip, many vertical stripes!

If you are talking about Sharpness btw - if your focus puller didn´t get it right, that isn´t exactly a camera misdesign. If you are talking about the the debayer strategy, which of the many available shrpening strategies in RC oder RA did you use?

tried any

So, do you recommend the Sony F23 then as maximum quality? And does the image gets better if you attach a codex?

i would use F23 definetly for nightshots. And in HD it's resolution is also better then red, because you loose much resolution if you downscale red material to HD. We compared several systems and algorithm.

*sarcasm*
And even worse on 65mm!!!
*sarcasm ends*

how many people has the knowledge and long practice in dealing with 65 mm, and can produce sharp picture fast and safe and cheap!??? Think about that! And imagine then what will be the reality!

Maybe an autofocus camera might be better for you? As alternative you can always reduce the active sensor area, then the nasty 35mm DOF goes away and focus is much easier with S16 sensorsize. Also you produce less "many data", which it seems is something you also would prefer.

i not a cameraman anymore i'm a post guy, i only analyse the quality of the picture i get to finish or to grade. And what i have seen until now is'nt that good as many claim here. show me the picture that poves the max resolution of the red. Who has posted a resolution chart shot with the red and is trustworthily!?

I believe at once that -you- didn´t reach D20 level - Steven Soderbergh and Peter Jackson however came to different findings, and even a modest smaller company as ours is able to get much more out of a red than out of a maxed D20. In resoluion, sharpness, noisefreeness etc.

many people believe to marketing. buy this camera and believe you are able to make films like steven! hahaha. May be, if you have the budged and the right team. Most have not!

While (for the first in your whole post) i agree, i have to tell you that 35mm -film- mostly worked like that... where you working in the industrie when everyone was using Avid AVR3 btw?

well i have edited longform 35 mm films not with avr3s because of the bad quality (i used avr6s with that you could proof the sharpness of the nose or eyes of the actor), and i used the biggest TV i could get. as i know you has not, so don't talk about thinks you havent realy done by yourself!

FCP isnt even 4K, but hey, in which NLE would we edit the D20 in -data- mode? Iridas Speedgrade?

There are so many issues to reach not 4k, also not 2k, so my final state of all i want to say is: try to make good HD, that would be much easier and cheaper and faster, then to deal with 4k. This camera has no 4k Resolution, and it's very difficult to get more the 2k out at the end of the workflow. I cannot tell you here all the experience we made: come to our next workshop: www.handsonhd.de there we will have several D20, RED, F23, in summary about 30 cameras and the capacity for more then 300 students for 7 day in june 2008 in Hannover.

 

[Ariana]

Mradeck, you clearly sound like an arri shill! Please, the propaganda is not really necessary, is it?

 

[reality]

Arri is looking to be bought and this is no doubt because their attempt in get into digital has failed. Everyone knows that the D20 is too noisy for practical speeds and that there's no way for them to save the project. Fraunhoffer did the D20 design which puts Arri in a bad spot for designing their own camera.

RED is good for arri as they'll make way more money selling accessories and lights to red users than trying to salvage a failed camera program.

 

[Shawn Bannon]

I think we have all seen low budget movies that look awesome and high budget movies that look like shit. We have also seen huge budget movies look awesome and low budget movies look like shit.

we've seen 35mm look like shit and 16mm look awesome. We've seen 35mm look awesome and HD look like shit.

anyways There is going to be some RED footage that looks like shit and some that looks fucking insane and awesome.

 

[mradeck]

Arri is looking to be bought and this is no doubt because their attempt in get into digital has failed. Everyone knows that the D20 is too noisy for practical speeds and that there's no way for them to save the project. Fraunhoffer did the D20 design which puts Arri in a bad spot for designing their own camera.

RED is good for arri as they'll make way more money selling accessories and lights to red users than trying to salvage a failed camera program.

everyone who has not worked with! knows many. have a look to the noise of red:
here compared with f23 top and red at the bottom:

http://www.hdforum.de/red/res-MB_compare.bmp

http://www.hdforum.de/red/res-MB_compare.mov-noise-screenshot-f23above-red4k-downunder.png


other test
http://www.hdforum.de/red/000021b-iridas-framecycler-analyse02-linebreak.jpg

http://www.hdforum.de/red/000021b-iridas-framecycler-analyse01.jpg

original picture posted here:

http://www.reduser.net/forum/showthread.php?t=8137

 

[laguun]

i have to add some insights to mr. radecks tests.

he mailed me offlist... and i have asked him from stopping that. in his mails, he stated quite interesting facts about his approach to resolution tests:

- they were not able measure focus - so they focused by the lcd.
- he "invented" the testcharts himself, as, let me quote "no one is able to design digital testcharts so far".
- he also was pretty generous and offered me his testcharts... which he designed in photoshop. my witty remarks regarding dithering, filtering and blending in this particular pipeline were met with "i can switch off filtering in photoshop". Btw, it seems he published his testcharts free of charge.
- he shared his interesting understandings regarding 1980ties state-of-the art debayer strategies etc.


therefore i have to regard his "measurements" as void - as he clearly has not enough expertise to measure anything more than his mistakes in the test setup.

An opinion shouldn´t replace knowledge imho. he might disagree.

We will luckily have to deal with that type of self proclaimed experts for some months still, and i think his findings are a valuable contribution for any educated reader.

 

[Dj Joofa]

It's being worked on now.

The issue you have with RGB in FCP is the codec, not FCP. A RGB codec must be written so that it presents a 32 float YCbCr option to FCP, and then it will use it. If not, then it will go down the 8bit RGB pathway.

Graeme

Graeme, I have a question. When you say 32 bit float YCbCr you mean each of the components Y, Cb, and Cr are 32 bits, or the total is 32 bits, i.e., something like 10 bits for each of Y, Cb, and Cr, and hence, 30 bits stuffed in 32 bits.

Thanks for any insight.

 

[mradeck]

Mradeck, you clearly sound like an arri shill! Please, the propaganda is not really necessary, is it?

Whats your problem? I have no relationship to any camera-manufacturer.

I always have problems with manufactures and owners, because i analyse and demystify. Not more. Is it that?

Can you tell us or show something interesting? So do it!

many owners doesn't want to believe the reality. They dream for the best tool for lowest money. that's also reality.

for the price the red is a good tool, i like to play with, in some things it's more like a toy, not very professional. I never see a d20 freezing like a red or losing buttons, or having problems with heat like red.
You can get good pictures out of red, if you know how to deal right. For that, many many people have to learn much! The possibilities to loose quality with the red are much higher than with other tools, doo to the lack of knowledge.