Build 15 & 16 news...

[DSPographer]

Okay, how about this -- how about no deBayer at all? A black and white 1280x720 patch ripped straight from the chip and put out pixel-for-pixel on the HD-SDI display? In black 'n' white. That would rock. Especially if you could mix in a "focus in red" feature, so the only thing that shows up in color at all is high-frequency detail.

If you mean to put out the bayer array data directly as black and white that would look strange. Grays might be OK but green patches look like a checkerboard and blue and red patches look like a fairly widely spaced pattern. I don't think it would be possible to tell if the image was sharp with that so you would be completely relying on the red detail overlay for focus but maybe i'm wrong.

 

[Jeff Kilgroe]

Hmm... I think it would work somewhat, but would still need some filtering. We would be talking about the luminance value recorded by each photosite or pixel, not the chroma. But I don't think the R, G and B photosites all see luminosity with the same sensitivity, hence why some sort of bayer filtering would still be necessary.

 

[Axel Mertes]

Okay, how about this -- how about no deBayer at all? A black and white 1280x720 patch ripped straight from the chip and put out pixel-for-pixel on the HD-SDI display? In black 'n' white. That would rock. Especially if you could mix in a "focus in red" feature, so the only thing that shows up in color at all is high-frequency detail.

Why not, a further option...

I am in to that ideas.

But a very simple linear deBayer of the 1024x600y window shown in the LCD should be working well I think. But 1:1 zoom is a highly demanded feature wish.

Jim, Jarred, Stuart, are you hearing us?

Axel

 

[Axel Mertes]

Hmm... I think it would work somewhat, but would still need some filtering. We would be talking about the luminance value recorded by each photosite or pixel, not the chroma. But I don't think the R, G and B photosites all see luminosity with the same sensitivity, hence why some sort of bayer filtering would still be necessary.

Good point.

So I hold up the idea:

A simple deBayer, linear interpolation of missing color values, just enough to judge on the sharpness of what we see in a 1:1 zoom.

Axel

 

[Adrian T.]

I have a wish for build 16:
Measuring right the signal. So the signal what is shown on the outputs is not what you really recording, I know. White is clipping before it does on the recorded image and so on. But will there a possibility to get a measure result via luma histogram or waveform monitor which show me the reality or at least very close the reality?

I'm asking for that for ages (and I'm not alone).

Histograms and traffic lights don't mean anything if they're not based on RAW data. We need to know when the RAW data clips, not when the RGB preview clips. It's up to post to translate the RAW data into something else.

Please!

(Rainer, I think this is also what you meant with your post :wink

 

[Brent J. Craig]

Histograms and traffic lights don't mean anything if they're not based on RAW data. We need to know when the RAW data clips, not when the RGB preview clips.

I asked this several weeks ago and did not get a reply. Are you saying that the histograms, false color and traffic lights are reading the Rec709 output signal and not the Raw data?!?

 

[Adrian T.]

I asked this several weeks ago and did not get a reply. Are you saying that the histograms, false color and traffic lights are reading the Rec709 output signal and not the Raw data?!?

Yup. Doesn't seem right, does it?

 

[Emery Wells]

I'm asking for that for ages (and I'm not alone).

Histograms and traffic lights don't mean anything if they're not based on RAW data. We need to know when the RAW data clips, not when the RGB preview clips. It's up to post to translate the RAW data into something else.

Please!

(Rainer, I think this is also what you meant with your post :wink

I concur. While we're no the topic of fixing the meters, how about rethinking the focus assist? I think everyone can agree that the visualization model for Graemes focus assist has not turned out to be all that useable. I don't know of anyone that in fact uses it. Im happy with the current visualization model in other displays which contours objects in focus with a red line. Its intuitive and it works.

 

[Mark Pedersen]

Our goal (you know how that goes) is Build 16 in 3 weeks or so. Plenty of time before NAB.

Jim

Oh you guys are on a roll!! GO team GO!

Very happy to see this happening guys. Awesome work! Just in time for my camera to ship. You guys told me months and months ago my camera would be full "feature enabled" and it looks like you've made it happen.

Congrats to everyone at Red!

And thank you!!

M

 

[Rainer Fritz]

I'm asking for that for ages (and I'm not alone).

Histograms and traffic lights don't mean anything if they're not based on RAW data. We need to know when the RAW data clips, not when the RGB preview clips. It's up to post to translate the RAW data into something else.

Please!

(Rainer, I think this is also what you meant with your post :wink

As you say big lebowski... Jim... please... the SI-2k from Silicon Imaging can handle this...

And one thing what they also deliver is the possibility to put in all SATA devices so I can put in this SSD drives:

http://www.hama.de/portal/articleId*158754/action*2563/

60MB/s read 46MB/s write speed @ 580 Euros

regards

 

[Pawel Achtel]

Remote Power On/Off in Build 16?

Remote Power On/Off in Build 16?

My wish for build 16 is ability to switch the camera on/off remotely. Usually I am not quick enough to open the housing underwater, press momentary power and close it before water gets in... hmy:
Assigning some pins on some connectors for remote power on/off would rock my boat...

Pawel
www.achtel.com

 

[Deanan]

The best solution for now is to rig up a solenoid to push the power button.

 

[laguun]

As you say big lebowski... Jim... please... the SI-2k from Silicon Imaging can handle this...

yeah, the raw histos are useful.

And one thing what they also deliver is the possibility to put in all SATA devices so I can put in this SSD drives:
http://www.hama.de/portal/articleId*158754/action*2563/
60MB/s read 46MB/s write speed @ 580 Euros
regards

interesting device - and yes, a general sata interface would be great for many applications, also as 500$ model as the c/f adapter.

 

[Axel Mertes]

Laguun,

buy no CF module and you got what you were asking for...

Axel

 

[Axel Mertes]

As you say big lebowski... Jim... please... the SI-2k from Silicon Imaging can handle this...

And one thing what they also deliver is the possibility to put in all SATA devices so I can put in this SSD drives:

http://www.hama.de/portal/articleId*158754/action*2563/

60MB/s read 46MB/s write speed @ 580 Euros

regards

Rainer,

vorsicht, ich hab das Ding mal hier gehabt (wir haben für die ne Produktvisualisierung gemacht) und es war nicht so schnell wie es auf der Website angegeben wurde. Vielleicht haben die noch was optimiert, aber damals kamen wir über ca. 32-40 MB/s nicht hinaus (was für vieles schon reichen könnte).

Gruß,
Axel

 

[Stuart English]

Media support

Media support

vorsicht, ich hab das Ding mal hier gehabt (wir haben für die ne Produktvisualisierung gemacht) und es war nicht so schnell wie es auf der Website angegeben wurde. Vielleicht haben die noch was optimiert, aber damals kamen wir über ca. 32-40 MB/s nicht hinaus (was für vieles schon reichen könnte)

When we test memory devices we often find they don't sustain the bandwidth of their published specifications. Hence most are not suitable for use with the RED-ONE camera.

 

[cinepost35]

The best solution for now is to rig up a solenoid to push the power button.

What solonoid would you think would do the job? This has been an issue on crane shoots. Having to commit a camera to a position you can't get to during filming and needing to recycle the camera for the odd crash.

 

[Stuart English]

Metering.

Metering.

Are you saying that the histograms, false color and traffic lights are reading the Rec709 output signal and not the Raw data?!?

Yes - now I'll put my helmet on. :red_bandana: Theory is good, but just to say that using REC 709 rather than linear (RAW) data for metering has several advantages, including interpretation of exposure points via luma and RGB histograms, zebras and monitors. A similar comment about focus, the tools are working, so don't expect to see major changes in these areas.

Having said that, feedback and suggestions are always welcome.

 

[Axel Mertes]

Stuart,

I may summarize some really demanded wishes once more:

- RAW histograms as further option
- RAW overexposure ample as further option
- 1:1 zoom (possibly with visible window basic deBayer) for even better focussing
- Autogained B/W display in EVF/LCD with colored focus overlay (the camera man want to do framing when shooting, not colors, colors are done before or afterwards).

We don't want to see major changes, but additions - further options - enhanced choice of how WE like to work.

Cheers,
Axel

 

[woodman]

I agree, measuring RAW is a must.

My workaround is: use false color/histogtam and sweep trough the different ISO options to have an idea of what is actually being recorded.

I don't find any use at metering processed signal. I am always telling my customers that using an Astro at the red's hdsdi outpput is like using an SD wfm on an HD camera: you're measuring the downconverter, not the camera. This is useless, unless you go straight to TV without any color timing.