Odyssey7Q

[freshTEA]

Dear ChromeBoy007

The output of the upgraded FS700 in HD mode is limited to 1080p60.

The output of the upgraded FS700 in 2K Raw mode is limited to 240 frames per second.

The output in 4K Raw mode is limited to 60 frames per second.
(The FS700 can write 4K to cache at 120 fps, but this is limited to around 4 seconds and is output at a slow rate.)

Thus, it depend on what source we use, plug other factors, to determine what maximum fame rate we can achieve in a specific mode.

I hope this helps.

Respectfully,


Dan Keaton
Convergent Design

Correct me if I'm wrong but just a reminder so that people don get lost in the information.

*The output of the upgraded FS700 in HD mode is limited to 1080p60.
-(7Q will start out delivering 1080p30 HD, and hopefully be able to deliver 1080p60 HD) - true 10-bit


If this is not correct let me know so I can take it away.

 

[chromeboy007]

evilthought...

As I've said elsewhere, I could care less about RAW. If I had the resources(storage space, acquisition pipeline, processing speed, throughput)to deal with 4K RAW, I'd be shooting an F5 or F55. The fact is, I shoot with an FS700. I accept 1080p as my format of choice. Given that, 444, 422 AND 10-bit are the minimum requirements I would like to be able to accept. If 4k RAW out of the FS700 external port can be demosaiced and windowed to 1080P, 10-bit, 422...I'd be very happy. No binning, no line skipping. I can accept a compression algorithm that is as clean as possible, Cineform, DNxHD or ProRes...again 10-bit 422 or 444. In fact, if an Slog2 profile could be applied to the encoding, then applied to the display, that would be ideal.This is where my post production workflow exists, and what I do. When 4K display becomes deriguer, I will rethink this philosophy, until then RAW is as rediculously useable as 3D, and just as faddish.

I agree. The demand for stepping up from AVCHD to a 422 codec is huge... it's a necessity with any external recorder. But the demand to go up yet another level is not anywhere near as high.

I personally would rather have overcrank at even 120fps at say ProRes422 than raw and not be able to overcrank, at any resolution. Obviously the more overcrank the better.

 

[DanKov]

So those who sent in their FS700's can't shoot 240 without the sony solution at the moment? That's a bummer. I will have to wait for the 7Q to be in my hands before sending the camera off.

 

[chromeboy007]

So those who sent in their FS700's can't shoot 240 without the sony solution at the moment? That's a bummer. I will have to wait for the 7Q to be in my hands before sending the camera off.

No no that can't be. After the upgrade the camera retains all the pre-upgrade functions and overcranks, right?

 

[capt chuck]

No no that can't be. After the upgrade the camera retains all the pre-upgrade functions and overcranks, right?

No worries, I have my upgraded cam in hand and I haven't lost any of the pre-upgrade functions or overcrank.

 

[Colinelves]

Actually it's a lot more problematical than that. It's that they're trying to extract 2K raw off of a 4K raw sensor, while not changing the field of view. And that leads to some skipping of rows and colums of pixels, and that can cause aliasing and moire patterns.


But I'll make you a bet that the 2K raw off the Epic is quite a bit more "solid" (i.e., no aliasing, no weird artifacts) than the 2K off the Sony will be. The Epic windows down to sample a native 2K off the center of the sensor, so there's no additional aliasing introduced. Sony instead decided to prioritize maintaining the same field of view -- and that means they have to basically point-sample the sensor, and that is what will cause the aliasing.

Thinking about it a bit more I suppose the issue is that, relative to a smaller 2K sensor, the photosites are further apart - (fewer pixels per inch) meaning that wider lines will cause aliasing. However, the flip side of this is that a windowed chip messes up the field of view of your lenses (making wide angles harder/more expensive to achieve) and increases the depth of field ( and directors these days love the shallow 5D look).

Plus, personally, I feel the aliasing is a bit of a storm in a teacup. Logically it's not going to be any worse than in the present SSM and I've never had a problem there. I mean, who the hell is going around filming fine lines at high speed? Everything I've used it for us far more organic than that. The only time I've really had issues is with aliasing is from the costume or art department, who really should know better by now than to go around buying stuff in gingham.

so yeah, the Epic, in all it's soft-muddiness will have less aliasing - big deal. It also messes your lens choices and has a less 'filmic' (ugh) look and it is four times the price and the chip is slow as my dead gran (which is a pain for slow mo). Which is to say: 95% of the time the FS700 will be the better choice and the one preferred by most clients - (although some clients will always for the the Red because "it is the best camera") That's even before you start taking about (aliasing free) 120fps 4K, which the epic can't touch (it also can't do 240 @ 2K, only 220)

Colin Elves
DP, FS700 owner,
London
www.colinelves.com

 

[brazzam]

*The output of the upgraded FS700 in HD mode is limited to 1080p60.
-(But the 7Q only have licence to output 1080p30 in HD mode) we will have to do with 8-bit in 60p as it looks now :/

I do believe that this is slightly inaccurate. The 7Q only has the licensing limitation on 4K recordings - any 4K recording cannot be higher than 30 fps. However, if CD uses the 4K camera output (which goes up to 60 fps) to make an HD recording, CD can down res that to record a 60fps HD image just fine. I believe that the 7Q with an upgraded FS700 will be able to record a 10-bit 60fps HD image once "Option 2" is implemented.

 

[freshTEA]

I do believe that this is slightly inaccurate. The 7Q only has the licensing limitation on 4K recordings - any 4K recording cannot be higher than 30 fps. However, if CD uses the 4K camera output (which goes up to 60 fps) to make an HD recording, CD can down res that to record a 60fps HD image just fine. I believe that the 7Q with an upgraded FS700 will be able to record a 10-bit 60fps HD image once "Option 2" is implemented.

Is this correct Dan?

God I hope your right brazzam, and that im wrong. But would we not then be able to get 120fps 1080p HD in true 10-bit? Since the fs700 can do 4 sec of 120fps in 4K mode..? (burst mode)
If it can do this I will break up with my girlfriend and move into a new apartment with Odyssey.

cheers mate

 

[Colinelves]

I wonder what will happen when you set the FS700 to 4K 60p out and hook it up to a O7Q set to 4K? Will it come up with a little message that reads "our current licensing agreement with Sony precludes us from recording this signal in this mode. We apologise for any inconvenience this causes." I think a message like that might prompt a few people to email Sony and ask "WTF?!? You guys are being d1ck$" rogar-BigGrin(DBG)

 

[jcs]

That's not the reason. You can't downsample 4K raw to 2k (or whatever) raw. You first have to demosaic it, but then it is no longer raw. See

http://www.dvinfo.net/forum/cinefor...ownsample-raw-camera-without-demosaicing.html

Here's what David Taylor (Cineform's CTO) said:

Jack, it is theoretically possible to resample Bayer data, but practically it is not really possible. The reason is that each of the Bayer pixel locations, although identified by convention as R, G, or B sites, are not pure in their color. In other words, each cell site has color bleeds that must be extracted through a demosaic algorithm. Color bleeds exist because of impure color masks at each pixel location. The color bleeding is not a problem by the way, instead it is a characteristic. But it requires a DeBayer (demosaic) filter to extract color purities from the individual cell sites. But of course when you perform a demosaic you create R, G, and B values at each pixel location. If you do NOT perform a demosaic, and just perform a more traditional down-res you get color aliasing that is visually objectionable.

I agree, it's not possible to resample the Bayer array in place, like a traditional RGB planar image. However, I believe it is possible, especially for the divide by 2 case if averaging down the R,G, and B photosite locations individually. The C300 averages down ~4K to ~2K all the green elements and takes R & B as is (no actual deBayering). It produces one of the nicest 422 images on the market, with a very sharp, detailed image and low aliasing. For the FS700, we could do the same thing, average down R, G, and B in the same way (separately) and store the result back in the Bayer grid. The end result is a low-pass filter on R, G, and B, and since the relative locations haven't changed after resampling, deBayering would not introduce any unusual color aliasing artifacts. The low-pass filtering would reduce aliasing and improve image quality (same effect as noted in the C300 white paper).

It appears the only use for RAW 2K is HFR- 240fps, etc. The HFR modes do indeed use line-skipping and binning (averaging), and everything above 60 FPS on the FS700 has noticeable aliasing artifacts which get worse as the frame rate goes up (the resolution also drops, and is not Full HD as claimed by Sony (Full HD is 1000+ lines resolution, h, v, and diagonal. See charts here)). This has nothing to do with 4K to 2K RAW conversion.

 

[brazzam]

But would we not then be able to get 120fps 1080p HD in true 10-bit? Since the fs700 can do 4 sec of 120fps in 4K mode..? (burst mode)

From what I understand, the 7Q (as of now) will only interact with the FS700 in the normal 4K mode and will currently not work with the 120fps burst mode, from what Dan has said. But who knows... Convergent Design seems pretty open to keep seeing what kinds of features they can add to the 7Q, so we can only guess what they'll eventually add via firmware upgrades.

 

[evilthought]

Here's what David Taylor (Cineform's CTO) said:


I agree, it's not possible to resample the Bayer array in place, like a traditional RGB planar image. However, I believe it is possible, especially for the divide by 2 case if averaging down the R,G, and B photosite locations individually. The C300 averages down ~4K to ~2K all the green elements and takes R & B as is (no actual deBayering). It produces one of the nicest 422 images on the market, with a very sharp, detailed image and low aliasing. For the FS700, we could do the same thing, average down R, G, and B in the same way (separately) and store the result back in the Bayer grid. The end result is a low-pass filter on R, G, and B, and since the relative locations haven't changed after resampling, deBayering would not introduce any unusual color aliasing artifacts. The low-pass filtering would reduce aliasing and improve image quality (same effect as noted in the C300 white paper).
.

C300 does not shoot lower resolution raw, so there is no way way you can claim Canon does any sort of raw to raw downsampling before debayering.

I am sticking with the fact that raw to raw downsampling isn't possible without introducing moire/aliasing issues.

 

[jcs]

C300 does not shoot lower resolution raw, so there is no way way you can claim Canon does any sort of raw to raw downsampling before debayering.

I am sticking with the fact that raw to raw downsampling isn't possible without introducing moire/aliasing issues.

Canon doesn't do deBayering at all in the C300. Two green 'grids' are averaged together (low-pass filtered), which creates 1920x1080 green. Red and blue are sampled directly, creating the remaining two color planes. The magic happens starting on page 7: http://cpn.canon-europe.com/files/p...EOS-C300-New-35mm-CMOS-Sensor-White-paper.pdf

If downsampling the green grid caused aliasing or color artifacting, the C300 would show it. I write image processing software in my day job- so far I haven't seen any math showing why low-pass filtering the Bayer elements independently could cause aliasing issues in the final deBayer step. Intuitively, the opposite should happen- if there's any aliasing in the 4K grid, after low-pass filtering there should be less aliasing in the 2K grid.

Again, the 4K to 2K aliasing issues are due to line-skipping with HFR (binning shouldn't be an issue for aliasing if sample-averaging two photosites horizontally, however resolution will be reduced (see charts posted earlier showing these effects)).

With 12-bit 4K sampled to 2K RGB, 2K RAW only makes sense for HFR, where the quality is reduced due to sensor sampling restrictions at those speeds. Unless the real-time deBayer for 4K is really weak, 4K deBayer then downsampling to 2K RGB should be higher quality than 2K RAW deBayered offline.

 

[Dan Keaton]

Dan, Will we be able to upload custom luts shooting log to the 7Q or would we have to manually adjust within the monitor. Also can that lut be exported to a second screen or village? Thanks Stan

Dear Elevated Visual,

Yes, we will support custom LUT's in the Odyssey7Q and yes, the image with the LUT applied can output to external monitors.

Please expect this feature to be released after the initial firmware release.

Respectfully,


Dan Keaton
Convergent Design

 

[Dan Keaton]

Correct me if I'm wrong but just a reminder so that people don't get lost in the information.

*The output of the upgraded FS700 in HD mode is limited to 1080p60.
-(But the 7Q only have licence to output 1080p30 in HD mode) we will have to do with 8-bit in 60p as it looks now :/

We are only promising up to 1080p30 fps at 10 Bits at this time.

We hope to be able to support up to 1080p60 in this mode (where we take the Raw image to obtain 10-Bit 1080p30 (or possibly 1080p60)).

This is not a licencing issue.

*The output in 4K Raw mode is limited to 60 frames per second.
(The FS700 can write 4K to cache at 120 fps, but this is limited to around 4 seconds and is output at a slow rate.)
-This is something you won't be able to use unless you have Sony's 4K solution.

Why do you say this? We may be able to process the 120 fps footage, which is output at a slower frame rate.

Since we have other priorities right now, our engineers are not concentrating on the 4K 120 fps cached mode at this time.

If this is not correct let me know so I can take it away.

Respectfully,


Dan Keaton
Convergent Design

 

[freshTEA]

We are only promising up to 1080p30 fps at 10 Bits at this time.

We hope to be able to support up to 1080p60 in this mode (where we take the Raw image to obtain 10-Bit 1080p30 (or possibly 1080p60)).

This is not a licencing issue.



Why do you say this? We may be able to process the 120 fps footage, which is output at a slower frame rate.

Since we have other priorities right now, our engineers are not concentrating on the 4K 120 fps cached mode at this time.




Respectfully,


Dan Keaton
Convergent Design

Its been corrected.
It was more of a statement and a question at the same time. Since I know there is a lot of questions and so many different threads going on, its easy to get the lost in the information (wich I apparently had) Thats why I stated "If this is not correct let me know so I can take it away."
The interest in your product now is really big so I think thats why there are so many questions and speculations.

Im sorry if my post was offensive or missleading that was not my intention. I was just trying to make sense out of all the speculations and information.

Thank you for clearing that out.
I'm happy that the 1080p60 /and hopefully 120fps burst isnt a licensed issue. More reason for the investment.

 

[Dan Keaton]

Its been corrected.
It was more of a statement and a question at the same time. Since I know there is a lot of questions and so many different threads going on, its easy to get the lost in the information (wich I apparently had) Thats why I stated "If this is not correct let me know so I can take it away."
The interest in your product now is really big so I think thats why there are so many questions and speculations.

Im sorry if my post was offensive or missleading that was not my intention. I was just trying to make sense out of all the speculations and information.

Thank you for clearing that out.
I'm happy that the 1080p60 /and hopefully 120fps burst isnt a licensed issue. More reason for the investment.

Dear FreshTEA,

I did not take any offense at all.

This is a very complex subject.

Respectfully,


Dan Keaton
Convergent Design

 

[sebasti]

Hi Dan,

I'm sorry this is bit off topic but I'm posting this here since you are actively checking this board. Although I may not go with FS700 as my next camera, I still need a good monitor and that's where Odyssey7/7Q comes in no matter which camera I choose.
1) Will Odyssey show on screen audio meters from the input signal?
2) After you implement multi-stream recording, are you planning to add 3D function same way as Gemini has it now?
3) Since I'm also thinking of getting Blackmagic's 4K camera, they are advertising the 4K output with 6G-SDI... have you guys thought about being able to record that to Odyssey? if you have I'm sure this needs 7Q. Wouldn't Odyssey be more future proofed if the connectors were actually this type, not just normal 3G-SDI? Shouldn't be too problematic to change the connector. Anyways, most likely I would just use Odyssey for HD proxy recording when shooting RAW internally... or for instant back up since BMCC is famous for dropping frames.

Cheers,
Sebasti

 

[Gene Sung]

Hi Dan K.


Since there is so much information on this thread, can I just sum up what I thought is coming with the 7Q upon release? I just need a refresher.


1) 10 bit 4:2:2 HD possibly up to 60fps
2) 2K RAW at HFR up to 240fps? Is 60fps and 120fps possible with 2K RAW?
3) Uncompressed 4K at 24 or 30 fps?
4) The usual 4:2:2 8-bit, only up to 30p since the SDI only does 60 Interlaced.


Thanks again Dan.

 

[Dan Keaton]

Hi Dan,

I'm sorry this is bit off topic but I'm posting this here since you are actively checking this board. Although I may not go with FS700 as my next camera, I still need a good monitor and that's where Odyssey7/7Q comes in no matter which camera I choose.
1) Will Odyssey show on screen audio meters from the input signal?
2) After you implement multi-stream recording, are you planning to add 3D function same way as Gemini has it now?
3) Since I'm also thinking of getting Blackmagic's 4K camera, they are advertising the 4K output with 6G-SDI... have you guys thought about being able to record that to Odyssey? if you have I'm sure this needs 7Q. Wouldn't Odyssey be more future proofed if the connectors were actually this type, not just normal 3G-SDI? Shouldn't be too problematic to change the connector. Anyways, most likely I would just use Odyssey for HD proxy recording when shooting RAW internally... or for instant back up since BMCC is famous for dropping frames.

Cheers,
Sebasti

Dear Sebasti,

It is nice to hear that you are interested in our Odyssey7 or Odyssey7Q.

Please note that the Odyssey7 is a great monitor, with a built-in single channel recorder, to record compressed, in HD.

The Odyssey7Q has more features, such as onion-skinning, supports higher frame rates, up to 120 fps, will support (later) compressed in RGB 4:4:4, and will support full uncompressed recording in HD and it is a four-channel monitor (Quad Split) and a four-channel live switcher. All of this is included in the Odyssey7Q, no extra cost options are needed.

Then, with extra cost options, the Odyssey7Q can support the FS700, ARRI ALEXA, Canon C500, and a group of specialty cameras, such as Weisscam, P&S Technik, Lux Media Plan, IO Industries, Indiecam, and possibly (later) one more. Also, the Odyssey7Q can, with another extra cost feature (due out in Q1 2014), support recording of up to four cameras simultaneously.

As a monitor the Odyssey7 is great, and this unit is due out in November 2013.

We expect to start shipping the Odyssey7Q in September 2013.

Now, to answer your direct questions:

1. Both the Odyssey7 and Odyssey7Q have audio level meters on screen.

2. We can add 3D support, but we intend to add this feature only if there is a demand for it.

3. We intended to build the Odyssey7 and Odyssey7Q with HD-SDI 6G support.

However, the HD-SDI 6G Transmitter and Receiver chips were not ready in time,
and the HD-SDI 6G SMPTE Standard has not yet been approved (as far as I know).

I hope this helps.

I have been concentrating on this thread, since I have been traveling with limited time to post.

(I am back in the office now, after completing one-lap of America. I drove 10,066.1 miles in 29 days.)

Respectfully,


Dan Keaton
Convergent Design