FS7: Raw to the Odyssey 7Q

Hey Mitch,

When you guys get the 7Q to be compatible with CineEI, will the 7Q be able to record the Metadata as well so you can bring the file in something like Catalyst Browse and see the settings that you had in the camera as opposed to the flat S-Log image? Like how it works with recording to the XQD cards?

It's not a huge deal if there's no metadata as I can just grade it quickly, but would be nice. Don't let the metadata coding, delay the gamma fix by any means.

And man.... I have to say I really miss Pro-Res after using XAVC for a while. XAVC is not bad, but it's so nice just to be able to double click and it opens in Quicktime instead of VLC Player (Which is a pretty terrible app).
 
gene_can_sing said:
Hey Mitch,

When you guys get the 7Q to be compatible with CineEI, will the 7Q be able to record the Metadata as well so you can bring the file in something like Catalyst Browse and see the settings that you had in the camera as opposed to the flat S-Log image? Like how it works with recording to the XQD cards?

It's not a huge deal if there's no metadata as I can just grade it quickly, but would be nice. Don't let the metadata coding, delay the gamma fix by any means.

And man.... I have to say I really miss Pro-Res after using XAVC for a while. XAVC is not bad, but it's so nice just to be able to double click and it opens in Quicktime instead of VLC Player (Which is a pretty terrible app).
Click to expand...

Hey Hey, some of us have been using and appreciating VLC for years to do things that are either not available elsewhere or hard to do. And it's free.

And Quicktime X is not Quicktime. It uses a different framework(api).



:-)
 
gene_can_sing said:
Hey Mitch,

When you guys get the 7Q to be compatible with CineEI, will the 7Q be able to record the Metadata as well so you can bring the file in something like Catalyst Browse and see the settings that you had in the camera as opposed to the flat S-Log image? Like how it works with recording to the XQD cards?

It's not a huge deal if there's no metadata as I can just grade it quickly, but would be nice. Don't let the metadata coding, delay the gamma fix by any means.

And man.... I have to say I really miss Pro-Res after using XAVC for a while. XAVC is not bad, but it's so nice just to be able to double click and it opens in Quicktime instead of VLC Player (Which is a pretty terrible app).
Click to expand...
I'm not sure if the metadata all travels through, and you need to understand that the signal is a different format from the one used on the F5/F55. I don't know if Catalyst Browse will read it at all.
 
Mitch Gross said:
I'm not sure if the metadata all travels through, and you need to understand that the signal is a different format from the one used on the F5/F55. I don't know if Catalyst Browse will read it at all.
Click to expand...

Not having catalyst is no big deal. It's nice, but far from a deal breaker.
 
crow said:
I've noticed something odd with our FS7 when used with our Odyssey 7Q. I'm plugging the Extension unit to the FS7, and the 7Q gets its feed from the RAW out of the extension unit. It doesn't look like Slog3 footage. It looks like the waveform has been stretched, with the blacks and whites getting clipped
Click to expand...

It almost seems like data/video levels debacle.

In Resolve is your clip that came from RAW 7Q set to data levels or video levels? If Resolve treated it as 'video levels' when in fact its 'data levels' it will do exactly what has been observed.
 
starcentral said:
It almost seems like data/video levels debacle.

In Resolve is your clip that came from RAW 7Q set to data levels or video levels? If Resolve treated it as 'video levels' when in fact its 'data levels' it will do exactly what has been observed.
Click to expand...

I was under the impression that the data/video switch in resolve only applied to the signal sent to a monitor. I tried this, but it doesn't seem to have affected the footage.
 
starcentral said:
There are three switches in Resolve actually.
Click to expand...

I went to the resolve manual, and sure enough you're correct! One switch for monitoring (in project settings:master settings), one for capturing from a deck (project settings:Capture and playback), and one for the clip (in Clip Attributes).

However, this isn't the problem we're seeing with the FSRAW to ProRes. The incorrect levels appear in the waveform of the Odyssey when we are recording. Just to be sure, I switched the data/video levels settings around in the Clip attributes, and while it changed the image I was seeing, it didn't resolve the issue.
 
Well since I doubt there is issue with 12-bit RAW stream from Sony perhaps then it's the 7Q doing some strange business on the signal (as concluded already by others)

I'm sure CD will get through these hiccups in short order
 
Wasn't this already explained? RAW is linear data, and for the O7Q to record it to ProRes it has to assign a gamma curve to interpret the raw. And apparently the current firmware's implementation of S-Log3 gamma is incorrect, to be fixed in future firmware.

I think this has already been discussed to death guys.
 
That's one explanation, but doesn't mean that the 7Q is handling levels properly.

Why does it bother you that others continue to theorize and discuss it?
 
starcentral said:
That's one explanation, but doesn't mean that the 7Q is handling levels properly.

Why does it bother you that others continue to theorize and discuss it?
Click to expand...

Because Mitch Gross, who works for the company in question, and has direct contact with the engineers has already explained this issue. I think "theorizing" when it's already been explained from the source leads to confusion, not clarity.


EDIT:

I'm not sure what I should take from your statement "doesn't mean that the 7Q is handling levels properly." Clearly, as per the official explanation for this bug, it isn't handling the levels in the properly defined manner. It's applying the wrong gamma curve, which Mitch has attributed to miscommunication between Sony and CD. If you're suggesting the fault is elsewhere in the O7Q's firmware or hardware, I guess "theoritically" that could be true, but it seems like a premature conclusion when a simpler explanation is at hand.

Sony raw is linear digital data. To record it to ProRes you're going to have to choose a gamma curve to apply. Doesn't it make sense that they just have the wrong math in there for the S-Log3 gamma curve and wait for the firmware?

When you hear wierd noises in the forest at night do you quickly jump to the notion that there must be a Yeti?
 
Last edited:
Viddovation said:
I'm not sure what I should take from your statement "doesn't mean that the 7Q is handling levels properly."
Click to expand...

In Josh's images I agree the symptom looks like a simple gamma shift, but if you look at the original poster's images and actually take the time to read his post (which by the looks of it from the first post you made on this thread you didn't) then you'd see that he had described the actual white and black points as having shifted.


Screen%20Shot%202015-01-23%20at%202.19.41%20PM_zps40k3aqcj.png



Screen%20Shot%202015-01-23%20at%202.19.27%20PM_zpseilbnpf2.png



My post about possible a data levels (0-1023) and video levels ( 64-940) debacle, and suggesting that maybe the 7Q is not handling levels properly was based on what happens exactly when you take a image in post using data levels and treat it as video levels. The contrast is increased hence I threw it out there incase it would help.

I also brought it up because CD had the same issue with another Sony camera and later acknowledged that they were not in fact handling the "levels" (not gamma) properly for it.
 
Last edited:
Gentlemen, please be aware that we are on this directly and it is a priority for us to fix. You may speculate as to the origins of the issue at hand all you want, but it has little bearing on the work at hand. We are correcting the situation and will release new information as it is completed.
 
starcentral said:
In Josh's images I agree the symptom looks like a simple gamma shift, but if you look at the original poster's images and actually take the time to read his post (which by the looks of it from the first post you made on this thread you didn't) then you'd see that he had described the actual white and black points as having shifted.


Screen Shot 2015-01-23 at 2.19.41 PM_zps40k3aqcj.png



Screen Shot 2015-01-23 at 2.19.27 PM_zpseilbnpf2.png



My post about possible a data levels (0-1023) and video levels ( 64-940) debacle, and suggesting that maybe the 7Q is not handling levels properly was based on what happens exactly when you take a image in post using data levels and treat it as video levels. The contrast is increased hence I threw it out there incase it would help.

I also brought it up because CD had the same issue with another Sony camera and later acknowledged that they were not in fact handling the "levels" (not gamma) properly for it.


Lastly James, this is an online community forum. Many people here ask questions, speculate, discuss and together help each other come up with answers. There is no reason for your sarcasm or attempt get other people not to speak or say certain things. It's rude, bullying, and really not in spirit of an online community. It's not the first time you've made comments on my posts and I certainly do not appreciate it.
Click to expand...

Let's step back for a moment and indeed remember that this is an online forum, wherein all we have is words - no auditory intonations, and it is possible for people to misinterpret the tone or intent of a message.

Such appears to have occurred here. Aside from the Yeti joke, everything I said was an honest opinion, and no sarcasm was intended anywhere. I don't see how I'm being rude or bullying by counterpointing, and I'm as entitled to my opinion as you are to yours.

If I have in fact mis interpreted the technical details here, that's one thing. But my intent was never to shut down an honest dialog. This forum is a valuable resource to everyone, and many lurkers read it long after we've stopped posting to a thread to find answers to their own technical problems. That's why I value clear and concise answers where they are available.

I stand by everything I've said, and I take some objection to how you've chosen to react to it. Hopefully this is a more minor misunderstanding than the tone of your message implies.
 
Are there any plans for ProRes 444 or better yet, 444 XQ for the new firmware? I'm sure we all appreciate your work and explanations.
 
We are addressing the issues brought up here. We believe we have this sorted and will provide more info as we lock things down.

There are plans for other codecs including compressed 444 in the near future. We're working steadily on it.
 
Hi everyone,

Looks like this tests were made recording 4K RAW to ProRes.

If I my ask on this thread, what about the HUGE magenta cast we have on the picture when recording 4K RAW ? Is it the FS7 fault or the way the Odyssey 7Q interprets the signal ?

Mitch, can you please give us a beginning of explanation ? I'm really puzzled about what's going on here... I managed to save a day of footage using Resolve but the process was quite painful.

I have the feeling internal XAVC recording is impaired with the same problem, although the magenta cast is barely noticeable and that it started after the 1.10 update. What you guys are thinking of this too ?

Thanks.
 
Convergent Design has found that the information in the metadata given to us by Sony is not properly aligning with other systems. This is not to rain down on Sony -- we have experienced similar issues with other camera systems upon introduction, and they all seem to need a bit of time to get things sorted out.

Resolve depends on this metadata to align the color matrix of the image. Because we transform the original FS RAW into Cinema DNG files, we are now reworking the metadata to set a proper color matrix. This should correct the issues with the RAW files and will be made available as soon as we complete and fully test the work. We are also looking at creating a translation program for RAW footage shot using the current Cinema DNG files that will adjust the metadata.

This, ladies and gentlemen, is what happens on the cutting edge.
 
You must log in or register to reply here.
Back
Top