My O7Q: first real world production encounter

[timmytimetravel]

cuervo or anyone else with a fw upgraded fs700 and a pix240. Have you tried recording the HD mode slomo from sdi using the inbuilt fs700 end trigger? Whilst we are missing a time window or pre-record buffer on the 7q, given the pix is 12bit 3gsdi, you should be getting 1080 60p uncompressed with an end trigger if it picks it up the trigger signal like the 7q does. It's a shame there is not more inexpensive recorders with 3Gsdi. Imagine something size and price of the hyper deck shuttle that was 3gsdi 12-bit. I'm assuming this end trigger signal in HD is still going to be 8bit in 10bit but at least HFR, uncompressed and end trigger so long as you have a 3G recorder?

 

[Stuart Keasley]

cuervo or anyone else with a fw upgraded fs700 and a pix240. Have you tried recording the HD mode slomo from sdi using the inbuilt fs700 end trigger? Whilst we are missing a time window or pre-record buffer on the 7q, given the pix is 12bit 3gsdi, you should be getting 1080 60p uncompressed with an end trigger if it picks it up the trigger signal like the 7q does. It's a shame there is not more inexpensive recorders with 3Gsdi. Imagine something size and price of the hyper deck shuttle that was 3gsdi 12-bit. I'm assuming this end trigger signal in HD is still going to be 8bit in 10bit but at least HFR, uncompressed and end trigger so long as you have a 3G recorder?

The PIX240 doesn't accept a 50/60p input. You can work around that by outputting the NEX as interlaced, however the pix will then understand that as PsF, and discard half the frames (ie your 200fps becomes 100 fps). We've tested that successfully on the pix240, ninja2 and samurai.

however the 7q will record HD HFR in exactly that way, record will activate on the trigger.

 

[timmytimetravel]

Bugger. Still not terrible though as a free option. Uncompressed 8-bit plus end trigger and 100fps with no7q (I had thought the blade and ninja 2 were non 3gsdi, need to check) plus depending on your recorder you could end up in prores right now. I presume there is no way the same thing would work on a shuttle?

with 7q you may not even need fs700 option to record that signal since it's coming from the hd? Plus you get an end trigger back. Probably don't care it's 8 bit until there is a hd from 12bit 4k option anyway.

 

[Mitch Gross]

Fair enough! Any thoughts on my previous question about the crop?

"Mitch, I wanted to quickly confirm something with you that I couldn't see in the 7Q Manual. With the FS700 at 2K Raw (And presumably 4K when that's operational) it's shooting 2048x1080, which is obviously wider than 16:9. The monitor, I realize is converting all signals to 1920x1080 before it displays it on the 720p OLED. I'm assuming this means it's disregarding (cropping off) the extra 128 pixels left/right of the frame? Is there anyway to have, as an option, a way to display the entire 2048x1080 image with black mattes on the top and the bottom?"

You are correct that we are chopping off the pixels on the sides (just for monitoring, we still record them). Right now we do not offer the letterbox option you ask for, but I will see if it is a possibility for the future.

 

[Stuart Keasley]

Bugger. Still not terrible though as a free option. Uncompressed 8-bit plus end trigger and 100fps with no7q (I had thought the blade and ninja 2 were non 3gsdi, need to check) plus depending on your recorder you could end up in prores right now. I presume there is no way the same thing would work on a shuttle?

with 7q you may not even need fs700 option to record that signal since it's coming from the hd? Plus you get an end trigger back. Probably don't care it's 8 bit until there is a hd from 12bit 4k option anyway.

You don't need a 3G recorder, you're just taking a standard 50i signal... I'd imagine the shuttle would work the same way, although I can't say for sure, we didn't bother testing that, it was about then that the a Odyssey popped up and changed our focus

And yep, you wouldn't need the fs700 option on the 7q for it to record HD HFR, it'll work straight out of the box.

 

[Gene Sung]

Hi Mitch,

I just got my 7Q and had only a brief moment to mess with it as I've been really busy. But going to play around some more. I have a few questions.

1) The 7Q look great. The colors are much, much closer to the FS700 LCD than my Small HD DP6 which nobody seemed to be able to calibrate to the FS cameras. As close as it is to the onboard LCD, the colors are just a bit off. Upon a quick first observation, it looks like the LCD has a bit more magenta while the 7Q is a little more green. I'm sure I can adjust it in the color settings, but is there an OFFICIAL CD 7Q color settings that you guys came up with to match the FS700? Like I said, it's pretty good and I can live with it, but it would be great if it were a perfect match.

2) The Histogram on the 7Q (even the smaller size option) is really big and takes up a good chunk of the lower corner. It's about 3x the size of the Sony Histogram. Is there anyway you guys can make it smaller and keep it in the lower corner? It seems to take up too much screen real-estate.

3) The Peaking is great. This is nit-picking and I'm not sure if this something that can be fixed since it's just peaking off the Sony LCD. But the info coming from the Sony LCD (ISO, Histogram, Shutter Speed, etc...) get's "peaked" in the red dots as well, making it harder to read since the text is covered in red dots. Is there anyway to have the Sony LCD info bypassed / exempt from the Peaking function?

The monitor fits on my shoulder rig just fine. It's a bit bigger than my old DP 6, but nothing I didn't get used to after 5 minutes. Actually, the larger screen size is pretty nice to view and worth the size penalty IMO.

Anyways, it's a great piece of kit. The monitor alone is worth the price and my first big 7Q project will be in early December. Can't wait to use it then. Hint, Hint.... on the compressed Codecs coming sooner than later

 

[Gene Sung]

Oh yes, BTW Mitch or Dan.

For the CD SSD cards, they did not come in any type of plastic storage case. The CD cards are propriety, so will they fit into any generic case? Anyone that you recommend?

 

[Mitch Gross]

Hi Mitch,

I just got my 7Q and had only a brief moment to mess with it as I've been really busy. But going to play around some more. I have a few questions.

1) The 7Q look great. The colors are much, much closer to the FS700 LCD than my Small HD DP6 which nobody seemed to be able to calibrate to the FS cameras. As close as it is to the onboard LCD, the colors are just a bit off. Upon a quick first observation, it looks like the LCD has a bit more magenta while the 7Q is a little more green. I'm sure I can adjust it in the color settings, but is there an OFFICIAL CD 7Q color settings that you guys came up with to match the FS700? Like I said, it's pretty good and I can live with it, but it would be great if it were a perfect match.

2) The Histogram on the 7Q (even the smaller size option) is really big and takes up a good chunk of the lower corner. It's about 3x the size of the Sony Histogram. Is there anyway you guys can make it smaller and keep it in the lower corner? It seems to take up too much screen real-estate.

3) The Peaking is great. This is nit-picking and I'm not sure if this something that can be fixed since it's just peaking off the Sony LCD. But the info coming from the Sony LCD (ISO, Histogram, Shutter Speed, etc...) get's "peaked" in the red dots as well, making it harder to read since the text is covered in red dots. Is there anyway to have the Sony LCD info bypassed / exempt from the Peaking function?

The monitor fits on my shoulder rig just fine. It's a bit bigger than my old DP 6, but nothing I didn't get used to after 5 minutes. Actually, the larger screen size is pretty nice to view and worth the size penalty IMO.

Anyways, it's a great piece of kit. The monitor alone is worth the price and my first big 7Q project will be in early December. Can't wait to use it then. Hint, Hint.... on the compressed Codecs coming sooner than later

1 The colors on the Odyssey7Q's OLED are more accurate than those on the FS700's built-in LCD. We wouldn't want to de-calibrate our display to match an incorrect one, and you probably don't want this either, as it is not what will end up getting recorded to files and sent to the edit.

2 First time I've heard someone complain that the histogram is too big. I think you're likely stuck with that one.

3 For peaking, we can only address the image that comes to us. The Odyssey has no way of knowing what is a part of the video image and what is overlays from the camera. Please note that if you have these menus on screen on the Odyssey7Q and record on the unit, these menus will also be recorded. You can adjust the display functions on the FS700 so that the SDI out feed is clean while the information remains displayed on the camera's LCD. Since many of the display functions are replicated and enhanced on the Odyssey7Q itself (peaking, exposure tools), you may wish to simply use our tools and not send such displays from the camera to the Odyssey7Q. One setup I really like is to use the first page of the STATUS CHECK on the LCD. This shows a large audio meter and other audio settings while still showing battery, recording and timecode status.

 

[Mitch Gross]

Oh yes, BTW Mitch or Dan.

For the CD SSD cards, they did not come in any type of plastic storage case. The CD cards are propriety, so will they fit into any generic case? Anyone that you recommend?

They are the same size as any standard 2.5" SSD, with the exception of the handle we add to the top of them. The overall dimension is about 2.75"x4.5"x.5". LowePro makes some nice little pouches for point & shoot cameras & cell phones which work fine for our SSDs. Take a look at models like the Volta 20, the Portland 20 and the Vail 10. There are others as well, and certainly plenty of other manufacturers of inexpensive little pouches. I find my drawers are full of them!

 

[Ted Herod]

You are correct that we are chopping off the pixels on the sides (just for monitoring, we still record them). Right now we do not offer the letterbox option you ask for, but I will see if it is a possibility for the future.

Thanks Mitch, I'd really appreciate it. It would be ideal to be able to see the exact frame that's being recorded.

 

[brazzam]

They are the same size as any standard 2.5" SSD, with the exception of the handle we add to the top of them. The overall dimension is about 2.75"x4.5"x.5". LowePro makes some nice little pouches for point & shoot cameras & cell phones which work fine for our SSDs. Take a look at models like the Volta 20, the Portland 20 and the Vail 10. There are others as well, and certainly plenty of other manufacturers of inexpensive little pouches. I find my drawers are full of them!

Thanks for the tip on those pouches. Personally, I'm looking for some hard-plastic, durable cases that I can put in my Pelican case with my O7Q. The pouches you recommended are nice and all, but I'd prefer a hard case if possible. I'd love to know if anyone has found a hard SSD case that can also manage the CD handle on top. I've done some searching online, and can't seem to find what I'm looking for.

 

[txdp]

Look at Pelican

 

[Gene Sung]

3 For peaking, we can only address the image that comes to us. The Odyssey has no way of knowing what is a part of the video image and what is overlays from the camera. Please note that if you have these menus on screen on the Odyssey7Q and record on the unit, these menus will also be recorded. You can adjust the display functions on the FS700 so that the SDI out feed is clean while the information remains displayed on the camera's LCD. Since many of the display functions are replicated and enhanced on the Odyssey7Q itself (peaking, exposure tools), you may wish to simply use our tools and not send such displays from the camera to the Odyssey7Q. One setup I really like is to use the first page of the STATUS CHECK on the LCD. This shows a large audio meter and other audio settings while still showing battery, recording and timecode status.

Got it, thanks Mitch. So we will have to basically have to work with 2 displays, the 7Q and the FS700 LCD which we need for things like ISO, Shutter speed, frame rates, etc.... There's probably no way the 7Q can display that info (ISO, Shutter speed, f-stop, etc..) on it's own without it getting baked into the image?

A minor workflow adjustment for me since I'm a recorder newbie. Not a big deal.

Thanks for the heads-up on the recorder colors. That's great you guys fine tuned the color so they are more accurate than the camera itself.

 

[blackdogdylan]

okay, little but of an update:

I've had a chance to do a lot more testing. here are some additional observations and questions for Mitch (et al.)
all of my initial points hold true.

I've been able to shoot a lot more 2K raw since then. I'll reiterate that recording the pass through from SDI out on the O7Q raw modes (2K & 4K) is not recommended.

looking at the picture on a PVM, I'm fairly certain the O7Q is only doing a half-debayer of the live image for viewing (CD engineers, please feel free to clarify this.)

it is more than adequate for viewing and focus, but it is not all that the camera and O7Q are capable of. in that half-debayer, the weaknesses of the camera system that we all know (aliasing in 2K modes, amongst other things...) become that much worse. additionally, currently the white balance in 2K is not correctly displayed and is obvious when switching the camera from 2K & 4K. 4K is accurate. 2K is not (even though it's recording correctly.)

now, this brings me to an important question for CD: will this debayer improve or is it the maximum that the O7Q can manage? I certainly understand if it is, since the unit is doing an enormous amount of processing to record and view that raw bitstream, but, for my purposes, if I am (for example) shooting 2K raw and looping out of the O7Q to a client monitor, I can already foresee the "is it going to look like that?" questions. in 4K, this is a lot less noticeable, but in 2K, it is quite significant. especially on wide shots.

now, good news is, that in Resolve, properly dealt with, the recorded 2K raw is VERY good. the resolution is superb and I think all the "on paper it should resolve less than 1080" fears were unfounded. I took a wide shot of the city and antenna on rooftops were clearly visible that could not have been more than a pixel tall. however, and this is a big however, the moire/aliasing issue (that is likely due to the camera's OLPF optimized for 4K frequency) is quite visible on high contrast, high frequency details. static resolution is great, but as soon as you start panning or tilting, all of those details start dancing (like those antenna I mentioned.) I highly doubt this is CDs problem but instead an artifact of the camera. I mean, there IS a reason why Sony is providing a 2K OLPF for their $30,000 F55. sorry, guys. our $8000 FS700s aren't going to do it better. and there's very little incentive for Sony to offer a fix for us.

I would also be curious to know how Sony is semi-baking in the white balance in the raw modes. in Resolve, when "as shot" is selected, it DOES look correct. but it shows as significantly higher color temps (I shot in 5000 and it became something like 8600 in Resolve.) it'd be great just to hear from CD what exactly Sony is doing & what Resolve is doing. is Sony's raw always natively pushing WB towards daylight (if I recall, the F65 does that too?) and as a tip to the uninitiated, keep an eye on that "as shot" in Resolve. if you make any real changes to exposure, etc, it defaults back to "custom" and skews it incorrectly. might be a bug or a "feature." not sure which.

but here's where I'm really hopeful. if the recorded 4K looks as much better as the full-debayered vs half-debayered 2K looks, we're gonna have one hell of a system on our hands. 4K raw would be nice (come on Sony) but I have a feeling 4K ProRes, or, I personally hope, 4K XAVC, will look spectacular with small file sizes and (I hope) none of that aliasing/moire I see in 2K.

anyway, just thought I would add a bit to my report in hopes of soliciting some answers from CD and helping the community at large to continue to improve upon the best bit of tech to arrive in years.

and I apologize I don't have any samples to show... since I sold my last external recorder, I didn't have a way to record what I was seeing off the O7Q over the weekend. I'll try to grab something the next time I have a recorder on rental.

- Dylan

 

[cuervo]

+1 for that XAVC suggestion.

 

[paulcurtis]

okay, little but of an update:

but here's where I'm really hopeful. if the recorded 4K looks as much better as the full-debayered vs half-debayered 2K looks, we're gonna have one hell of a system on our hands. 4K raw would be nice (come on Sony) but I have a feeling 4K ProRes, or, I personally hope, 4K XAVC, will look spectacular with small file sizes and (I hope) none of that aliasing/moire I see in 2K.

- Dylan

Great update and +999 on the 4k path. Just look at a 4K still from the camera and look at the only 4K sample on the CD website - there are no issues with aliasing, coloured noise etc,. It's what the 700 can do, anything else is seriously compromised (because of how it is delivered).

Out of curiosity on the high speed 2K RAW are you seeing what looks like wavelet compression artefacts around high contrast areas in addition to aliasing?

cheers
Paul

 

[tscav]

+1 for that XAVC suggestion.

Is that a realistic option - does anyone know?
Hopefully it's not just wishful thinking - if so Cineform RAW would be another very interesting prospect, (albeit wishful thinking too) - wouldn't it?.

 

[blackdogdylan]

Great update and +999 on the 4k path. Just look at a 4K still from the camera and look at the only 4K sample on the CD website - there are no issues with aliasing, coloured noise etc,. It's what the 700 can do, anything else is seriously compromised (because of how it is delivered).

Out of curiosity on the high speed 2K RAW are you seeing what looks like wavelet compression artefacts around high contrast areas in addition to aliasing?

thanks, Paul. funny enough, I haven't even tried SSQ mode with the O7Q, so I can't really say! however, as one more note, I connected the camera (via HDMI) and the Odyssey (via HDMI) so that I could easily toggle between the results on my 42" LCD. first off, yes the camera DOES in fact output a picture on HDMI at the same time that it spits out raw down the SDI.

secondly, on a less critical display, I have to say the 4K debayered 1080 output from the O7Q looks pretty good. maybe I was being too harsh when I said "don't record it." perhaps I should amend it and say, "it will probably get better and will likely be inferior to what is being recorded to SSDs, but you must decide if it works for you."

but, most importantly, I pointed the camera at some wine glasses that were "pinging" with some bright highlights. looking at the HDMI output from the camera, that nasty "black line" artifact is plainly visible on the highlights/contrast edges. toggling over to the output from the O7Q however (processing the 4K raw stream) the "black line of death" vanishes! I'm sure Noah (and the rest of us) will be happy to hear that it finally appears we'll be able to be rid of that- thanks to CD.

let you know if I get to test some highspeed.

 

[Mitch Gross]

okay, little but of an update:

I've had a chance to do a lot more testing. here are some additional observations and questions for Mitch (et al.)
all of my initial points hold true.

I've been able to shoot a lot more 2K raw since then. I'll reiterate that recording the pass through from SDI out on the O7Q raw modes (2K & 4K) is not recommended.

looking at the picture on a PVM, I'm fairly certain the O7Q is only doing a half-debayer of the live image for viewing (CD engineers, please feel free to clarify this.)

it is more than adequate for viewing and focus, but it is not all that the camera and O7Q are capable of. in that half-debayer, the weaknesses of the camera system that we all know (aliasing in 2K modes, amongst other things...) become that much worse. additionally, currently the white balance in 2K is not correctly displayed and is obvious when switching the camera from 2K & 4K. 4K is accurate. 2K is not (even though it's recording correctly.)

now, this brings me to an important question for CD: will this debayer improve or is it the maximum that the O7Q can manage? I certainly understand if it is, since the unit is doing an enormous amount of processing to record and view that raw bitstream, but, for my purposes, if I am (for example) shooting 2K raw and looping out of the O7Q to a client monitor, I can already foresee the "is it going to look like that?" questions. in 4K, this is a lot less noticeable, but in 2K, it is quite significant. especially on wide shots.

now, good news is, that in Resolve, properly dealt with, the recorded 2K raw is VERY good. the resolution is superb and I think all the "on paper it should resolve less than 1080" fears were unfounded. I took a wide shot of the city and antenna on rooftops were clearly visible that could not have been more than a pixel tall. however, and this is a big however, the moire/aliasing issue (that is likely due to the camera's OLPF optimized for 4K frequency) is quite visible on high contrast, high frequency details. static resolution is great, but as soon as you start panning or tilting, all of those details start dancing (like those antenna I mentioned.) I highly doubt this is CDs problem but instead an artifact of the camera. I mean, there IS a reason why Sony is providing a 2K OLPF for their $30,000 F55. sorry, guys. our $8000 FS700s aren't going to do it better. and there's very little incentive for Sony to offer a fix for us.

I would also be curious to know how Sony is semi-baking in the white balance in the raw modes. in Resolve, when "as shot" is selected, it DOES look correct. but it shows as significantly higher color temps (I shot in 5000 and it became something like 8600 in Resolve.) it'd be great just to hear from CD what exactly Sony is doing & what Resolve is doing. is Sony's raw always natively pushing WB towards daylight (if I recall, the F65 does that too?) and as a tip to the uninitiated, keep an eye on that "as shot" in Resolve. if you make any real changes to exposure, etc, it defaults back to "custom" and skews it incorrectly. might be a bug or a "feature." not sure which.

but here's where I'm really hopeful. if the recorded 4K looks as much better as the full-debayered vs half-debayered 2K looks, we're gonna have one hell of a system on our hands. 4K raw would be nice (come on Sony) but I have a feeling 4K ProRes, or, I personally hope, 4K XAVC, will look spectacular with small file sizes and (I hope) none of that aliasing/moire I see in 2K.

anyway, just thought I would add a bit to my report in hopes of soliciting some answers from CD and helping the community at large to continue to improve upon the best bit of tech to arrive in years.

and I apologize I don't have any samples to show... since I sold my last external recorder, I didn't have a way to record what I was seeing off the O7Q over the weekend. I'll try to grab something the next time I have a recorder on rental.

- Dylan

Wow Dylan, great post. Let me try to answer where I can.

Our current deBayer was in fact designed for monitoring, not recording, but it is definitely more than a half-deBayer. DeBayer algorithms have many, many parts to them and there are lots of different approaches that can be taken. Our current one is actually pretty good, but we do plan to improve upon it. We have an engineer dedicated to it right now. He wrote his doctoral thesis on the subject.

We are also lookig at an option for a deBayer which is optimized for the FS700's 2K mode. This would be designed to alleviate some of the issues people have seen. This is still in a developmental stage so I can't promise how well it will work or even if we will ever release it. Trying to fix an issue later down the pipe is like trying to take salt out of your soup. Not too easy, but we're giving it a try.

A couple of notes on the white balance.
Sony bakes in some degree of information to be more efficient in the use of the RAW stream. The info is still RAW and you have a ton of flexibility in post, but there is a modest expectation that you had at least a bit of a clue as to what you were doing when you shot it. By making this assumption, Sony can dedicate more bits of information to areas you will more likely find more useful, such as shadow detail.

Also on the issue of white balance, we actually discovered a mistake in our monitoring deBayer's color controls. It looks like we got our color temperature conversion backwards. Not a big deal unless you go really far out of adjustment, but we'll fix it in our upcoming firmware release.

 

[Mitch Gross]

+1 for that XAVC suggestion.

We never rule anything out.