FS100 w/ Nanoflash Testing

[Kholi]

Figured I'd go ahead and start a different thread about this, instead of continuously hijacking the PP thread. Pooling thoughts and experiences here so that I can reference them later and basically share the info with anyone else interested in this combo. I'm sure there's a lot on the net, but I couldn't really find anything here or elsewhere that illustrated what I needed to know.

I'm testing for a feature film that requires two very different looks. One's extreme low/available light and the other traditional narrative, both combined, but again very different.

After the first few days of testing I'm pretty impressed with the performance of this camera in both low-lit and day-lit situations. I used a beta FS100 a good month or so before it came out and liked it, but had a few disappointments with the image. It looks like they were mostly cleared up in the production model.

The choice to shoot Nanoflash instead of internal is simply trying to eck out as much as possible, even if it may be negligible to some. And, the fact that there will be a lot of movement, and I've already seen the Nanoflash 220Mb/s recording hold tighter than the Internal (which is still really good).

Sooo more info as I gather it.

 

[Kholi]

I'm using Kyle McGonaghy's profile so far--

Strapping Profile:

Black Level: +10
Gamma: Cinematone1
Black Gamme: High, +7
Knee: 100%, -1
Color Mode: Cinematone2, Level 8
Color Level: -2
Color Phase: -1
Color Depth: R -2, G +3, B +4, C +7 M +2, Y +7
WB shift: LB -1, MG/GR: -2, R -2, B 0
Detail: -2

I'm experimenting with Kyle's modified settings out to nanoFlash, recording the 220Mb/s setting because I wanted to see how far I could actually push the image down in camera (underexpose) and bring it up in post, and how working like that compared to using the internal AVCHD (which like everyone says is already REALLY freakin' good.)

Seems like the FS100 is clean enough at low ISOs to where you could do this and not suffer a major penalty, not sure though:

]

And for highlights, I let some slightly go on skin to see how it would turn out:

So far I'm pretty pleased. Still scrubbing footage, but the noise is not offensive working at extremes like this, and the color's great.

(If it's not yet realized, I am a newb to FS100 still xD)



I took the footage to my colorist, put it through the DaVinci theater and got to see it at 18 feet output through a Christie. Very very nice look, especially with the footage posted from the stills. Props to Kyle's PP again, the colorist went and stood up at the screen to look at the noise etc and openly stated he really liked what he was seeing as far as noise and especially skin goes.

The issue was pushing the footage into anything more than just adjusting contrast and light saturation really started to show the weaknesses in how the image was captured, specifically the blue channel. At 18 Feet the XDCam 422 stuff really began to break up if you tried to go anywhere with it.

We're narrowing it down to the gross fader that I was using, definitely not underexposing as drastically as I had been, and trying to shoot out a picture profile that is basically standardized with at most black gamma and black level adjustments.

And, not using a standard white balance setting, instead balancing to a grey card.

I've got the camera for a bit so I'm gonna try to do another test with Kyle's profile, and just the standard Cinematone setup with nothing changed.

 

[Kholi]

As an side, I also had the colorist look at this image:

This was done with Kyle's profile, but set to -15 black level and 0 gamma.

The blue channel on scopes looked a lot healthier, the noise was even less offensive, but he still had a hard time separating things. Couldn't really adjust specific areas of the image (of course this isn't raw, so not expecting miracles.)

We all did conclude, though, that there was a ton of potential and it's likely due to the few things that I stated above, like the weird fader cast, picture profile settings, etc.

We're doing another color session monday, I'm gonna grab a lot of stuff over the weekend to test out with different profiles.

Does anyone already have an idea of which profile seems to love being touched (graded) the best out of the box? has anyone taken their stuff to a colorist to analyze the profiles?

 

[Kholi]

Kay I looked at the files from the Nano and the AVCHD, on a tripod, there's barely any difference at all between the two on a tripod as far as detailed is concerned. Hats off to the internal compression, it's quite amazing.

When moving, I've been able to pinpoint differences in detail between the two, but you'd have to really look for it (or be super anal about it) to even see it. One big thing that I've noticed AVCHD can't handle well are fast changes in exposure or exposure ramping, and this flavor is no different.

Here's the only reason why I am going to use the Nanoflash regardless of that (probably dual recording at all times), and it applies to underexposing and black gamma:

Original (+5 Gamma) - Zebras @ 85, only the reflections on the cars are clipping (click for the larger PNG screenshots):


Internal Recording (AVCHD):


Nano Flash XDcam HD422 220Mb/s:


It's pretty apparent, even in the JPGs, the Noise (FPN) and macroblocking. Which kinda doesn't matter because this is a very special case, extreme method of working that most people won't bother with.

On the XDCam nano footage, it's all fine grain/noise which is EASILY removable. Even a chroma denoise and leaving the luminance would be very pleasant.

On the topic of Black Gamma, Nirv's information has turned better results for me in the noise department and pushing extremes. Also, when combining that info with Kyle's picture profile, and adjusting Black Levels -7 versus -15, Zebras set to 95 so only the highest highlights clip on the LCD, the boosted image is frighteningly clean at 500ISO and very very heavy ND (Hoya ND8 + Cheap Fader combination, F4).

I'm still looking for a way to post some stills, but for now I can only do PNG screenshots. When I can find a good frame to post I will do so.

Thanks for all of the info in this thread, now it's a matter of balancing underexposing too far and trying to find some more detail when NDing way too much. X_x;

Kay, here's a quick set with a -15 adjustment on Kyle's settings. The differences in recording formats aren't nearly as obvious here, and it's still a rather extreme (purposely done) push. There's a difference, for sure, but again, not as crazy (png's on click):


Original -15 Boost (zebras showing up between the trees @ 85)


AVCHD boosted to spot noise etc:


Nano boosted to spot noise etc:


Still very impressed at how clean this camera's image is, especially when you're basically starving it of light. I can't see any of this being an issue with proper exposing before the lens.

From the other thread.

 

[Everts]

Great thread.... but what do you mean by "the XDCam 422 stuff" ?

 

[Kholi]

Great thread.... but what do you mean by "the XDCam 422 stuff" ?

y000 Everts.

I'm doing dual recording, so AVCHD internal and the Nanoflash, which records xDcam HD422 8-bit (edit) 220Mb/s out from the HDMI.

Also, I had no idea Fs100 could output 4:4:4 to a recorder that could actually see it. We're going to test that in a few days as well, though not sure where it'll lead it'll be a fun experiment. =D

 

[Boon]

thanks Kholi for the test. can't wait to see more.
I'm thinking of buying a FS100 for documentary work.

 

[wannabeapilot]

Would it be possible for you to post a small raw sample of the same scene.

Recorded internally to avchd, and externally to the nanoflash?

I love the look of xdcam compressed footage vs avchd footage.

I've been considering buying a nanoflash for a long time, and it would be extremely helpful to see how the raw files look when recorded with xdcam out of the fs 100.

I could compensate you for your time as well.

 

[Kholi]

thanks Kholi for the test. can't wait to see more.
I'm thinking of buying a FS100 for documentary work.

No problemo. Hopefully there'll be something in there that points you toward an answer.

Would it be possible for you to post a small raw sample of the same scene.

Recorded internally to avchd, and externally to the nanoflash?

I love the look of xdcam compressed footage vs avchd footage.

I've been considering buying a nanoflash for a long time, and it would be extremely helpful to see how the raw files look when recorded with xdcam out of the fs 100.

I could compensate you for your time as well.

No need for compensation! Is there something specific you were looking for? I typically shoot handheld, but at some point I'll drop the camera on sticks and try to get a wide static shot, exterior, to compare the two.

Or would something like the bike clip up there work?

 

[Everts]

y000 Everts.

I'm doing dual recording, so AVCHD internal and the Nanoflash, which records xDcam HD422 8-bit (edit) 220Mb/s out from the HDMI.

Also, I had no idea Fs100 could output 4:4:4 to a recorder that could actually see it. We're going to test that in a few days as well, though not sure where it'll lead it'll be a fun experiment. =D

Kewl looking forward . Thanks for doing this !

 

[wannabeapilot]

No problemo. Hopefully there'll be something in there that points you toward an answer.



No need for compensation! Is there something specific you were looking for? I typically shoot handheld, but at some point I'll drop the camera on sticks and try to get a wide static shot, exterior, to compare the two.

Or would something like the bike clip up there work?

just a wide static shot on sticks, outside would be fantastic!

thanks so much!

 

[Kholi]

just a wide static shot on sticks, outside would be fantastic!

thanks so much!

If I can get out early enough tomorrow before I go to a commitment I'll pull a wide static from both units. Can't say it'll be more than two or three seconds because the file-size coming from the Nano's HUGE. But, will do.

It's LA, there's not much interesting here for a wide.


Anyway, despite the profile stuff, and even learning that all the profile tweaking's doing more to hurt the image when going to a grade/color, I still like Kyle's profile straight out of camera, and even messing with it on my own.

(click for screenshot png)

kyle's out of camera


posted


Tomorrow I'll shoot Dylan's profile, Kyle's, and everything OFF as my final tests, white balancing to a grey card.

One thing I've noticed is that the NanoFlash footage is pretty nice for denoising. When the noise is drastic, well, only so much you can do about it but in scenarios where it would be moderately offensive. Pretty sweet.

 

[Gary W]

Great Kholi, looking forward to seeing what you come up with.

 

[blackdogdylan]

Great Kholi, looking forward to seeing what you come up with.

I am, as well!

 

[wannabeapilot]

much appreciated Kholi!

 

[Kholi]

So here's where I landed after a good week or more of testing the FS100 and NanoFlash. I'm prefacing that I generally use RED for any major jobs, I'm used to a RAW workflow and it may affect how I see the FS100's final image. This is all about the image for me, and I'm doing this to find two different looks for my next feature film, which will be a 10K challenge project (US $10,000).

the good:

Lowlight, insane. Like that needed to be said. Being able to roll out with so little lighting is pretty crazy, and I know a lot of the situations I put the camera in RED would not have produced such clean results. Yeah, you can denoise etc., and perhaps get close but I sincerely do not believe you can achieve the same great lowlight results. Close, but not exactly the same.

Gotta hand it to Sony, they really do make lowlight champs.

Pretty healthy amount of usable DR which can't be understated either. I've read eleven (11) stops, and I believe it.

The NanoFlash combo produces an image with a fine grain that's actually pretty pleasing to my eyes. Where the very well done implementation of AVCHD falls short (detailed images with movement, exposure ramping, 4:2:0) the Nano picks up and does a great job holding everything together. I've also noticed that the AVCHD can get "macroblocky" while the Nano @ 220Mb/s stays pretty solid in general.

Great stuff.

Jello versus anything beneath it (I'm guessing AF100 is similar though), definitely a winner. My shaky hands appreciate this.

the neutral:

Color's not bad, not the best either, but nothing that can't be worked with if you get everything close in camera. There's a green bias out of the gate, which can be corrected with a simple tweak.

the not so good:

We tested picture profiles galore, bouncing from Franks well-known PPs, Kyle's PP (which I actually find pretty nice if you want straight-out-of-camera footage and you aren't tweaking much more than contrast), PPs off, Dylan's REC709/PRO PP (which is also pretty nice, great for tweaking), so on and so forth.

Right now, the consensus is that OFF or something like what Dylan's using is the best way to go if you plan on doing any serious grading after acquisition. The image simply does not want to be pushed around very much beyond light adjustments. Now, I prefaced this with the fact that I'm so used to shooting RED that I may be spoiled. But, even very modest grading (Resolve DI theater and a serious colorist) it would not really move.

We DID conclude, however, that part of this may have been due to IR contamination. IN doing sharpness tests between NDs, I took the NDs completely off of the camera and just stopped down, I made a point to note this to the colorist as we were going through the footage, playing back on a 60" plasma and the 18 foot screen, he then promptly began to adjust this image and immediately saw different results when trying to separate colors, mids highs etc.

It wasn't drastic, but it was a positive difference. This issue brings me to the camera's most positive node... which also happens to be a bit of a negative...

Good GRIEF do you have to ND the living hades out of the camera to get exposure. It's fine if you like shooting at say T8 on exteriors, then you can use less ND, but man does it like to fight you for proper exposure. This also rolls right into the endless topic of highlights and how hard it is to tame them. Balancing the image is very tough in natural lit scenarios (exteriors with minimal grip gear, or bright windows nearby, bright sources in frame) and if a practical or something in the shot gets WAY too hot, it becomes a blob of over-exposed pixels edged by a neon color.

Similarly, not bringing down any highlights in camera, you run the risk of things becoming neon. Now, you can sort of massage this in post by desaturating highlights above a certain value, but it never really completely removes the problem.

Even using profiles that had adjusted slopes, this didn't help once you hit the DI session and you begin to balance highlights, colors, etc. It just fell apart again. Lastly...

I'm super ... super particular about the way cameras render motion.

Motion meaning cadence, or how anything in the frame is represented when moving in the final image. Sony has always, to me, been the weakest choice for film-like motion (which sounds silly to even say...). 24P doesn't look like 24P to me, or at least not like Panasonic's 24P, RED's 24P and especially not Alexa's 24P.

With Canon, I resorted to using a 1/45 shutter to fix the issue. I've had friends with similar sensitivities to video motion at twenty-four frames remark that "it looks right, how did you do that?". On the GH2, I use 1/40th shutter and it looks perfect to me.

Nothing that I could figure out on the FS100 would get it to work. And, I took the camera to set, shot b-roll on well lit narrative scenes (a feature film), matched it against Epic shooting the same thing, projected it in a finishing suite at 18 feet (best way to do camera tests for a feature I must say!) and nothing worked. It's just the Sony look.


the conclusion:

I'm still going to use the FS100 Nano Flash combo as one of the two looks that I need in this feature. It did extremely well in lowlight, and ODDLY enough the greatest problems I have with the image are perfect for what this second look needs. It means I won't have to shoot 30P, I can shoot 24P for the aesthetic and still be able to differentiate the looks.

In addition to that, being able to use it as a prop camera when it's all built out like that will look cool in the traditional narrative portions.

However, I won't be using it for the traditional narrative sections. These problems will just kill me when I look at footage, and yeah we can all talk about content being more important than say blown out highlights or weird motion, and maybe it is after it's all said and done... while I'm working, I'd rather be as confident as possible in the tool(s) that I'm choosing/using so that I can focus on said content.

At 18 feet of real-estate, testing my camera choices (still have one more to do) the FS100 had a lot of positive things going for it, but the other options still looked more pleasing to my eyes for a feature film.

I think the combo, and even the camera alone, is a very excellent choice for a lot of people and I'm amazed at how much you get for less than 4K used even without the Nanoflash. Like I said, I'll still be using it, as it provides something that I can't get out of any other camera in this *price point/package, when you weigh all aspects together.

Still owe some footage, I'll compile as soon as I can and will also upload the widest I was able to get running around. May take a bit, still testing.


*price point meaning the 4K range.

 

[Chris Johnston]

Hey Kholi, not to hi jack the thread but for the two pictures posted above, what was done in post? (just a quick laymans terms explanation) the difference is staggeringly different. looks awesome.

 

[wannabeapilot]

great write up Kholi!

no hurry on the footage, i actually just snagged a nanoflash off ebay for a great price

 

[Kholi]

Hey Kholi, not to hi jack the thread but for the two pictures posted above, what was done in post? (just a quick laymans terms explanation) the difference is staggeringly different. looks awesome.

No problem, not a hijack at all.

In Resolve, it would be as simple as:

1. Gamma and Lift
2. Balance Mids toward warm
3. Balance highs toward cool
4. very slight balance of shadows toward cool
5. Skin key with the keyer, apply secondary
-5a. Mid lift
-5b. highlight lift
-5c. Highlight balance a slight bit warmer (to offset the global cool on her face)
-5d. Slight soften for skin/beauty
6. Power mask over her hair (for contrast, milky hair is super ugly IMO... it comes from shooting beauty products for years)
7. Master Curve balance of overall highs, mids, shadows.

You can do all of this in Colorista II.

Step 5 is unnecessary if you actually just use a bounce card for the face. haha

great write up Kholi!

no hurry on the footage, i actually just snagged a nanoflash off ebay for a great price

Okay cool. You won't regret it.

 

[Kholi]

Here are some additional notes:

White Balancing to a grey or white card seems to be essential. I tested by carrying a small 4x4 square of eighteen grey around and would choose a stock daylight white balance as one preset, dial another, and white balance to the grey card. Under various conditions, the eighteen grey always looked the best out of camera and the parade in the DI theater was a slight bit better in the blue channel consistently.

The highlight clipping thing, when it's a sky and you've got branches or leaves, it becomes an issue with the idea that you might want to key the sky and do a full on replacement. The problem is that it doesn't look pretty. So you have to really just make a choice in camera about what you want to do, which is fine, but sometimes the flexibility is important.

With the NanoFlash, I mentioned this before, it was pretty easy to at least bring exposure up a few notches without a major noise penalty. It just hurts the color a bit because you're going under in a severe way when trying to balance highlights that you can't control on a small budget. It's not bad! I didn't mind the look at least. Just something to keep in mind.

Stressing... and this should've been obvious to me but I just didn't think about it... IR ND, all the way.

Will keep dropping info etc., and remember ,this is just me so feel free to disagree and discuss even! I don't mind, it's just my take on it.