Hangs4Fun

13 minutes ago, Llaasseerr said:

Absolutely no way it would get anywhere near 38.42, that's Alexa territory. But I was expecting more than 12 considering the DR test that Gerald Undone did.

You make a good point that maybe this footage was shot slightly overexposed and got exposure adjusted down, or maybe he did something else wrong in post. So it's really too early to say.

yeah, I for one will be publishing a ton of test footage when my A7SIII comes in.  I'm hoping Atomos gets their firmware released in time.  But even if it doesn't the new All Intra should be decent to look at.  The real test will be when there is a 16bit recorder that we can see footage from, since that will be the full potential of this camera.

my guess is that the footage has some overexposure in it.  S-Log3 can be tricky in mirrorless to expose properly, there is none of the cool tools you have in the pro Sony cameras.  If they didn't find 32% gray properly and bump up 2 stops from there, then the results we see are what I would expect.  Since most people (rightfully so) shied away from using S-Log3 in 8bit mirrorless camera's they are rusty on how to properly expose it. 

I will bet you lunch that they got zero at middle gray and then added 2 stops from there (instead of starting at 32%).  Phillip Bloom has some downloadable footage from his time with the A7SIII on his Vimeo channel.  Haven't had time to see what he put there.  He did have the Atomos RAW option (though pre-release).

16bit RAW to SSD $/GB costs vs 10bit All-I internal to CFexpress Type A

I'm really looking forward to using the RAW output.  I know they are outputting in 16bit RAW over HDMI, and my Ninja V will get a firmware update that will allow me to capture that output as 12bit ProRes RAW HQ upto 4K/60p.  Though I am sure there will be a recorder coming out soon that will record the full 16bit RAW (maybe Ninja V2)  Just getting the proper SSD's for the Ninja V to handle this new RAW is costing more money, though I can't complain.

Besides not being able to capture internal RAW, the other main reason I am not considering getting CFexpress Type A cards is the $$ per GB costs with that compared to external SSD costs.  The difference is significant, over 13 times more expensive.  There is only 2 options that require the CFexpress Type A cards, with many options being supported by V60 and most of the ones I will be shooting in requiring V90.  Beyond V90, I will just use external RAW.

If you compare CFexpress Type A $/GB vs Atomos Supported SSD's rated to handle 12bit RAW ProRes HQ $/GB you get:

• $2.50 per GB for CFexpress Type A
• $0.19 to $0.25 per GB for Atomos approved SSD for RAW on Ninja V

I just used the Ninja V https://www.atomos.com/drives supported media page, and looked up the cost of 2TB and 480GB OWC Mercury Extreme Pro 6G drives and divided the latest prices by the gigs and compared that to the Sony TOUGH card prices ($399 for 160GB).  I'm probably going to have to wait 2 years or more for the CFexpress Type A prices to go waaaaay down before I would ever consider using them.  Plus you can't even do RAW to them, so my plan is to use V90 Sony G cards for main shots, Sony M V60 for longer run stuff, and special stuff RAW out to SSD.  Unless someone pays me extra to do all RAW.

But $0.19 versus $2.50 makes CFexpress Type A 13.16 times more expensive per GB (I've had my Ninja V since the end of 2018, so I'm not counting that cost).

3 minutes ago, Llaasseerr said:

I know what it does to the log curve if you apply the wrong interpretation, and I would never do that. I'm putting this out there in case the tag is wrong in preproduction. It will be clear when ProRes RAW footage is available to review along with an Slog3 clip.

I realise that it's going to clip quite far below the Slog3 latitude since that was designed for Venice, F65 et al.  But 12 is too low on a curve with a max value of  38.42. I can get a max linear value of 12 shooting HLG on my X-T3.

The "stretched" max value in video levels it's still ~24 which is about half a stop below the Slog3 max, and more like what I would expect. The max value in Slog2 is about 13.75,  so if the DR really taps out at 12 then Sony could have just used that instead of Slog3. And in log, the current max value is 0.87 so this is more like V-logL vs V-log.

 

I hope they haven't taken notes from Canon's Cripple hammer.  Would it be expected to get close to 38.42 in that published clip, what was the expectation?

Are you thinking that it is pre-release firmware issues and/or the footage itself?

10 minutes ago, cpc said:

The sun is so bright that you'd need significant underexposure to bring it down to below clip levels (on any camera). And these images don't look underexposed to me. A clipping value of 0.87106 is still very respectable: on the s-log3 curve, this is slightly more than 6 stops above middle gray. With "metadata ISO" cameras like the Alexa the clip point in Log-C moves up with ISOs higher than base, and lower with ISOs lower than base. But on Sony A7s cameras you can't rate lower than base in s-log (well, on the A7s you can't, at least), so this is likely shot at base s-log3 ISO 640.

I any case, the s-log3 curve has a nominal range of around 9 stops below mid gray (usable range obviously significantly lower), so this ties up with the boasted 15 stops of DR in video. You can think of the camera as shooting 10 - log2(1024/ (0.87*1024 - 95)) bit footage in s-log3. That is, as a 9.64 bit camera. 🙂

I will take that all day long, if I'm being honest.  I'm so used to crappy 8bit from Sony Mirrorless (or you would say 7.38 bit, lol)  Very cool, I have never calculated the bit range like that before.

5 minutes ago, cpc said:

With the middle point mapped as per the specification, the camera simply lacks the highlights latitude to fill all the available s-log3 range. Basically, it clips lower than what s-log3 can handle.

You should still be importing as data levels: this is not a bug, it is expected. Importing as video levels simply stretches the signal, you are importing it wrong and increasing the gamma of the straight portion of the curve (it is no longer the s-log3 curve), thus throwing off any subsequent processing which relies on the curve being correct.

What if this is an on-set exposure issue?  I think the assumption being made is that the S-Log3 footage is properly exposed.  If your goal was to protect even the sun rays, then setting exposure levels is critical.

So my main argument here is, we need to know what the exposure settings were when shot to draw conclusions here.  Better to do tests with varying levels done and see if this is really the camera, the NLE, or just the recording conditions not matching the post production goals

20 hours ago, ajay said:

I believe it has the same tracking as the A9, A9II, A7RIV and the newer crop-sensor cameras. Touch-to-track a subject. (Using the touch screen.) I wish the end of September would get here.

On 8/18/2020 at 2:47 AM, noone said:

PB in the video between your and my post says eye AF is disabled in clearzoom.

So I asked Sony PRO Support about this and here is the official Sony response:

"The CIZ works in 4k and 1080, AF works in both modes but is wide area only, the touch tracking is disabled while in CIZ"

My guess is this could be resolved in a future firmware upgrade.  Most likely the Active Tracking functions either need increased processed and/or the way they get data (each frame) to determine where the object being tracked is in the frame, then get focus info and set lens focus position to that distance (the A7SIII can do this 120 times per second btw).  With CIZ being a subset of the sensor output, the Active Track may need a new function written to consume this new data feed (each frame of the subset of the sensor output from CIZ).  

Keep in mind, with video, CIZ can be at multiple levels.  So when the algorithms determine what the subject is at 1.1x and then you "zoom" to 1.3x, there is going to have to be some new functions created the determine where the subject is.  They could have some easy lookup tables to convert pixel locations based on which zoom level you are at, but none of that is there yet.

I have not doubt that the dual BIONZ XR processors would be able to do this, and probably will have less to process to support active subject tracking with a smaller frame, but the pixel readout stream that goes through processing needs to be tweaked to work with current active tracking functions or a new function written.  I'm pressing Sony on this, as they could absolutely do this.  And many of us will be using CIZ when shooting video as our "zoom lens".

4 hours ago, Santoso said:

I agree. While playing around with the A7SIII Slog3 footage that has been posted makes me drool, the A9 will give you a fairly malleable image at Standard (or Neutral) -3-3-2. 

yes, agreed.  It's a great photography camera (I primarily got for fast action: sports and BiF), and for other types of photography where an a7iii or other standard mirrorless camera works, the a9 does that in its sleep, its real strength is in fast action especially if your subject is all over the place with lots of distractions around).  I didn't get it for video.  But in preparation for buying the A7Siii, I sold my A7Riii over a year ago thinking I will just use my a9 for video until the A7Siii.  Now, mind you, it had been over 3 years since the A7Sii had come out, so I thought for sure I would be without a proper mirrorless suitable for video (which was my A7Riii).  Dang did Sony take their time on the A7Siii.  I was this close to buying the S1H and some lenses.  The R5 rubbed me the wrong way right from the beginning, the way they pushed it like it was primarily a video camera.  It's basically Canon's version of the A7RIV and should have been pushed as a high res photography camera (with AMAZING BiF capabilities btw) ... oh, and.. it has good video capabilities with occassional 8K options and high quality 4K options. 

Luckily, Sony released the A7Siii, I was about to do something drastic, lol...

I can not wait to take my first 4K/120p footage in 4:2:2 10bit in an All Intra codec (if I can find a single CFExpress Type A card).  I also plan to do some 16bit RAW output to my Ninja V (in 12bit ProRes RAW HQ).  But unless customers are willing to pay for that RAW, my plan is to just use the internal codecs at 4:2:2 All Intra.  Will be the first time I can actually shoot in S-Log3 in a mirrorless AND still push the grade around without it falling apart.

4 hours ago, wobba said:

Great info. If you don’t mind sharing, I would be interested to also see your A9 settings.

Believe me, I WISH I could do Picture Profiles on my a9, but unfortunately, the Sony cripple hammer has kept the picture profile options from the a9 series.  So with the a9 it is what it is when it comes out of the camera.  I will say though, that the a9's video quality straight out of camera is pretty dang good (for a 4:2:0 8bit camera)

3 minutes ago, Hangs4Fun said:

holy crap, IKR!!   NOT gonna happen, over Sony's dead body, lol.   Can you imagine their fastest camera, also having strong video adjustability?

I will be honest with you though, Straight out of the camera, that A9 shoots beautiful 4K footage.  First time I ever felt no need to muck with the settings on the camera.  I usually spend a good bit of time coming up with custom Picture Profiles (when I am going to shoot B roll to go with an FS7ii or FX9, I shoot them against a color checker card like the Passport and scope out the levels of the camera I want to "copy its' colors", then take the B Roll cam (a7Riv, a7iii, a6600, etc) and point at the same card in the same lighting and adjust the colors via the picture profiles until I get the lines to match up as close as possible.  btw, this process takes a loooooooooooooooong time, and you have to go back and forth in the picture profile settings sometimes.  I can see why Andrew charges for those settings.  Any way, once I get them nailed down, I can then shoot with the "color science" of the main camera. 

With the a9, I can't do this obviously, but I still profile it under the same conditions, so I know about how much off it is from the main camera, which makes color correcting a little easier.

I am not concerned that the a7Siii doesn't have "S-Cinetone" colors or the "Venice color science", because I will just make my own picture profile, that essential gives me just that.  NOT to mention, this will be the very first mirrorless where I felt we could actually use S-Log3.  So I'm 100% confident that I will be able to match the colors from the a7Siii with the FX9 or even the Venice (sadly I don't do much with that beautiful beast).

Here is a similar technique that I loosely described above.

For example the below picture profile settings, came from Paul over at "extrashot", he used similar techniques to arrive at these settings for the A7 mark III to get as close to the FX9's s.Cinetone as possible (limited of course due to the 4:2:0 8bit nature of the a7iii's sensor, but close enough for government work, lol):

**A7 Mark III Picture Profile settings to simulate FX9 and s.Cinetone:

Black Level: -10
Gamma: Cine2
Black Gamma: Range Middle, Level -7
Knee: Mode Manual, 100%, Slope 0
Colour Mode: Still
Saturation 0
Color Phase: -1
Colour Depth:
R: +2
G; +1
B: -1
C : 0
M: +1
Y: 0
Detail: Level -7
**to get as close as possible, A7iii is 8bit camera, soooo, not gonna be perfect 😉

If Paul doesn't beat me to it, I will share the settings I come up with color matching the A7Siii with an FX9's  s.Cinetone (I signed up for prioritized shipping from Sony PRO support and had my pre-order confirmation email in hand within the first 10 secs of 10am on 7/28, so I should be one of the first with a production model).

Changes are Paul will beat me to the punch, since he owns an FX9, and I will have to arrange to borrough one.  He uses almost the same exact technique as me, so his results would be pretty much identical.  We both try to do it as scientifical as possible, which is why our results comes out almost the same (but there is still a "gray" area when doing the picture profile adjustments, which is very touchy)

3 hours ago, thefactory said:

I enjoy the A9 a lot. I would love just one more update on the a9, its never going to happen. PICTURE PROFILES. 

holy crap, IKR!!   NOT gonna happen, over Sony's dead body, lol.   Can you imagine their fastest camera, also having strong video adjustability?

I will be honest with you though, Straight out of the camera, that A9 shoots beautiful 4K footage.  First time I ever felt no need to muck with the settings on the camera.  I usually spend a good bit of time coming up with custom Picture Profiles (when I am going to shoot B roll to go with an FS7ii or FX9, I shoot them against a color checker card like the Passport and scope out the levels of the camera I want to "copy its' colors", then take the B Roll cam (a7Riv, a7iii, a6600, etc) and point at the same card in the same lighting and adjust the colors via the picture profiles until I get the lines to match up as close as possible.  btw, this process takes a loooooooooooooooong time, and you have to go back and forth in the picture profile settings sometimes.  I can see why Andrew charges for those settings.  Any way, once I get them nailed down, I can then shoot with the "color science" of the main camera. 

With the a9, I can't do this obviously, but I still profile it under the same conditions, so I know about how much off it is from the main camera, which makes color correcting a little easier.

I am not concerned that the a7Siii doesn't have "S-Cinetone" colors or the "Venice color science", because I will just make my own picture profile, that essential gives me just that.  NOT to mention, this will be the very first mirrorless where I felt we could actually use S-Log3.  So I'm 100% confident that I will be able to match the colors from the a7Siii with the FX9 or even the Venice (sadly I don't do much with that beautiful beast).

8 hours ago, ajay said:

Crap! That's what I have. Hopefully this will be resolved before I get my camera. Thanks for the info and if you don't mind, please be sure to let us know what you find out as far as a resolution to this. I'm also a PRO Support member. Maybe I will contact them as well.

8 hours ago, MeanRevert said:

Good info, thank you.

So any V90 card can handle all the modes including slow mo and All I?

The modes that you need CFexpress Type A cards for are:

• XAVC S-I 4K at 120p
• XAVC S-I HD at 240p

But you can still do XAVC S-I 4K 10bit 4:2:2 at 24/30/60P on a V90 card. 

I can't afford the CFexpress Type A cards right now, so I will be doing 4K/120p in either XAVC HS 4K or XAVC S 4K (both have max bitrate of 280Mbps), but my guess is since H.265 can squeeze twice the image quality in the same file as H.264, that I will likely use the new XAVC HS 4K code for 120p (I ALWAYS use 422 ProRes Proxy files for editing any way).  And unless a customer wants to cough up some extra money, I have no plans on doing the above 2 at All-Intra.

VERY interesting thing was I learned there was also a 1,200Mbps  frame rate on the A7SIII (I had only heard 600Mbps till then), was S&Q mode for XAVC S-I 4K at 120P.   The stated Playback speed was 240mbps, but the Write speed was listed as 1,200Mbps (holy crap that's gonna be expensive!!).  I DOUBLE checked with PRO Support that this was not a typo and that everywhere else had published 600Mbps.  I saw the 1,200Mbps on official Support docs in the Zoom session (I may have snapped a couple of photos of the screen with my phone under the visibility of my laptops camera, wink-wink)

What surprised me, was the number of modes and options you could be in where the V90 was an option (many of those had Max bitrates of 550Mbps to 600Mbps). 

21 hours ago, MeanRevert said:

So which SD / CFExpress cards you guys gonna use with this cam?

I plan on mostly Sony TOUGH G cards (and a handful of original Sony UHS-II G cards (SF-G128) since those are V90 cards (well, actually HOLD that thought).  Also will be using some larger Sony TOUGH M cards which are V60 cards.  Most of my internal recordings will be on the V60's, with my B Roll and Slowmo on the V90 cards.  Will go external RAW for some, in which case, I have aquired several SSD's for my Ninja V (only one of which is an Atom X)

OK, back to the V90 statement.  

**WARNING**  For anyone who invested in those awesome Sony UHS-II G series cards like the SF-G128, SF-G64, SF-G32, they are currently NOT reporting themselves as V90 cards to the A7SIII (even though they clearly have the performance of a V90 card).  I brought this to Sony PRO Support today during a Zoom meeting, and proved that those cards generate the following error (currently) on the A7SIII:

"To perform shooting with this setting, use a memory card higher than SDXC V90 or a memory card higher than CFexpress VPG200 Slot 1"

He brought it to Engineer's attention today and will get back to me on what Sony is going to do about it.  Apparently those original G cards, just didn't have the label in the metadata of the media to say it is a V90 which is what the A7SIII's software does when you switch from mode to mode.  When you stick a memory card in, it handshakes with it and puts in temp memory info about the memory.  Then as you switch modes they have logic that ensure you have a V60 or V90 or even CFexpress Type A. 

Will pass along more once I hear back from PRO Support.  Might have to hack the info in the memory card to use it like a V90, not sure yet..

22 hours ago, IronFilm said:

Who remembers the original URSA Mini announcement from years ago? If I recall correctly, they did say it had a gyroscope built in to record that data for VFX work in post. 

Of course, in the final launch that feature got dropped. But maybe as this gets more popular, such as the FX9, then we'll see more and more having it. Time for an URSA Mini Pro G3 with the promised gyroscope??

If BMD puts it in the URSA Mini Pro G3 then it is certain to be supported in Resolve as well!
 

Heck, even my tiny little Sony RX Mark II has the gyro info embedded in the mp4 from about 2 years ago (that was their attempt to go against superior digital stabilization from GoPro and DJI Osmo Action), but the concept is the same.  For larger full frame, I actually really like that gyro data to use.  When you apply it, it literally is instantaneous, but the crop is hard to swallow.  Can't wait to see where this technology goes down the road.

On 8/14/2020 at 3:51 PM, MeanRevert said:

It looks like some are 4:2:0 and some are 4:2:2, however.

true, but in EVERY single format and framerate, 4:2:2 10bit is an option.  In fact,

1. XAVC HS 4K you get both 10bit 4:2:2 and 10bit 4:2:0
2. XAVC S 4K you get both 10bit 4:2:2 and 8bit 4:2:0
3. XAVC S-I 4K you ONLY get 10bit 4:2:2
4. XAVC S HD you get both 10bit 4:2:2 and 8bit 4:2:0
5. XAVC S-I HD you ONLY get 10bit 4:2:2
   ** 1-3 all support upto 120fps, and 4-5 supports upto 240fps

Just spent over an hour today in a zoom session with Sony PRO Support, I saw it personally, as I asked that (typed the above from some official slides he showed me (that I may have snapped a photo of, lol)

So, a better way to say what you said is:  The Sony A7SIII allows you to shoot ANY option on the camera in 10bit 4:2:2 color and also gives you the option to drop down to 10bit 4:2:0 on XAVC HS 4K and the option to drop down to 8bit 4:2:0 on the 2 XAVC S modes.  But every single option on that camera lets you shoot in 10bit 4:2:2

On 7/28/2020 at 12:06 AM, IronFilm said:

Sony doesn't have a track record if giving big improvements via new firmware updates to their old cameras. (unless forced to, like what happened with the F5)

Thus I reckon this is highly unlikely the a7mk3 would get such an update. 

Maaaaybe if we were discussing Fuji instead, this would be a slim possibility. Perhaps. 

My Sony a9, has had 6 major firmware releases over a 3 year period, and some of those firmware upgrades literally brought it some amazing features and usability (eye AF, animal eye AF, intervalometer, video eye AF, etc, etc..  They do these firmware upgrades based on the processor, sensor, and body capabilities.  

1 hour ago, TheBoogieKnight said:

At 4k/30 the R IV is reading more data (5.9k oversampled) from the sensor than the S III at 60p and the R IV has a limit of 170 minutes (three times longer). Sure a much bigger sensor area is being read so maybe it's thermal limitations of the sensor. If it is then cool-down times at least should be acceptable.

Isn't the R IV a 4:2:0 8bit though?  Also, I still think we are going to see XEVC-S based on H.265, so also more heat there as well for processing (now that being said, the R IV would generate heat during down sampling, so could be a mute point).  We will know more tomorrow, hopefully I can sleep (sad, I know) lol

I'm still hopeful for XEVC-S and s-Cinetone

3 hours ago, Django said:

The design is definitely uninspired. The Sony exec claiming just a few weeks ago “ Everything is new” and “exceeding expectations” was clearly trolling.

I would’ve really liked one new killer design feature like a E-ND or something.

"Everything is new" as in everything is new to that a7S likely (new stacked sensor (that gets almost global shutter like read out speeds without the DR hit), new codec (hopefully XEVC-S), new color science (from venice cinealta line), new EVF (almost 10million 😮 ), new features (no recording limits, 10bit 4:2:2 internal, faster fps upto 120 in 4K, no overheating, possibly new menu system), new processor (to handle new H.265 based codec), new Z100 battery (from a9 and recent a7's), new body design (mostly from a7r4), new dual memory (and quite possibly two options for each of the dual memory slots), new fully articulating screen (from zv-1 and a million other Sony camcorder designs since the beginning of time, lol), etc..  I would definitely give them a pass on saying "Everything is new" ... to the a7S line, if this all is the case...

Ending on you E-ND comment, I seriously hope so and that we have seen the eND of having to use ND filters thanks to eND, lol... 😉 

1 hour ago, androidlad said:

It's called XAVC RS.

I've never heard of that.  I'm familiar with XAVC-I and XAVC-L, and a little bit of XAVC-S in my usage of my a9, but never heard of RS?  In fact, I've never seen a Sony codec have anything other than a single character in that last position.  Any chance you can back that up with a link to something official that describes more about it?  Otherwise I claim your RS is BS, lol

On 7/24/2020 at 4:41 PM, Oliver Daniel said:

Where did this image come from with the LCD screen? Not posted on SAR. 

Anyone else curious if we are getting XEVC-S in the a7S3 or not?  When I zoom in on this leaked image, right where I could have confirmed that they are still using XAVC-S or using the new XEVC-S codec, the pre-production version in the leak appears to have XXXX's there and instead of a single letter for the level (like S, I, or L, obviously on an a7 it would only be S), there appears to be 2 letters there.  Maybe it is XAV CS or XEV CS and the low res of the frame capture is just not good enough to tell.  But from what I can see, they have the codec "X'ed" out.

Right behind the new stacked sensor, I really looking forward to the codec, which is hopefully XEVC-S paired with poor mans Venice color science from the CineAlta line.  

1 hour ago, Trek of Joy said:

I thought it was odd, the different record button really stood out.

yeah, that top record button makes zero sense on an a7 chassis.  They added it to the top of the ZV-1 because it has no forward sloping shutter release like a7 and a9 chassis.  Since the a7S3 is supposedly a video camera first, then technically the shutter release is more about starting and stopping video recording.  And having it already sloped forward, plus it is already a big button, you already for the perfect location to engage recording (not to mention I use all my custom button and especially the C1 and C2, don't want to loose one of them when there is no need.

What I would rather see than a top record button is some menu work.  First a redesign of the menu based on video workflows AND the ability to touch the navigation (I know sounds space age right, lol... smh).  Then some simple changes that would greatly help us when we are getting content solo without a recorder/monitor like the Ninja V.  So, when you have the fully articulated screen facing forward and are recording, instead of a tally light physically on the body, how about just putting a red box around the whole frame of the display (like Ninja V does when recording).  Also how about making the audio levels REALLY big and easily visible (with a high water mark that changes every 5 or so seconds).  I think Sony's focus Peaking and focus boxes are pretty decent already, but how about some real scopes and tools to help us set and monitor exposure?   These are all things that don't require changes to the chassis and button layout, just simple coding that could also be shared to some of the other models through firmware upgrades.

On 6/29/2020 at 10:55 PM, Jonathan Grijalva said:

Before too many people get overly hyped on the redesign, consider that they introduce a new lens mount.... then you would have to buy all new glass! 

Either way, it's cool to see new tech coming out, but it's unbelievable how excited people get about 8k. This to me is pointless. I have the S1h and am soooo happy with it, because it has all the functions you need for video. Like for real. It makes everything so seamless and smooth when shooting. I think if camera companies want to really entice videographers, focus more on really important features like more dynamic range, better tools for exposing correctly (waveform/false colors), better button customization, and so on. But no, these companies really just seem to focus on resolution and people eat it up. Remember how much you want that 8k raw when you buy your C-fast cards.... whoa boyyy that's a lot of storage you'll need, and C-fast cards aren't cheap! I think 2020 will be a great year for tech, and I'm looking forward to seeing the footage from both of these cameras regardless. Thanks for the info! 

Agreed on a huge boom in memory cards to support all of this crazy 8k and 12k RAW output, that was the main driver for me to buy stock in a few of the companies that produce them.  Don't get me wrong, I love RAW, but only shoot with it if I am getting paid extra (or it's a special pet project to me), but rarely am I seeing interest in RAW production requirements (granted, take that with a grain of salt, I'm a small production company and pick up commercials and corporate work here and there.  I'm sure there is more demand for the RAW workflow in cinema and music videos, neither of which I have had the opportunity to do yet).