Sony FX30 BRAW Crop With Video Assist 12G Models?

Hi guys, Elijah here! I am super excited that Blackmagic Video Assist 12G models will start to support BRAW on some Sony cameras, as that’s what I’ve shot on for years. I have an FX30 currently, and I’m wondering in advance whether it’s worth purchasing a Video Assist in order to shoot BRAW on it. One of the things that bugs me about the Atomos recorders (besides ProRes RAW) was that when filming in ProRes RAW on the FX30, there’s an additional ~1.33x crop on the footage, for whatever reason. This is only the case with the FX30; there’s no crop on the FX3, FX6, etc.

This brings me to my question: do you know where in the signal chain this crop occurs: within the FX30, or within the recorder upon the translation of the 16-bit linear raw to BRAW or ProRes? I’d love to shoot BRAW as I think it could significantly upgrade production value and general image quality in my work, but I’m not sure it’s worth it to me if I can’t use my camera’s full sensor (which is already a significant crop away from full-frame by nature of APSC). I’m not wanting to crop in any more on an already “small(er)” sensor.

Another way of looking at this might be this: does the FX30’s sensor output RAW data across the entire array of pixels, or is it only a subset in the middle? And furthermore, if the entire array is outputting RAW, will the Video Assist be able to reinterpret the entire sensor’s array as BRAW, or will it have to crop in like the Atomos? If the Video Assists are able to interpret the entire sensor’s pixel array as BRAW, that would be a HUGE win over the Atomos recorders.

Here’s why I ask: when shooting ProRes RAW with the Atomos recorders, the crop actually results in a whole stop of dynamic range being lost (plus increased grain), from what I’ve seen in different tests online. I want to purchase a Video Assist even just as a monitor, but the crop on the FX30 with the Atomos when shooting ProRes RAW has been a dealbreaker for me, and if the case is the same with the Video Assist 12G models, that’s an equal dealbreaker because the result leads to slightly poorer image quality with more flexibility (not what I’m needing in my specific use-case). I understand firmware on the cameras has a while til release (September 2025, last I checked), so all good if it’s not possible to tell yet! But I figured this channel might have been the best place to ask! I’m also curious to see if others have had this same question; I’ve not found anything anywhere else on the internet in regard to this! Let me know if there’s a better route to go about asking this question!

Thanks for your time reading/responding, and I’m looking forward to hearing your thoughts on this!

—Elijah

Hello!

I've been shooting a lot with FX30 and Atomos in Prores RAW. The cropping happens because RAW data is read pixel by pixel. This is not only the case with Sony cameras, but other companies like Panasonic, Canon, Nikon, RED.

The second reason why it's 4.7K (4672 x 2628) and not, say, 6K (6192 x 4128) is that there is simply no way to read every pixel at high bit rate from a given sensor. This is confirmed by the fact that the 100/120fps high-speed shooting mode utilizes an even smaller sensor area. The sensor and its pipelining are simply not fast enough. It might be possible to use the sensor with a lower color calculation depth, but then Sony wouldn't be able to deliver a high level of quality.

I remind you that when recording RAW, the camera does not use most of the processing pipeline, i.e. it does not perform demosaing (i.e. conversion from Baer scheme to standard RGB), color interpretation, noise reduction, distortion correction and other operations. The 12-bit raw data received from the sensor is packaged in a special way and transmitted to the recorder via HDMI.

Do you know where in the signal chain this crop occurs: within the FX30, or within the recorder upon the translation of the 16-bit linear raw to BRAW or ProRes?

Within the FX30. Its sensor is roughly 6K, so in order for a BMD or Atomos recorder to record the full area of the FX30's sensor in a raw codec they would need to receive a 6K raw signal from the FX30. That's certainly possible—there are a lot of cameras that can output a ≥4K raw signal to an external recorder—but for whatever reason Sony just doesn't allow the FX30 to do that over its HDMI port. It's enough of a limitation that I wouldn't really recommend recording ProRes Raw or BRAW with this camera.

The better option, IMO, would be to record 4K ProResHQ from the 10-bit 4K scaled output, which is still a substantial improvement over the internal X-AVC codec.

Here’s why I ask: when shooting ProRes RAW with the Atomos recorders, the crop actually results in a whole stop of dynamic range being lost (plus increased grain), from what I’ve seen in different tests online

I don't wanna take this thread off topic, but I am not a fan of the way IMATEST has impacted the way we as a community discuss dynamic range. It's like, no, the FX30 doesn't have less dynamic range in its cropped raw mode than it does in its scaled X-AVC mode; you're just bypassing the in-body compression and noise reduction. Now you get to use even better noise reduction in post, which means you're giving yourself a lot more flexibility to optimize your images from shot to shot.

It's like, if you compare eating a cake from a store (X-AVC) to eating handfuls of flour (BRAW/ProRes Raw) then of course the pre-made cake is gonna be better on paper. But if you know how use the raw ingredients to make your own cake (e.g. grading, noise reduction, film grain emulation, lens corrections, etc.) then you can produce something that you like a lot more.

Nice explanation, Alex!
Seeing so much instant noodle products around here, I always compare it to those vs. chef cuisine.

From what ive read braw from sony cameras will be limited for now to the burano and the fx6, The fx 3 maybe. no mention of the fx30.

Ric Marty

The better option, IMO, would be to record 4K ProResHQ from the 10-bit 4K scaled output, which is still a substantial improvement over the internal X-AVC codec.

Can that be done with the Video Assist? If so that would make me much more likely to consider using Sony for video in the future.