RAW vs. other recording formats

jmone said:
I shoot a bit using S&Q at 1 and 5fps and it crops as well.
Click to expand...

Interesting. I hadn't noticed that but will take a look when I can. The only time I observe cropping is when I push the frame rate to 120. I never notice any other cropping going on. Obviously, you'd see cropping with breathing compensation, but I never use that mode.
 
Just checked a project. I had to in post zoom the "normal" in by 130% to match the frame with the 5fps S&Q take.
 
You are right. It's breathing comp. Turned it off, and the only crop was at 100 and 120fps in S&Q. Turned breathing comp off. Never realized how much of crop breathing comp did!
 
Ha, ha. I wish I'd refreshed my browser to see your correction before I pulled out my camera to double-check that there is no cropping befow 100 fps. There isn't. I was going to suggest you may have accidently been using Clear Image Zoom, but breathing compensation makes more sense, especially if the cropping is around 30%. HFR cropping is only about 10%.

And while I had the camra out, I also checked to see if there is cropping at 100 fps, and there is, even though the CineD article says there isn't. They are wrong on that point.
 
Doug Jensen said:
... but the difference is extremely hard to quantify if the footage was shot well in the first place.
Click to expand...

And that's the crux of it, today. So many people in this industry now just can't do the basics and "have" to have RAW and LOG to salvage the majority of their footage. Whereas people that know how to shoot, can/are using RAW and/or LOG to maximize their images and eek that last "1/10" out, so many others are using it just to get a "useable"/"passable" image that is barely as good as someone that knows what they're doing is getting straight out of camera with a burned in "709 look".
 
Yeah - I missed one of your posts as I too did not refresh in time! It seems that Breathing Comp is turned off when S&Q mode is enabled. So just tested some more of the Cropping @ 100 and 120fps in S&Q with Breathing Comp On and Off. From what I can see is if you have Breath Comp turned on it crops in on "normal" mode (I still can't get over by how much!) but when swapping to S&Q at 100 or 120fps it first crops (as expected) in then turns off Breathing Comp so crops out.... and you end up with the same frame as "normal". So (for me at least on the 28-135 @ 28mm). No wonder I was confused seeing different framing at <100fps but not at 100/120fps. I'm leaving that Breathing Comp off!
 
Run&Gun said:
And that's the crux of it, today. So many people in this industry now just can't do the basics and "have" to have RAW and LOG to salvage the majority of their footage. Whereas people that know how to shoot, can/are using RAW and/or LOG to maximize their images and eek that last "1/10" out, so many others are using it just to get a "useable"/"passable" image that is barely as good as someone that knows what they're doing is getting straight out of camera with a burned in "709 look".
Click to expand...

Absolutely right.
 
jmone said:
Yeah - I missed one of your posts as I too did not refresh in time! It seems that Breathing Comp is turned off when S&Q mode is enabled. So just tested some more of the Cropping @ 100 and 120fps in S&Q with Breathing Comp On and Off. From what I can see is if you have Breath Comp turned on it crops in on "normal" mode (I still can't get over by how much!) but when swapping to S&Q at 100 or 120fps it first crops (as expected) in then turns off Breathing Comp so crops out.... and you end up with the same frame as "normal". So (for me at least on the 28-135 @ 28mm). No wonder I was confused seeing different framing at <100fps but not at 100/120fps. I'm leaving that Breathing Comp off!
Click to expand...

Bokeh control and Breathing Comp are gimmicks. There's not a chance in hell I'd ever use either one.
 
...& worse still you lose 2-3mm at the wide end. Can aways crop in post but can't get back what has been cropped in camera, and the 28-135 is not that wide to start with.
 
Doug Bee said:
Raw means unprocessed in terms of workflow.

If you can set a white balance and ISO then those are the main tenants of a raw workflow. While a high bit depth log image lets you bend the image a lot, I personally like a slider that has a white balance number and iso number on it.

And bits. 12bit log is needed to manipulate without breaking the image.
Click to expand...

No numbers matter. Only image quality.

You can bend an image colour balance in post or make it brighter or darker.

Ive been through this with a resolve developer. That whoever you make a change the destruction to the image is the same. Add red or pull a colour temperature slider.. it is the same thing.

Of course a 12bit log or raw file than can take more bend than 8 or 10 bit.

(the amount of bend before break is not well understood on the log v raw thing)
 
Run&Gun said:
And that's the crux of it, today. So many people in this industry now just can't do the basics and "have" to have RAW and LOG to salvage the majority of their footage. Whereas people that know how to shoot, can/are using RAW and/or LOG to maximize their images and eek that last "1/10" out, so many others are using it just to get a "useable"/"passable" image that is barely as good as someone that knows what they're doing is getting straight out of camera with a burned in "709 look".
Click to expand...

Of course there are total idiots who cant shoot. And there are those looking to squeeze the best image from the available resources.

Someone like me 'knows how to shoot' (the old school way) - and that would be a closed road, some security guards, a genny truck and operator, an a couple of 5k outside shooting through a 12x12 sheet with two mpre guys stopping it from blowing away.

But I also know that it is not going to happen for reasons of budget. so because I cant control scene dynamic range with lights (the 'proper' way) having some stops in terms of log or raw will help me get by with my puny ap600 light.

Again a sky with church spire.. you could do it the 'proper way' with a grad nd (which would make the spire dark too) or you could draw around the sky + spire in post and pull it down (darken the sky) creating a better result than the grad nd which is a fixed thing that cannot account ofr the spire or shape of the horizon.
 
morgan_moore said:
No numbers matter. Only image quality.

You can bend an image colour balance in post or make it brighter or darker.
Click to expand...

I think the tools numbers matter. Because each post tool uses arbitrary numbers otherwise that don't translate between platforms.

It's the same in still image processing. You can make it warmer or cooler on a slider, or you can type in 2800K or 6500K. They are the same result.

But from a process point of view I'd rather know that it's warmer by a number that translates into something absolute, and know how different it is to the color temp was that was metered in the scene, not just broad ideas like warmer or cooler on a slider that gives the same IQ.

Is it an actual stop darker than what I metered or exposed on the day or is it 20 points of adobe make it darker slider which is the same as 12 points of C1 make it darker? It might not matter to you, but I need to know what's changed and by how much so that I know what to do differently next time.
 
Yes indeed.

WHat I meant was if you can make it warm with a red slider or a colour temp slider (in resolve) the file will look exactly the same, not create banding or suchlike - ergo the numbers dont matter.*

If you want to communicate across platforms or image sets having a consistent language (numbers!) to do that may well matter.

Phrases like 'give it a kick' 'a steep S' 'a mild kick' 'its a bit cold' may not have consistent interpreatation.

--

In my years of raw shooting - motion and stills Ive mainly used 'click on a grey card' to find grey and then not looked at what numbers that led to.

If you need to have a 'rule': " process all shots taken under 5000k lighting at 6000k" then such numbers matter.


--
*Obviously you cant output a cold jpg and fix that without creating banding.. you must be accessing the raw data.
 
jmone said:
If you have not read it, this test of the FX6 (and A7SII) of Log Vs ProRes Raw was interesting - Sony FX6 Lab Test – External ProRes RAW vs Internal XAVC-Intra | CineD

Internal Log was objectively better.
Click to expand...

I don't think that conclusion is entirely correct. It seems that internal log has some noise reduction applied (even on the fx6 with noise reduction turned completely off). And that basically the in-camera NR performs better than what they can do with the PRR in Resolve, thus internal log has slightly better DR and underexposure latitude.

But there's an advantage in color quality with the PRR.

3 stops below base exposure...
Both the ProRes RAW and the internally recorded XAVC-Intra samples look very similar at this stage (but colors get muted somewhat on the internally recorded images),
Click to expand...

And at 5 stops underexposure...

What needs to be mentioned though, the 12bit ProRes RAW files do retain colors in a much nicer way, despite the noise.
Click to expand...

I'm not saying this merits shooting raw over internal log. But it is a notable difference. It's not clear to me if they observed superior color quality in general, or just in their underexposure test. If the latter, then that implies that shadow colors are superior with PRR.
 
Doug Bee said:
I think the tools numbers matter. Because each post tool uses arbitrary numbers otherwise that don't translate between platforms.

It's the same in still image processing. You can make it warmer or cooler on a slider, or you can type in 2800K or 6500K. They are the same result.

But from a process point of view I'd rather know that it's warmer by a number that translates into something absolute, and know how different it is to the color temp was that was metered in the scene, not just broad ideas like warmer or cooler on a slider that gives the same IQ.

Is it an actual stop darker than what I metered or exposed on the day or is it 20 points of adobe make it darker slider which is the same as 12 points of C1 make it darker? It might not matter to you, but I need to know what's changed and by how much so that I know what to do differently next time.
Click to expand...

I'm with you. I hate relative value adjustments versus absolute metrics that give me greater understanding of what's happening, better repeatability, etc. Either way can get the job done but it's a great feature of raw processing when you can essentially dial in your camera settings anew.
 
ahalpert said:
I don't think that conclusion is entirely correct.
Click to expand...

You are probably right, but after reading their results (and as Doug pointed out), it became apparent it is not worth the hassle. Even with their errors, it seems to be the only non-subjective comparative test of PRR vs SLOG on the FX6 I could find. I've found little to complain about Sony's Log footage except if it gets right to the edge. I've some footage from my India trip where the sky was way hazy (from burn offs). I could pull it to blue in resolve and push the brightness around a bit but tended to get some faint sky banding in the 4:2:0 render. No idea if RAW would have helped with that or not.
 
You must log in or register to reply here.
Back
Top