F5: Why is the F5 more expensive than the FS7?

You know...I used to like Math...

Would you rather have...

Uncompressed 10-bit, 4:2:2 or Compressed 10-bit 4:4:4?
 
If it's HDCAM-SR or SRMaster compression, I'd definitely choose the compressed 4:4:4.

EDIT: Although, it does partially depend on the source. From an F3, 4:4:4 doesn't gain you as much as it does from an F23 or an F35.
 
I'd rather have 3:1-5:1 compressed RAW because it's smaller than uncompressed YCrBr and can even be smaller than compressed (ProRes 4444), and I think the quality of RAW is better.

If I had to answer the actual question, I'd pick compressed. "Un"compressed is too big to work with, and chances are during post workflow the decision will be made to convert it to ProRes anyway.
 
Last edited:
starcentral said:
I'd rather have 3:1-5:1 compressed RAW because it's smaller than uncompressed YCrBr and can even be smaller than compressed (ProRes 4444), and I think the quality of RAW is better.

If I had to answer the actual question, I'd pick compressed. "Un"compressed is too big to work with, and chances are during post workflow the decision will be made to convert it to ProRes anyway.
Click to expand...

That's a bit of a camera dependent response though...
 
Yes, no RAW possible on the F3 that's for sure.

If you can handle the whopping file sizes from the F3 uncompressed, I did shoot some tests with it once using the hyperdeck shuttle and the uncompressed images by far had little, if any, artifacts like I saw when comparing it to the same shot in XDCAM 420 (internal cards) and even ProRes 422 (via hyperdeck).

Uncompressed from the F3 is very very nice looking.
 
Just going back on something....if the camera is not a true 4:4:4, I don't see a point recording in that.

I would think 4:2:2 HQ or uncompressed to the Shuttle (super cheap alternative) would make more sense.
 
starcentral said:
If you can handle the whopping file sizes from the F3 uncompressed, I did shoot some tests with it once using the hyperdeck shuttle and the uncompressed images by far had little, if any, artifacts like I saw when comparing it to the same shot in XDCAM 420 (internal cards) and even ProRes 422 (via hyperdeck).

Uncompressed from the F3 is very very nice looking.
Click to expand...


Which is why more lightly compressed 4:4:4 codecs like SR444 and ProResXQ do such a nice job. They get you comparitively manageable file sizes (at HD/2K) for minimal compression artifacts - particularly when you're getting 'true' 4:4:4 from a downsampled 4k sensor like the F5/F55.

You get VERY high overall image quality from these formats. That said, the total lack of artifacting in compressed footage sure is a wonderful thing - just very hard to justify the storage costs for.
 
Also, for cameras like the F23 working from a prism or the F35 from an RGB stripe, RGB444 is pretty equivilent to raw - it's a log curve instead of linear, but it's essentially data that maps back to specific photosites. So the advantage of "raw" is partially dependent on how the sensor functions in the first place.
 
DLD said:
There's no visual difference, really. Maybe, if one blows up some charts, he can tell. Alexa65 is a $300,000 camera and it's still 17.6.
Click to expand...

True.

The F35 really was a brilliant and unique design. Too bad Sony moved way from the striped sensor, but I'm not sure they could have pulled that off with 4k and the S35 format.
 
NorBro said:
Just going back on something....if the camera is not a true 4:4:4, I don't see a point recording in that.

I would think 4:2:2 HQ or uncompressed to the Shuttle (super cheap alternative) would make more sense.
Click to expand...


It's almost 444 (as all single chip cameras are except the F35). I really wish I could find the explanation by Mitch Gross where he elaborated on the advantage of recording 444 regardless.

I guess it depends on what you are shooting and what your budget is. Personally I want every bit I can hold of.
 
NorBro said:
But 4:4:4 is 10-bit just like 4:2:2? (in F3)
Click to expand...

No. In terms of recording YUV 4:2:2 assigns a single color value for each pair of pixels in the recorded image, but luma (brightness) values for every pixel. With 4:4:4 every pixel is assigned it's own RGB color/brightness value. The effect is that 422 YUV only has 50% of the color information that 444 does. A camera like the Sony F3 has a sensor that is collecting data that falls between the two standards in such a way that by recording 4:2:2 you're losing some information, and by recording 4:4:4 you're recording more information than you actually collected. Hence, a benefit to recording 4:4:4, but not as much so as with a camera that truly gets 4:4:4 from the sensor.
 
Harry Lime said:
... The F35 really was a brilliant and unique design. Too bad Sony moved way from the striped sensor, but I'm not sure they could have pulled that off with 4k and the S35 format.
Click to expand...
Not with this current F5/55/65 generation, where it was more important for them to get to 4K resolution. Their replacements should have it in the top of the line, at the very least, especially since the sensor sizes should increase as well.
 
You must log in or register to reply here.
Back
Top