ProRes vs ProRes HQ

[Jeremy Garchow]

The ProRes whitepaper explains a little more for you but probably not enough.

https://images.apple.com/finalcutstudio/resources/white_papers/L342568A_ProRes_WP.pdf

Gary Adcock has gone on record saying that for live capture ProResHQ is best, but if you are doing any software converion (from 8 bit sources), ProRes SQ is best due to the way Quicktime handles the rounding up to the extra bits.

https://forums.creativecow.net/readpost/98/870346

Jeremy

[Gary Adcock]

Stealing my thunder again eh jeremy???

herb,

I did some extensive testing on this, not finding enough difference to proclaim one better than the other from a visual point of view.

on the other hand- the software converted camera native files to ProRes (NOT HQ) consistently played better on the 5 systems I tested them on. It seems that when you are doing a software conversion to PRHQ files the playback needed to have a higher performance CPU to maintain the playback without dropping frames.

Couple of things to remember- camera captures are 8bit – ProRes is 10bit.

ProRes HQ is forcing the data to completely fill out the entire 10bit space that is 4X as levels of gray as 8bit. (256 vs 1024 levels) where as the “plain” version of ProRes allows the 8bit camera file to just float inside the 10bit color space – only adding the extra information to the video file when it is needed, rather than forcing all of those 768 additional levels of gray to be occupied all the time like the PRHQ version does.

Make Sense?

gary adcock
Studio37
HD & Film Consultation
Post and Production Workflows
Inside look at the IoHD

[Sean Oneil]

That is one excellent post. Thanks for sharing your findings!

Sean

[Herb Sevush]

Gary –

I appreciate your trying to enlighten me, but i still have a few dark clouds overhead.

not finding enough difference to proclaim one better than the other from a visual point of view.

Is this only true for conversions to ProRes, or even when shooting to ProRes? And if it is true for both, why would Apple even bother confusing the issue?

On the other hand I’m quite happy to learn that I can pretty much ignore HQ for all my editing needs.

Thanks.

Herb Sevush
Zebra Productions

[Jeremy Garchow]

[gary adcock] “Stealing my thunder again eh jeremy??? “

Hey, at least I gave you the credit!

🙂

[Chris Poisson]

Gary,

I’ll second Sean’s comment, excellent, well informed post.

Have a wonderful day.

[Sean Oneil]

[Herb Sevush] “On the other hand I’m quite happy to learn that I can pretty much ignore HQ for all my editing needs.”

I think if you have a 10-bit source then HQ will be better.

Sean

[Gary Adcock]

[Herb Sevush] “Is this only true for conversions to ProRes, or even when shooting to ProRes?”

well since you cannot shoot to prores…..

actually think of it this way,

capturing from a true 10bit source – “LIVE” output from a camera, HDcamSR, high end frame sequences – go with PRHQ

HDV and DVCPROHD – you will not see any difference in the encode to plain ProRes until you get to the 5 or 6th generation of re-compression.

Herb it is all about using the bits in the best manner- if you have lots of CPU power (Octo w/ oodles of RAM) you are not going to see much, if any playback issues. However if you are on an older machine- the RT benefits from using standard vs PRHQ are amazing.

Chances are you are not shooting with native 10bit camera, standard ProRes is smaller in size and easier on the CPU to Process than the HQ version. Less hassles for you and your computer.

gary adcock
Studio37
HD & Film Consultation
Post and Production Workflows
Inside look at the IoHD

[Chris Coote]

I have been working a lot with Prores and Prores HQ in HD.
I have noticed that HQ stands up a lot better to further rendering in FCP than
non HQ…even with just one render pass.
It’s especially noticeable in areas of high movement…HQ shows a lot less noise and artifacting.

[Gary Adcock]

[Sean ONeil] “That is one excellent post. Thanks for sharing your findings!”

thanks to both of you for the comments.

I have spent about 3 months tearing prores apart and figuring out the software conversion issue was about 2/3’s of that time. Most of my comparisons were done thru a hardware scope so that I could look at the info at a quantitive level rather than “it looked better”

I am talking to Tim about posting the WP on the Cow – I am working out a clearance issue for one of my images. it is kinda long at about 3000 words.(9 pages with graphics)

gary adcock
Studio37
HD & Film Consultation
Post and Production Workflows
Inside look at the IoHD