ProRes HQ vs HDV

[Robert Rogoz]

Ok, sorry about newbie question. I looked through some previous posts, but did not find clear answer.
I am using JVC GY-HD100 as a camera shooting 30p. I was going to use 24p setting, but on the first day I was suffering from weird drop-outs. I am shooting something I can't reshoot, so I switched to 30p. I am using 16x9 format. I am also going to use some of the shots from 2 years ago, shot on DVX100 in 4x3, 30p. I captured the footage in HDV setting. However I was advised to use ProRes HQ setting, as I am going to show this film on big screen. Since I upgraded the firmware and had camera head cleaned I had no drop-outs and I would like to shoot the rest in 24p.
So here are the questions:
1. Is the quality of capture and edit in ProRes QH is that much higher compared to HDV?
2. If I decide to re-capture it in ProRes QH, should I also re-capture DV footage in ProRes?
3. How would the process of output to a tape (JVC GY-HD100) and to DVD via compressor look like?
4. Is there any sense to shoot 24p, since it's going to be only about 1/4 of the footage? (or is it simply just going to complicate the post workflow without much benefit?)
Any advice appreciated
Sincerely- Robert Rogoz

[Christopher Ruffell]

Originally Posted by Robert Rogoz

1. Is the quality of capture and edit in ProRes QH is that much higher compared to HDV?

Much, much higher. Consider ProRes and ProRes HQ "the best" codec out there. When I used to shoot HDV, I converted everything I shot to ProRes and it was worth it. Converting your HDV footage to ProRes won't 'fix' your HDV footage (it'll look the same) but it'll hold up better in the editing and post-production process.

[Chad Dyle]

Originally Posted by Christopher Ruffell

Much, much higher. Consider ProRes and ProRes HQ "the best" codec out there. When I used to shoot HDV, I converted everything I shot to ProRes and it was worth it. Converting your HDV footage to ProRes won't 'fix' your HDV footage (it'll look the same) but it'll hold up better in the editing and post-production process.

Chris,

When you say "post-production process", are you talking about anything other than general editing, transitions.....? We have recently moved over from Sony Vegas where all we did was edit HDV. On the Mac side, I've been playing around with converting everything to Pro Res upon capture, but the storage is killing me. I have multiple people editing for me and trying to wrangle all of this data has become very stressful. Is it harder on the Mac to edit HDV? One of my editors is using an iMac and the other has a Quad-Core Mac Pro. I'd love to switch back to HDV to save the space, but would I be sacrificing that much with what we do?

[Noah Kadner]

Much higher quality with ProRes with of course some additional storage requirements over HDV but much higher image quality and ability to handle concatenation- rendering through generations. These are apples to oranges in comparison.

Noah

[Gary Nattrass]

One alt is that I edit in HDV and then do my final master in pro res 422, I find it is a more robust master file that I then produce compressed files from it.
I also go 25p in the final export to pro res.

[Dino Leone]

Thanks for bringing this up. I've been wanting to switch my HDV workflow for a while. This thread will definitely make me switch.
I also found a good article that explains how to easily capture into ProRes:

Capturing HDV Into ProRes Via FireWire | Moviola ? Training for Avid, Apple & Adobe | Production & Post Production Rentals & Sales | Los Angeles, Hollywood, Burbank

near the end of the article are the data rates of ProRes listed (between 12 and 22 Mbytes/sec, depending on the settings). It is a bit steeper than regular HDV which is around 5 Mbytes/sec if I remember correctly, right?

Best,
Dino

[Andrew Kimery]

Robert Rogoz,

Since you are mastering back to HDV tape there's no benefit, AFAIK, from transcoding into another codec since you will just have to transcode again back to HDV. An exception to this would be if you wanted to do your CC in Apple Color then you would need to transcode because Color doesn't like HDV. But, again, I don't know how well the footage will hold up getting compressed back to HDV for output to tape.


-A

[Justin Ferar]

You really can have the best of both worlds on a Mac...

Shoot and capture HDV.

In your sequence settings change the compression setting to ProRes and in your render settings change over to 10 bit.

There is no advantage to digitizing all your footage as ProRes because you are essentially doing the same thing in the workflow I just described!

Or just edit everything in HDV and then drop into a ProRes sequence and render all night which is fine too.

With regards to JVC- the only advantage to digitizing as ProRes is that you can digitize a whole tape as one clip, which is not possible with JVC HDV. It breaks up the clips at each start/stop.

[Aric Mannion]

Originally Posted by Justin Ferar

You really can have the best of both worlds on a Mac...

Shoot and capture HDV.

In your sequence settings change the compression setting to ProRes and in your render settings change over to 10 bit.

There is no advantage to digitizing all your footage as ProRes because you are essentially doing the same thing in the workflow I just described!

Or just edit everything in HDV and then drop into a ProRes sequence and render all night which is fine too.

With regards to JVC- the only advantage to digitizing as ProRes is that you can digitize a whole tape as one clip, which is not possible with JVC HDV. It breaks up the clips at each start/stop.

I used to think this too, but lately I have been capturing footage with "apple intermediate codec" instead of HDV and I have noticed the difference. The problem with hdv is not just that it is destructive whenever you cut or use a dissolve because of the GOP frames, it actually looks noticeably worse than apple intermediate codec. I shoot a lot of fast action, and I recently asked in a thread why every thing that is red has awful artifacting and harsh interlace lines. For a while I was convinced it was Final Cut or Quicktime not displaying HDV properly (the tape looked great on TV), but now that I can compare the HDV footage next to the apple intermediate codec footage I can see that my problems were from HDV compression during capture.
How could rendering HDV as Pro Res fix the artifacting damage that was done to the footage during capture? The only way to prevent your quality from getting worse during capture, is by capturing in high quality in the first place. I hope to be corrected if I am wrong, but this was my experience.

Seriously, I am not a technical guy and this knowledge is from my eyes. If I am misunderstanding something technically -I'd like to be set straight.

[Jason Livingston]

Aric, what you seem to be missing is that the HDV compression is done by the camera before the data is written to tape, not during capture. The video on the tape is already compressed in HDV, the damage has already been done.

When you capture HDV by firewire, you are doing a bit-for-bit exact copy, like copying a file. There is zero additional loss during capture because the video isn't being compressed again. (In fact, since compressing HDV is very processor intensive, it isn't even possible to compress HDV live without specialty hardware.)

However, when you transcode HDV to ProRes or AIC (or any other lossy codec), there IS some loss. That's because ProRes and AIC are lossy codecs too. It is mathematically provable that there will be some loss in image quality whenever you transcode to a lossy codec. The loss might be very small, maybe even invisible to the human eye, but there will be some loss. If the HDV captured to AIC looks better to your eye than the original HDV, then there must be something else going on, perhaps something wrong with your system or your workflow, because mathematically that can't happen.

If you have a source of video that is not already HDV compressed (using a HD-SDI, HDMI, or component connection from a live camera, or some other non-HDV camera), then capturing it to ProRes will give you much less loss than capturing that source to HDV. But once it's on HDV tape, the damage is already done.

Justin's suggested workflow does indeed give you the best of both worlds. Video that is unchanged stays in HDV with the exact same quality as recorded by the camera. Parts that are changed are rendered in ProRes, minimizing the loss that occurs.

[Gary Nattrass]

One side note to add to my own workflow is that I am working totally off compact flash now with the S270 and Z7 cameras. I use clipwrap to wrap and load the rushes into final cut pro, edit in 1080i 50i and then output to pro res 422 at 25p.

Seems to work really nicely and the tape is just there as a belt and braces back-up but having done lots of jobs so far I havent had a single problem recently.

[Arlen Sahakian]

Originally Posted by Gary Nattrass

One side note to add to my own workflow is that I am working totally off compact flash now with the S270 and Z7 cameras. I use clipwrap to wrap and load the rushes into final cut pro, edit in 1080i 50i and then output to pro res 422 at 25p.

Seems to work really nicely and the tape is just there as a belt and braces back-up but having done lots of jobs so far I havent had a single problem recently.

so ur saying that theres no need to capture the HDV to AIC or Pro Res just output it pro res
ok but ur shooting and editing 1080i 50i whats the point outputing it 25p ?

[Brian David Melnyk]

i remember a year ago capturing HDV as pro res because my initial install of FCPS2 left out the HDV codecs (not funny to figure out...) anyway, after i solved the mystery, i recaptured HDV as HDV, then opened QT movies of both, zoomed far in on the same frames, and i remember the pro res looked worse off. it had more jagged edges, i think due to the compression...
so i capture/edit HDV and only render pro res, and it seems to hold up nicely. also, the idea of loading 13 hours of footage for a project in pro res seems like a scary amount of space for a questionable benefit...
correct me if i am wrong!

[Gary Nattrass]

Originally Posted by Arlen Sahakian

so ur saying that theres no need to capture the HDV to AIC or Pro Res just output it pro res
ok but ur shooting and editing 1080i 50i whats the point outputing it 25p ?

Sorry forgot to add the 25p output is just for the internet TV files a tape master at 1080i 50i is also produced.

So in effect I am shooting and keeping everything at the same format 1080i 50i until final output which is then pro res 422 1080i 25p for files and 1080i 50i for tape.

The 25p looks nicer but I do not wish to shoot 25p on the camera as I prefer to do this in the controlled post prod environment.

[Aric Mannion]

Originally Posted by Jason Livingston

Aric, what you seem to be missing is that the HDV compression is done by the camera before the data is written to tape, not during capture. The video on the tape is already compressed in HDV, the damage has already been done.

When you capture HDV by firewire, you are doing a bit-for-bit exact copy, like copying a file. There is zero additional loss during capture because the video isn't being compressed again. (In fact, since compressing HDV is very processor intensive, it isn't even possible to compress HDV live without specialty hardware.)

However, when you transcode HDV to ProRes or AIC (or any other lossy codec), there IS some loss. That's because ProRes and AIC are lossy codecs too. It is mathematically provable that there will be some loss in image quality whenever you transcode to a lossy codec. The loss might be very small, maybe even invisible to the human eye, but there will be some loss. If the HDV captured to AIC looks better to your eye than the original HDV, then there must be something else going on, perhaps something wrong with your system or your workflow, because mathematically that can't happen.

If you have a source of video that is not already HDV compressed (using a HD-SDI, HDMI, or component connection from a live camera, or some other non-HDV camera), then capturing it to ProRes will give you much less loss than capturing that source to HDV. But once it's on HDV tape, the damage is already done.

Justin's suggested workflow does indeed give you the best of both worlds. Video that is unchanged stays in HDV with the exact same quality as recorded by the camera. Parts that are changed are rendered in ProRes, minimizing the loss that occurs.

I did my test by looking at HDV footage and AIC footage in after effects. But after I wrote my comment I dragged some HDV footage into a Final Cut sequence with AIC footage in the timeline. That was a better test, and I could no longer see the difference between the footage.
I was mistaken when I said I could see the difference between footage captured as HDV and footage Captured as AIC when rendering them as prores, AIC, etc...
But I'm still not totally sold, can someone clarify a couple of things for me?

1. what's the difference between pro res and apple intermediate codec?
2. I mostly do green screen work, I believe you when you say capturing HDV vs Pro Res may be invisible to the naked eye. But in my case of capturing footage to final cut, and then keying that footage in after effects: What is the best workflow?

I think a lot of times people shrug things off if it works for them, but we all don't do the same type of work. In the case of green screen work, how should I capture from my HDR-FX1?