ex1 SD solutions and HD delivery options questions

To get 4:2:2 you have to use HD-SDI which demand a lot of extra computing 'hardware'. I wouldn't bother going to 4:2:2 unless I plan to do a lot of extensive green screen, compositing and other CG work.

There is also a product development that would allow 4:2:2 recording into a small portable device, but it is not a cheap solution. The best thing that you can do to decide is either to

a) rent the cameras that you are interested in.
b) go watch some sample clips from the A1, HVX, EX1 ....etc.
 
In post, why are you downconverting after editing? To preserve quality? Would it be equally possible to downconvert first then be able to editing with a smaller file that should be less stress on you machine and thus faster to work with?
Is the downconvert a slow process that you wouldn't want to do with all your footage?

My questions more to how I sell this workflow to SD clients than to my own use.

- Lenny
 
Lenny,
Yes you got the concept right. You can either down-convert before or after post depending on their hardware. It is a little lighter work to the machine to edit SD vs HD regardless of the codec (DVCPROHD, HDV, XDCAM EX ...etc).

Selling to the client is easy. Tell them that you can give them the master in HD also. I am sure everyone want s to see their events(wedding,bdays ...etc.) in all its HD glory in the future.
 
Personally I have found that working in HD then down converting in compressor gives a better final project tha downconverting first and editing in SD. Plus, if a client wants a HD copy further down the line, its a piece of cake.
 
Steve Shovlar said:
Personally I have found that working in HD then down converting in compressor gives a better final project tha downconverting first and editing in SD. Plus, if a client wants a HD copy further down the line, its a piece of cake.
Click to expand...

Absolutely! Not once have I used DVCPro50 on our HVX. When we got it I thought we'd use DVCProHD for filmmaking and DVCPro50 for corporates. Like I say haven't used DVCPro50 at all, if we're doing a corporate we shoot 720pn for max recording time on a P2 card. For filmmaking 1080p. Even though DVCPro50 is a good codec, it's a bit redundant. Shoot the highest rez possible and it gives compressor more to work with to create a SD mpeg2, and in my tests it looks better than shooting SD and compressing to SD.

For low light filming, the HVX is downright terrible due to the image noise in the darker areas of frame. Especially if you stick a 35mm adaptor on the front of it, and a 35mm lens.

I'm considering getting an EX1 as soon as we're finished on current projects.
 
Dxmetal said:
If you go with the Canon A1, plan to spend another $1k on the canon hv20 or hv30 to use as a transfer deck (if you shoot 24p). Unless, you want to use your A1 as a deck and wear out the head.
Click to expand...

It is true that running the cam as a deck adds extra wear to the heads. It is also true that these heads are designed to run for a very long time. Way more than the hours youre likely to bput on the cam. Once every 20 hours I run the head cleaner for 5 seconds. And get a proper clean once a year.

Steve Shovlar said:
Personally I have found that working in HD then down converting in compressor gives a better final project tha downconverting first and editing in SD. Plus, if a client wants a HD copy further down the line, its a piece of cake.
Click to expand...

Absolutely, there is simply no need to do it any other way for the final output. Its just the most simple method. No need to complicate things further. The latest compressor does a great job IMO. The output process takes awhile but the results are worth it.
 
I have just a few more questions. I'm pretty sold on this camera but workflow wise:

1) does it seamlessly go between one card to another when filming? for example if i fill up card 1 while still filming, will it switch to card 2 without losing a frame? If so can you pop out card 1, put in another card, and just keep going that way?

2) I have an old MacBook Pro (not intel). Will I be able to use the card slot on this computer to pull files off a card during a shoot? Is it just drag and drop, or do i need to use the Sony File Manager prog, and if so does that require an intel Mac?

Thanks once again for all the help.
 
1) Yes, the process is seamless and you can do your card rotation like that.

2) The SXS card will appear as a drive letter and you can just copy and paste everything to a folder on your laptop or an external HD. If you want to view the clip though, you need the Sony transfer software. I am 98.99% sure the sony transfer software works with mac intel only.

mjdorris said:
I have just a few more questions. I'm pretty sold on this camera but workflow wise:

1) does it seamlessly go between one card to another when filming? for example if i fill up card 1 while still filming, will it switch to card 2 without losing a frame? If so can you pop out card 1, put in another card, and just keep going that way?

2) I have an old MacBook Pro (not intel). Will I be able to use the card slot on this computer to pull files off a card during a shoot? Is it just drag and drop, or do i need to use the Sony File Manager prog, and if so does that require an intel Mac?

Thanks once again for all the help.
Click to expand...
 
Wow, thats even better ! yeah, just confirmed 100% that Sony xdcam transfer program is indeed powerpc and intel mac friendly !

Your laptop should be fine as a "dumping machine" as well as viewing the clips. I dont remember if the powerbook has express card slots or not. However if you have or can get an express card adapter that is even better. You can buy 1 more SXS card and have the following workflow/rotation.

i) Shoot with 2 SxS cards. When one is full, hand in the "full" card to the assistant.

ii) Get the 3rd card (blank SXS) from the assistant and continue shooting.

While you are shooting.... the assistant could

i) download the footage to the laptop or straight to 2 external HD for backup & extra security.
ii) erase the card's content
iii) have the card ready for the next rotation.

Enjoy the XDCAM EX1 and post some footage to share. If you want a full HD resolution camera from Panasonic, the cheapest one would be the hpx3000 $48k without lens.
 
Ok, so I'm about to purchase this camera. Was just getting my finances in order. I have a new 2.8ghz dual core IMAC. I would assume that with a 35mbps datastream that it would easy handle the edits. Or would I want to spend the cash on a Mac Pro? I would just imagine that this would speed things up in after effects and motion, but I'm not going to suffer horribly on the IMAC am I?

Thanks
 
It's not the megabits that cause the problem, it's the type of compression. DVCPRO-HD is 100 megabits, yet a 2.8ghz imac could probably handle three or even four streams in realtime.

XDCAM-EX is long-GOP, which is much more processor-intensive. Even so you should be able to get at least one stream in realtime on that iMac, I would guess. The real delay would come when you finish your edit and you go to conform, that can take a long time.
 
mjdorris said:
Hi All,

I currently use a borrowed DVX100b for shoots. I'm considering stepping up my game and purchasing an HD camera. Originally I was attracted to the HVX for it's SD and HD capabilities, and the option to work with tape and solid state. As far as I can tell the EX1 doesn't shoot SD. Does anyone having any experience using the Ex1 and down converting for SD delivery? Is it a reasonable idea to shoot in 720p or 1080 and downconvert? I still have clients that ask for SD, but more and more are asking for HD.

Another quick question--what is the typical way that people are delivering HD product? Create a master dvd folder and then take it to a blue ray capable duplicating house? Excuse the noobishness.

Thanks,
Matt
Click to expand...

it's NOT A PROBLEM to go from HD to SD. A very large percentage of the content that you see on standard definition NTSC broadcast television originated on a higher resolution format (like HD or Film).

OVERSAMPLING IS GOOD :grin:

Lots of data made small isn't an issue. Not lots of data made big IS an issue.

I think it's not so much about the resolution as it is about the shape. 16:9 vs. 4:3.
 
Barry_Green said:
It's not the megabits that cause the problem, it's the type of compression. DVCPRO-HD is 100 megabits, yet a 2.8ghz imac could probably handle three or even four streams in realtime.

XDCAM-EX is long-GOP, which is much more processor-intensive. Even so you should be able to get at least one stream in realtime on that iMac, I would guess. The real delay would come when you finish your edit and you go to conform, that can take a long time.
Click to expand...


Thanks for the info Barry. I did some research into what long GOP was, but couldn't turn up many results. Does long GOP just mean there are more frames packed in than normal GOP?

I was hoping to get through this without buying a new computer too, but then again, the new Mac Pros are pretty exciting.
 
Well, there's intraframe compression and there's interframe compression (which uses a GOP).

Intraframe compression means that all frames are treated discretely and individually. This is the type of system that every professional recording system has traditionally ever been, from DigiBeta to MPEG-IMX to HDCAM to DVCPRO-HD to DVCPRO50, including DVCAM and DVCPRO, and also as used by the highest-quality recording/mastering formats around, HDCAM SR and D5-HD and AVC-Intra.

Then there's a new development called interframe, which was developed to try to cram more footage into a tinier bandwidth. Interframe was developed for low-bandwidth delivery systems like DVD and ATSC broadcast television. It has been added in camcorders primarily because the HDV tape and XDCAM disc had very limited bandwidth transfer rates, and using long-GOP compression was the only way they could fit a high-def recording onto limited-bandwidth media.

The idea of long-GOP means that it's using a long Group Of Pictures as one compressed entity. There's also short-GOP, as used by CineForm (IIRC, CineForm uses a two-frame GOP) and JVC uses a medium-length GOP of six frames in their HDV 24P. The general idea is that they don't encode each frame individually, instead they only track the changes between frames and encode those. That way, if something remained unchanged from frame to frame it would be simply copied over, rather than re-encoded (as would happen in an intraframe codec).

In theory long-GOP is a great idea, in practice it works very well on static scenes and can be impacted (or even severely hampered) by scenes where a lot changes from frame to frame. This isn't necessarily a fault of long-GOP itself, but moreso with the tiny bitrates that have been allocated to existing long-GOP systems. Long-GOP with 100 megabits would probably be just wonderful, long-GOP at 19 megabits leaves a lot to be desired!

When going to editing, long-GOP is much more processor intensive, around 7x as processor-demanding as intraframe would be. With intraframe the NLE can select any frame and access it directly. With long-GOP, that's not possible because frames do not exist in their entirety. You have to uncompress a series of frames, applying the changes from frame to frame, to get to the one you want.

As computers get faster and faster, this becomes less of an issue.
 
You must log in or register to reply here.
Back
Top