Advantage to SDI capture? was: Couple of questions

[Chris Hurd]

Originally Posted by Floris van Eck

Or can I output through SDI and record to my computer from tape when the footage was already record on tape?

Yes you can... in fact that was the whole point of the posts from Thomas and myself just above, at http://www.dvinfo.net/conf/showpost....13&postcount=6 and http://www.dvinfo.net/conf/showpost....13&postcount=8

Is there any other way to record uncompressed footage, let's say firewire?

Not through FireWire. FireWire is limited to a maximum throughput of just 400Mbps (or 800Mbps if it's FireWire 800). Remember, uncompressed is about 1.5Gbps, so there's no way FireWire can accomodate that. It is possible to capture uncompressed through the component video output terminal, but that's an analog connection and it raises an entirely different set of concerns (which have already been discussed extensively on this site in our XL H1 and HD110 forums).

I understand that the editing of uncompressed footage is faster, but what about the required storage space, is it also x times higher?

Yes the required storage space is higher by several orders of magnitude, you'll need terrabytes of it, but keep in mind the point Thomas was making in the post just above yours. It doesn't *have* to be uncompressed; it can simply be a less compressed codec than HDV, such as DVCPRO HD or JPG2000 for example.

The genlock and timecode outputs are only useful for multi camera shooting, or also if you use only one camera for 100% of your work?

The sole purpose of Genlock is to sync multiple devices together so if you're shooting only with one camera then it's pretty much worthless to you. As far as TimeCode is concerned, it depends on whether you choose to record audio in the single system style (that is, on camera) or double system style (separately from the camera). For double system sound, the TimeCode jack is a must.

[Thomas Smet]

DVCPROHD can take up to around 12.5 MB/S for hard drive space and needed bandwidth. While this is higher then the 3.125 MB/S for HDV it is much much lower then using uncompressed which could be as low as 125 MB/S or much higher. Basically DVCPROHD uses 1/10th the amount of space and speed as the lowest form of uncompressed which is a very good balance between quality, speed and file size.

Another thing to think about is that these are 1080i cameras. While DVCPROHD may be a little soft for 720p video because it only uses 960x720 pixels the 1080i form isn't as bad. With 1080i DVCPROHD it uses 1280x1080 pixels compared to HDV using 1440x1080 pixels. There really isn't that much of a difference between the two so capturing to DVCPROHD with 1080i HDV gear in my opinion is much better then using DVCPROHD with a 720p camera. If I had a 720p camera I'm not sure if DVCPROHD would be good enough for me. I have no problem with DVCPROHD at 1080i however.

On the PC side the new MJPEG codec from Blackmagic seems to do around 11 MB/S for normal video. This number can change based on how complex the video is because it will adjust the datarate to maintain the level of quality. If the scene is blowing trees over busy water ripples then the datarate may be higher. If you are shooting a solid color such as a bluescreen shoot then the datarate could be half of that.

So yes the files are larger but not really that much larger.

[Stu Holmes]

Originally Posted by Floris van Eck

I understand that the editing of uncompressed footage is faster, but what about the required storage space, is it also x times higher?

Unless someone corrects me on this (entirely possible!) AFAIK, uncompressed "HDV level" footage is approx. 540Giga-bytes per hour. That's about 40times as big as 'normal' HDV footage.

[Barlow Elton]

Originally Posted by Floris van Eck

So the conclusion is that the SDI output is not worth the $3000 extra for most of us? I rarely expect to capture live footage. I am not sure, but to capture uncompressed one would need a fast computer or harddisk with you in the field? Or can you still output to tape and capture through SDI?

My experience is mostly based on HDV and Final Cut Pro, and from my own SDI experiences I can tell you that if you don't intend to ever capture the uncompressed signal for acquisition or in a switched TV studio environment or greenscreen stage (i.e. live camera signal, NOT HDV tape playback through SDI) then you just don't need to spend the extra money.

Chris is right on the money about the jackpack being more than SDI--timecode i/o and genlock are very important for a bunch of professional reasons, but if you intend to use one of these new Canon cameras just like you used prosumer DV cameras in the past, only now in HD, then you don't need SDI. Just use firewire and HDV editing.

Thomas noted a number of good formats to convert HDV footage to for "faster editing", and there are lots of good reasons to go the intermediate route, (which I did exclusively in the past) but I've found native HDV editing within FCP to be fairly responsive, and it saves a lot of disk space.

What I tried to explain though is that it's fairly easy to take an HDV timeline (at least in FCP) and change the codec and do a final render in a less compressed format and save all the time and hassle of capturing raw footage to an intermediate with it's larger/faster disk requirements. It also avoids the entire "conform to HDV, rebuild and recompress GOPs" etc. This allows you to make a final edit that retains all the quality of the original footage without using up so much disk space overall, and also renders graphics, color correction and other effects noticeably better too.

Does this make sense?

[Bill Pryor]

Good info there, but would you mind getting more specific for the HDV newbie. We're using FCP now, and I've shot HDV for others but have never taken any HDV into my system, though I am interested in doing so if this new A1 camera looks as good in the flesh as it does in person. I read a lot of different things about how to edit HDV in FCP and all of it confuses me.

Is what you're saying like this: You capture in HDV...edit in HDV and then change the render to something else? Or do you start a timeline in something else after you capture but before you edit? 'Scuse my ignorance--I'm still in a DVCAM world.

[Barlow Elton]

Originally Posted by Bill Pryor

Is what you're saying like this: You capture in HDV...edit in HDV and then change the render to something else?

Yes, that's it. If your deliverable is something other than HDV it helps to preserve the best quality possible. (given you render and output in a better codec)

[Floris van Eck]

Originally Posted by Barlow Elton

What I tried to explain though is that it's fairly easy to take an HDV timeline (at least in FCP) and change the codec and do a final render in a less compressed format and save all the time and hassle of capturing raw footage to an intermediate with it's larger/faster disk requirements. It also avoids the entire "conform to HDV, rebuild and recompress GOPs" etc. This allows you to make a final edit that retains all the quality of the original footage without using up so much disk space overall, and also renders graphics, color correction and other effects noticeably better too.

Does this make sense?

So basically what you are saying is that I can shoot in regular HDV, and then convert to a different format, which gives me faster render times / editing and a better image quality but relatively low disk usage? Can you tell me a bit more about the Cineform codec? What does it do exactly?

[Jay Stebbins]

So, if you don't need the genlock or timecode of the G1, but still would like the option of capturing uncompressed video; can you use the component out on the A1 and digitize the analog signal with a capture card. Would the data-rate be any where near the 1.4g of the HD-SDI, or the quality? Or has the signal coming out of the component already gone through the HDV compression?

[Barlow Elton]

Originally Posted by Floris van Eck

So basically what you are saying is that I can shoot in regular HDV, and then convert to a different format, which gives me faster render times / editing and a better image quality but relatively low disk usage? Can you tell me a bit more about the Cineform codec? What does it do exactly?

I'm saying edit all the way with HDV. When you have a finished edit, copy and paste the HDV timeline into a custom timeline with your choice of codecs. For ultimate quality I use the Sheer codec (which cost me $150 but is nearly 1/3rd the bit rate of uncompressed, yet actually the same quality) but PhotoJPEG and a few others are pretty good too.

CineForm for now is a Windows-only format. It works with Premiere Pro and Vegas as far as I know. It costs $500 for a version that will allow you to capture via firewire in real time and converted to their nearly lossless codec.

Steven Dempsey uses this system. He could tell you more about it than I could.

[Barlow Elton]

Originally Posted by Jay Stebbins

So, if you don't need the genlock or timecode of the G1, but still would like the option of capturing uncompressed video; can you use the component out on the A1 and digitize the analog signal with a capture card. Would the data-rate be any where near the 1.4g of the HD-SDI, or the quality? Or has the signal coming out of the component already gone through the HDV compression?

The live signal from the camera head (non-tape playback signal) is uncompressed analog from the A1. It can be captured with the appropriate hardware, and at least with the XL-H1, the signal is virtually the same as SDI.

This will offer higher quality if compressed to a better format, but the question is if it's worth the trouble.

I would say in a studio greenscreen shoot...absolutely, but in the field...not really, until some company creates a firestore-like device that will allow trouble free recording in the field with a portable device.

[Jay Stebbins]

I have been looking around for the answer to this on the firestore. Do you lose any quality by recording to the Firestore? And how reliable have they been for the XLH1? 24F? I know what I want to know is buried in these forums. But where?

The component uncompressed on the A1 (possibly) sold me on that camera. I am going to give the A1 & HDV a shot. I just feel better knowing I have access to an uncompressed feed from the camera if needed. (20mm shots of the moving ocean, moving trees, and moving clouds; my grand vista shots.) I fear that the amount of data in those shots might overwhelm the HDV codec. As everything in scene is small, detailed and moving. I have no problem acquiring the pieces to capture an uncompressed signal in the field. (well, I do have to wait for a special Expresscard 34 to be created) But these shots will happen in the Spring.

First however, I will let HDV try to capture what I see...

Thank you very much, without this forum, this project would never happen.
Jay Stebbins

[Paul Matwiy]

Originally Posted by Jay Stebbins

I have been looking around for the answer to this on the firestore. Do you lose any quality by recording to the Firestore? And how reliable have they been for the XLH1?

The files are recorded as m2t MPEG2 files. Quality will be identical to that laid down on tape while not being susceptible to tape dropouts. I beta tested the Firestore a couple of years ago and found them to be extremely reliable. I'm planning on upgrading the one I won to be able to take the HDV signals.

[Gene Crucean]

Originally Posted by Paul Matwiy

I'm planning on upgrading the one I won to be able to take the HDV signals.

How is this possible? And does that mean you could buy a really cheap firestore and upgrade it to work with other cameras like a Z1 or P2?

[Chris Hurd]

I'm sure that Focus Enhancements would be the first to inform you that there is no such thing as a "really cheap" FireStore, but yes, you can purchase an upgrade for the FS-4 or FS-4 Pro in order to make it HDV compatible. Cost of the upgrade is $299.

[Jim Feldspar]

Originally Posted by Thomas Smet

DVCPROHD can take up to around 12.5 MB/S for hard drive space and needed bandwidth. While this is higher then the 3.125 MB/S for HDV it is much much lower then using uncompressed which could be as low as 125 MB/S or much higher. Basically DVCPROHD uses 1/10th the amount of space and speed as the lowest form of uncompressed which is a very good balance between quality, speed and file size.

Another thing to think about is that these are 1080i cameras. While DVCPROHD may be a little soft for 720p video because it only uses 960x720 pixels the 1080i form isn't as bad. With 1080i DVCPROHD it uses 1280x1080 pixels compared to HDV using 1440x1080 pixels. There really isn't that much of a difference between the two so capturing to DVCPROHD with 1080i HDV gear in my opinion is much better then using DVCPROHD with a 720p camera. If I had a 720p camera I'm not sure if DVCPROHD would be good enough for me. I have no problem with DVCPROHD at 1080i however.

On the PC side the new MJPEG codec from Blackmagic seems to do around 11 MB/S for normal video. This number can change based on how complex the video is because it will adjust the datarate to maintain the level of quality. If the scene is blowing trees over busy water ripples then the datarate may be higher. If you are shooting a solid color such as a bluescreen shoot then the datarate could be half of that.

So yes the files are larger but not really that much larger.

"With 1080i DVCPROHD it uses 1280x1080 pixels compared to HDV
using 1440x1080 pixels."

This sounds great but I'm interested in shooting dramatic shorts and
using 24F. Can i get the benefits of 1080i with 1280x1080 and shoot
at 24F or am I completely stupid because the "i" means no "F"?