Capturing using firewire or capturing using capture card? The difference?

[Dom Stevenson]

Thanks for that Bill.

A couple of years ago i worked for a guy who was running an early version of Premiere to edit DV CAM stuff on. It was a nightmare and despite my attempts to get him to put his hand in his pocket and get a Final Cut system he refused. He didn't want to spend the money. After 10,000 crashes and several expensive cock ups he eventually decided to take the plunge. Only now he couldn't spend enough, and insisted we needed a top of the range Blackmagic card to capture DV at high quality. I told him it was a waste of money and that there was nothing to be gained by the considerable expense.

He won. Hey, it's his money.

To my knowledge he has never had the slightest benefit from having this £1500+ gizmo in his computer.

And your post confirms this.

[Steve Oakley]

Originally Posted by Bill Davis

You get NO benefits from a dedicated card if you're working with HDV.

If that's all the data rate you're ORIGNIATING there will be NO advantage to doing anything other than simple firewire capture - and you will NEVER need a more robust capture card. Period.

So your need for a dedicated card STARTS and ENDS with your ability to originate, work with, monitor and output HIGH DEF. Period.

There's a sorta-maybe exception to this is if you're preparing SD or HDV footage for HD broadcast. Then it might make sense to prepare your titles and graphics in high def over SD footage - and a card will help you do that on a somewhat older computer - but that's a pretty specialized game.

Again, if you aren't finishing to HD, there's NO need for any cards.

If not, forget about them.

FWIW.

this really is not correct.

1. kona, BM and MXO cards handle the scaling involved between the codec's native res of 1440 to 1920 output res. that takes the load off of the CPU/GPU. DVCpro100, HDV, HDcam all reduce the res.

2. capture via SDI in DV may produce somewhat better looking images due to the VTR processing the chroma channels. SDI -> uncompressed or lossless may well indeed produce better results than DV native capture. it all depends on the camera used and the quality of shooting.

3. capture to 10 or 12 bit color space can very much indeed improve things when doing color correction, especially when pushing those color corrections. while the source it only 8 bit, using a higher color space will reduce banding and noise when pushing things hard. of course you generally don't need to do this if the shooting was good in the first place but sometimes life isn't perfect and you have to deal with bad shots. 10/12bit color space also allows for more finesse when doing even modest corrections. if you haven't done this, you need to try it. I would also say that FCP is a less then great place to do this. combustion, smoke, flame, even AE if you must are better places to do this. this process doesn't add anything, it simply reduces the rounding errors that happen when you work entirely in 8bit color space. if you look at the image ona wave form and work entirely in 8bit, you can see bands of no pixels. if you go to 10/12bit space, those bands are reduced or eliminated depending on how you process the footage.

4. a I/O card is a requirement to monitor HD simply for color managed output. SD and HD color space are different. monitoring in the correct color space is critical for color correction. one reason I like the MXO for my laptop.

5. An I/O card will provide realtime downconvert to SD which is something I use to make DVD-R's for preview almost daily.

6. using SD material in HD projects is very common these days. an IO card won't help here in uprezing, but if you have betaSP sources, you need something to bring it in and I'd rather pull it in via 8/10 bit uncompressed or lossless then loop thru a DV deck to convert it to DV any day. That will produce much better results. fact is, the worse you source, the more you need uncompressed/lossless to preserve what is there. compressing to DV only makes it worse and harder to fix.

[John Miller]

The ONLY benefit of using the hardware described for HDV (or DV) source material is that it is dedicated to the job and, therefore, operates in real time as well as freeing up other computer resource(s).

Mathematically, everything that can be done with the dedicated hardware solution can be done with a computer ingesting DV or HDV material. The hardware solution is, after all, nothing more than a dedicated digital computer running specific and highly optimized algorithms. Whether the right tools exist on a given computer is another matter.

This applies to material that is already in the DV or HDV format. Obviously, live source material will benefit from the SDI route assuming that the live feed from the camcorder is the original uncompressed material.

The real consideration here is the cost vs. benefit of investing in the hardware vs. additional computers + corresponding software.

[Chuck Spaulding]

Wow this conversation covers about all the misconceptions about HD...

There are several things to consider in developing an efficient workflow, one is obviously how you want to capture your footage.

Uncompressed HD is 1920x1080, it is 4:2:2 8 or 10 bit with a data rate of approximately 192MB's/sec. [that's mega BYTES] The variable for the data rate is the frame rate.

There's also 1280x720 which is also 4:2:2 8 or 10 bit with a data rate which I believe is about 20% less than 1080.

For compressed HD there's DVCProHD which is 1280x1080 4:2:2 8 bit with a data rate of approximately 100 Mb's/sec [that's mega BITS][the DVCProHD codec is also a constant bit rate]. For 720P - the DVCProHD is resolution is 960x720.

HDV is 1440x1080, it is 4:2:0 8 bits with a data rate of 25Mb's/sec.

These are acquisition standards. There are also codecs that can be used as a digital intermediate or "post" production format.

ProRes which is a wavelet VBR codec that comes in various resolutions, 8 bit and 10 bit and is 4:2:2 with two HD data rates, medium quality 140Mb's/sec and HQ 220Mb's/sec. Keep in mind that Sony HDCam has a bit rate of 140Mb's/sec, so if your going to finish on tape the medium quality is fine. Also you should be aware that this is NOT a lossless codec and not only is it a MAC Quicktime only codec its a FCP only codec. [This may have changed but last time I attempted to use it with the Adobe CS3 Production Suite it did not work] and it does not support an alpha channel.

There's the Cineform NeoHD codec which is similar to the ProRes codec and anyone whose interested can check it out at [http://www.cineform.com/] but ther is a sginificant difference -- CineForm QT files are compatible on both Windows and MacOS and is compatible with Adobe CS3; Apple Final Cut Pro, and Motion. This is also a LOSSY codec.

SheerVideo is a LOSSLESS codec that is platform and application agnostic which also does support an alpha channel. This is a great choice if your shooting uncompressed HD.

Aja and BlackMagic video cards can capture HDV, HDSDI and component HD and convert it to either DVCProHD, ProRes or the Cineform NeoHD in realtime. This can have a significant impact on your workflow.

First, for anyone who's been editing HDV they know the time it takes to conform before writing to tape. Although this eliminates the conforming time all of these compressed codecs have larger file sizes comsuming more disk space and requiring more bandwidth [FW800 and eSata are generally fast enough].

And although a capture card can't add more quality than what was in the original, converting from a long gop 4:2:0 (HDV) to an intra frame 4:2:2 is much better for any kind of image processing, color correcting, or compositing.

I use Motion and Color correct quite a bit so when I edit HDV I almost always capture to either DVCProHD or ProRes and not only is the quality better but the projects are much more efficient (smooth).

Anyway, I hope this helped.

[David Knaggs]

Originally Posted by Chuck Spaulding

Wow this conversation covers about all the misconceptions about HD...

HDV is 1440x1080, it is 4:2:0 8 bits with a data rate of 25Mb's/sec.

Just to clear up another misconception ...

There are TWO types of HDV: HDV1 and HDV2.

Chuck is referring to the interlaced format, HDV2 (Sony and Canon).

HDV1 is a progressive scan format, 1280 X 720, 8-bit, 19.7 Mbps (JVC). It is compressed with MPEG-2 and the first JVC HDV cameras used short GOP (6 frame). The later, more advanced models are using a longer GOP, I believe.

For more info, see the article on this link by Chris Hurd (under the section "What is HDV?"):

http://www.hdvinfo.net/

[Bill Davis]

There's already too much needless complexity here.

I'll just say a few simple thing.

HDV is an absolutely wonderful format for budget video producers. You get a LOT of bang for your buck. If that's what you're working in, EMBRACE it. But don't attempt to tart it up and act like something it's not.

IF you're working with HDV - stop worrying about the perfect back end workflow and also don't waste your money on fancy computer add-on cards.

Spending hundreds or thousands to achieve a few percentage points of better color correction is a WASTE of time and resources IMO.

Trust me, NOBODY's going to sit down after viewiing your work and say...

"Nice effort - however it was ruined by the fact that the color of the glass of Bordeaux on the table in shot 152 was slightly off..."

You can get PERFECTLY acceptable color - and a PERFECTLY acceptable and watchable video program out of a stock HDV camera and a stock editing system WITHOUT ANY ADD ON CARDS. Period.

Spend that money on other things - like audio or lighting gear or self-education - that will have a REAL impact on the professionalism of your work output.

The rest of this discussion is like arguing about a blemish on the back of a supermodel's knee.

It might well exist, but it's largely irrelevant in the overall scheme of things.

[Jim Fields]

This is what I read in all of the replies..

Blah Blah blah, GOP, 4:2:2, 4:2:0, blah, Pro Res, 24P, blah blah.

I take the KISS method, if I shot the event ( I shoot weddings) in HDV 1080i I then use a firewire cable and import using...get this.....easy setup HDV 1080i 60.
The funny thing about doing that, is i am working in the native format i shot in. I know I know it sounds really stupid of me to do something so simple, but, I do like to Keep It Simple Stupid.

I know what your saying though, just weddings, thats all I am doing, and your working on a budgeted feature, well, I know alot of people who shoot weddings at 24P, and try to replicate a film look, and in the end they take 3 times longer than I do to finish an edit.

I have a BM card so I can capture HD recording off of my Dish DVR, my camera never touches the BM card. I dont worry about GOP Formats, Pro Res, etc. I import in the NATIVE format I shot in (and the funny thing is, your not really shooting 24P, try using film, thats 24P) and I have no issues editing.

But what about the coloring? Well, your shooting with Video, on a camera that probably cost less than 15,000.00 . Get a camera that runs 50K or more and you will get good color compared to a Sony V1u, or a P2. Otherwise, your NLE should work fine when it comes to Color Correction, or even use Color from Apple.


I will bet my Mac Pro against anyone in DFW that I will get the same results using the KISS method, import over firewire, cut it, export, than trying to use some goofy system with Capture cards, Pro Res, etc. And get it done alot faster.

[Dom Stevenson]

Another excellent post Bill.
The voice of sanity.

Jim, I'm getting great results with FCS2's Color and love working with the programme.

I would however like a Matrox MXO box to go with my 23" apple display, so i guess that's next on my shopping list.

Anyway, great thread by all involved.

[Andrew Kimery]

Originally Posted by Bill Davis

You can get PERFECTLY acceptable color - and a PERFECTLY acceptable and watchable video program out of a stock HDV camera and a stock editing system WITHOUT ANY ADD ON CARDS. Period.

I agree w/the spirit of your post, but w/o any additional I/O hardware you can't monitor your HDV reliably. A BM Intensity card plugged into an HDTV from Best Buy
would work better than judging the footage using the just Canvas window or Digital Cinema Desktop function.

-A

[Bill Davis]

Originally Posted by Andrew Kimery

I agree w/the spirit of your post, but w/o any additional I/O hardware you can't monitor your HDV reliably. A BM Intensity card plugged into an HDTV from Best Buy
would work better than judging the footage using the just Canvas window or Digital Cinema Desktop function.

-A

Andrew,

On this we're going to have to disagree.

EVERY (and I use the term EVERY advisedly) consumer television I've ever seen has been "factory tuned" for one thing and one thing ONLY. To make the picture POP in a home viewing situation. I've personally never found a consumer television that sports anything even remotely resembling the accurate colorometry of a broadcast monitor.

So, no matter how expensive your "card solution" is - if you're running it into ANYTHING less than a certified CALIBRATED professional high-def broadcast monitor, you're completely at risk for WASTING all your efforts at color adjustment.

And I've got to believe a person legitimately constrained by finance to working with a sub-$3k camcorder is pretty dang unlikely to be sitting in front of even a starter true High Def monitor (Sony BVM-A14F5U perhaps) that costs nearly TWICE that.

And if NOT, how can they justify buying the card when they can't even be sure they'll SEE the color changes they're hoping they're making accurately!!??

Makes no sense to me at all.

FWIW

[Dom Stevenson]

There are some very good articles (particularly by Shane Ross) explaining how the 23' Apple cinema display or equivalent, along with the Matrox MXO, while not quite perfect, is far and away the best bang for buck HD grading set up on the market today. If you own the monitor already, as i do, it's a great budget package. Shane also points out why the BM Intensity option falls short, but you'll have to read the article yourself if you haven't done so already.

Here it is.

http://library.creativecow.net/artic..._shane/MXO.php

D

[Andrew Kimery]

Originally Posted by Bill Davis

Andrew,

On this we're going to have to disagree.

EVERY (and I use the term EVERY advisedly) consumer television I've ever seen has been "factory tuned" for one thing and one thing ONLY. To make the picture POP in a home viewing situation. I've personally never found a consumer television that sports anything even remotely resembling the accurate colorometry of a broadcast monitor.

So, no matter how expensive your "card solution" is - if you're running it into ANYTHING less than a certified CALIBRATED professional high-def broadcast monitor, you're completely at risk for WASTING all your efforts at color adjustment.

And I've got to believe a person legitimately constrained by finance to working with a sub-$3k camcorder is pretty dang unlikely to be sitting in front of even a starter true High Def monitor (Sony BVM-A14F5U perhaps) that costs nearly TWICE that.

And if NOT, how can they justify buying the card when they can't even be sure they'll SEE the color changes they're hoping they're making accurately!!??

Makes no sense to me at all.

FWIW

I'm not saying it's a good option, I'm saying an inexpensive card and an inexpensive TV is better than nothing (nothing being only monitoring your footage in FCP's canvas window or using Cinema desktop).

You said, "You can get PERFECTLY acceptable color - and a PERFECTLY acceptable and watchable video program out of a stock HDV camera and a stock editing system WITHOUT ANY ADD ON CARDS. Period" and I don't think that's an accurate statement because there is no way to send the HDV signal out of your computer for monitoring w/o an additional card or box. As inaccurate as an Intensity card plus consumer HDTV is it's more accurate and helpful than no external monitoring option at all, IMO.


-A

[Craig Parkes]

I've produced a short film on Mini DV, that we had the good fortune of ending up being able to grade on a Baselight grading machine ( http://www.filmlight.ltd.uk/products/baselight ).

I can tell you without a doubt, that there is a DIFFERENCE when it comes to a firewire versus SDI workflow.

The film was edited in DV, captured via firewire, cut on Avid Xpress.

As DV is a compressed format, if we had gone to a Baselight grade from that with DVCAM tapes, we would have been stuffed, we would have been stuck in 8 bit realm, the grading machine would not have had the latitude to add the depth of colour it did etc.

What we did was we ingested into Avid Media Composer at 1:1 in an online edit over SDI, output to Digibeta, and then had that sucked into the Baselight machine.

The additional latitude in the grade this gave us was amazing. If we had stayed in DV all the way there was no way we would have access to these pro tools also, it doesn't fit the workflow.

You don't gain anything in terms of picture quality, certainly - a digibeta master ungraded and a DVmaster ungraded will look identical.

However, if you then bring back that Graded digibeta master, and copy it to DV, the difference is stark and plain to see. Generational loss and less colour space has a marked effect.

Your choice of workflow should be determined by your mastering format first and foremost.

If you are hoping to go film out, then you want to keep the pipe as fat as possible all the way as long, no exceptions.

If you are going to HD Broadcast, then you want to either being doing an offline/online procedure or be working at or above your final HD Master format the entire way through.

If you are going to festivals/high end SD Broacast 10bit SD video at 1:1 should be your goal for ingest and output, again an offline online workflow is fine, but if you never get to 10bit 1:1 in the edit, you'll also never get their in the grade, and for FX and titles, and this can honestly REALLY make the difference.

If you are doing Wedding Videos, and your master is likely to be SD DVD's, and you aren't going to be grading them heavily anyway, then keeping an entirely firewire based, cut the acquisition format workflow makes sense.

To use the pipe analogy - The size of the pipe should also be relevant to the size of the bucket at the end of the pipe. If your pipe is wider than your bucket you really are just wasting water.

Going to a lot of trouble to get the highest quality images possible into your machine then not doing significant high end grading or titles/fx and chucking it all onto DVD at the end would be a complete waste of time.

However, if your take is "I don't know why ANYONE would do that" then you haven't thought about people with the bigger bucket at the end of the pipe, and you certainly haven't seen the difference in results in terms of effects work and colour correction this can lead to.

[Chuck Spaulding]

Originally Posted by Jim Fields

This is what I read in all of the replies..

Blah Blah blah, GOP, 4:2:2, 4:2:0, blah, Pro Res, 24P, blah blah.

I will bet my Mac Pro against anyone in DFW that I will get the same results using the KISS method, import over firewire, cut it, export, than trying to use some goofy system with Capture cards, Pro Res, etc. And get it done alot faster.

First of all DVCProHD and ProRes are FREE!, both codecs come with FCP6x. Secondly you can convert HDV to DVCProHD or ProRes in REALTIME with the BM intensity Pro, a $295 card, and thirdly, any image processing is done at 8 or 10 bit 4:2:2 [depending on the format you chose to capture with] which provides a much broader color gamut and requires NO CONFORMING.

This is simple, it is not expensive and I will bet that this workflow is MUCH FASTER than editing HDV natively.

This is not an endorsement of HDV, in fact the first thing I tell everyone I work with is get out of that format as soon as practicable. But HDV is definately finding its way into just about every level of production.

Also, this is not an either or proposition. Many of the projects I work on mix uncompressed, compressed HD and HDV footage together. That's what the ProRes codec was designed for.

If your a single artisan doing wedding videos using the same HDV cameras to produce a SDDVD, fine stay HDV. But if your collaborating with other artisans with a variety of formats and/or you want to master to blu-ray you need to really understand the blah blah blah don't you think?

[Bill Davis]

You said, "You can get PERFECTLY acceptable color - and a PERFECTLY acceptable and watchable video program out of a stock HDV camera and a stock editing system WITHOUT ANY ADD ON CARDS. Period" and I don't think that's an accurate statement because there is no way to send the HDV signal out of your computer for monitoring w/o an additional card or box. As inaccurate as an Intensity card plus consumer HDTV is it's more accurate and helpful than no external monitoring option at all, IMO.


-A[/QUOTE]


Maybe we just both have disagreements on what real users want to do verses what they need to do.

I have no statistical evidence of the following, it's just my gut feeling.

Most of the people who are worried about this stuff are actually working on what I heard one pro call "resume films."

This is to say they have an idea. A passion. Friends and associates who will support them. They have acquired a decent basic understanding of the production process. But they lack the funding, the track record or the connections to work inside a full-scale production but their passion drives them to do the best they can with what they can afford.

Which is GREAT. But the idea doesn't stop there...

It's a resume film because if it manages to open the right doors, they'll JETTISON HDV in the blink of an eye - as soon as they can afford to - and would LAUGH at the idea of doing Feature 2 in HDV if Feature 1 succeeds.

THINK ABOUT THIS...
Why do you want to waste time learing how to MAX OUT a tool that you're gonna want to DUMP at your first opportunity???

That's the definition of a BIG waste of time, IMO.

All I'm saying in this thread is that the things that will make someone's RESUME FILM succeed have NOTHING to do with color grading or color space or the amount or kind of signal compression between the shoot and mastering. Nothing. Nada. Bupkis. Period.

I know this is true because a cavalcade of genius filmmakers from DW Griffin to Charlie Chaplin worked with gear and a workflow that was CRAP compared to any camcorder or NLE you can name today.

And in my opinion, as harsh as it is, spending significant time learning color grading is pretty useful *only* IF YOU WANT A CAREER AS A COLOR GRADER!

If you want to actually have a hope of someday making excellent FILMS - put away the distractions like becoming your own colorist - and go study Abnormal Psychology or Drama - or Religion or Music.

Because the actual useful understanding is that films and videos aren't made in cameras or NLE's at all. They're made about 5 inches behind the occupied eyepiece.

Feed your head.

And go on a diet about anything that has to do with technology that you KNOW you'll outgrow as you get better.

And that includes pushing HDV to it's uber-max limits.

For most people, in most real-world situations - it's a diversion - not a smart career move.

But that's just my opinion...

YMMV