HDSDI vs FIREWIRE

I'm digitizing from a panasonic 1400 DVCPRO HD 720p24. I've been told that the HDSDI capture method I'm using is lower quality than firewire. I called AJA video and they said that HDSDI is the exact same quality as firewire. How can that be? Is that true? I'd rather use HDSDI so I can return the deck and use one we don't have to rent.

FCP 6.0.1
XRaid 2.7 TB

It is true...same quality. HOW? It is a digital signal that is being fed from the HD SDI port.

You're fine.


www.shanerosseditor.com

Listen to THE EDIT BAY Podcast on iTunes
[itunes.apple.com]

But, my tech argues is that the HDSDI signal becomes uncompressed to full on HD (albeit in digital form) when it goes out of the VTR. His argument for this is that monitors don't have a way to uncompress DVCPRO footage, and since you can plug the HDSDI cable into a monitor... it must be a standard uncompressed, no-codec signal. Therefore, when that uncompressed signal goes back into the AJA card, DVCPRO compression is reintroduced to the footage, essentially adding a level of compression that doesn't exist when you hook it up straight firewire.

His argument makes a lot of sense, which is why I trust him even though everyone else says the opposite.

Dan

The DVCPRO HD codec in on the tape...so when that signal is being fed out via HD SDI, it is the DVCPRO HD codec...for that is the best it can get. Now, the KONA hardware can choose to capture that as DVCPRO HD, or capture it as uncompressed HD, or as DV...or whatever. The card chooses what compression to capture as. If the signal is DVCPRO HD, and you choose to capture as DVCPRO HD, then you'll get DVCPRO HD.

NOW...if the signal was coming DIRECTLY from the camera and avoiding tape altogether, THEN the signal would be an uncompressed signal...because it hasn't hit the tape yet. THEN if you captured DVCPRO HD vs Uncompressed, you'd see a difference.

But you aren't...you are taking a DVCPRO HD compressed signal, and capturing it as DVCPRO HD. If you had a unit with firewire and HD SDI you could do a test and see that there is NO difference. I did that test (as did a few other editors I know who work with this format), which is why I know that it is the same quality.

But this is a bit moot isn't it? You don't have a deck with firewire so the ONLY option you have is via HD SDI. So no matter what, that is how you have to capture.

The AJA techs know their stuff...Trust them. And your engineer isn't wrong either. But from what I was told and can tell from my tests, the DVCPRO HD codec isn't converted to uncompressed via the HD SDI port, then recompressed when you capture. An HD SDI DVCPRO HD compressed signal is sent out, that the Kona sees and captures.

Now, that is the extent of my technical knowledge here...I am a creative guy for pete's sake!


www.shanerosseditor.com

Listen to THE EDIT BAY Podcast on iTunes
[itunes.apple.com]

The DVCPRO HD codec in on the tape...so when that signal is being fed out via HD SDI, it is the DVCPRO HD codec .... the DVCPRO HD codec isn't converted to uncompressed via the HD SDI port, then recompressed when you capture. An HD SDI DVCPRO HD compressed signal is sent out ...

don't think so Shane. the data on the tape is video footage compresed using the DVCPRO HD codec ... but when its played out over HD-SDI it is being decompressed using the same codec. the signal coming in via the HD-SDI connection is an uncompressed digital video signal.

as to whether the you can discern any difference in quality between capturing the raw compressed HD data via firewire versus capturing the decompressed HD via SDI, as DVCPRO HD, well maybe you can't ... and if so would be a testement to the quality of the AJA product, and may also explain why AJA themselves would attest that it is the exact same quality, but its not the exact same data.

as ever, i reserve the right to be wrong ... especially so far from my comfort zone!

I reserve the right to be wrong too. But this was how it was explained to me.


www.shanerosseditor.com

Listen to THE EDIT BAY Podcast on iTunes
[itunes.apple.com]

Alright, so I put a call in to the Panasonic Broadcast Customer support people. An engineer will call me tomorrow and hopefully they'll give me some sort of definitive answer that I'll post here. In response to your post, Shane, we are renting an AJ-1400 deck with both firewire AND HD-SDI, because our house deck only has HD-SDI. So, part of the reason I started this was to figure out a way to save money by not renting the deck. However, I have to admit, at this point curiosity is what's really driving me.

To my knowledge a single decompress-recompress of a digital signal is virtually transparent. Any degradation of the signal would probably only be noticeable by running comparative analytical tests of the "before" and "after" images. If you have a workflow that necessitates several decompression/recompression or transcoding processes from capture to delivery, you will start to see noticeable quality loss at some point.

I haven't seen any studies of the DVCPro HD codec yet, but according to EBU research, DVCPro 50 (SD, 4:2:2@50 Mb/s) can handle 7 generations of compression before degradation becomes noticeable at a critical viewing range (4 times the height of monitor). The DVCPro HD codec uses a heavier compression, but I wouldn't think it's too far behind DVCPro 50 in terms of robustness.

So I don't think using HD-SDI for capture would matter, unless the rest of your workflow requires a lot of recompression (and it shouldn't). Looking forward to hearing about what the Panasonic engineer has to say about it.

Geir.

Haven't heard back from the from the engineer yet, but I did capture the same frame via SDI and Firewire. I actually did the same section of firewire twice. I noticed that the two firewire captures were identical when zoomed in 800 percent. The HD-SDI was slightly different from the Firewire. I wouldn't call it a quality difference to be perfectly honest. In a blind test I wouldn't be able to say one was better. But the important thing to note from this test is that the HD-SDI is DIFFERENT from the Firewire. Whether there is a quality loss or not, the fact is the HD-SDI doesn't end up with the same picture. This *might* be because I use the AJA capture settings when I capture via HD-SDI... and somehow AJA interprets things differently. For that I'll have to wait for Panasonic to call. But either way, I am now convinced (given that I know Firewire is a strict data transfer) that HD-SDI is not the same as Firewire.

More later.

Curious, try this:
Lay the HD-SDI footage on a track above the FW footage. Sync them up to the frame and make the top layer's composite mode Subtractive. Identical images would result in a completely black frame.
What do you get?

ak
Sleeplings, AWAKE!

OK....take a GANDER at this thread at the Cow:

[forums.creativecow.net]

Take no notice as to how WRONG I AM about the codecs that BM and AJA use. I was WRONG. So...
Nothing to see here...move along.


www.shanerosseditor.com

Listen to THE EDIT BAY Podcast on iTunes
[itunes.apple.com]

> OK....take a GANDER at this thread at the Cow:
> [forums.creativecow.net]

Holy frijole...that argument got heated. It's like a religious war over there.

Awfully interesting topic, though.


www.derekmok.com