Worthwhile to add Convergent's nanoFlash to the JVC?

[Mike Schell]

Originally Posted by Jim Boda

That's why I would like to see it re-tested. Both clips are HDV compressed. For some reason they didn't do a comparison to regular HDV captured in it's native format. The extra analog conversion is NOT going to make HDV look better. It gives the look of having the backfocus slightly out of focus and the blacks are a little brighter.

We know that the JVC 250 is best option for the NanoFLASH...but, I would like to see a comparison w/ a 100 series (since there are more 100's out there) converted to HD-SDI and then compared to what the camera gets off tape in it's native HDV format.

The current test does make you wish that JVC had a HDMI out on the camera...if in fact, the current component analog outs go through a D/A conversion.

Hi Jim-
After we get the initial Flash XDR units shipped this next week, we'll go back and do more tests with the analog conversion. I agree, it's difficult to access the quality without using the exact setup: analog converter + nanoFlash.

I can absolutely guarantee you that the analog outputs from the JVC camera go through a D/A conversion. That's the only possible way to get analog video out, since all the internal processing is done in the digital domain.

[Bill Ravens]

Several comments, re: the intrinsic value of "uncompressed", CF storage, even if comparative video is not yet available...

1-My EX1(which has HD-SDI port) suffers quite noticeably from long form GOP motion artifacts. The effect of compressing, even at 35mb/sec, are very noticeable in the form of image blurring, especially on full frame pans. I think this is an inherent weaknes of HDv compression format, because my HD110 suffers from the same motion artifacts. The conventional wisdom is that these artifacts are negligible, due to ordinary motion blur. I, wholeheartedly, disagree. I'm very anxious to see whether I-frame compression mitigates compression artifacts for full frame pans.

2-I routinely use m2t compression on a Firestore FS4HD. These units are quite expensive(I own two of them, and rely on a hard disk a a fair bit of battery power, to record data). By comparison, the SxS SS storage on my EX1 is quite a bit more convenient for replaying clips, uploading to my NLE, and marking circle takes. Unfortunately, SxS cards are prohibitively expensive and they require a very expensive card reader on my desktop workstation.I quite look forward to relatively inexpensive CF card storage for the convenience. I have a ton of CF cards laying around, already, mostly of the 8Gb 133x variety.

3-My understanding is that the Flash recorder is also a playback device. The ability to export rendered 100Mb/sec files to an HD-SDI capabale monitor is very enticing.

[Mike Schell]

Originally Posted by Bill Ravens

Several comments, re: the intrinsic value of "uncompressed", CF storage, even if comparative video is not yet available...

1-My EX1(which has HD-SDI port) suffers quite noticeably from long form GOP motion artifacts. The effect of compressing, even at 35mb/sec, are very noticeable in the form of image blurring, especially on full frame pans. I think this is an inherent weaknes of HDv compression format, because my HD110 suffers from the same motion artifacts. The conventional wisdom is that these artifacts are negligible, due to ordinary motion blur. I, wholeheartedly, disagree. I'm very anxious to see whether I-frame compression mitigates compression artifacts for full frame pans.

2-I routinely use m2t compression on a Firestore FS4HD. These units are quite expensive(I own two of them, and rely on a hard disk a a fair bit of battery power, to record data). By comparison, the SxS SS storage on my EX1 is quite a bit more convenient for replaying clips, uploading to my NLE, and marking circle takes. Unfortunately, SxS cards are prohibitively expensive and they require a very expensive card reader on my desktop workstation.I quite look forward to relatively inexpensive CF card storage for the convenience. I have a ton of CF cards laying around, already, mostly of the 8Gb 133x variety.

3-My understanding is that the Flash recorder is also a playback device. The ability to export rendered 100Mb/sec files to an HD-SDI capabale monitor is very enticing.

Hi Bill-
Long-GOP MPEG2 (the type of compression used in HDV) is intrinsically more prone to motion artifacts than I-Frame only compression. Long-GOP uses both spatial (I-Frame) and Temporal (P,B Frames) to compress the video. The P and B frames are created by compressing the differences from the initial I-Frame. In a 15-Frame GOP, there can be considerable differences from the first to last frame, especially in a high-motion shot.

However, as the bit-rate is increased, this motion blurring can be reduced substantially, by applying additional MPEG2 tricks, which amount to basically inserting a new I-Frame in the middle of the GOP.

So, I would recommend that you test both I-Frame only and high bit-rate (100 Mbps) Long-GOP MPEG2 before making a final decision. I have been pleasantly surprised at how well the Long-GOP compression holds up to high motion conditions. But, everyone has there only definition as to acceptable quality.

Yes, you are correct, both the Flash XDR and nanoFlash have HD-SDI playback capability. I think there will be many applications for capturing video directly from you NLE into nanoFlash and then using the nano as a very high-quality playback device (client review, museums, projectors, etc)

[Brent Kolitz]

Originally Posted by Tim Polster

I also agree a test using a 250 model with a Nano recording SDI out, a Nano recording component out (with converter) and recording to tape at the same time would be very useful.

This would show all options and we could see how much the Nano is affecting image quality.

Agreed -- very much looking forward to this test, as this will really determine whether it's worth it to pursue this path. Not sure whether the component out circuitry on the HD100 and HD200 series are identical, otherwise I'm sure the HD100/110 users will want that tested as well.

The converter will also make a difference in the comparison, but it seems that the new BMD Analog-SDI would be the one people might use...