OT: Let's discuss LP vs SP recording

Comments

corug7 wrote on 3/10/2006, 8:25 PM
The VX-1000 used the same CCD set, I believe, as the VX3.

==============================
Instead Sony joined the shrink the CCD trend like the others did.
==============================

Boy, that TRV-600 took a great picture for a single chipper.
craftech wrote on 3/11/2006, 5:35 AM
The VX-1000 used the same CCD set, I believe, as the VX3.

==============================
Instead Sony joined the shrink the CCD trend like the others did.
==============================

Boy, that TRV-600 took a great picture for a single chipper.
Message last edited on 3/10/2006 8:30:43 PM, by corug7.
===========
I love reading the User Reviews for products on Amazon.com
You can always tell who is older. They invariably have the same complaint:

"I bought this camera thinking it was going to take such great pictures and it doesn't even take pictures as good as my old one did"

And YES, I do remember the 3-chip Hi8 VX3. Amazing image. Crystal clear.
I also remember when the standard CCD size for a single chip Hi8 camera was 1/3 CCD. The industry took advantage of the relative lack of "generation loss" with DV cameras by deciding to gyp the consumer by doing something I can't stand........ Adding useless bells and whistles for advertising purposes and short changing the basics to save on manufacturing costs.

John
farss wrote on 3/11/2006, 6:18 AM
Don't blame it all on the manufacturers, I think we get the cameras we deserve (the 'we' excludes us).

Few years agon I'm in a local DSE store (that's Dick Smith Electronics), they have a few palmcorders on display, one of them the very capable PC100, best camera ever built in that form factor and it's around $1,000 off, end of line runout. The sales guy is trying to convince someone it's the best camera but he's not going to buy it, he wants the new model that's around $1K more.

I try to help the salesman but still the guy's not convinced, see, he wants the new one with USB, don't give a damn about the picture quality or the better build, nope, just that damn USB port.


The again I've intercut footage from a PDX10 (3x1/6" CCDs) with a DSR 570 (3x1/2" CCDs) from stage shows and I was pretty amazed at how well the PDX10 held it's own. With extreme lighting the 570 shines. We did one show and some of the numbers were just slashes of light with dark skined talent wearing white darting in and out of the lights (everything on manual though) and for that I was surprised that anything could capture it but the 570 came through.

All the Hi8 footage I've worked on though looks a little too soft to me, it's got some sort of appeal I'll admit but there's something missing, still I'm comparing it to cameras like the PD150.

Bob.
Stonefield wrote on 3/11/2006, 9:42 AM
This whole LP vs SP is fascinating. Being of the old school analog thinking myself, I never even considered the possiblities of recording in LP. My workflow is to record, capture and store the tape away never using it again. ( Archive only ) All back ups are on some kind of hard disk.

I'm gonna do some experiments with LP recording.

Good thread....

Stan
GenJerDan wrote on 3/11/2006, 10:26 AM
Do the cameras do any kind of error-correction encoding when they lay down the bits?

I could understand why not, if they don't. Something like Viterbi would at least double the space used on the tape....but wouldn't it be nice to have?
farss wrote on 3/11/2006, 2:15 PM
There is some CRC data in DV but not much.
HDV has much more error correction. However by it's very nature when HDV's mpeg-2 looses it, it goes down big time.
Now here's the really oddball bit, there's more error correction applied to DV out the anaolgue ports of a deck than the firewire ports. Several times I've been able to recover cleaner vision from a tape going S-Video from the DSR-11 into an ADVC-300 and then into VidCap than I could going from the VCR direct into VidCap.

As to the SP V LP thing, it really depends on your workflow. If all the tapes are only for your use and are played back in your gear where you know what works and what doesn't you're probably going to be OK, sort of like driving to the local shops without your seatbelt on.
Once you start having to send tapes outside, that's when LP becomes a disaster waiting to happen.
Bob.
RalphM wrote on 3/11/2006, 7:18 PM
When necessary, I'll run my VX2000's in LP. Have even done it with 80 minute tapes. No problems, but I do it only when necessary.

I've received several tapes recorded in LP on other cams, even other brands. Once got 6 hours of tape recorded LP on a JVC camcorder. No problem playing it back on the Sonys.

Of course, the expensive solution is to record to a FireStor or one of the other HDD solutions, or to a laptop if it is a stationary shoot.

RalphM

Steve Mann wrote on 3/12/2006, 11:32 PM
Farss - you said two things that I have to question...

"DV has just too many dropouts"
Compared to what, and how do you measure this? Every digital format - EVERY digital media has dropouts. It's up to the error-correction built into the protocol to recover from the dropouts. Some modes do a better job than others, but it is exceedingly rare for anyone to actually see dropout errors in DV.

"There is some CRC data in DV but not much."
DV has an extremely robust error correction that goes way, way beyond simple CRC. Every frame is written to tape ten times. Each stripe is individually CRC'd and the ten stripes are further error coded in temporal. It's called a matrix error correction mode. As I said, uncorrected data dropouts (not to be confused with dropped frames - an entirely different problem) are exceedingly rare.

Please don't refer to marketing hype as your source of the "DV has more dropouts" myth. It may have more dropouts than, say Beta Digital or full-size DV. But, if all of the data dropouts (remember, all digital recording media experiences data dropouts) are corrected with a huge margin, then the argument means nothing.

Steve
farss wrote on 3/13/2006, 12:42 AM
OK,
I should have been specific, DV has too many visible (and audible) dropouts compared to DVCAM.
Another caveat, I almost exclusively work with tapes from others, oftenly shot by whoever could be found on the day, under adverse weather conditions of high heat or moisture, and on gear that could use more maintenance.
DVCAM and D8 seems to fare better than DV but as I said that's only my own experience. I'm pretty good at finding those little artifacts in the audio and the twinkles in the vision. It's more of a problem also with the stuff I work on when 99% of what's on the tape is used. Of course no one has ever complained so maybe I'm being overly sensitive to the issue.

What I'm really curious about is these 'dropouts' rarely showup in the analogue outputs from the deck, just in the captured files.

Given what you've said about the level of error correction in DV really makes me wonder what's going on, is there less error correction on the DV feed from the VCRs, is it some time constraint or what.
The fact that I've been able to get clean vision and/or audio by capturing from the analogue outputs of a VCR when the 1394 capture had serious problems is a real mystery.

Bob.
riredale wrote on 3/13/2006, 10:02 AM
n19093:

Don't mean to be nitpicky, but I don't think DV's error correction technology extends to the writing of a frame 10 times to tape. As described here, an individual frame is comprised of 10 tracks or stripes on a DV tape for NTSC, and 12 stripes for PAL.
corug7 wrote on 3/13/2006, 11:02 AM
Farss,

It makes perfect sense that the DV output would show more dropout than the analog outputs. You are basically transferring 1s and 0s. It stands to reason that there would need to be an analog conversion of the signal in order for the playback device to detect a need for correction.

I could be way off base on this, but I'm just considering how error detection works on SP, 1", and other analog formats. Also, most prosumer recording devices won't allow audio amplification through the DV inputs. Seems like that would be for the same reason.
farss wrote on 3/13/2006, 12:47 PM
Huh?
When I read data from a HDD error correction is done in the HDD's logic, same goes for CD ROM and CD R. Logically one would assume the same amount of error correction is applied to the 1s and 0s being transferred from tape via 1394 as is applied to the 1s and 0s before they go into the D->A converters in the VCRs.

Most prosummer DV decks do not have audio input level controls, true, most have some form of limiters / AGC.

SP isn't a digital format, error correction on playback includes time base correction and dropout compensation. The latter may or may not be done in the VCR, if not then you need access to the RF from the heads to detect that signal has been lost. This isn't lossless error correction as far as I know, the process involves patching video from another line into the missing part, more like error compensation that correction, well then again its is called "dropout compensation", keep having to do that over generations and eventually you just get garbage, same doesn't apply to error correction in the digital domain, although the audio CD spec uses a system to smooth out errors if error correction fails.

Bob.
corug7 wrote on 3/13/2006, 8:43 PM
Bob,

I did a little research to try to defend my statements, and (gulp) it looks like I might not be able to. Honestly, I'm aware of most of what you wrote, I just figured there must be some kind of dropout compensaton applied during the D/A process. I guess the other option is that you are losing about 50 lines of resolution (Here in NTSC land anyway) in the D/A conversion to S-Video and the "softer" image is missing a portion of that dropout when it is re-encoded to DV.

I've seen the same issue using IEEE 1394 out to a standalone DVD recorder. The image from 1394 had better color saturation and was much sharper, but I was able to see the "twinkles" you speak of occasionally. S-video provided a softer picture, and almost all signs of dropout were absent in the final MPEG-2 recording.