DV vs Digital 8

8 replies [Last post]
Wilf Parsons
Joined: Apr 20 1999

does anyone have any idea of how digital 8 compares to DV for picture quality etc? What about transfering DV onto digital 8 - does this involve an extra layer of compresssion and deos that mean a noticeable loss in picture quality? Also will firewire cards like DV300 work with the tiemcode from a digital 8 player?

Joined: Apr 12 1999

There's no theoretical reason why Digital-8 quality should be any different to DV; however, current Digital-8 cameras have Hi8 front-ends, so the resolution is currently limited to about 400 lines, compared to 500 or so for the best DV cameras.

Transferring DV to Digital-8 shouldn't be a problem if you can get a Digital-8 camera with Firewire in; the compression is identical, it's just the storage medium which is different.

Don't know about timecode, but I'd guess that would be the same as DV too.

One big issue with Digital-8 is whether Sony will support the format for long; DV is likely to continue for years, and you'll probably be able to get Hi8 tapes for years, but there's some doubt as to whether Digital-8 will be popular enough that you'll still be able to buy cameras in a few years.
Of course in that case you could transfer everything over to DV, but that's a pain.

P4-3.06/2GB RAM/2500GB IDE/SATA. Avid Media Composer, Liquid Edition, Premiere 6, Lightwave, Vue 6, eyeon Fusion 5. DV and HDV editing/compositing.

Joined: May 5 1999

Digital 8 claims 500 lines resolution but have read that tests showed up to 480; that's still quite a bit better than 400 lines on Hi8.

I wouldn't worry too much about Sony stopping production of Digital 8 camcorders - it'll be a while more. It'll be a problem if the D8 machine cannot be repaired and if you can't buy another D8 cam to play back your collection of D8s.

Seems like an ideal upgrade for Hi8 users like myself with a large collection of Hi8, but stopping short of mini DV upgrade.

Joined: Apr 12 1999

The tests I've seen showed more like 400 lines; 380, AFAIR.

P4-3.06/2GB RAM/2500GB IDE/SATA. Avid Media Composer, Liquid Edition, Premiere 6, Lightwave, Vue 6, eyeon Fusion 5. DV and HDV editing/compositing.

Joined: May 5 1999

Video Camera (April 1999) has a good coverage on Digital 8. Their test of the DCR-TRV310E. To quote the relevant paragraph (page 41):

"The similarity in picture resolution, texture and character between D8 footage shot on the TRV-310 and mid-range DV are so small as to be almost insignificant.

Our sample managed an excellent 480 lines without any trouble, colours are vibrant and lifelike, accurately aligned and with good contrast range."

Unless they got their measurement wrong... :)

Joined: Apr 12 1999

Hmm, there's a comparison of test patterns between a TRV900 and D8 camera at:


The D8 image does not look very good at all... though it is much cheaper than the TRV900. Presumably Sony are starting to fit better camera elements now, or their PAL cameras were always better than NTSC.

As for getting their measurements wrong, in my experience all these measurements vary widely; I've seen figures for the TRV900 ranging from about 460 lines to 520+. They can't all be right, unless there's a wide variation between cameras.

P4-3.06/2GB RAM/2500GB IDE/SATA. Avid Media Composer, Liquid Edition, Premiere 6, Lightwave, Vue 6, eyeon Fusion 5. DV and HDV editing/compositing.

Joined: May 5 1999

Had a look at the page - interesting except that I wouldn't expect the TR7000 to compare with the TRV900. TR7000 is the lowest entry model D8 without even an LCD panel while the TRV900 is one of Sony's highest end 3CCD consumer mini DV. It's like comparing a Ford Escort with a Ferrari!

I don't know whether there's problem with test results in general - a lack of scientific test controls? Or people using tests without knowing what they're really doing. I certainly hope that the manufacturing process is reasonably consistent with strict quality controls nowadays.

I suppose IMHO, as long as you like what you're seeing, who cares about the tests! Actually, I have just bought a Sony DCR-TRV410E with DV In and Out. After using Hi8 for a while, I'm very happy with it. Never jumped into the mini DV bandwagon caused I can't use my Hi8 collection.

Now, to decide whether I should get the DV300 or Canopus DVRaptor. Any suggestions?

[This message has been edited by kkchin (edited 08-05-99).]

Joined: May 21 1999

I read an article in your "Computer Video" magazine month?? says that D8
camcorders use (10:1) compression ratio vs  miniDV (5:1). 

Maybe this is when recording from D8 cam only !!

Can anyone explain this compar to John's TEST?

In an article wrote by John in the rec.video.desktop:
I don't know if this has been reported before, but just so you know... my
experiments indicate that (1) Digital 8 and MiniDV store exactly the same
bits, and (2) lossless copying between these formats over many generations
is possible.

I made a simple 30-second video clip in Premiere (incorporating two
high-resolution still frames, a photograph of cats, and a scanned-in
resolution test pattern) and wrote it va firewire from my PC to my TRV900
MiniDV camera. I read it back into the computer via firewire and extracted
the two stills in Premiere in BMP format, calling this "1st generation".

I then copied that 30-second segment on the tape from the TRV900 over
firewire to the TR7000, giving me a second generation copy on Digital 8,
and then back to the TRV900 for a 3rd generation on MiniDV, and back
etc. up to the 9th generation (on the TRV900 at this point) which I then
uploaded back to the PC again, and again extracted the two stills. The 9th
generation looks identical to 1st generation. I did a binary file
comparison on the BMP files and they are bytewise identical- every single
bit is the same. Looks like I had no tape dropouts along the way!  So, my
conclusion is that this works- the firewire transfer is actually lossless-
and you can use the Digital8 models as an inexpensive digital VCR for
editing purposes.

By the way there *is* a difference between the "original" images and the 1st
generation due to the lossy DV codec, just like JPEG is different from an
original BMP, but you no doubt knew that already :).


[This message has been edited by Samuel (edited 28 May 1999).]

Joined: Mar 7 1999

First, the piece is CV which you refer to - about D8 compression being 10:1 was wrong - but this was what Sony said, and we merely reporting it.

As for your tests, I've no explanation why you suffered no loss but it is normally the case that you can see a change after about five copies or earlier if you are using FireWire cables that don't have those big lumpy bits at each end, which cut down on interference.

There is not, as best as we can make out and proper error correction, whether you are capturing from camcorder to disk or from disk to camcorder. Instead of errors being corrected, they are masked by interpolation.

Bob C