Canopus chroma subsampling during capture/editing?
 

Guys,
according to a Canopus ad (page 3 in the attached file) its DVStorm hardware DV codec ‘encodes analog signal into DV format and transfers data to PC’. This is OK, understood. Chroma subsampling is not mentioned.
Then Canopus claims that its software decoder ‘changes DV video to YUV 4:2:2’ at the editing stage. Wikipedia says that terminology wise digital transformations should be 4:2:2 Y'CbCr not analog YUV. The picture then shows that after editing the data are encoded into compressed DV format. Is it possible that data is edited in analog format between these two DV encodings taking into account they use analog YUV term? Or just the explanation may be a little bit confusing for rookies like myself and they simply compress edited DV further?
On this forum you guys say that capture should preferably be ‘lossless AVI 4:2:2’ while Canopus provides for 4:2:0 in PAL which is also fine. Should it be understood that DVStorm captures in 4:2:0 and then decodes to 4:2:2 ? Is it good or bad in itself?

Canopus was a poster child of BS in marketing. They often skated the line of truth, and often fell and scraped a knee in the land of malarkey.

Canopus ceased to exist by 2005, acquired by Grass Valley. GV continued to sell some Canopus items, namely the terrible ADVC-110 boxes, but it was likely nothing more than unsold NOS (new old stock). Canopus was essentially liquidated for the underlying tech, though some ghosts of Canopus still exist (naming conventions of products, like ADVC).

Back in the mid/late 90s, you mostly had two choice for NLE (non-linear) cards: Matrox or Canopus. Remember, back then, most editing was still analog and LE (linear). Canopus was first to adopt DV for non-shooting, which is not why DV existed. It was never intended as a conversion format, only shooting. I remember seeing the DVstorm and DVstorm2 in B&H ads back in 2000, and I drooled over them. The cards offloaded video tasks (in a world before GPUs existed) to the NLE cards, and worked in tandem with Premiere 5 (Matrox) or Edius (Canopus). Because I was a Premiere user, I wanted the Matrox RT2000, then RT2500, then RT.X100. But I never bought those, and instead settled on the game-changing ATI AIW -- cheaper ($300 vs $1k), better quality AVI (non-DV, but lossless), hardware-software Hybrid MPEG encoding, and it worked outside of NLEs. Even after the AIW, I still had some infatuation with Canopus and Matrox. Mostly Matrox because had some hardware MPEG going on, but it was pricey at $1k+ (three AIWs, with change left over!).

The idea that Canopus could magically turn ingested 4:1:1 DV into 4:2:2 was outright nonsense, more damned lies from Canopus.

An irony with Canopus is that the pro products used a much better DV encoder chipset than the consumer ADVC units. While still 4:1:1 and lossy, it doesn't rape the signal as badly as those dumb ADVC DV boxes.

Most of that gear is OS locked to Win2000/XP, NLE-locked to certain ancient (now worthless/useless) versions of Premiere or Edius. I may have drooled over the Canopus/Matrox ads 20 years ago, but I wouldn't give those cards the time of day now. (The same applies to some women in my past. :laugh:)

If I were to stray into non-AIW legacy video tech, I'd be far more tempted to look at Osprey, and not anything from Canopus or Matrox.

Thanks, Lordsmurf. In all honesty I've been much tempted to buy a used DV Storm 2 card for a good price from a good source (a studio) which could provide me with all the necessary software, installation and usage guidance / advice which is of paramount importance for a novice (frankly speaking I'm still a little bit scared by the complexity and intricacies of the right digitising process). However spending days going thru' this and other forums, common sense talk me out of this purchase.
I still wonder whether the upscale 'change' of DV video from YUV 4:2:0 to YUV 4:2:2' is possible from the physical (electronic) point of view and if so how much loss it entails and whether it is worth the effort they exerted.

Converting 4:2:2 to 4:2:0 is okay but the opposite is usually disastrous. Now with new playback devices people who capture analog video don't even have change the chroma sub to 4:2:0 during encoding, it is left intact at 4:2:2 in the h.264 codec and the size difference is only like 5% increase, most devices I've tried play it just fine (smart TV's, iPhone, Samsung tablets, Media players ...).