SPX 800 vs. SDX 900

[Jan Crittenden livingston]

Most of the guys that use the DVCPRO50 know that if they are going to do a lot of color correction they tend to move to an uncompressed SDI transfer into a 10 bit system. This makes it very stable for color correction. The EBU and SMPTE did an in-depth study some years back where in analysis, the Digi-Beta and DVCPRO50 were very similar generations out.

Hope that helps,

Jan

Jan Crittenden Livingston
Product Manager, DVCPRO, DVCPRO50, AG-DVX100
Panasonic Broadcast & TV Systems

[Graeme Nattress]

Jan, you’re absolutely right, but now that we can get 12bit captures on our digital cameras, and we can do quite severe colour correction, exposure compensation etc. without any kind of visual loss, we’re still waiting for the video industry to move beyond antiquated 8bit. I just got a Canon 20D digital SLR, and it’s quite shocking how are ahead the stills camera industry is compared to video. I know that capturing moving images is a whole lot different to capturing stills, but it really makes you think. We appreciate that Panasonic is getting rid of tape for us, and that move alone will improve the flexibility of video no-end, but I guess what we’re saying is that’s just the first step, and that progress must continue on to larger chips, lower noise and better dynamic range.

Just my thoughts…

Graeme

– http://www.nattress.com – Film Effects and Standards Conversion for FCP

[Jan Crittenden livingston]

[Graeme Nattress] “I just got a Canon 20D digital SLR, and it’s quite shocking how are ahead the stills camera industry is compared to video.”

Mostly I think the problem is the amount of data to be recorded is huge with video. If you think that the record times are short on a 8GB card with DVCPRO50, imagine exponentially changing the amount of data. Going from 256bit at 8 bit to 1024 at 10, or at 12 which is 4096, it changes everything. 8 Bit is/has been a compromise that has been working. As memory capacity continues to increase at costs that are acceptable, it may well be easier to do this in the future. But that is a future idea.

Best,

Jan

Jan Crittenden Livingston
Product Manager, DVCPRO, DVCPRO50, AG-DVX100
Panasonic Broadcast & TV Systems

[Graeme Nattress]

Ah, but compression technology, even in the intra-frame kind that Panasonic rightly prefers, has advanced more than enough to compensate for the small increase in data needed to go from 8bit to 12bit. Although, even getting NLE manufacturers to do 10bit properly is hard work, so it’s a bit chicken & egg. But as the HVX200 will be the popular camera to get people used to using memory, not tape, a future popular camera could be the catalyst to get decent >8bit performance on NLEs.

Graeme

– http://www.nattress.com – Film Effects and Standards Conversion for FCP

[Toke]

Jan, you are right that cc’ing with 10b helps to reduce 8b color banding, but in the future it’s not enough.
EBU/SMPTE study was done years ago and their target (if I recall correctly) was only (crt-)sd-tv.
Now technology and industry is changing; more and more same people and same equipment are being used to produce sd, hd and (digital) cinema.
I’m not sure about economics in US, but in smaller countries there’s a huge increase in air-time, so the cost of content needs to be lower and lower as the money flow from end-users is not increasing significantly. Production equipment is also getting cheaper all the time, which increases competition, because you don’t need so much capital to start a business.
And like I said, end-users displays’ quality is getting better all the time and so is demand for picture quality. With hdtv revolution people are really starting to demand a real movie theather quality at their homes.
With new distribution tech (eg. AVC/h.264 and increased color depth) it is possible to give that PQ to homes.

So what does this all mean as a content producers need from camera manufacturer?
We need to have cameras that use modern tech in capturing (p2, cf, hdd, pd, etc.) which is cheap, fast, and easy compared to old tape/film tech. This is something that Panny is in leading edge.
Qualitywise we need cameras that are able to produce quality which can carry as far as digital cinema distribution. This means among other specs more color depth. Getting rid of awkward tape standards helps in this.
And moneywise we need cameras that are affordable to use in all range of productions. Any serious photography hobbyist can afford to buy Canon 5d (at least in couple of years) and have same quality level than best professionals in the world. So the key to make affordable high quality moving picture camera is to make it a mass product. Only with that you can raise the quality/price ratio. How long does it take, that we can buy 5d equivalent in moving picture cameras?

About technical details, Graeme might be right in that we need bigger chips to produce >8b colors, but already all manufacturers are adverising how much more bits they have in internal dsp’s.
So it’s more like getting those bits out of that camera. This also needs a little change in manufacturers ENG-ideology, that they let only “ready-made” picture (gamma, color, resolution etc. corrected) out of their cameras.
Bitrate increase is not a problem. If you now have 720p24@8b dvcprohd for 40Mbps, getting 720p24@10b would only be 50Mbps. (And I bet that there will be very little difference in real resolution with 1080p with hvx200.)

I’m not asking for Dalsa or Genesis for thousand bucks and I’m happy to accept a lot of compromises.
But the thing that I’m not so willingly accepting is that there’s no development for a long time at all, like in video color depths.

So a good compromise might be 1/2″ or 2/3″ 1-chipper with 2048×1080 resolution and a little bit more low quality & resolution area around that chip so that you could view more than “active” frame like in film cameras. You could get raw 12b bayer picture out or switch to gamma, color and resolution corrected 720@10b (or even 1080p) picture for hasty situations.
Pricetag this for $10k and most of professionals would be interested and for $5k ten times more of prosumers would buy it right away.

More color depth would also mean easier (= faster = cheaper) lighting on the set. Every cinematographer knows how hard it is to make high quality pictures with 8b colors.

And again, I’m not trying to bug anything or -body or saying that Panny sucks, I’m just offering my opinions and visions for discussion and even future designing for free of charge…