Kind of like this?

[Copernicus]

Spiff,

I fully agree that you have excellent compression algorithms out there. All I'm saying is that it is all about compromises, I believe that it is preferrable in some situations to have 2K uncompressed RAW originals at your disposal than 4K 6:1 compressed material that you never will be able to fully decompress (especially if the 2K camera is 2-4 times cheaper...). It all depends on your preferred public and on what type of screens you want to present your material. Also, see it as a copyright thing that gives you full freedom to do what you want with your own originals...

/Cop

 

[Steve White]

So long as the full bandwidth of the writeable media is used, I'll be happy, since this is the bottleneck on portable systems.

If the system has a 2k sensor, and the media data rate is enough to do lossless RAW, then by all means, do it. In this situation, I would rather them use a higher resolution sensor, and lossy RAW compression.

 

[lordtangent]

2592 x 1944 CMOS on a cheap ass Aiptek. I guess 2592 x 1944 is pretty pedestrian now? I wonder how good that sensor is? Is it as clean as the Canon sensors, or the "Mysterium"?

Cineform is working on a low power hardware verison fo thier codec. Coupled with something like this 2592 x 1944 CMOS, a decent lens and basic OS and you have a camera. Red could even subcontract Aiptek to do the development and keep thier guys on the Red One.

I would really like to see is something like this 2592 x 1944 CMOS with LinLog. It would be great to finally lose the "clamping so hard they look like white streaks when motion blurred" spec hits. I think LinLog might be albe to do it.

 

[MikeHedge]

you guys have all seen this???

http://www.visionresearch.com/index.cfm?sector=htm/app&page=products&prodcatid=8

 

[MikeHedge]

new Casio video
http://www.youtube.com/watch?v=G2PEaCtOPtE