Then you haven't used an HV20 with a 35mm adapter!

I predict that more and more consumer camera sensors are going to be hampered by onboard consumer-grade "transport" (codec/recording medium) than by the actual capabilities of the capture electronics.

The greatest appeal of this product is to free low budget camcorders from having their signals mangled by HDV.

If the "Cinfeform SOLID" did come in @ $2k, then my kit would still cost HALF of Sony's cheapest "FULL" HD camera, the EX1.

Mangled by HDV? That's going a bit too far :)

Putting aside the wonderful advantages of the "Cineform Solid" (i.e. 4:2:2, 1920x1080, instant pulldown removal into Cineform's great wavelet codec, bypassing HDV mp2 audio compression), HDV is still really really good IMO.

Some of us do not have the money to, ALL AT ONCE, upgrade to this level of performance. I'm really interested in the idea of modular upgrades, that will allow users to incrementally, over time, improve the quality of their video production, without having to throw away their initial investment and start over again. I might not have $8000 to spend today, but I might have $2000-3000 a year for the next 3 or 4 years to spend on incremental upgrades.

So, for example, in Step 1, you buy a decent HDV camera with an interchangeable lens. (The Sony Z7, for example?)

In Step 2, you buy a new lens, or start experimenting with 35mm still-camera optics.

In Step 3, you upgrade the recording system, by adding something like this Cineform Unit.

Even if you end up upgrading the camera itself, you should be able to re-use lenses or external recorders you purchased in Steps 2 & 3.

This is the way it always used to work with 35mm SLRs, and to some extent, still does with digital SLRs.

I don't see how, objectively, HDV can be considered "really really good." Now, of course, I use it. I have to. It is manageable, acceptable, and certainly better than nothing, but honestly it's not "really really good."

1. It was antiquated before it even became available on the first camera. Many more modern, efficient codecs existed back then, and now the disparity is even greater.

2. It throws away the bulk of the color information.

3. It filters out a lot of low contrast detail.

4. It's not hard at all to get it to show visible macroblocking and other artifacts.

5. It doesn't hold up to any sort generational work.

So, yes, I stand by the loaded term of "mangled." I monitor my HV20 shoots via live HDMI, and it's always a little sad to see what is left over once recorded via HDV.

Maybe Craig referred to a 35Mps mpeg compression?

That isn't in the HDV spec is it?

35Mbps is an XDCAM data rate.

XDCAM holds up far better than HDV under every circumstance I've tested. It also has uncompressed audio.

Still XDCAM is in many ways HDV's big brother, so maybe that's what Craig meant?

Still I wouldn't call XDCAM "really really good" either.

I'd say that XDCAM was usable, the same rating I give DVCPRO HD and DV.

I call HDV, "not entirely unusable."

Oh well, this whole post is gonna be off topic. If you want to talk about it more than this, then lets make a new thread somewhere and copy some posts over there.

I sure haven't- but I have to say that I am unimpressed with the results of 35mm adaptors I have used with other cameras.

The main thing they add is a nice shallow depth of field.

You still see the limits of the low end sensors.

Well, you forget your own earlier comment. I think a lot of cameras are hampered by low end glass as well as being mangled by the onboard codec.

Still, sensors will continue to improve very rapidly. So will electronics and storage. As a result I expect we'll see better and better recorder sections and sensors.

Glass however is glass. Optical physics and engineering are very well understood and relatively mature sciences. It will be hard to make quality lenses for a long time. It will take a revolution in manufacturing techniques, like the arrival of a nanotech general assembler- but I stray far off topic.

Let me just say that sensors will improve fastest, the recorder sections will follow fairly rapidly, but glass will lag far behind.

So... we agree generally.

Yes. You get what you pay for though.

The HV20 will still exhibit a relatively narrow exposure latitude and you'll still be looking at 8-bit color going out the HDMI port. Oh, and you'll still have 1440x1080 HDV resolution sensors. And the same gain characteristics.

Even a tool as cool as Cineform SOLID can't fix that for you.

You still have a great bit of kit though, and you can certainly upgrade the camera and get a better sensor without worrying about either the glass or the recorder. Something like the Iconix might start looking very attractive- though I think this little bugger is rather expensive:

http://www.iconixvideo.com/products.html

I have never understood why Canon and Nikon haven't made video sensors with a full frame size to attach to their 35mm lenses.

In any case Cineform is making it easier for someone else to come along and do just that.

Eventually I expect to just plunk down my cash on a RED and actually have a real 35mm sized sensor with prime 35mm PL mount glass. (Cooke S4... droooool)

Useable...Unuseable....it all depends..... Its funny how people like to gripe about HDV's artifacts and try to measure it up to other "more professional" platforms but the fact is it has been used on some professional shoots in recent years (the tv show JAG comes to mind) successfully. I would say that is a perfect example of its useability.

From what I have been reading...HDV had a bad reputation from the start with its early blockiness issues and other artifacting problems....but to be honest its current implimentation (especially in a cam like the HV20) has come a long way and is much better than when JVC first came out with it.

I think it is very useful...even more today that it ever was.

You may get your wish sooner than you think from the Nikon-series "Live view" option over HDMI.

At this point it doesn't look like the "Live-view" is real-time on the D3, but at some point in the near future I'm sure it will be. So at that time you'll basically have a down-converted 1080i signal over HDMI that you could feed into the SOLID from your full 35mm-sized Nikon sensor for very low cost. It probably won't be as optimal in use as a dedicated cinema camera, but it would be an option.

Sorry to hijack this thread, but actually the HV20 has a full 1920x1080 sensor. That's why so many people have adopted lugging a PC around in order to use the Blackmagic Intensity to capture this full-rez 4:2:2 signal, and which is why the Cineform "SOLID" is such a perfect mate.

Yes, there are many things about this cheap little camera that don't live up to professional standards, but for sub-$1k it's a unique beast. For the truly indie filmmaker, it is a godsend. All the money saved can be spent on lighting and everything else that shows on screen.

We all want a RED in the price range of a good DSLR, and the industry will get there, but the big Japanese hardware manufacturers will do it kicking and screaming, so in the meantime we're left with trying to hack their last generation technology (i.e. MPEG-2, tape, etc.) Luckily there are people like Dennis out there, with product ideas like the "SOLID" to do just that.

Who is Dennis?

Sorry, I meant David! Too much tritophan is muddling my brain. ;-)

I disagree that on camera recorder sections will be improving as fast as you suggest. The reason we have MPEG2 and AVC in cameras is a lack of imagination on the part of the camera designers, using distrubtion formats to increase record length, resulting in quality issues is why we are have this thread. There hasn't be a truely development new camera format (putting asside our own CineForm RAW) since DV, and it varients DVCPRO-50 and DVCPRO-HD. The issue has been post support, the catch-22 of having a new format in camera before the NLE is s ready. In CineForm's case we developed a post production format, so it is now ready to move into cameras, with wide NLE support already in place.

Just a couple of thoughts after reading all the above

1) M2T

Everybody seems to hate it, but honestly I think it isn't so bad for its originally intended purpose - which is to compress the video to the point where it could fit on a miniDV tape and still look rather good when played back. I think it has succeeded quite well. The key point is the part about "LOOKING good when played back". In other words it uses a perceptual model of human vision to discard information to which the human eye is less sensitive.

This objective is absolutely not the same as the objective of pulling great keys or holding up well to mutligenrational editing, etc. That's what we have Cineform for. Of course I'd be happier if there were a direct to Cineform option in camera, but that's because I want to edit the footage, not because I think it looks all that horrible if I just want to play it back. Not perfect, but not all that bad either, IMHO

2) Resolution of the HV20

Nothing against the HV20. I'll probably even get one one of these days. But I think there's a point being missed whan discussing its resolution. A three-chip camera has a sensor for each of the primary colors. The only way I know to get a three color image from a single chip camera is to dedicate some of the pixels to each color, thus cutting the effective resolution. There are computational games one can play to minimize the loss in resolution, but there is no free lunch. I think a three-chip "pro" camera will have an advantage here.

Eerie!

I just posted on the subject here:

http://www.dvinfo.net/conf/showthread.php?p=780806

Opened a new thread since didn't want to deviate this thread from its intended purpose.