HV20 vs HG10 - Video quality?

[Serge Victorovich]

Chris, at bitrates above 10Mbps difference between mpeg2 and h264 ~30%
above 15Mbps ~20%...and so.
In high bitrates above 30Mbps differences very small.
20Mbps of h264 may be equal in visual quality to 25Mbps of mpeg2.
If h264 codec so good at 15Mbps why we can not see realisation this into expensive camcorders above $2-3 grands?

AVCHD with max 15Mbps is for low cost camcorders.
HDV 25Mbps is better for another category.
DV100, AVC-intra or Cineform codec is better an what we need :)

[Thomas Smet]

AVCHD is about 2 or 3 times better at compress ger given bitrate when compared to mpeg2. With that said if you must look at it in terms of numbers that would mean 15mbits/s AVCHD could equal mpeg2 HDV at around 30 mbit/s to 45 mbits/s. The thing with interframe compression however is that you can't just look at numbers to figure out how it will look. There are so many factors that go into how the image is compressed that it is very hard to make a prediction.

One area that makes AVCHD better then mpeg2 is the block sizes it uses to process. Mpeg2 is locked into 8x8 pixel blocks where AVCHD can adjust the size of the blocks based on how complex the scene is. This does help reduce the amount of blocking and for the most part results in a more natural looking compressed image.

With all that said however a lot of people usually make the mistake of assuming all encoders are equal which is not ture. With software mpeg2 encoders people know that some are much better then others in terms of speed and quality. The hardware encoder chips in camera are no different. Hardware encoder chips have to be able to encode and decode the video in realtime. Some are better then others and it all comes down to how expensive and complex the encoder chip is in the camera. Decent realtime mpeg2 chips have been around for a little while now so they are much easier and cheaper to build. AVCHD is a little tough right now because it is so complex to encode that the hardware chips going into consumer cameras I would say are not very good. I wouldn't expect the same level of quality from a all day long AVCHD software encode to match what the camera could do in realtime. It will of course get better with time as better encoders come out.

So as Chris said the only way to know for sure how it will look is to compare the video from the two cameras. Running numbers is in no way at all going to give you the answer.

[Chris Hurd]

Originally Posted by Thomas Smet

Running numbers is in no way at all going to give you the answer.

Since you brought that up Thomas...

I'm compelled to point out for the understanding of everybody who visits this site, that I'm strongly opposed to the practice of "just running the numbers" as a way of finding answers, or worse, influencing others to make a purchase decision. To that extent it's important for folks to know that generally speaking, I tend to strongly resist the types of posters who insist on doing that to the degree that I'll sometimes show them to the door. If somebody wants to run the numbers, fine; just don't do it here on this site please -- let me clarify that I mean "running the numbers" while ignoring all other factors. It's usually okay to explore the numbers here as long as it's only a part of the bigger picture -- and the bigger picture, mainly, is how the image actually looks. And until a particular camcorder is actually in your hands, there's no way to properly evaluate it by examining just the numbers alone. I find that to be a highly misleading practice and unfortunately it's spelled the end of a few posting careers here before.

The primary thrust of what we're trying to accomplish here at DV Info Net is that while it's always a good idea to understand the numbers, the main thing to consider when evaluating a format or a particular camera is to actually look at the image, first and foremost. Then evaluate the ergonomics, workflow and feature sets. Numbers and tech specs, etc. fall into a distant last place behind these other far more important considerations.

Don't just run the numbers -- look at the image. Enough said -- now back to our regular programming,

[Ian Holb]

Originally Posted by Adam Perry

also, take into account that if you wanted to back up all that HDD footage, you would need a big external drive, or tons of DVDs. I'd prefer just archiving them on DV tapes that I don't tape over. that's a factor to me. I'll wait for the HV40 or 50, haha.

I'm going to run the numbers on this:

500GB hard drive = $100,
1 MiniDV tape = $4

There is 13GB in one MiniDV tape, therefore (500GB / 13GB = 38.5) a 500GB HD holds roughly 38.5 tapes.

39 MiniDV tapes at $4 each cost $156.

Conclusion: It costs less to store footage on hard drives than on tape. In addition, footage is instantly accessible and ready to edit. How long and how much would it cost in time spent searching and capturing footage off tapes?

For archiving and cost-effectiveness, hard drives are the way to go. Backing up footage is a simple process preserving bit and timecode data, whereas duplicating tapes re-writes timecode, unless you own a pro deck that can preserve timecode.

[Chris Hurd]

Originally Posted by Ian Holb

1 MiniDV tape = $4

Well... I don't know that I would use a $4 tape to record HDV. I'd use a tape specifically formulated for HDV, which costs quite a bit more than $4. But you've certainly made your point.

[Serge Victorovich]

Cost 1miniDV for me $2. In vacation tape is better (for me).
I choose flash or ssd and tape for backup any data similar to DLT.
A few year ago when HDD cost more than tape ($ per GB ) a smart guys did backup of hdtv(mpeg2ts) recordings from hard disk to miniDV or Digital8 tapes.

[Chris Barcellos]

Ian:

So in your cost analysis, you haven't accounted for cost of the tape to capture in the first place. Are you saying you would use tape over and over again. I think that risks loss and drop out more than using a cheap brand of tape for one single run.

I agree that access is so superior on hard drives but, long term storage of hard drives becomes another issue that has been addressed before. But is it as dependable as raw tape ? Tape, to me, still seems to be the most stable for archiving the raw footage.

[Chris Ames]

Originally Posted by Chris Barcellos

Ian:
But is it as dependable as raw tape ? Tape, to me, still seems to be the most stable for archiving the raw footage.

True, tape is more dependable. However it will expire eventually. The ideal solution for me is what I have now. I always record on new miniDV tapes. And I always move the good stuff to my external RAID 1 (mirror).

I just use my HV20 to capture home movies and memories. So I don't have a significant amount of volume and the footage I do have you can't put a price on.

Chris
http://dewde.com/videos

[Luc Fender]

Interesting, didn't think about. Eventually you have to store it somewhere long-term. DVD is not dependable long-term. So tape is certainly convenient. There are some other alternatives like external tape backups but certainly more work and hardware+interfaces keeps changing.

Also in max quality you only have 5 hours recording which I can imagine could not be enough sometimes so it's a hassle then to dump it to someplace to free it up.

Re the bit numbers I agree H264 has more advanced compression techniques. But it does appear that most 12Mbps AVCHD though don't quite approach the quality in fast moving recordings. But I think 15Mbps shoud be enough to fix that (+a good encoder). So we'll have to wait for the reviews.

Assuming for now 15Mpbs makes it very comparable I still think the biggest difference for me is the audio quality. I have hopes it wouldn't emit the bad tape motor noise (and not too fond of bigger external mic solutions).
Hard-disk can be noisy so hopefully they picked a quiet one.

I also notice they advertise again the superrange steadyshot. One thing I wished would be better is the amount of correction (since most mini-dv camcorder correct much more). So I hope Canon adds one day an additional option for more aggressive correction (e.g. walking where I don't care if I get slight resolution loss).

[Ian G. Thompson]

Originally Posted by Chris Ames

True, tape is more dependable. However it will expire eventually. The ideal solution for me is what I have now. I always record on new miniDV tapes. And I always move the good stuff to my external RAID 1 (mirror).

I just use my HV20 to capture home movies and memories. So I don't have a significant amount of volume and the footage I do have you can't put a price on.

Chris
http://dewde.com/videos

LOL...I just watched "The Prank Before Christmas".....that's wrong man...LOL....the question is.......R U still employed??? That was pretty good.

[Chris Ames]

Originally Posted by Ian G. Thompson

LOL...I just watched "The Prank Before Christmas".....that's wrong man...LOL....the question is.......R U still employed??? That was pretty good.

Ian:

Thx man. Yeah we are all still employed. But our boss (the victim) quit 3 months later!

Chris
http://dewde.com/videos

[Ian Sturtivant]

It always amazes me that the guys with big multi-terabyte RAID arrays never factor in the cost of keeping the things powered up. I can store a DV tape for years at a cost of $0. But your NAS server will use power 24/7, if it's to be left accessible to you.

Don't get me wrong: I am no Luddite. I had an Secure Digital camcorder last year before I got really interested in this hobby, now I have an HDV cam, and last year's mpeg-2 files are stored, and archived, on every single PC I own, and every back-up drive also. Every HDD is born to die, and I'm less paranoid about my DV archive. ;)

[Luc Fender]

By the way the [reference removed] review gives completely mixed messages. On one hand they said they were impressed with the compression, and therefore motion artifacts can be hardly seen. On the other hand they say for stationary objects you see resolution loss (their theory: compression issues).
Then again their resolution tests showed to the eye slightly higher resolution for HG10!? Looking at the 2 picture crops it's clearly not just visible so doesn't make sense.

Stationary scenes should be able to take much more compresssion to look good. Unless the VBR drops to the lowest point and then looses detail.
Looking at clips at watch-impress I don't see much of a resolution loss if at all compared to HV20. Also his clip riding with his bike down the street shows that it does a pretty good job.

[Ken Hodson]

Originally Posted by Thomas Smet

With all that said however a lot of people usually make the mistake of assuming all encoders are equal which is not ture. With software mpeg2 encoders people know that some are much better then others in terms of speed and quality. The hardware encoder chips in camera are no different. Hardware encoder chips have to be able to encode and decode the video in realtime. Some are better then others and it all comes down to how expensive and complex the encoder chip is in the camera.

Very well said.
This area of modern cam comparison is where so many people go wrong. My favorite example is in the case of the new JVC HD200/250 HDV cams. They capture to the exact same 19.7 Mbs HDV codec as the previous HD100/110 models yet at twice the frame rate (30p vs 60p) they offer even better video quality then the previous cams capturing at only 30p. That is the power of the encoder.
As has been pointed out there are simply too many variables that add up to image quality to simply do a simple bit-rate comparison.

[Chris Hurd]

Originally Posted by Ken Hodson

there are simply too many variables that add up to image quality to simply do a simple bit-rate comparison.

As I keep on telling people, it's not the about the bit-rate. It's about the efficiency of the encoder.