I do LOVE my G30. I've only had it for a month and my first pro-shoot with it is this Thursday. I had read somewhere else in these forums about disabling the smart resampling - good to know!!! Thanks.

So then, OldSmoke, do you shoot mostly in 60p or 60i? MP4 or AVCHD? I am still figuring out what to shoot and when... I have enough hard drive to handle the highest settings at all times - but just don't think it's necessary to shoot at 36Mbps 60p when capturing my cats fighting :) -- I generally shoot at 17Mbps 60i for home movies. Thoughts?

60p all the way simply because I have all the options with it. I can convert it to 30p and 60i even PAL formats look ok. I just shot in MP4 at the maximum setting. I don't see much difference between AVCHD and MP4 and I don't have any test equipment to check which one is better. I do however feel the MP4 is easier on to edit in Vegas.

My two issues with the camera is the build in and not controllable ND filters. I shot up in the mountains in Colorado with all the snow and opted for an external ND filter. That really helped to preserve the footage at that conditions. It helps to keep the aperture at F4 which is it the sweet spot I feel. The other is the zoom ring which doesn't respond as I am used to from my Z5U. It is not like a "direct" drive, servo response is too slow and not linear with your motion. I do a lot of sport filming and a good manual zoom ring is important to me; I will try out the FDR-AX100 when it comes out and then make a decision which one to keep.

For weddings the only advantage to shooting 60p is where you might want to do smoother slo mo, like maybe the bride walking down the aisle or first kiss. Shooting 60p to deliver normal speed 30p is silly, the delivered frames are identical, just uses twice your storage. At the same bitrates, there's no quality difference between G30/XA20 AVCHD and .mp4, the .mp4 is easier to work with but there's no difference on the timeline, they are both CABAC H264 video. (Yes, the lack of manual ND filters is annoying, but if you are happy with f4 that's where they try to keep it, I don't miss a zoom ring since you can set the zoom rocker to variable fast speed or do touch panel/joystick zooms and keep the ring for focus/aperture, one issue for me is the slow start up time, but overall a really great camera)

"60p all the way simply because I have all the options with it."

60p all the way here as well.

The problem with 60i or any type of interlaced video is that it has to be de-interlaced for online use and I have never been happy with the results of de-interlacing.

I do mostly Church videos where the lighting is not always great and at times quite poor and that can make interlaced look much worse. Theater video can be all over the place and that is also bad for interlaced.

As for weedings, (GRIN!) the poor lighting there is not good for interlaced either.

Anyway, I tried all kinds of formats and 60p gives me the most consistent best results.

As for mixing cameras, even with the same model camera I would never want to mix interlaced with progressive. I sold and will sell the rest of my interlaced cams and be 100% same model cams and 100% 60p.

I would think that for those cameras that have a 30p mode but encode the data file as 60i, that a deinterlace of that should be essentially perfect (blend mode).

Yes/No/Maybe?

Maybe. You're describing PsF mode. Theoretically its "perfect", should be identical to 30p, but a whole lot of other issues come into it.

People shooting 60p also should realize that at the G30/XA20 highest 60fps 35Mbps MP4 bitrate the 30p equivalent is around 17 Mbps, significantly below the 24Mbps best 30p rate. In simple terms, if you want top quality images or to do any post processing, you are better off shooting 30p at 24Mbps than 60p at 35 Mbps.

Nobody asked, so I guess I will . . .
What is the intended use for your video?
How will it be presented?

" In simple terms, if you want top quality images or to do any post processing, you are better off shooting 30p at 24Mbps than 60p at 35 Mbps.

I am not sure that is true. Encoding 60p doesn't take twice as much space as 30p. Where I agree is that in low light situations you get longer exposure, 1/30 vs. 1/60.

Did you actually test the 30p quality vs. a 60p converted to 30p?

Encoding 60p doesn't take twice as much space as 30p.

That is certainly mostly true for encodes from our PCs. "mostly" in that it depends on the actual content being recorded. Typically we get a lot more inter-frame compression at 60p in relation to 30p.

Cameras do not have the compute power or time to do much, if any, inter frame compression. Cameras must encode in real time.

Of course 60p doesn't take twice as much space as 24Mbps 30p, since the 60p is encoded at 30p 17Mbps equivalent. Shoot identical 30p content at 17bps, it will be closer to twice the space. Of course there's more inter-frame compression. Of course 17Mbps quality is lower than 24Mbps - not that you'll probably notice in ordinary viewing but you probably will in post , for example, if you're trying to color match footage from two cameras.
Shooting 60p or 30p should be a conscious decision in each case. If you just decide "60p all the way", you need to be aware of what you're giving up.

I have three Sony HDR-PJ710V cameras that give me the choices of 60i, 60p and 24pee. No 30p and/or mp4 on this model. I briefly had a Sony camera that did mp4 and its quality was plain awful!

60i is out because of the interlace, 24pee is out because it stinks, so 60p is the only real choice for me.

As for 24pee, what many do not understand is that it is NOT actually what film looks like. Film is shot at 24 Frames Per second, it goes through a projector at 24 FPS but because of the projector's shutter, it is actually viewed as if it were 48 FPS or even 72 FPS.

So if one really wants a true film look with video they should choose either 48 FPS or 72 FPS.

The only reason for 24 FPS with film is because of the high costs of film and it is the lowest frame rate that will also work with audio. Otherwise the industry would have gone with a lower FPS.

Just to clear up some confusion in this thread ...

First, the only thing which determines file size is bitrate. The frames per second (24, 25, 30, 60) and the interlacing have nothing whatsoever to do with file size.

Second, it is true that real film is projected in a way where the shutter opens and closes several times for each frame in order to increase the flicker rate so that people don't get headaches while viewing. Each time the shutter is closed, there is no light on the screen. Thus it is quite correct that video shot at 24 frames per second is going to look a little different than 24p video because when we view 24p video, there is no time whatsoever during which the screen is blank.

However ...

It is wrong to say that a "true film look" would be achieved by shooting at 48 fps or 72 fps. This is wrong because all those extra frames are taken at points in time between those frames that would result from shooting at only 24 fps. You would actually get a very smooth, "video feel" if you shot at 72 fps.

But ...

You have come across an idea that might be interesting, namely, shoot at 24 frames per second, but then store that in a 144 fps "wrapper," repeating each frame three times, and interspersing that with three frames of black. That actually might do a good job of simulating what film looks like on the screen, when projected with a three blade, equal angle shutter, like the one on the left in this picture:

"First, the only thing which determines file size is bitrate"

I seriously doubt that and you can easily check that by rendering the same clip out of Vegas with same bitrate but different frame rate; the 60p file will be bigger then the 30p file.

Film projectors have a shutter for the same reason film cameras must have one, pulldown. 35mm projectors from memory have only a two blade shutter and the angle is pretty big so for quite a period of time the screen is black.

Digital projection is quite different, no need for a shutter. That means way less powerful lamps are needed. A typical film projector in a cinema uses a 2KW or 5KW Xenon arc lamp with all the attendant problems of cooling and dealing with the ozone they create.

The main driver for higher frame rates for cinema is the larger angle of view, 24p is OK for art house sized screens but for 4K and 8K in the latest extreme screen cinemas the industry seems to heading for 120fps to 150fps. The BBC is also pushing those high frame rates as the standard for all 4K acquisition.

Bob.

I seriously doubt that and you can easily check that by rendering the same clip out of Vegas with same bitrate but different frame rate; the 60p file will be bigger then the 30p file.Yours is a common misconception, but I can guarantee, with 100% certainty, that what I stated is true: bitrate is the one and only thing which determines the size of the final render.

One reason why so many people get this wrong is that there are many posts in various forums about doing things to video and then measuring the size of the resulting file. For instance, over at doom9.org, many people will apply a noise filter and then measure the file size in order to see how much smaller the file becomes. You would think that if what I say is true, then they should see absolutely no difference in file size.

However ...

These people are using a "constant quality" encoder, something that Vegas doesn't have. This encoder measures the instantaneous results of the encode and then increases or decreases the overall bitrate in order to maintain some vaguely-defined measure of quality. The key thing to note is that these encoders are changing the bitrate, and that is why the file size is different.

The variable bitrate encoder in Vegas also changes the bitrate at various parts of the video, but if you do a 2-pass encode, Vegas ensures that the "average bitrate" exactly equals the number you set.

BTW, if you decide to do a test in Vegas to see if what I am saying is true, you have to use either the 2-pass VBR, or the CBR encoding. The single-pass VBR, at least for MPEG-2 is not accurate and will produce different results with different video, even if you keep the length of each video the same. This is the result of the algorithms Sony/MainConcept uses, and not the result of the theory being wrong.


Film projectors have a shutter for the same reason film cameras must have one, pulldown. True, but the reason projectors have a multi-bladed shutter, which opens and closes the shutter several times for each frame of film, is to increase the flicker rate in order to avoid giving the audience headaches. This is the key idea in my previous post.

[I]"True, but the reason projectors have a multi-bladed shutter, which opens and closes the shutter several times for each frame of film, is to increase the flicker rate in order to avoid giving the audience headaches. This is the key idea in my previous post."[/I]

Sure but the flicker is to some extent a product of the fact that the projection of the film has to be blanked whilst the film is pulled down through the gate. As a projector (well 35mm ones at least) use a fast pulldown system that's 90deg. So that we notice this 24Hz flicker less the image is also blanked out by a two (or more) blade shutter so the flicker frequency is 48Hz, which can still be seen if you looks for it. Also flicker can also be reduced by having the shutter disk spin faster giving a quicker light/dark/light transition but doing that costs money and only a few projectors did this. From my reading even 3 blade shutters are uncommon on 35mm projectors as that means there's no light for 270deg and that would really push up costs to theatre through needing even more powerful lamps.

Now on a LCD or plasma screen and for digital projection there's no pulldown and no need to block the light going to the screen so no flicker and as I said less powerful lamps are needed as well.

On the other hand there's also the issue of motion judder but that's a different issue.

Bob.

Well I agree that for CBR it should be the same it isn't for VBR, I tested that.

Anyway, the argument was a different one; 30fps at 24Mbps has better image quality then 39Mbps 60p which one might think because you would need 48Mbps to get the same quality in 60p. That is where I y that is not true because it doesn't take twice the bitrate to encode at 60p vs 30p.
The main difference is in the longer exposure time with 30p and how you can use it to your advantage.

"First, the only thing which determines file size is bitrate. The frames per second (24, 25, 30, 60) and the interlacing have nothing whatsoever to do with file size."

That is 100% correct.

"It is wrong to say that a "true film look" would be achieved by shooting at 48 fps or 72 fps."

Maybe I should have said effectively true film look. As in how one sees it not what it actually is. It is not 24 fps that one is actually seeing with film. Even though there is no shutter with video, watching a video at 24 fps with all that motion stutter can give me a headache.

Doing a Church video for instance, you have a person preaching and waving their hands all over the place and with 24p, you see a lot of stutter in their hand movements. So it does not matter how slow one pans or whatever, the stutter in the hand and other movements by the person speaking looks bad to me.

Finally, with the cameras that I have, 60p gives me the best results. With other cameras that may be different. One needs to do a lot of serious testing to find out what works best from start to finish with the cameras they have as well as everything else. This includes ones workflow.

What works best for me, may not work best for someone else.

I prayed and asked for guidance and that is what I go by for me.

"it doesn't take twice the bitrate to encode at 60p vs 30p"
Not exactly, but if you want the same quality per frame in your 60p as your 30p, it pretty nearly does.

[I]"Doing a Church video for instance, you have a person preaching and waving their hands all over the place and with 24p, you see a lot of stutter in their hand movements. So it does not matter how slow one pans or whatever, the stutter in the hand and other movements by the person speaking looks bad to me."[/I]

That's certainly a problem however just where it's coming from and why confounds me at times.

I shot a show last week at 25p and playing it back directly from the camera to a HDTV the stutter is very bad. Looking at it on Vegas's secondary monitor it's fine and so to are the SD 50i DVDs I made from it on the same model HDTV.

Bob.

Rainer

Again I say no it doesn't. The temporal difference between two frames in 60p is less then in 30p and more information can be taken from the previous frame to assemble the next frame or you can say less information is required to assemble the next frame. If you make an image sequence then yes, it will take twice as much information but a video stream will not be twice as large. If what you say is true we would not be able to have high speed cameras running 1000fps in full HD.

Oldsmoke, to summarize, if you are delivering 30p footage, unless you are doing slo mo your action will be as smooth and your image quality from the G30/XA20 will be better if you shoot 24Mbps 30p rather than 60p. If you have 60p delivery options or are shooting 60p to future proof your footage of course you will shoot 60p and in all probability the 60p footage displayed at 60p will look better than 30p displayed at 30p. I'm mainly saying be aware of what you are doing and have valid reasons for doing it.