HVR-HD1000 vs. DSR-250P

[Adam Gold]

Originally Posted by William Hohauser

The CMOS chip in the HD1000 is a progressive imaging chip. When you select 1/30th shutter in the HD1000 you get 30p distributed over a 60i signal. This is similar to how a camera like the Panasonic DVX100 distributes a 24p image over a 60i recording except it's a lot simpler with 30p. Both interlace fields have half of the same frame.

The manual for the HD1000, like the lack of f-stop and db increments in the viewfinder, is quite perfunctory and written more like a consumer camera manual. Not surprising since the camera is a consumer camera in a better body.

Thanks for the clarification, and for anticipating my next question about whether 1/30th is a good workaround to 30p.

And even the Pro manuals from Sony leave much to be desired, as if they are written and translated by someone who knows neither Japanese nor English.

[John Bosco Jr.]

Originally Posted by William Hohauser

The CMOS chip in the HD1000 is a progressive imaging chip. When you select 1/30th shutter in the HD1000 you get 30p distributed over a 60i signal. This is similar to how a camera like the Panasonic DVX100 distributes a 24p image over a 60i recording except it's a lot simpler with 30p. Both interlace fields have half of the same frame.

That doesn't make a bit of sense. The CMOS chip might be progressive, but that's where it ends. The signal is processed as interlaced, and adjusting shutter speed isn't going to make it a progressive signal.

According to your reasoning, 1/60th shutter would be 60p, 1/24 = 24p, 1/8 = 8p? I don't think so. All you do when you slow the shutter is make the camera more light sensitive at the expense of motion artifacts. This is an interlaced camera only, like the HC7.

There is also a big question mark of whether it has a progressive chip in it at all. Sony says it's not a progressive chip. Some are speculating that it is based on other Sony pro division cameras. I hardly think they would have that resolution-poor cinema effect if they could get progressive out of the camera.

If you want to say it gives a film-like look slowing down the shutter, then fine, but don't sell us that it is progressive. Let's stick with the facts.

[Chris Hurd]

Originally Posted by John Bosco Jr.

That doesn't make a bit of sense.

Actually it makes perfect sense.

The difference between interlace and progressive is the 1/60th-second separation between fields which causes interlace artifacts. If he shoots in 60i with a 1/30th-second shutter speed then he's dragging the shutter to the point where both fields are exposed at the same time, eliminating the time difference between the two fields of any frame. Yes it's a work-around to get 30p but it's called that for a reason -- it works.

The signal is processed as interlaced...

Of course it is; it has to be. This is an HDV format camcorder. The signal *must* be written to tape as interlaced field by field, or it wouldn't be HDV. But shooting 60i at 1/30 removes the time difference between every two pairs of fields, so it would look as if it were shot in 30p. And that's what he's going after: a look.

In HDV all frame rates -- even from natively progressive cameras such as the JVC Pro HD line -- are written to tape as interlace. That doesn't stop the recorded signal from remaining progressive. When the time difference between pairs of fields is zero then it's progressive even though it's still broken into fields.

According to your reasoning, 1/60th shutter would be 60p, 1/24 = 24p, 1/8 = 8p?

No, that's not what he's saying at all. This works only when it's 60i shot with a shutter speed of 1/30th second. It also works across the oceans when it's 50i shot at 1/25th to get 25p.

The only other instance I can think of would be shooting 24p with a shutter speed of 1/12th second to get 12p but who would want to do that. The 30p / 25p trick is much more useful.

There is also a big question mark of whether it has a progressive chip in it at all. Sony says it's not a progressive chip.

There is no question mark. It is a CMOS chip therefore by definition it is a progressive chip, as *all* CMOS chips are natively progressive. Sony says that the signal is not processed as progressive. That is the difference. What happens at the chip and what happens at the DSP are two completely separate things.

...but don't sell us that it is progressive

He is not "selling" it as progressive. He is offering a very common work-around that has been discussed on this site in numerous other camera boards here for years, ever since the XL1S was something new. I'm having a hard time believing that someone else is having a hard time believing this. 30p from 60i at 1/30 (or 25p from 50i at 1/25) is one of the oldest tricks in the book.

[William Hohauser]

Thank very much, Mr. Hurd. Couldn't have responded better myself especially since I was tending to a cute response like:

Even if you keep your shirts and pants in different drawers it doesn't mean you're not dressed when you put them on.

Did that make any sense?

What does make sense is the confusion over the 1/30th shutter since in the past interlaced cameras (especially the Sony cameras) would not do a progressive sweep in the shutter mode. Cameras like the PD-170 throw out one field and hold over the first (which is 240 lines horizontal). This makes a resolution poor image filled with aliasing. This is probably due to the chip being designed for interlaced video. I have not worked with any of the Sony CCD HDV cameras but from the complaints I've read, a similar method is being used for long shutter speeds and "cinema" effects.

The CMOS chips come out of the photography division where interlacing has never been used (as far as I know). A blatant assumption, I deduce that the HD1000 is actually taking progressive images at 1/60 but half the information is thrown out to fit the requirements of 60i video.

[John Bosco Jr.]

Originally Posted by Chris Hurd

Actually it makes perfect sense.

The difference between interlace and progressive is the 1/60th-second separation between fields which causes interlace artifacts. If he shoots in 60i with a 1/30th-second shutter speed then he's dragging the shutter to the point where both fields are exposed at the same time, eliminating the time difference between the two fields of any frame. Yes it's a work-around to get 30p but it's called that for a reason -- it works.

I admit I was incorrect in stating my point. I should have said, "Yes, it's progressive only with a slow shutter which introduces motion blur." Common shutter speed for native 30p is 1/60 sec and 1/48th sec for 24p. Mostly, slow shutter is used to increase low light sensitivity.

Originally Posted by Chris Hurd

Of course it is; it has to be. This is an HDV format camcorder. The signal *must* be written to tape as interlaced field by field, or it wouldn't be HDV. But shooting 60i at 1/30 removes the time difference between every two pairs of fields, so it would look as if it were shot in 30p. And that's what he's going after: a look.

I understand that; however, Bill assumed the hd1000 scans video progressively because of the progressive chip. Unlike the Sony V1u, it does not. They could have easily done that if they wanted to, but Sony decided not to.

Originally Posted by Chris Hurd

In HDV all frame rates -- even from natively progressive cameras such as the JVC Pro HD line -- are written to tape as interlace. That doesn't stop the recorded signal from remaining progressive. When the time difference between pairs of fields is zero then it's progressive even though it's still broken into fields.

Are you sure about that? HDV 1 standard allows 720p record, so I could have sworn that the progressive images are recorded natively to tape. HDV 2 is interlace and doesn't allow for progressive recording. That's why progressive has to be wrapped in the 60i stream.

Originally Posted by Chris Hurd

He is not "selling" it as progressive. He is offering a very common work-around that has been discussed on this site in numerous other camera boards here for years, ever since the XL1S was something new. I'm having a hard time believing that someone else is having a hard time believing this. 30p from 60i at 1/30 (or 25p from 50i at 1/25) is one of the oldest tricks in the book.

He's not? Well, it seemed like it to me. I apologize for the misunderstanding if that is the case. He could have said, "I have a work-around solution to get 1080 30p from the hd1000." I just don't want misinformation to get out, and I thank you for setting me straight on my misinformation.

[William Hohauser]

You get a progressive frame from the camera which taking a progressive image. This is not "misinformation" unless you can prove beyond an isolated Sony statement that I am wrong. This forum has a lot of members who dissect manufacturer statements and report on the true specs and functions of the equipment in question.

It seems that the misunderstanding is how interlaced works. Until progressive television became a reality, relatively recently, 60i was called 30 frames a second (or 29.97 to be exact). The interlacing was a method of reducing the bandwidth needed for broadcast transmission and later videotape recording. Two half resolution fields from alternating odd and even horizontal scan lines (240 lines) combined made a decent image with good resolution (480 lines). Interlacing depends on the phosphors of a standard CRT television to make it all work. The first even field is still glowing when the second odd field is projected on the tube. Nobody counted the fields although technicians were very aware of the fields when maintaining VTRs or timing equipment. 60i is 30fps. It could be said that NTSC uses a viewable image area of 240 progressive, 60 frames. But since the receivers (CRTs) can't show that and since each field is from alternate lines of the original 480 frame (odd/even) it wouldn't be accurate at all.

When HDTV finally started development for real over 25 years ago, the argument was whether we should stick with interlacing for a higher resolution (1080) or go for progressive with a lower resolution (720) but better motion detail. Remember that bandwidth is still calling the shots here. That argument has not been settled and we are entering a era where multiple resolutions and frame rates can live side by side sort of peacefully. 24p can live on a 60i recording and be extracted. 30p can do the same.

Let's also take into account that 30p is not standard for sporting events or live news. It's a pleasing, film-like frame rate good for drama, music and documentaries. JVC had to create a camera that shot in 60p to satisfy productions that wanted a more "live" feel then the HD100 could provide.

I mixed the HD1000 in 30p with a JVC HD100 (which shoots 30p natively) and had excellent results with no resolution hits from the HD1000. My resolution tests with the HD1000 show that you get the same still resolution with the 1/30th shutter as the 1/60th shutter. This does not happen with several other 60i HDV camera models. Perhaps I did the tests wrong, perhaps you have a HD1000 camera and can do some tests and post the results. I would be happy to see them and figure out what I did wrong if I did wrong. This is how science works.

[John Bosco Jr.]

Originally Posted by William Hohauser

You get a progressive frame from the camera which taking a progressive image. This is not "misinformation" unless you can prove beyond an isolated Sony statement that I am wrong. This forum has a lot of members who dissect manufacturer statements and report on the true specs and functions of the equipment in question.

It seems that the misunderstanding is how interlaced works. Until progressive television became a reality, relatively recently, 60i was called 30 frames a second (or 29.97 to be exact). The interlacing was a method of reducing the bandwidth needed for broadcast transmission and later videotape recording. Two half resolution fields from alternating odd and even horizontal scan lines (240 lines) combined made a decent image with good resolution (480 lines). Interlacing depends on the phosphors of a standard CRT television to make it all work. The first even field is still glowing when the second odd field is projected on the tube. Nobody counted the fields although technicians were very aware of the fields when maintaining VTRs or timing equipment. 60i is 30fps. It could be said that NTSC uses a viewable image area of 240 progressive, 60 frames. But since the receivers (CRTs) can't show that and since each field is from alternate lines of the original 480 frame (odd/even) it wouldn't be accurate at all.

When HDTV finally started development for real over 25 years ago, the argument was whether we should stick with interlacing for a higher resolution (1080) or go for progressive with a lower resolution (720) but better motion detail. Remember that bandwidth is still calling the shots here. That argument has not been settled and we are entering a era where multiple resolutions and frame rates can live side by side sort of peacefully. 24p can live on a 60i recording and be extracted. 30p can do the same.

Let's also take into account that 30p is not standard for sporting events or live news. It's a pleasing, film-like frame rate good for drama, music and documentaries. JVC had to create a camera that shot in 60p to satisfy productions that wanted a more "live" feel then the HD100 could provide.

I mixed the HD1000 in 30p with a JVC HD100 (which shoots 30p natively) and had excellent results with no resolution hits from the HD1000. My resolution tests with the HD1000 show that you get the same still resolution with the 1/30th shutter as the 1/60th shutter. This does not happen with several other 60i HDV camera models. Perhaps I did the tests wrong, perhaps you have a HD1000 camera and can do some tests and post the results. I would be happy to see them and figure out what I did wrong if I did wrong. This is how science works.

I own a Sony HC7. I had an interest in the shoulder mount HD1000, so I consulted with a Sony technician, not a sales rep.

When I questioned about the difference between the two cameras, I was told by the technician that they were essentially the same. The HD1000 has a few minor tweaks and in a more professional layout. When I questiioned whether or not the HD1000 added progressive scan, I was told no. The technician told me that, like the HC7, the HD1000 employs interlace scanning.

I can accept Chris Hurd's explanation that all CMOS chips are naturally progressive, but I cannot deny the guys who make the camera that it is interlaced. In fact, the new HC9 adds the progressive shutter feature for stills only. Why would Sony go to the trouble of faking progressive scanning if the camera already scans progressively? Your test could have shown a higher resolution still because of the higher resolution sensor of the HD1000.

Also, I know what interlace is. It was mainly developed as a way to avoid flicker because of the lack of bandwidth of NTSC to broadcast a full 60 frames. And, yes, I know why 720p60 was added. If by your "live feel" means sharper and smoother images, then you are correct.

I realize now my error when I suggested that the 1/30th shutter solution was not progressive. It is 30p only with a slow shutter. As I mentioned in an earlier comment, that introduces motion blur. For some, that's a good work-around when using an interlace camera. Yes, I'm implying that the HD1000 is an interlace camera only. That is not a bad thing; it's just not what it could have been, and that's a progressive scanned camera with true 30p and 24p.

I admited my error; now it's time you admit yours, and stop the misinformation that this is a progressive scanned camera. It's just simply not true, unless you have more information than the guys who make it.

[William Hohauser]

If you can prove that the two fields of a frame shot by the HD1000 with a 1/30 shutter are from different points in time as it is with 60i, I will admit to spreading "misinformation". If the two fields are odd and even lines from the same 1080 line frame then it's a progressive frame, there is no question about that.

The motion artifacts in 60i are a result of each field being a different point in time. If they were from the same point in time, the frame would be progressive with a 1/60th shutter much like what the JVC HD100 does in it's default settings. And as Mr. Hurd relates, the JVC apparently records the frame interlaced to tape (I didn't know this) but it's still progressive in creation.

By the way, I think our little disagreement is giving a number of people a decent lesson in TV technology.

[John Bosco Jr.]

Originally Posted by William Hohauser

If you can prove that the two fields of a frame shot by the HD1000 with a 1/30 shutter are from different points in time as it is with 60i, I will admit to spreading "misinformation". If the two fields are odd and even lines from the same 1080 line frame then it's a progressive frame, there is no question about that.

The motion artifacts in 60i are a result of each field being a different point in time. If they were from the same point in time, the frame would be progressive with a 1/60th shutter much like what the JVC HD100 does in it's default settings. And as Mr. Hurd relates, the JVC apparently records the frame interlaced to tape (I didn't know this) but it's still progressive in creation.

By the way, I think our little disagreement is giving a number of people a decent lesson in TV technology.

I'm not denying that you are getting a progressive frame using 1/30th sec shutter. In fact, if you carefully re-read my previous post, you will see that I conceded you are getting 30p by using the 1/30th sec slow shutter method.

Originally, I failed to understand this logic until Chris Hurd correctly explained to me that slowing the shutter down in half exposes both the even and odd fields from 60i. However, I correctly added that in doing this, you inherit the drawback of motion blur. Try setting your JVC 100 to 30p with a 1/30th shutter (normally the shutter would be 1/60th). You will see motion blur. We're not talking about interlace artifacts; we're talking about the result of slowing the shutter speed when shooting moving objects.

Slowing the shutter to 1/30th sec to achieve 30p is not for cameras that feature progressive scan; it is for cameras that use interlace scanning, providing they have the capability to set the shutter speed to 1/30th sec. The trade off is progressive video with the expense of a certain degree of motion blur. For some, this is acceptable to achieve a so-called film-like look.

If the HD1000 was able to scan progressively, it would be able to get 30p and 24p without having to set the shutter speed to 1/30th sec, like the Sony V1 through pulldown, for instance. The HD1000 is an interlace-only camera. Again, that's not my opinion; that's a fact from the guys who make it.

I hope this explanation helps you understand that the HD1000, unfortunately, does not feature progressive scan. By the way, I agree with you; us knocking heads on this topic has produced, like you said, "... a decent lesson in TV technology." I know I have learned a thing or two from this thread.

PS - I'm not completely sold on Chris Hurd's comment that all HDV frame rates are recorded to tape as interlace. HDV 1 standard allows for High definition recording of 720p in 24p, 25p, 30p, 50p and 60p. It allows standard definition recording of dv in interlace. I think Mr. Hurd is in error about this, but I'm not 100 percent sure, yet. There could be some confusion about this because the HDV 2 standard only allows for interlace recording (1080 60i), so 24p and 30p (or 25p Europe) recorded from a progressive scan camera must wrap the progressive images in the 60i stream.

[Jaser Stockert]

sorry to bring up an old thread, but how does the hd1000u's 60i 1/30 shutter compare to a true progressive camera such as the canon hv30? will there be resolution loss such as deinterlacing 60i footage in a nle? thanks.

[Bryan Daugherty]

Originally Posted by Chris Hurd

...The difference between interlace and progressive is the 1/60th-second separation between fields which causes interlace artifacts. If he shoots in 60i with a 1/30th-second shutter speed then he's dragging the shutter to the point where both fields are exposed at the same time, eliminating the time difference between the two fields of any frame. Yes it's a work-around to get 30p but it's called that for a reason -- it works....

Chris,
I have been operating my HD1000u this way from the beginning and edit in my NLE as a progressive project (1080p) and never have I heard this explained so clearly before. I knew it worked but I didn't know for sure why. Thanks for your clarification, and to William for pointing it out in his previous post.