Confused about CF25 and shutter speed at DVinfo.net

Garry Matthews · January 3rd, 2005, 05:07 AM

I'm in the PAL half of the world, and I'm wondering, what is the difference between cineframe25 mode, and just using the 1/25th shutter speed? From a totally non-scientific point of view, to my eyes the two settings seem to produce the same effect on the screen.
I mean, with 1/25 shutter, the lens stays open for both field scans of a frame, so there would be no interlacing, right?

I can see how CF24 does do something in the FX1 (2:3 pulldown), but does CF25 do anything in FX1E?

Tim Ashbrooke · January 3rd, 2005, 04:50 PM

I did some testing on the weekend at the beach with CF25. I can dial up the shutter speed and still keeps that film-look cadence happening. I don't know what the difference is with the shutter speed set to 1/25th as opposed to 1/25th in CF25 mode?? Maybe it could be the same as Canon's Frame mode, where there's still two fields for each 'frame' but those two fields are the same progressive pic? If that is that case, is it still full 1080 res?

Enlightenment is very welcome!

Toke Lahti · January 3rd, 2005, 06:35 PM

<<<-- Originally posted by Garry Matthews: I mean, with 1/25 shutter, the lens stays open for both field scans of a frame, so there would be no interlacing, right? -->>>

Very good question and still a mystery for me also.
I'd see only one solution: fx1(e) has "progressive ccd's".
If ccd's would be read interlaced, the resolution should drop to half when using shutter speeds that are longer than recording time of one field.
To my experience resolution does not drop to half, so I keep z1 in my shopping list and wish that somebody creates a modification to it, that a real progressive picture can be taken out of it...

Chris Hurd · January 4th, 2005, 09:53 AM

<< I'd see only one solution: fx1(e) has "progressive ccd's". >>

Absolutely incorrect; I'm not sure why some folks insist on clinging to this myth. All of the Sony broadcast division people I have spoken with, including vice presidents and engineers and others who know what they're talking about, insist that it is the same interlaced CCD block shared by the FX1 and they have no reason at all to lie about it.

Chris Hurd · January 4th, 2005, 10:00 AM

Tim:

The Canon frame mode you're referring to is actually a Panasonic technology. And all camcorders, regardless of whether their CCD's are true progressive scan or not, are recording interlace (separated fields) to the tape. They have to, otherwise you wouldn't be able to play the video back on a standard NTSC or PAL monitor. The difference with true progressive scan (or Panasonic frame mode or Sony CineFrame) vs. interlace CCD's is that when those separate fields are recorded to tape, there's no longer a 1/60th of a second acquisition delay between those two fields. Instead, there's now a 1/30th of a second acquisition delay between every other field. You still have two fields per frame going to tape as the DSP breaks the progressive feed from the CCD block into interlace. The temporal motion of progressive scan is preserved intact as there is no time delay between fields for any given frame. Now there's a longer delay between frames, which is the signature of Frame mode, CineFrame and true progressive scan CCD's.

Interlace recording of interlace CCD's:

Frame 1 Field A > 1/60th sec. > Frame 1 Field B > 1/60th sec. > Frame 2 Field A > 1/60th sec. > etc.

Interlace breakdown of progressive CCD's (also Panasonic frame mode, Sony CineFrame):

Frame 1 Field A, Frame 1 Field B > 1/30th sec. > Frame 2 Field A, Frame 2 Field B > 1/30th sec. > etc.

Heath McKnight · January 4th, 2005, 11:10 AM

Though we shot in CF 24 for my short film (clips and jpegs coming soon) and we used a shutter of 1/60, I'm not going to recommend it for something bigger, ie, a feature. The 20-30% vertical resolution loss isn't worth it.

But keep in mind, I'm VERY happy with the look of our short film, 3 Sisters done in CF 24.

heath

Toke Lahti · January 4th, 2005, 12:33 PM

<<<-- Originally posted by Chris Hurd :Absolutely incorrect; I'm not sure why some folks insist on clinging to this myth. All of the Sony broadcast division people I have spoken with, including vice presidents and engineers and others who know what they're talking about, insist that it is the same interlaced CCD block shared by the FX1 and they have no reason at all to lie about it. -->>>

Then how do you explain that resolution does not halve with slow shutter?
Or how do these honest sony people explain it?

Or are you claiming that resolution does halve?

When Steve Mullen says that every ccd is after all progressive he's wrong?

Why do you think broadcast cameras have more lines in their ccd's than they feed out or record to tape?
(Hint: interpolation, line pairing, progressive readout...)

Example: set "interlaced" PAL camera to exposure time 1/12 s. Exposure time for one field is 1/24 s. Camera has to record a field in every 1/50 s. So what does it do?
Records the same field twice consecutively.
Effect: only half resolution.

So it is impossible for "interlaced" camera to have more than half resolution with slow shutter than with normal or fast shutter.

Barry Green · January 4th, 2005, 12:56 PM

Quote:

Then how do you explain that resolution does not halve with slow shutter?

On the FX1? Resolution does halve when you go into CF30 (i.e., you get about 540 lines of resolution, it's not half as compared to the interlaced picture, which is around 800 lines). I didn't test resolution on slow shutter mode, but I would expect similar.

Chris Hurd · January 4th, 2005, 01:01 PM

Toke Lahti:

My only "claim" is that there is no difference between the CCD blocks in the FX1 and Z1, and that they produce 1080i video. The "i" stands for interlace. Despite what people want it to be, it can only do what it designed to do: interlace.

The idea that there is some sort of consipracy at work here, that the camera's manufacturer would purposefully misrepresent this and is not being "honest" about this, is ridiculous on its face. If you wish to pursue such a fanatasy, we have a dedicated forum for that here called Area 51. The rest of this site, including this board, is reserved for the discussion of accurate technical information. Sometimes certain people do not understand this concept and insist on forcing misinformation into these discussions. Sooner or later they must part ways with us. Please don't let that happen to you.

The burden of proof regarding how this camera's resolution is or is not preserved in slow shutter speeds does not lie on my shoulders. My interest lies only in pointing out factual techical information. The fact is that this camera is made for 1080i.

You are lifting Steve Mullen's statements entirely out of the context in which he made them. He will be the first to tell you that there is much more involved, that the limitations of the bit depths of the CCD block's analog/digital converter and the camera's digital signal processor also play a big factor in whether or not a camera produces true progressive scan or interlace video. You are attempting to oversimplify things; in reality it's much more complicated than that.

I will repeat myself and advise you yet again to please get into the spirit of the proceedings here.

Toke Lahti · January 4th, 2005, 06:35 PM

So I'm still waiting for _somebody_ who yet owns the camera to measure the resolution with slow shutter speeds and without/with cf.

Amazing that so many people here has the camera, some are even willing to dismantle it to parts (great!), but nobody seems to be interested measuring its acuity.

I made a quick hands on test with fx1e and there seemed to be no resolution loss with slow shutters.

Chris, forgetting about myths, conspiracies and fantasies, if you have deeper knowledge of camera technology and engineering than I do, how do you explain this no resolution loss with slow shutter?

As long as nobody states otherwise, I believe the feeling which I got with my own eyes, camera and native hd resolution monitor.

So this is no speculation, it's something that I have experienced and if somebody says that slow shutters with no half resolution with interlaced ccd is possible, I'd like to hear explanation how.

Believe me, I'm not trying to bug anybody here, I'm trying to find answers to very simple questions.

Based on those answers I should be able to decide if I should buy z1 in feb or maybe sumix 2/3" hd box cam in march. And I can't afford to buy both...

Tim Ashbrooke · January 4th, 2005, 08:33 PM

Will I lose resolution when I use CF25 mode? or will i still have the full 1080 lines?

Chris Hurd · January 4th, 2005, 09:11 PM

Toke

<< forgetting about myths, conspiracies and fantasies, if you have deeper knowledge of camera technology and engineering than I do, how do you explain this no resolution loss with slow shutter? >>

I make no claim to having a deeper knowledge of anything. There are some folks who will happily support me on that claim and they will even go so far as to say that I have a deep knowledge of nothing. However. The way we operate around here when it comes to speculation and supposition and surmising, is that we base it from the manufacturer's published specifications. They are in no position to misrepresent themselves. Occasionally they make mistakes -- for example, Canon USA has inaccurately listed the physical dimensions of the XL2 CCD area measurements. But typo's will happen. What does not happen is for a manufacturer to lie outright about what their product is or is not.

The burden of proof does not rest upon me, but upon the party who makes the claim that something is not what it seems. And that happens only very rarely.

<< Believe me, I'm not trying to bug anybody here, I'm trying to find answers to very simple questions. >>

Glad to hear that; thank you.

Heath McKnight · January 4th, 2005, 11:22 PM

<<<-- Originally posted by Tim Ashbrooke : Will I lose resolution when I use CF25 mode? or will i still have the full 1080 lines? -->>>

Tim, yes, you will. Between 20% and 30%.

heath

Barry Green · January 4th, 2005, 11:55 PM

<<<-- Originally posted by Tim Ashbrooke : Will I lose resolution when I use CF25 mode? or will i still have the full 1080 lines? -->>>

You will lose resolution, yes. But you never have 1080 lines in the first place. The FX1 in interlaced mode delivers an observed vertical resolution of about 775 lines, and Steve Mullen's article illustrates why it should theoretically deliver about 820... 775 is probably close enough to 820 to be within the margin of error when shooting a resolution chart.

When switching to CF25 you lose resolution, but not half. It goes down to about 575 observed lines as shown on a resolution chart.

Toke Lahti · January 5th, 2005, 03:26 AM

<<<-- Originally posted by Chris Hurd: The way we operate around here when it comes to speculation and supposition and surmising, is that we base it from the manufacturer's published specifications. They are in no position to misrepresent themselves. -->>>

So, what's sony's official specs for resolution with slow shutter?
This might be also just about semantics. When sony says that camera is interlaced, what does it mean?
Recording to tape interlaced?
Processing the image in dsp in interlaced mode?

Nobody in Sony says that pd150/170 is a progressive camera. Nevertheless it records progressive frames with slow shutter.

January 3rd, 2005, 05:07 AM	#1
Garry Matthews New Boot Join Date: Jan 2005 Location: Melbourne, Australia Posts: 5	Confused about CF25 and shutter speed I'm in the PAL half of the world, and I'm wondering, what is the difference between cineframe25 mode, and just using the 1/25th shutter speed? From a totally non-scientific point of view, to my eyes the two settings seem to produce the same effect on the screen. I mean, with 1/25 shutter, the lens stays open for both field scans of a frame, so there would be no interlacing, right? I can see how CF24 does do something in the FX1 (2:3 pulldown), but does CF25 do anything in FX1E?

January 3rd, 2005, 06:35 PM	#3
Toke Lahti Major Player Join Date: Oct 2004 Location: Helsinki, Finland Posts: 158	Re: Confused about CF25 and shutter speed <<<-- Originally posted by Garry Matthews: I mean, with 1/25 shutter, the lens stays open for both field scans of a frame, so there would be no interlacing, right? -->>> Very good question and still a mystery for me also. I'd see only one solution: fx1(e) has "progressive ccd's". If ccd's would be read interlaced, the resolution should drop to half when using shutter speeds that are longer than recording time of one field. To my experience resolution does not drop to half, so I keep z1 in my shopping list and wish that somebody creates a modification to it, that a real progressive picture can be taken out of it...

January 4th, 2005, 09:53 AM	#4
Chris Hurd Obstreperous Rex Join Date: Jan 2001 Location: San Marcos, TX Posts: 27,368 Images: 513	<< I'd see only one solution: fx1(e) has "progressive ccd's". >> Absolutely incorrect; I'm not sure why some folks insist on clinging to this myth. All of the Sony broadcast division people I have spoken with, including vice presidents and engineers and others who know what they're talking about, insist that it is the same interlaced CCD block shared by the FX1 and they have no reason at all to lie about it. __________________ CH Search DV Info Net \| 20 years of DVi \| ...Tuesday is Soylent Green Day!

January 4th, 2005, 10:00 AM	#5
Chris Hurd Obstreperous Rex Join Date: Jan 2001 Location: San Marcos, TX Posts: 27,368 Images: 513	Tim: The Canon frame mode you're referring to is actually a Panasonic technology. And all camcorders, regardless of whether their CCD's are true progressive scan or not, are recording interlace (separated fields) to the tape. They have to, otherwise you wouldn't be able to play the video back on a standard NTSC or PAL monitor. The difference with true progressive scan (or Panasonic frame mode or Sony CineFrame) vs. interlace CCD's is that when those separate fields are recorded to tape, there's no longer a 1/60th of a second acquisition delay between those two fields. Instead, there's now a 1/30th of a second acquisition delay between every other field. You still have two fields per frame going to tape as the DSP breaks the progressive feed from the CCD block into interlace. The temporal motion of progressive scan is preserved intact as there is no time delay between fields for any given frame. Now there's a longer delay between frames, which is the signature of Frame mode, CineFrame and true progressive scan CCD's. Interlace recording of interlace CCD's: Frame 1 Field A > 1/60th sec. > Frame 1 Field B > 1/60th sec. > Frame 2 Field A > 1/60th sec. > etc. Interlace breakdown of progressive CCD's (also Panasonic frame mode, Sony CineFrame): Frame 1 Field A, Frame 1 Field B > 1/30th sec. > Frame 2 Field A, Frame 2 Field B > 1/30th sec. > etc. __________________ CH Search DV Info Net \| 20 years of DVi \| ...Tuesday is Soylent Green Day!

January 4th, 2005, 11:10 AM	#6
Heath McKnight MPS Digital Studios Join Date: Apr 2003 Location: Palm Beach County, Florida Posts: 8,531	Though we shot in CF 24 for my short film (clips and jpegs coming soon) and we used a shutter of 1/60, I'm not going to recommend it for something bigger, ie, a feature. The 20-30% vertical resolution loss isn't worth it. But keep in mind, I'm VERY happy with the look of our short film, 3 Sisters done in CF 24. heath __________________ My Final Cut Pro X blog

January 3rd, 2005, 04:50 PM	#2
Tim Ashbrooke Regular Crew Join Date: Apr 2003 Location: Grafton, NSW, Australia Posts: 38	I did some testing on the weekend at the beach with CF25. I can dial up the shutter speed and still keeps that film-look cadence happening. I don't know what the difference is with the shutter speed set to 1/25th as opposed to 1/25th in CF25 mode?? Maybe it could be the same as Canon's Frame mode, where there's still two fields for each 'frame' but those two fields are the same progressive pic? If that is that case, is it still full 1080 res? Enlightenment is very welcome!

January 4th, 2005, 12:33 PM	#7
Toke Lahti Major Player Join Date: Oct 2004 Location: Helsinki, Finland Posts: 158	<<<-- Originally posted by Chris Hurd :Absolutely incorrect; I'm not sure why some folks insist on clinging to this myth. All of the Sony broadcast division people I have spoken with, including vice presidents and engineers and others who know what they're talking about, insist that it is the same interlaced CCD block shared by the FX1 and they have no reason at all to lie about it. -->>> Then how do you explain that resolution does not halve with slow shutter? Or how do these honest sony people explain it? Or are you claiming that resolution does halve? When Steve Mullen says that every ccd is after all progressive he's wrong? Why do you think broadcast cameras have more lines in their ccd's than they feed out or record to tape? (Hint: interpolation, line pairing, progressive readout...) Example: set "interlaced" PAL camera to exposure time 1/12 s. Exposure time for one field is 1/24 s. Camera has to record a field in every 1/50 s. So what does it do? Records the same field twice consecutively. Effect: only half resolution. So it is impossible for "interlaced" camera to have more than half resolution with slow shutter than with normal or fast shutter.

January 4th, 2005, 01:01 PM	#9
Chris Hurd Obstreperous Rex Join Date: Jan 2001 Location: San Marcos, TX Posts: 27,368 Images: 513	Toke Lahti: My only "claim" is that there is no difference between the CCD blocks in the FX1 and Z1, and that they produce 1080i video. The "i" stands for interlace. Despite what people want it to be, it can only do what it designed to do: interlace. The idea that there is some sort of consipracy at work here, that the camera's manufacturer would purposefully misrepresent this and is not being "honest" about this, is ridiculous on its face. If you wish to pursue such a fanatasy, we have a dedicated forum for that here called Area 51. The rest of this site, including this board, is reserved for the discussion of accurate technical information. Sometimes certain people do not understand this concept and insist on forcing misinformation into these discussions. Sooner or later they must part ways with us. Please don't let that happen to you. The burden of proof regarding how this camera's resolution is or is not preserved in slow shutter speeds does not lie on my shoulders. My interest lies only in pointing out factual techical information. The fact is that this camera is made for 1080i. You are lifting Steve Mullen's statements entirely out of the context in which he made them. He will be the first to tell you that there is much more involved, that the limitations of the bit depths of the CCD block's analog/digital converter and the camera's digital signal processor also play a big factor in whether or not a camera produces true progressive scan or interlace video. You are attempting to oversimplify things; in reality it's much more complicated than that. I will repeat myself and advise you yet again to please get into the spirit of the proceedings here. __________________ CH Search DV Info Net \| 20 years of DVi \| ...Tuesday is Soylent Green Day!

January 4th, 2005, 06:35 PM	#10
Toke Lahti Major Player Join Date: Oct 2004 Location: Helsinki, Finland Posts: 158	So I'm still waiting for _somebody_ who yet owns the camera to measure the resolution with slow shutter speeds and without/with cf. Amazing that so many people here has the camera, some are even willing to dismantle it to parts (great!), but nobody seems to be interested measuring its acuity. I made a quick hands on test with fx1e and there seemed to be no resolution loss with slow shutters. Chris, forgetting about myths, conspiracies and fantasies, if you have deeper knowledge of camera technology and engineering than I do, how do you explain this no resolution loss with slow shutter? As long as nobody states otherwise, I believe the feeling which I got with my own eyes, camera and native hd resolution monitor. So this is no speculation, it's something that I have experienced and if somebody says that slow shutters with no half resolution with interlaced ccd is possible, I'd like to hear explanation how. Believe me, I'm not trying to bug anybody here, I'm trying to find answers to very simple questions. Based on those answers I should be able to decide if I should buy z1 in feb or maybe sumix 2/3" hd box cam in march. And I can't afford to buy both...

January 4th, 2005, 08:33 PM	#11
Tim Ashbrooke Regular Crew Join Date: Apr 2003 Location: Grafton, NSW, Australia Posts: 38	Will I lose resolution when I use CF25 mode? or will i still have the full 1080 lines?

January 4th, 2005, 09:11 PM	#12
Chris Hurd Obstreperous Rex Join Date: Jan 2001 Location: San Marcos, TX Posts: 27,368 Images: 513	Toke << forgetting about myths, conspiracies and fantasies, if you have deeper knowledge of camera technology and engineering than I do, how do you explain this no resolution loss with slow shutter? >> I make no claim to having a deeper knowledge of anything. There are some folks who will happily support me on that claim and they will even go so far as to say that I have a deep knowledge of nothing. However. The way we operate around here when it comes to speculation and supposition and surmising, is that we base it from the manufacturer's published specifications. They are in no position to misrepresent themselves. Occasionally they make mistakes -- for example, Canon USA has inaccurately listed the physical dimensions of the XL2 CCD area measurements. But typo's will happen. What does not happen is for a manufacturer to lie outright about what their product is or is not. The burden of proof does not rest upon me, but upon the party who makes the claim that something is not what it seems. And that happens only very rarely. << Believe me, I'm not trying to bug anybody here, I'm trying to find answers to very simple questions. >> Glad to hear that; thank you. __________________ CH Search DV Info Net \| 20 years of DVi \| ...Tuesday is Soylent Green Day!

January 4th, 2005, 11:22 PM	#13
Heath McKnight MPS Digital Studios Join Date: Apr 2003 Location: Palm Beach County, Florida Posts: 8,531	<<<-- Originally posted by Tim Ashbrooke : Will I lose resolution when I use CF25 mode? or will i still have the full 1080 lines? -->>> Tim, yes, you will. Between 20% and 30%. heath __________________ My Final Cut Pro X blog

January 4th, 2005, 11:55 PM	#14
Barry Green Barry Wan Kenobi Join Date: Jul 2003 Location: North Carolina Posts: 3,863	<<<-- Originally posted by Tim Ashbrooke : Will I lose resolution when I use CF25 mode? or will i still have the full 1080 lines? -->>> You will lose resolution, yes. But you never have 1080 lines in the first place. The FX1 in interlaced mode delivers an observed vertical resolution of about 775 lines, and Steve Mullen's article illustrates why it should theoretically deliver about 820... 775 is probably close enough to 820 to be within the margin of error when shooting a resolution chart. When switching to CF25 you lose resolution, but not half. It goes down to about 575 observed lines as shown on a resolution chart.

January 5th, 2005, 03:26 AM	#15
Toke Lahti Major Player Join Date: Oct 2004 Location: Helsinki, Finland Posts: 158	<<<-- Originally posted by Chris Hurd: The way we operate around here when it comes to speculation and supposition and surmising, is that we base it from the manufacturer's published specifications. They are in no position to misrepresent themselves. -->>> So, what's sony's official specs for resolution with slow shutter? This might be also just about semantics. When sony says that camera is interlaced, what does it mean? Recording to tape interlaced? Processing the image in dsp in interlaced mode? Nobody in Sony says that pd150/170 is a progressive camera. Nevertheless it records progressive frames with slow shutter.