what watching the 50" HDTV at 1m distance taught me at DVinfo.net

Piotr Wozniacki · April 27th, 2008, 01:55 AM

I have recently installed the Panasonic HDTV, 50" monster over my editing seat (and the regular 24" LCD, also full HD). Using Vegas or Edius, I am now able to monitor my EX1's blown-up picture with virtually pixel-to-pixel precision. And I can tell you the quality of images this machine produces is absolutely and ultimately marvellous! In fact it's so good I can hardly see the need for Flash XDR's 50 or 100Mbps, 4:2:2 any more:)

Did I ever say I liked the picture quality from my old V1E better? Really? Well, I am taking it back now!

However, this close inspection also let me find that even though I never noticed it before using the 24" HD LCD, I do have an ever so slight back focus problem. My FW version is 1.03; what would you do mates: send it to Sony for upgrade (Gosh, I was counting on the ability of doing it myself); or try the "Gerald procedure" first? I have some important projects soon, so - considering the Prime Support servicing turn-around might take long, and doing it myself might render it worse instead of perfect - this is quite a dilemma.

Any new facts that might help make my mind up?

Gerald Loidl · April 27th, 2008, 02:29 AM

Piotr,
I had my camera at Sony Primesupport 3 times and they could not fix the backfocus issue. It always took at least one week turnaround - the first time it was in for 3 weeks...
They even could not fix it with the new firmware. I guess being in Europe you will have to send it to the same service station I had to. There are numerous reports from european owners that they got it back even worse.
I would try the repair procedure by yourself. It´s almost impossible to make it worse if you follow my procedure. I once tried to make the backfocus worse for test purposes and it wasn`t easy...
I successfully "repaired" Sony´s loaner camera with FW 1.03, as did many others.
Go for it! If anything goes wrong - you can still send it in. You can´t loose anything ;-)

cu,
Gerald

Piotr Wozniacki · April 27th, 2008, 02:42 AM

Thanks Gerald,

I'll probably do as you say. After all, there have been no reports on any important changes that the 1.05 fw brings over the 1.03...

Being lazy though, I'll probably just display the Siemens star on my 50" plasma, instead of printing 12 A4 cards, as you did. As far as I know, this is how Adam Wilt did his adjustment. My only reservation is related to the fact the the plasma is tilted down some 15 deg, so I'll have to take special care in order for my EX1 to look square at it.

Do you think it'd work?

Gerald Loidl · April 27th, 2008, 02:51 AM

Never tried it with a screen - but it worked for Adam.
Make sure that no other things are in the view of the camera which could irritate the automatic setup procedure.
I think the most important thing for the software is that it is a high contrast image with sharp edges it can focus on.
I think Adam did the setup in a completely darkened room with only the light of the monitor.

Good luck - let us know!

Piotr Wozniacki · April 27th, 2008, 02:53 AM

Quote:

Originally Posted by Gerald Loidl

I think Adam did the setup in a completely darkened room with only the light of the monitor.

Good luck - let us know!

This is what I'm planning, too - dark room with just the plasma shining...

I'll certailny report back on the results!

Bruce Rawlings · April 27th, 2008, 03:43 AM

I used the 'Gerald' 12 x A4 system and it worked pefectly. It was re assuring to see the auto system lock on to the wide shot and focussed the charts. I think the 1 metre test is interesting as I have a 50" Pioneer plasma and an 730s HDCAM with £10k lenses set to BBC lineup. The EX1 is looking very good when comparing the two but wide angles do highlight that you do get something extra for your HDCAM money!

Mike Stevens · April 27th, 2008, 07:45 AM

Quote:

Originally Posted by Piotr Wozniacki

This is what I'm planning, too - dark room with just the plasma shining...

I'll certailny report back on the results!

Piotr:

Please explain how you are doing this test. When I tried I was unable to see if focus was retained zooming out as the subject focused on gets smaller I can no longer see just how good the focus was.

Did you focus on the screen so the camera was loking at itself? Did you use Peaking? Expanded focus?

Mike

Benjamin Eckstein · April 27th, 2008, 07:56 AM

I tried it on a 50" plasma a while back with a Siemens Star but it was hard to do for all three ND settings. I cannot remember exactly what happened, but I think it had a hard time figuring out focus from the screen and it ended up getting stuck halfway through the adjustment and essentially crashed. I had to turn off the camera and turn it on again (which freaked me out that it was mid adjustment, although nothing bad happened). Anyways, maybe it will work for you.

B

Don Greening · April 27th, 2008, 08:46 AM

Quote:

Originally Posted by Piotr Wozniacki

After all, there have been no reports on any important changes that the 1.05 fw brings over the 1.03...

The tech guy at our local Sony repair facility told me that the latest firmware upgrade v.1.0.5 is supposed to fix the back focus issue when using the in-camera ND filters. This means that the back focus routine should be done with no ND filter engaged and when finished the new firmware will compensate for the ND settings to keep the back focus accurate.

- Don

George Kroonder · April 27th, 2008, 09:19 AM

I'm not a big proponent of using a display as a BF target, although I'm not saying it won't work. Using a dark room with essentially a backlit target can be effective when an "uncluttered" wideangle view is not available to prevent ambiguous autofocus.

Basically when you display a siemens star target on a display, you're testing that displays resolution. This is also the case when you print a chart, only you're testing the printer. However a good laserprinter has a much higher resolution then your screen.

It is worse when you use scanned charts and print those, or use inkjet printers (with bad paper). These will, in my humble opinion, not produce adequate target material.

A low quality/low resolution target will require you to put it further away from the lens to 'compensate' for the poor representation. Conversely you can shoot a good quality target from closer up.

I've attached two screenshots, one of a target I Googled, another of a 'digital original' both at 6400% (max. Acrobat zoom). You can clearly see the difference between a scanned copy and an original.

If you use a good quality target, you should be able to put that 2-3m away from the lens and get good results. You can test if the autofocus has trouble focussing on the target by using PUSH AF (in AF manual mode) when wide. If it focuses ok, it will do ok in the BF adustment.

You can use the same setup to check for BF problems. If you can attach a big(ger) high-res monitor to the camera output that would be a boon. Best to check focus at full res. Same setup used for manual BF adjustment (not on the EX1 as it has none).

Using a printingpress properly printed original has my preference, but I've made a few derivative originals (not scanned) I print off on my 1200dpi laser printer up to A3 with good result.

George/

Sean Donnelly · April 27th, 2008, 09:33 AM

I agree. A decent photo printer (epson R2400,etc.) has an effective resolution of 360 pixels per inch. A 50" 1920x1080 plasma still only has as many pixels as a 24" 1920x1080 display, they're just larger. A 50" display is roughly 45" in the horizontal direction, giving you a resolution of 42.667 ppi. Less than 1/6th of a chart printed on a home photo printer. For black and white graphics and text, a decent laser printer will resolve approximately 600ppi. The Putora sharpness indicator is my personal favorite, which uses a very high resolution pattern specifically designed to cause interference patterns on a digital image making it VERY easy to see focus.

Steven Thomas · April 27th, 2008, 09:47 AM

Thanks for sharing your thoughts George.

You made an interesting point about verifying using "push" auto focus before running the test.

Also, having said that, I wonder if it may be smart to put the camera in 1080 60i (or PAL 50i) mode before performing the BF calibration.

Auto focus work a lot better in interlace mode since the sensor is updated twice as fast.

Eric Pascarelli · April 27th, 2008, 10:01 AM

I agree about the printers, monitors and such, but I don't think an autofocus routine (and hence the EX1 backfocus routine) really needs a good Siemens chart to do its thing.

The Siemens "focus bullet" chart is great for the human eye to see perfect focus (especially in an optical system) because it "snaps" into focus nicely. But does the EX1, which looks for contrasty edges, really need this? The very sharpest parts of the Siemens chart occupy such a small area of the frame that I doubt the camera even acknowledges them for its focus routine (especially at full wide at 3 meters, A4 sized chart).

When I first got the instructions (from a Sony dealer in Canada) for doing the Auto FB adjustment, no mention was specifically made of a Siemens chart - just a backfocus chart of some undefined sort (perhaps Sony has a proprietary one?). Anyway, a large chart with smaller features at the center and larger features at the edges would seem to be the best thing. A room sized Siemens chart might fit the bill.

I haven't tried this, but I bet some crisp black tape on a white background would work just as well or better than the Siemens focus bullet. A clean tape edge has a very high "resolution" (the sharp transition between black and white) as far as autofocus is concerned. For my next Auto FB, I'll put a focus bullet in the middle and some larger, crisp black features around it.

And I'm sure a plasma will work fine as well. Regardless of its resolution for displaying a Siemens chart, there's probably enough there for the camera to focus on - the pixels themselves in closeup and the edges of the screen at full wide.

George Kroonder · April 27th, 2008, 10:19 AM

Hi Eric,

True, I've asked myself the same questions (and you're probably right). But I don't know the inner workings of the FB ajust function. Adam Wilt was adamant about using a siemens star target and I believe it is just good practice. For all I know it detects interferance from the chart.

Somewhere there exisits a Sony directive on how the FB Adjust rig/setup needs to be. I'd love to have that clarity.

Until then I just stick to what I know to work, regardless of camera and lens.

George/

P.S. Steven, as the 'scene' is static I don't think the mode matters and the routine may just override anything anyways.

Sean Donnelly · April 27th, 2008, 02:09 PM

That's true Eric, the auto FB doesn't neccesarily benefit from the added resolution. I've had very good luck doing the adjustment off of a piece of letterhead with printed text on it taped to a blank wall. All that the autofocus does is look for edge contrast, and I'm guessing because of the nature of the CMOS sensors it only looks in the horizontal direction (just like the Peaking function).

I don't know if it would actually work with the auto function, but this is the chart we normally use to judge sharpness:http://www.ascmag.com/store/product....cat=333&page=1

I think what Sony is using must be something like the Digiprime backfocus tool. it is essentially a backlit siemens star behind an optical element. It works very well.

April 27th, 2008, 01:55 AM	#1
Piotr Wozniacki Inner Circle Join Date: Aug 2006 Location: Poland Posts: 4,086	what watching the 50" HDTV at 1m distance taught me I have recently installed the Panasonic HDTV, 50" monster over my editing seat (and the regular 24" LCD, also full HD). Using Vegas or Edius, I am now able to monitor my EX1's blown-up picture with virtually pixel-to-pixel precision. And I can tell you the quality of images this machine produces is absolutely and ultimately marvellous! In fact it's so good I can hardly see the need for Flash XDR's 50 or 100Mbps, 4:2:2 any more:) Did I ever say I liked the picture quality from my old V1E better? Really? Well, I am taking it back now! However, this close inspection also let me find that even though I never noticed it before using the 24" HD LCD, I do have an ever so slight back focus problem. My FW version is 1.03; what would you do mates: send it to Sony for upgrade (Gosh, I was counting on the ability of doing it myself); or try the "Gerald procedure" first? I have some important projects soon, so - considering the Prime Support servicing turn-around might take long, and doing it myself might render it worse instead of perfect - this is quite a dilemma. Any new facts that might help make my mind up? __________________ Sony PXW-FS7 \| DaVinci Resolve Studio; Magix Vegas Pro; i7-5960X CPU; 64 GB RAM; 2x GTX 1080 8GB GPU; Decklink 4K Extreme 12G; 4x 3TB WD Black in RAID 0; 1TB M.2 NVMe cache drive

April 27th, 2008, 02:29 AM	#2
Gerald Loidl Regular Crew Join Date: Feb 2008 Location: Austria Posts: 122	Piotr, I had my camera at Sony Primesupport 3 times and they could not fix the backfocus issue. It always took at least one week turnaround - the first time it was in for 3 weeks... They even could not fix it with the new firmware. I guess being in Europe you will have to send it to the same service station I had to. There are numerous reports from european owners that they got it back even worse. I would try the repair procedure by yourself. It´s almost impossible to make it worse if you follow my procedure. I once tried to make the backfocus worse for test purposes and it wasn`t easy... I successfully "repaired" Sony´s loaner camera with FW 1.03, as did many others. Go for it! If anything goes wrong - you can still send it in. You can´t loose anything ;-) cu, Gerald Last edited by Gerald Loidl; April 27th, 2008 at 02:30 AM. Reason: added info

April 27th, 2008, 02:42 AM	#3
Piotr Wozniacki Inner Circle Join Date: Aug 2006 Location: Poland Posts: 4,086	Thanks Gerald, I'll probably do as you say. After all, there have been no reports on any important changes that the 1.05 fw brings over the 1.03... Being lazy though, I'll probably just display the Siemens star on my 50" plasma, instead of printing 12 A4 cards, as you did. As far as I know, this is how Adam Wilt did his adjustment. My only reservation is related to the fact the the plasma is tilted down some 15 deg, so I'll have to take special care in order for my EX1 to look square at it. Do you think it'd work? __________________ Sony PXW-FS7 \| DaVinci Resolve Studio; Magix Vegas Pro; i7-5960X CPU; 64 GB RAM; 2x GTX 1080 8GB GPU; Decklink 4K Extreme 12G; 4x 3TB WD Black in RAID 0; 1TB M.2 NVMe cache drive

April 27th, 2008, 09:19 AM	#10
George Kroonder Major Player Join Date: Jan 2008 Location: Utrecht, NL \| Europe 3rd Rock from the Sun Posts: 612	I'm not a big proponent of using a display as a BF target, although I'm not saying it won't work. Using a dark room with essentially a backlit target can be effective when an "uncluttered" wideangle view is not available to prevent ambiguous autofocus. Basically when you display a siemens star target on a display, you're testing that displays resolution. This is also the case when you print a chart, only you're testing the printer. However a good laserprinter has a much higher resolution then your screen. It is worse when you use scanned charts and print those, or use inkjet printers (with bad paper). These will, in my humble opinion, not produce adequate target material. A low quality/low resolution target will require you to put it further away from the lens to 'compensate' for the poor representation. Conversely you can shoot a good quality target from closer up. I've attached two screenshots, one of a target I Googled, another of a 'digital original' both at 6400% (max. Acrobat zoom). You can clearly see the difference between a scanned copy and an original. If you use a good quality target, you should be able to put that 2-3m away from the lens and get good results. You can test if the autofocus has trouble focussing on the target by using PUSH AF (in AF manual mode) when wide. If it focuses ok, it will do ok in the BF adustment. You can use the same setup to check for BF problems. If you can attach a big(ger) high-res monitor to the camera output that would be a boon. Best to check focus at full res. Same setup used for manual BF adjustment (not on the EX1 as it has none). Using a printingpress properly printed original has my preference, but I've made a few derivative originals (not scanned) I print off on my 1200dpi laser printer up to A3 with good result. George/ Attached Thumbnails Last edited by George Kroonder; April 27th, 2008 at 10:28 AM.

April 27th, 2008, 10:19 AM	#14
George Kroonder Major Player Join Date: Jan 2008 Location: Utrecht, NL \| Europe 3rd Rock from the Sun Posts: 612	Hi Eric, True, I've asked myself the same questions (and you're probably right). But I don't know the inner workings of the FB ajust function. Adam Wilt was adamant about using a siemens star target and I believe it is just good practice. For all I know it detects interferance from the chart. Somewhere there exisits a Sony directive on how the FB Adjust rig/setup needs to be. I'd love to have that clarity. Until then I just stick to what I know to work, regardless of camera and lens. George/ P.S. Steven, as the 'scene' is static I don't think the mode matters and the routine may just override anything anyways. Last edited by George Kroonder; April 27th, 2008 at 10:22 AM. Reason: K>C, add PS

April 27th, 2008, 02:51 AM	#4
Gerald Loidl Regular Crew Join Date: Feb 2008 Location: Austria Posts: 122	Never tried it with a screen - but it worked for Adam. Make sure that no other things are in the view of the camera which could irritate the automatic setup procedure. I think the most important thing for the software is that it is a high contrast image with sharp edges it can focus on. I think Adam did the setup in a completely darkened room with only the light of the monitor. Good luck - let us know!

April 27th, 2008, 03:43 AM	#6
Bruce Rawlings Major Player Join Date: Feb 2008 Location: Cardiff, Wales, UK Posts: 410	I used the 'Gerald' 12 x A4 system and it worked pefectly. It was re assuring to see the auto system lock on to the wide shot and focussed the charts. I think the 1 metre test is interesting as I have a 50" Pioneer plasma and an 730s HDCAM with £10k lenses set to BBC lineup. The EX1 is looking very good when comparing the two but wide angles do highlight that you do get something extra for your HDCAM money!

April 27th, 2008, 07:56 AM	#8
Benjamin Eckstein Major Player Join Date: Jan 2007 Location: Boston, Massachusetts Posts: 380 Images: 1	I tried it on a 50" plasma a while back with a Siemens Star but it was hard to do for all three ND settings. I cannot remember exactly what happened, but I think it had a hard time figuring out focus from the screen and it ended up getting stuck halfway through the adjustment and essentially crashed. I had to turn off the camera and turn it on again (which freaked me out that it was mid adjustment, although nothing bad happened). Anyways, maybe it will work for you. B

April 27th, 2008, 09:33 AM	#11
Sean Donnelly Regular Crew Join Date: Jan 2008 Location: New York, NY Posts: 188	I agree. A decent photo printer (epson R2400,etc.) has an effective resolution of 360 pixels per inch. A 50" 1920x1080 plasma still only has as many pixels as a 24" 1920x1080 display, they're just larger. A 50" display is roughly 45" in the horizontal direction, giving you a resolution of 42.667 ppi. Less than 1/6th of a chart printed on a home photo printer. For black and white graphics and text, a decent laser printer will resolve approximately 600ppi. The Putora sharpness indicator is my personal favorite, which uses a very high resolution pattern specifically designed to cause interference patterns on a digital image making it VERY easy to see focus.

April 27th, 2008, 09:47 AM	#12
Steven Thomas Trustee Join Date: Sep 2005 Location: Gilbert, AZ Posts: 1,896	Thanks for sharing your thoughts George. You made an interesting point about verifying using "push" auto focus before running the test. Also, having said that, I wonder if it may be smart to put the camera in 1080 60i (or PAL 50i) mode before performing the BF calibration. Auto focus work a lot better in interlace mode since the sensor is updated twice as fast.

April 27th, 2008, 10:01 AM	#13
Eric Pascarelli Major Player Join Date: Nov 2007 Location: Los Angeles, CA Posts: 498	I agree about the printers, monitors and such, but I don't think an autofocus routine (and hence the EX1 backfocus routine) really needs a good Siemens chart to do its thing. The Siemens "focus bullet" chart is great for the human eye to see perfect focus (especially in an optical system) because it "snaps" into focus nicely. But does the EX1, which looks for contrasty edges, really need this? The very sharpest parts of the Siemens chart occupy such a small area of the frame that I doubt the camera even acknowledges them for its focus routine (especially at full wide at 3 meters, A4 sized chart). When I first got the instructions (from a Sony dealer in Canada) for doing the Auto FB adjustment, no mention was specifically made of a Siemens chart - just a backfocus chart of some undefined sort (perhaps Sony has a proprietary one?). Anyway, a large chart with smaller features at the center and larger features at the edges would seem to be the best thing. A room sized Siemens chart might fit the bill. I haven't tried this, but I bet some crisp black tape on a white background would work just as well or better than the Siemens focus bullet. A clean tape edge has a very high "resolution" (the sharp transition between black and white) as far as autofocus is concerned. For my next Auto FB, I'll put a focus bullet in the middle and some larger, crisp black features around it. And I'm sure a plasma will work fine as well. Regardless of its resolution for displaying a Siemens chart, there's probably enough there for the camera to focus on - the pixels themselves in closeup and the edges of the screen at full wide.

April 27th, 2008, 02:09 PM	#15
Sean Donnelly Regular Crew Join Date: Jan 2008 Location: New York, NY Posts: 188	That's true Eric, the auto FB doesn't neccesarily benefit from the added resolution. I've had very good luck doing the adjustment off of a piece of letterhead with printed text on it taped to a blank wall. All that the autofocus does is look for edge contrast, and I'm guessing because of the nature of the CMOS sensors it only looks in the horizontal direction (just like the Peaking function). I don't know if it would actually work with the auto function, but this is the chart we normally use to judge sharpness:http://www.ascmag.com/store/product....cat=333&page=1 I think what Sony is using must be something like the Digiprime backfocus tool. it is essentially a backlit siemens star behind an optical element. It works very well.