Difference in monitor vs projector displays

[Ed Overstreet]

This post is a follow-up to an earlier post which sparked some lengthy research by Jim Robertson (JRR) and myself; see the earlier post for the background:

http://www.picturestoexe.com/forums/index....st=0#entry59039

The original post was specific to one projector, the Canon SX50, in displaying a particular PTE show. However, we have now discovered that we encounter similar projection problems, and the same inconsistency between monitor and projector display, with a Panasonic PT-LB104 projector as we did with the SX50. The only difference is that the SX50 displays what can be described as a “wipe effect” during certain pan and zoom animations, while the Panasonic displays what can be described as “jitters” on the same animations. Both effects render the animation displays unacceptable. Our tests revealed the problem is not specific just to one particular panorama pan, nor unique to panorama or panorama-format images. Neither projector has difficulty displaying all other PTE 5.5 shows or animations we have tested on it, other than those animations that have the parameters we have tentatively identified as being problematic (see below).

We have created and run a test show to investigate these differences, and have tested it on four different computers (and hence four different video cards), both projectors, and several monitors, in most of the possible hardware combinations. The problem always occurs with both projectors, and never with the monitors we tested.

We have investigated and eliminated as possible causes about 20 factors, including all of those previously mentioned by other Forum members in the above-mentioned thread.

We have written PDF documents that describe the problem, how we conducted our tests, the specifications of the hardware we used, what we found, what we conclude and why we conclude it. These documents, along with two different versions of our test show (including the EXE file, the PTE file, and all the JPGs used in the test show) may be downloaded from the following link:

http://www.mediafire.com/?ymwmyurgvik

We conclude from our results that the projection problem is partly, if not entirely, related to panning and/or zooming outside the opening image, something which will necessarily always happen with a panorama-format image when displayed on a 4:3-format device and often even on a 16:9-format device. The percent of zoom used in opening the image to be panned or zoomed, and the speed of the pan or the zoom, also seem to be related to the projection problems in some cases.

The fact that we can’t get these animations to project smoothly, even though the same computer that ran the projector can display the animations smoothly on a monitor, is a considerable cause for concern for us. For one thing, it means that for such animations (and maybe for other animations we haven’t yet explored or discovered), we can’t trust the monitor display to warn us when something won’t look good on projection in front of an audience. Many producers don’t own projectors, but present at photo clubs or other venues, using equipment owned by the club or someone else. They cannot easily pre-test shows on a projector. Any inconsistency in animation smoothness between monitor display and projection, even when using the same computer and video card to run both devices at the same resolution, is both baffling and a significant concern (and embarrassment at times).

Please see the details included in the zip folder available at http://www.mediafire.com/?ymwmyurgvik if you wish to test our shows and findings.

At their request, I am also providing our test information to technical support at Canon Canada, with whom I had corresponded earlier concerning the problems we had encountered initially with the Canon SX50. If I learn anything from them that I think would be of interest to this Forum, I will report it in this thread.

I wish to stress that no fingers should be pointed specifically at Canon or at their SX-50 projector. The problem is broader than one model of projector.

Ed Overstreet

[JRR]

There might be a potential problem downloading the file Ed referred to in the above post from: http://www.mediafire.com/?ymwmyurgvik

Ed used the ZIP utility in WinXP and that retained two folders that had the test PTE shows in them. However, I use WinZip and that removes the folder separation in the file and threw everything in together in the ZIP file when I opened.

That might create problems when running the two show (1024x768 and 1400x1050) as the image files have the same name but are different sizes.

So to be safe, if anyone hits that problem with the download, we have broken the original zip into two files (both have the pdf files in them). One for the 1024x768 show and another for the 1400x1050 show.

They are available from:

1024x768

1400x1050

Sorry for any confusion

[Guest Yachtsman1]

Hi Ed

Tried your 1024x768 show. It isn't an exe file so I don't know how it was operated. The only critism I could level at it was the speed of the zooms and pans, much too fast. I saw no moire effect, jittering etc, the JPEG I checked however was a big one, 1261kb much larger that what would normally be used in PTE. Maybe it's the layers in PS increasing the file size??? I also ran it on my Nobo X17 xga projector, showed fine on that. That's my threepenneth

Yachtsman1

[xahu34]

... Maybe it's the layers in PS increasing the file size??? ...

Hi Yachtsman1,

If you have a layered image in PS and export (save) a copy in JPEG, the copy will automatically be flattened. So, the layered structure of the original will not have any influence on the copy's file size.

Best regards,

Xaver

[Guest Yachtsman1]

Hi Yachtsman1,

If you have a layered image in PS and export (save) a copy in JPEG, the copy will automatically be flattened. So, the layered structure of the original will not have any influence on the copy's file size.

Best regards,

Xaver

That's not my point, single large files have problems on some systems, particularly when animated, (panning & zooming). 1261kb for one image is 10 x more than needed.

Yachtsman1

[JRR]

Yachtsman:

Thanks for running a test.

Yes some image files are larger than recommended, but our point is that ALL the transitions/animations run smoothly on a monitor, but the projectors that we tested had difficulty with some. Especially with panning outside the opening image portion on screen.

[Guest Yachtsman1]

Yachtsman:

Thanks for running a test.

Yes some image files are larger than recommended, but our point is that ALL the transitions/animations run smoothly on a monitor, but the projectors that we tested had difficulty with some. Especially with panning outside the opening image portion on screen.

Hi Jim

Have you tried sizing your images at a maximum of 150ppi, There is another thread running about projectors I suppose you've read it?

Yachtsman1

Just thought of another possibility, where are you getting your projector signal from, VGA or Serial port?

[Lin Evans]

Hi Eric,

The dpi (ppi) is really immaterial. This is simply an EXIF header which tells a print device the print density intended for print. It really has zero effect on display images whether that be projector or screen.

You could have the dpi set at 10 dpi or 1200 dpi and the image will be absolutely the same. Image size is determined by the number of pixels in the horizontal and vertical aspect. This seems to be one of the least understood issues in imagery.

To demonstrate this, I'm going to post two images which you can take into photoshop and test for yourself. One will have the dpi set at 10 dpi and the other at 1200 dpi. You will see absolutely zero difference. The difference in photoshop is only how large the prints would be "if" you used the EXIF header to determine this. The image quality and display size are identical.

First the 10 dpi (ppi) image:

http://www.learntomakeslideshows.net/sampl...friend10dpi.jpg

Next the 1200 dpi (ppi) image:

http://www.learntomakeslideshows.net/sampl...iend1200dpi.jpg

Best regards,

Lin

[Ed Overstreet]

Just to reinforce Jim's point -- yes the panorama file is larger than what is normally recommended, but if one is to use a panorama stitch that fills the screen top to bottom and therefore extends beyond the boundaries of the image frame in the O&A window, that is inevitable. And, as Jim points out and as we demonstrate in the test show and report in our findings, our monitors have no trouble at all running the animations not only on the 1.2 mb pan stitch but also on Jim's 3mb Lions B shot. Please take a close look at these in the documentation, and try them on your monitor, and I think you'll see what we mean. The issue is NOT the file size, either in kb/mb nor in pixel dimensions, at least not in our tests. The issue is that neither of the two projectors we test can do what the our monitors can do, on the same computers we tested.

The projector was attached to the computer, in every case, using the same connection port that one uses to attach a monitor to the computer, don't know what it's called, sorry. It's blue in colour on most systems. So the projector is getting its signal from the same port as does the external monitor.

Also, with reference to Yachtsman's query, the show is in fact an EXE show, please look again in the zip folder and subfolders. We included both the EXE show and the PTE and JPG files, the latter so people could go into the show itself and see exactly what we did, if they want.

Some of the pans are perhaps fast to some taste (though a 7-second pan is what the music called for in my original show, so that's what it was), but one of them is a 22-second pan, which I think is plenty slow enough, and as we report in the results, the 22-second pan didn't look any smoother on projection than did the 7-second pan -- though both looked very smooth on all our monitor tests. The speed of the pan is NOT the issue either, at least not in the panorama-stitch file, except perhaps in slides 5 and 6, especially slide 5 where the pan was 1.5 seconds just to test for the effect.

As others have mentioned above, layers are irrelevant, all these files are JPGs and hence have no layers in them, though the original PSD files from which the pan stitch was saved obviously had layers. That is also irrelevant to the results; we encounter exactly the same pattern of results if we take a single image and produce a panorama crop from it and pan through that. Whether layers ever formed any part of the image at any stage of its creation is irrelevant to the results.

[fh1805]

Hi Ed,

I've just downloaded your 1024x768 sample. I'll try it through my laptop/projector combination this weekend if I get time. Just a thought, though, arising from reading the "README.FIRST" document.

You state that the problem seems to be connected with images that are subject to a pan that proceeds beyond the edge of the initially displayed image. What happens if you place a 1024x768 "windowed mount" (with, say, an 800x600 hole) as an extra object sitting above everything else in the panned image? Does that eliminate the wipe and jitters? By windowed mount I mean the digital equivalent of the card with a rectangle cut out of it that a print photographer would use to display their print image.

If the problem is resolved then it does look like it is a consequence of panning an image that has extreme edge pixels and pixels from beyond the extreme edge participating in the pan. If the problem remains then I'm not sure what we will have proved (if anything).

regards,

Peter

[Ed Overstreet]

Thanks for the suggestion, Peter. I don't have access to either projector myself, and access to the SX50 is very difficult to schedule. However I'll pass this on to Jim and see if he can try that on his projector (Jim, I think I understand what Peter is suggesting and will create a short show with Slide 3 and the added object super-imposed throughout the animation, maybe also do it with Slide 4 and the slower pan on the same image).

A couple of other minor technical points that Jim and I neglected to mention in the documentation (because it didn't occur to us), that Canon Canada has asked me about (they are still looking at the files but were very prompt in getting back to me, quite impressive I must say).

All of our tests involved running the EXE file off the relevant computer's hard drive; no we did not run these tests from a CD, a DVD, or a USB device.

In all tests the projector was connected to the computer using the cable supplied by the manufacturer with the projector, which if I recall correctly was maybe one metre long or perhaps shorter. We tested the Canon projector with its cable and the Panasonic with its cable. The cables were connected to the 15-pin video port (that's what Dell says the thing is called, I looked it up once I found my manual) on the computer and the corresponding port on the projector. No we did not test with any other cables, but I very much doubt the problem would be the cables we were using. 1) Both projectors have a problem, and they aren't using the same cables (and these are after all the cables the mfr ships with the projectors). 2) Neither projector has any problem smoothly displaying any other animations we've thrown at it, except the ones illustrated in our test shows. 3) In my experience, when a cable fails, the failure is a lot more catastrophic and a lot less subtle (relatively) than what we're seeing. Thought I should mention these points here, in case these questions occur to anyone following this thread.

Jim has sent a separate query to Panasonic tech support, with links to this forum thread and to our zip folder. As far as I know he hasn't heard back from Panasonic yet.

[cjdnzl]

Ed, I wonder if timebase differences could be a factor, e.g. the frame repetition rate from the computer is different to the frame rate of the projector.

A way to test this would be to use a composite video cable to feed the projector. I think using composite video would produce the computer output at TV frequencies which should be fully accepted by the projector. The video will be lousy at TV definition, but it might give you a pointer as to what's going on. That's of course if your laptop has a CV outlet -a small circular socket. My Dell lappy has this.

Colin

[Picsel]

Hello Ed

When you use your projector, is your video card manager set for "laptop + projector" or "projector alone"? If the projector resolution is larger than laptop one you should choose "projector alone" and set the resolution for the projector, otherwise you may encountered some video problem. Example if my laptop is 1024x768 and projector is 1280x720 I must choose "projector alone" and set the resolution display to 1280x720. The same happens with 1440x900 laptop and 1920x1080 LCD HDTV, etc, etc..And in any case you must set the resolution display for the projector and not for the laptop. But I am sure you already knew that, just in case of...

Daniel.

[Ed Overstreet]

Thanks for the replies and further suggestions.

Jim will be testing Peter's suggestion on his projector some time this weekend; I'll report what he finds on this forum. The frame makes no difference on my CRT monitor and laptop (the animation is smooth as it always was), but that's no surprise. The key is what will happen on projection.

It is not easy for Jim and me to test the Canon projector, because it has to be used at the rec centre and we have to book a room at a time that is convenient for both of us and at which rooms aren't already booked for other purposes, which is not easy. Some time in the next week or so we will re-test a few things on the Canon projector, but for now will have to limit ourselves to working with the Panasonic. We'll save up the suggestions already made and any more than come to this forum in the next few days that seem to be worth pursuing, then test them all at once on the Canon. Canon wants us to test a couple of ridiculously large files that probably were taken with Canon's new full-frame camera and haven't been resized; they're 5600+ pixels on the long dimension and one is 13 mb which even my trusty laptop and CRT monitor choke on, but we're also testing some more reasonable downsizings of their test files (1024 pixels on the long dimension file sizes around 400-600 kb). They gave us one portrait-format and one landscape-format image to play with, and we're panning them vertically in, through and out of the frame (similar to the animation on slide 17 of the posted test show, only with a vertical pan instead of a horizonal pan).

We've never tried testing with what I think Dell Canada calls an S-video cable connection to the projector. My laptop has such a port, but Jim's doesn't and he doesn't have a cable or connection of that sort for his projector. I don't know whether the club Canon and desktop have the appropriate connector ports and cables; I don't think we've ever used that sort of connection (whether CV or S-video or whatever one calls it). If we find a port and cable of that type that we can test with the Canon whenever we get to it next, we'll try that.

The question about the Display Properties setting for "projector only" vs "laptop plus projector" is something we were careful about; the problem isn't from that setting, as far as I can tell and remember.

[Guest Yachtsman1]

Just had a quick look on canon uk site, the sx50 isn't listed, so I checked the nearest which is the sx60 and appears to be a home cinema projector with the facility to project analogue sxga & xga signals, however the analogue signal spec is lower than my dedicated xga spec. I've copied a few of the technicalities for the experts.

Yachtsman1

http://www.canon.co.uk/For_Home/Product_Fi...dex.asp?specs=1

http://www.brochures.canon-europe.com/pdfs...-1222871566.pdf

Cinema at home

The dedicated Home Cinema mode optimises the projector settings for watching films in your living room. Contrast is boosted to 2000:1 for incredibly rich colours, to provide maximum viewing enjoyment. The double frame-rate drive of the XEED SX60 LCOS panels ensures smooth reproduction of moving images, without any annoying flicker.

High definition and advanced connectivity

The SXGA+ resolution also ensures faithful reproduction of High Definition sources (720p/1080i). A connection via DVI (Digital Visual Interface) provides optimum image quality from digital sources, including HD images. The XEED SX60 supports the HDCP standard, for high quality playback of encrypted digital content.

Native Resolution

1400 x 1050 (SXGA+), 1470000 pixels

Digital RGB Compatibility

D-SXGA+ / D-SXGA / D-SXGA / D-SVGA / D-VGA

Analogue RGB Compatibility

UXGA / SXGA+ / SXGA / XGA / SVGA / VGA

HD / Component Scan Systems

1080i / 1035i / 720p / 575p / 575i / 480p / 480i

Standard, Presentation, Movie, Home Cinema, sRGB

Standard, Presentation, Movie, Home Cinema, sRGB

[TonyFalla]

I think Colin is making a good point about Frame Rate, it is something the TV manufacturers make a big thing about. I notice that Sony say that some of their latest TV's have a frame rate of 200. I find that information for this is difficult to find for Monitors and Projectors. I suspect however that a Monitor will have a high frame rate like a TV, it may be that a projector has a much slower frame rate so if your monitor changes the picture lets say 75 times in 1 sec, but the projector only does it 25 times in 1 sec, then movement shown via the projector is not going to be as smooth as the monitor

Tony

[Guest Yachtsman1]

Had another look through this thread and in my opinion it would appear both projectors mentioned are dedicated home cinema units. I just downloaded this statement from Projector Points web site

"Most projectors are capable of taking inputs from both a PC and video sources, such as a DVD player or Set top box. However, most are also optimised either for PC use or for home cinema. Very few truly excel at both. "

The Panasonic PT-LB104 mentioned is not listed on their UK web site, googled gave me a series of oriental links.

Yachtsman1

[fh1805]

Eric,

Not sure what point you are trying to make...

If you are suggesting that the problem stems from these projectors being "home cinema" projectors then, surely, they ought to be capable of projecting "moving" images extremely well, shouldn't they?

If this is a factor in the problems seen by Ed and Jim then perhaps the "frame rate" or "refresh rate" is going to be a significant factor. I've become involved in the problem diagnosis activity with Ed and Jim. Once we have something new and certain to report, we'll be back. For now, we continue to test, analyse, test again, analyse again, etc.

regards,

Peter

[Guest Yachtsman1]

Eric,

Not sure what point you are trying to make...

If you are suggesting that the problem stems from these projectors being "home cinema" projectors then, surely, they ought to be capable of projecting "moving" images extremely well, shouldn't they?

If this is a factor in the problems seen by Ed and Jim then perhaps the "frame rate" or "refresh rate" is going to be a significant factor. I've become involved in the problem diagnosis activity with Ed and Jim. Once we have something new and certain to report, we'll be back. For now, we continue to test, analyse, test again, analyse again, etc.

regards,

Peter

Hi Peter

You know me, when I see a problem I like to arrive at a solution. Just been watching a film on TV and during the commercial I thought of a simple test to my theory.

Make a DVD of the problem sequences, then when they get access to the projectors again, connect a dvd player as opposed to the laptop, to the projector and run the sequence. If it is still the same that blows my theory, I think. If you read the specs of the projectors mentioned non cinema portion is way below mine, which is nothing special EG Contrast, & Lumens.

Regards Eric.

[fh1805]

Eric,

Your suggestion of burning to DVD and connecting a DVD player to the projector is, for me, a non-starter at this stage. It introduces far too many new variables: VideoBuilder software or whatever to create the AVI or MPEG file, the DVD player, the cables, etc.

I haven't finished my part of the testing yet but, to give a sneak preview: I, too, get problems on all the same slides as does both Ed and Jim when I play the exe on either my Acer laptop connected to my Dell projector or on my old Fujitsu laptop connected to the same Dell projector. But, whilst testing last night, I found a setting on the Fujitsu that let's me "turn off" the laptop's monitor. When I did this, all the problems vanished and the exe projected perfectly. Tonight I need to confirm that this result is repeatable and I also need to find the equivalent setting on the Acer to see if it also resolves the problem on that system.

Watch this space, as they say!

regards,

Peter

[Ken Cox]

ERIC

that is the obvious test to run - when i was first making exe's to dvd i would take the test dvd's to a av shop and play on their dvd/beemr and project on a 100" screen in their viewing room

Al Robinson would do it both ways at home and hawk would take dvd to his daughters hd ws /dvd player and test there

you use the computer to make the show but take it out of the mix to eliminate the idisycrosies of the computor

ken

[Ed Overstreet]

Latest update from our end:

Jim tested the frame effect that Peter suggested, on his laptop and Panasonic projector. The frame made no difference to the problem we're seeing. The panorama continues to display badly behind the frame, while the frame itself remains stable on the screen and projects well throughout the animation.

To clarify my earlier comment -- on all our tests with our two laptops, only the projector display was turned on, the laptops' integrated LCD monitors were turned off during the tests. With the tests involving the club's desktop (which showed the same problems as both laptops, with both projectors), only the projector was connected to the 15-pin video port, no monitor was ever connected to the desktop during the tests.

Jim ran pan animations through the images provided to us by Canon, with the same results that we've been seeing all along (except that, as reported earlier, the very large original files generally displayed badly on monitors as well as on the projector -- the versions of the images that I down-sized displayed well on monitors but not on projection, in a pan-through).

Given the reality of how we project images at our club, and the limited tolerance of many of our club members (certainly including me) for the complexities that can involve, the DVD-as-an-intermediate-output option is a non-starter from our standpoint. However anyone on this forum who wants to test that option is welcome to download the files at the links given above, have a try, and let us know what they found.

[Guest Yachtsman1]

Hi Ed

I think you are missing my point, I'm suggesting the projectors in question are designed specifically for home cinema, projecting TV/Film images, not photographic AV work. However it's up to you. Looks like your tests have eliminated the leaving off of the laptop monitor. So where next? My recent expeiences with Panasonics support for my 32in LCD TV were abysmal, hope you get on better.

Regards Eric

(Yachtsman1)

[Barry Beckham]

Ed.

I have not read all the replies to your original post as they have built up rather quickly, but I discovered what you have spoken about a couple of years ago and it does effect what PTE slide shows I project when doing a demo.

I use the same desktop PC for both the creation of the slide show and then to demo those shows to camera clubs, Its a long story why I do this, but it suits the way I work and I find laptops inadequate unless you spend your entire whealth on one. The PC I use has respectible enough power to create slide shows and handle my Raw files through Photoshop.

It will play any slide show including animated ones perfectly on the monitor I use.

However, take that same PC to a demo and hook it up through a Digital Projector (without a monitor, I demo from the projector image) and now some, but not all of the animation can show a jerkyness that just destroys things for me.

I always assumed it was the limitaions of projectors and in time they will improve. I can't really understand why a show displays perfectly on the monitor, but not so perfectly through a projector. I suspect the techno boffins will have some ideas, but to be honest I am not interested unless there is something I can do to get a perfect result through my projector.

I am fairly experienced in putting together slide shows and what I do is keep my images to a minimum size, so its not anything like that which is causing the problem. I also notice this on both my projectors. One Epson and one Mitsubishi and I consider both these good projectors aside from this issue.

[fh1805]

I have completed my tests and reported the details back to Ed and Jim via e-mail. This is a summary of my findings.

I ran the tests using my Dell MP3300 digital projector (native resolution = 1024x768) connected to:

- My Acer TravelMate 5510 laptop, native screen resolution = 1280x800 (the system automatically resets this resolution to 1024x768 when it detects the projector)

- My Fujitsu-Siemens Amilo laptop, native screen resolution = 1024x768

The Acer has the ATi Mobility Radeon X1300 graphics with 128MB of native memory and 256MB of shared system memory. The Fujitsu-Siemens has the Intel 82852/82855 GM/GME Graphics Controller with 64MB of total memory. (Details taken from Dxdiag reports)

All testing was done using sequences stored on a USB memory stick.

(N.B. The Acer and the projector are used to take my own sequences out for audiences. They have never shown any sign of being unable to handle anything that I throw at them).

I started with the Acer. Ed’s sequence displayed OK on the laptop’s LCD monitor. But when projected there was a significant amount of “visual interference”. You know how you get a line of “snow” horizontally across the TV screen when you play a VCR tape that isn’t tracking quite right? Well that’s what the interference looked like: a line of “snow” that came in from the bottom of the screen and moved slowly up to about one fifth way up and then vanished. It was most easily observed well over towards the left hand side of the images but it did extend across the full width.

I disconnected the Acer and hooked up the Fujitsu laptop. Same result. The sequence played fine on the monitor but gave the same “snow” when projected. Having now seen the problem several times I realized there was a timing pattern involved. Each line of ”snow” appeared at approximately one second intervals (I had no means of doing accurate timings but, as an experience AV worker, I reckon I can count seconds just as accurately as any NFL quarterback!). Quite what this timing pattern tells us, I have no idea!

Up until this point I had been running both laptops with their screens active at the same time as the projector was active. This meant that I could look from one to the other to check whether the interference was present on both. When it was present on the projected image, I could and did also see it on the monitor image at the same time. When the projector was disconnected then the problem did not appear on the monitor.

I used the appropriate Fn key to turn off the monitor on the Acer laptop: the interference was still visible on the projected image. I used the appropriate Fn key on the Fujitsu to turn the monitor off: the interference vanished! All pans and zooms now played perfectly!

I repeated all these tests the following day and got exactly the same results.

As a result of my tests I’m prepared to say that the fault is not in PTE code, is not in the PTE project setup and is not in the projector. These factors were common to all tests and in one specific arrangement, the sequence ran cleanly. The only component to be changed was the PC.

Therefore, in my opinion, the problem is caused by the computer and is not attributable to either the projector or the exe file.

In my opinion, the root cause lies somewhere in the computer hardware/software (excluding PTE) and most probably in the graphics card or its drivers.

regards,

Peter