Equipment display differences that make no sense to me

[Ed Overstreet]

Here’s another puzzle for the technical folks on this forum (at least, those who are a lot more knowledgeable about things technical than I am).

I have a PTE show (created it in 5.1 but I don’t think that matters) that opens with a left-to-right pan through a rather large panorama stitch that I created in Photoshop from about six overlapped 10 megapixel RAW images. The file is rather large, at 4832 by 1050 pixels and a JPG size of 1145 kb. Yes, I know this is pushing the limits. But — everything that follows it in the show is a normal-sized JPG (1400x1050 is normal in my shows, generally 300-400 kb at the compression ratios I use).

Here’s the puzzle. At home, when I run this show on my 4-year-old laptop onto a 9-year-old CRT monitor at 1600x1200 res, the pan is silky-smooth in spite of the file size. At the photo club, on a much newer and generally more powerful desktop CPU powering a very good Canon 1400x1050 projector at that resolution, the pan looks awful. It’s jerky, a wavy line descends the screen from top-to-bottom as the pan progresses, and the image loses sharpness during the pan, though it regains sharpness when the pan ends. The rest of the show runs just fine, as it does at home.

My laptop has an ATI Raydeon card with 128 mb of memory; the club has an Nvidia Geoforce card, I think (anyway it isn’t ATI), with at least as much video memory as my card has. My system has 1 GB RAM and a 2 GHz clock speed, the club’s computer doubles the RAM and probably has a faster clock speed. Both systems use Windows XP. My system uses SP2; I'm not sure whether the club upgraded to SP3 or is still using SP2, if that would matter. (You'd hope if anything SP3 would produce better results than SP2, but that logic has, at times, escaped Microsoft, from what I infer from other comments about SP3). Neither system is networked; both are stand-alone systems.

So any ideas as to why this effect looks great on my laptop/CRT combo and looks pretty awful at the club on the better-desktop/projector combo, and whether there’s anything to be done at the club to improve the appearance of the pan?

I can’t use my laptop on the club’s projector, it would be much too disruptive to have to swap between computers just to run one 8-minute show during a 100-minute program with 18 other shows in it. None of the other shows have as ambitious a pan, though a few do use those frequent smallish pan-and-zooms, for which personally I rarely see any point, with no ill effects on the screen. Even if the logistics of switching computers for the projector were feasible (they aren’t), I wouldn’t want to do anything to jeopardize the other 18 shows. It’s not worth the risk. (I have successfully tested the other 18 shows on my laptop and monitor, as I'm the coordinator for the evening, but I haven't tried running them on the projector through my laptop, there won't be time, and there's no guarantee that other shows that run fine at home won't at the club if we use my laptop at the club, whatever happens with my own show.)

It is possible to get the panorama image to a smaller JPG file, by drastically increasing the compression ratio, down to about 450kb which might run more smoothly. However, this is a very detailed landscape with a lot of blue sky, and I have found that any JPG compression yielding a file size much less than 1.1mb has quite obvious and unacceptable (to me) banding in the sky. I can’t get the file size any smaller without compromising the image quality in the sky to a degree that I would find unacceptable. And the opening pan is absolutely essential to my show, I’m not giving it up.

If no one can think of a quick fix in terms of system settings or something else easily checked and done with the club’s computer or projector, I’m just going to run the show as-is and take my lumps, hoping that the audience, if it notices the jerky pan or cares about it, will have forgotten all about it by the end of the show, the rest of which is silky-smooth and (if I do say so, many audience members already have) is very impressive. But it really baffles me why an older, less-powerful computer running the same show on the same operating system (albeit on a CRT monitor) would perform better than a newer, more powerful desktop computer running a high-end digital projector, both at roughly the same screen resolution (1600x1200 for the CRT, 1400x1050 for the projector). The show’s images were all sized for 1400x1050, so if anything you’d think the CRT display would be more problematic because on that display the images have to be up-scaled on the fly, as I set PTE to run the show full-screen and 100% of the slide to show main images. On the projection system, there’s no need for image resizing during playback, since the images are exactly the same pixel dimensions as the projector setting (and the projector is running at its native resolution). Yet, as I say, on the projection system, that pan runs very poorly indeed.

We ran the test that showed the problem this Monday, "show time" for the club is December 2, and I'm out of town for five days next week, and I have no opportunity to get back onto the projection system before the 2nd, though I will have maybe half an hour before the crowd starts to arrive, but after setting up the equipment, to experiment with a limited number of possible fixes. Not asking for much, eh? But I thought I'd try asking anyway, sometimes someone suggests something seemingly tangential that solves a problem (e.g., the character-name-length business with my screensavers earlier this week). Nothing ventured, nothing gained.

[xahu34]

Hello Ed,

Is it a REALiS SX50 projector? I made similar comments in this forum, some time ago, concerning the SX50. This projector seems to react (in the way as you describe it) on (small?) changes of the frame rate, while monitors do not behave in the same way.

Best regards,

Xaver

Munich

[cjdnzl]

It's possible, as Xaver says, that the projector is the problem, but it could also be the graphics card in the desktop. Not all Nvidia cards are good for PTE when panning etc. I had an Nvidia 9250 card with 128 MB of ram in my 3.00 GHz desktop, and it stuttered on playback. I replaced it with a 6600GT which works well for PTE.

I would first check the card in the laptop. Try running the desktop into a monitor, to see if the condition exists with a monitor. If so, graphics card. If not, projector.

Colin

[xahu34]

Try running the desktop into a monitor, to see if the condition exists with a monitor. If so, graphics card. If not, projector.

Hi Colin,

I don't think that life is that easy. As long as the graphics card provides a constant frame rate (60 fps), everything will be ok, even with the Canon SX50. So it is very likely that the bad effects have their origin on the graphics card.

The SX50 sometimes mixes frames (upper part newer, lower part older), with the effect of a line (wave) running to the bottom until synchronization is ok again. With this particular projector, the said effect (while running long panoramas programmed with PTE and m.objects) occurred with notebooks, with a WinXP PC with a weak card (Nvidia 8500), but in the same way with a gaming PC (Vista) having two linked cards of type 9800. I have a short show which behaves nicely after a reboot, showing the effects on the SX50 during the second round. The positions where the effects occur actually vary during a looped presentation, maybe they depend on the internal resource management of the graphics card.

Montiors (and other projectors?) seem to show a different (better) behavior. When I use a big monitor with my PC (8500GT) everything remains smooth, but there are some sudden jerks, but no mixed images with running lines.

So, what I would recommend: If you plan a presentation, test the exact configuration (computer, projector/monitor) you want to use during the presentation.

Regards,

Xaver

[Conflow]

Xaver, Ed and Colin,

Xaver there is an 'Opto-Electronic Effect' called Cogging which is a derivitave of the well known

'Lissajous Pattern' effect which we use to determine phase-angle of various electrical signals.

You probably know about it...as you have touched on Image synchronisation in your Post.

This can be 'accidentally created' with any Pan & Zoom effect in Digital Photography when a

certain set of circumstances are accidentally set up, particularily with Image Panning.

The "attachment" below describes how it 'accidentally' happens ~ but at this stage I don't want

to go into the reasons why as it is quite complex.

Its more than likely what's happened with Ed's Image Panning...I also add that this may not be

clearly evident on a PC.Monitor but when the effected Image is expanded some 20x~30x it will

become very visible indeed.

Brian.Conflow.

[uuderzo]

When I went to theatre to present the show I attached the laptop to the theatre digital projector.

All pannings were awful and i feared the show would be a catastrophe.

Fortunately we carried our projector, the one we used to test the show.

After connecting the same laptop to our projector the show went silky smooth.

I don't know the cause, but I suppose that projector makes the difference.

Regards. Umberto.

[Ed Overstreet]

Thanks very much for the feedback.

Yes, sadly, the projector is indeed a Realis SX50. I wasn't aware there was this problem with that projector. Unfortunately changing the projector won't be an option at this date, for me in that venue. I guess I'll just have to grit my teeth and live with it. Doesn't sound like there's anything much I can do about it It certainly doesn't seem likely that wiring my laptop to that projector would help, even if that were an option that I'd be willing to try (which I'm not, as I mentioned, it would be too disruptive that evening).

Thanks for the diagram Brian. My pan is a horizontal traverse, which according to your diagram shouldn't cause "clogging," so I don't think that's the issue. It's probably the Canon.

Good suggestion Xaver about trying the show on that computer wired to a monitor, which we do have for it but don't use very often. There probably won't be time to check that before the presentation, and even if there were I don't think the club would have time or want to change the graphics card (if that turns out to be the issue, instead of the projector) before the presentation. But that's something we definitely should check out for future reference and possible action.

[xahu34]

Yes, sadly, the projector is indeed a Realis SX50.

I do not kow, but you may ask Canon if there exists a firmware update???

Regards,

Xaver

[Ed Overstreet]

I do not kow, but you may ask Canon if there exists a firmware update???

Regards,

Xaver

Another useful suggestion I checked Canon Canada's website just now; there's nothing listed on their Downloads page for any digital projector, but I've posted a query to them. If there is an update, I'll reply here to that effect and post the link, if there is one. Thanks!

[Conflow]

Ed,

O.K if its not due to Transv-Pan effects it could be Raw or JPeg Compression variations across the 6 Images taken.

Now I hear a 'howl' that Raw is not compressed ~ yes it is! ~ it undergoes primary compression within the Camera so that

all Image's-Data fall's within the measurement range of the Photo A/D Converter (Analog to Digital Converter) in the Camera.

(Its at that stage that the internal 'Digital-Exif' data is created ~ not to be confused with JPeg Exif data).

Now you wrote:-

"I have a PTE show that opens with a left-to-right pan through a rather large panorama stitch that I created in Photoshop

from some six overlapped 10 megapixel RAW images.

Ed, with respects, you are trying to create a 'Panorama' with no compression ratio correction and as far as I am aware

PTE was never designed to handle that type of 'Pan' job....it was designed to 'Pan across single Images' of known Format.

You already know when one makes a 'Panorama' it requires 'special stitching Software'...O.K Photoshop might do that job, but its

been my experience that this is one 'heck' of job even for a (stand-alone) Panorama Program and they cost a fortune because

they use special techniques to linearise variations in compression-ratios, contrast and brightness.

Any variations in the above mentioned between those 6 Images will never give a homogenuous Image no matter how much you

edit the final Image because the 'digital control signals' are all different ~ Been there done that and got burnt!!

Why not try out (1) Single Photograph ~ Pan that according to the PTE Instructions and then see what happens. At least you

will be able to arrive at a logical conclusion as to whether its the:- PC or Projector or the technique you used !

It makes sense to me....

Brian.Conflow.

Below is a snapshot one of the least-expensive Panorama-Programs $80

and a Link to its Website. Just 'Google' Panorama-Photography and

you will see what I mean.

LINK: http://www.panoramafactory.com/

[xahu34]

Brian,

I got the impression, that the German software "Panorama Studio" is the better and cheaper solution (www.tshsoft.de/en/index.html, 35 Euro, 43 Dollars). The Photomerge function in Photoshop CS3 also does a good job if you use the option "cylindrical".

Regarding the Canon SX50, you can leave all your theories aside. In my former photo-club, we tested this projector thoroughly. Any kind of pan (e.g. vertical pan of an upright image, horizontal pan of a stitched panorama, Lin Evans' Cliff Palace Show, ...) lasting 15-20 seconds or longer resulted in the said running line effect. The effect was independent of the movement's speed, at slow speeds it was less distracting but it was always there (at varying positions). During shorter animations, the effect appeared sporadically.

Regards,

Xaver

[Ed Overstreet]

Brian,

I got the impression, that the German software "Panorama Studio" is the better and cheaper solution (www.tshsoft.de/en/index.html, 35 Euro, 43 Dollars). The Photomerge function in Photoshop CS3 also does a good job if you use the option "cylindrical".

Regarding the Canon SX50, you can leave all your theories aside. In my former photo-club, we tested this projector thoroughly. Any kind of pan (e.g. vertical pan of an upright image, horizontal pan of a stitched panorama, Lin Evans' Cliff Palace Show, ...) lasting 15-20 seconds or longer resulted in the said running line effect. The effect was independent of the movement's speed, at slow speeds it was less distracting but it was always there (at varying positions). During shorter animations, the effect appeared sporadically.

Regards,

Xaver

The pan effect in my show lasts about 7 seconds. I do recall running this show some months ago on the same projector and don't recall seeing the effect, or as much of an effect, at that time, which is consistent with your last sentence about the effect being sporadic during shorter (than 15-20 seconds) animations. Essentially the same effect did occur twice last Monday, during two launches of the show about an hour apart (the first launch was just to set the sound level for the amplifier, when I noticed the pan problem; the second was during the program dry-run-through of my show and 18 others, ten of which ran before mine did). So on Monday night there was some consistency, but not over six months (or maybe it's my memory that isn't consistent, but I think I'd remember something like what I saw on Monday if it had happened in April). For whatever difference that makes, probably not much.

The pan stitch was in Photoshop CS3. I don't own a panorama camera, so that's how I do panorama stitches. The issue for me is what kind of proector or projection system will do what I want with my pans. Sounds to me like the answer is "something that isn't a Canon SX50." Unless of course there's a firmware fix out there for it, on which I'm still waiting a reply from Canon ...

I don't know whether any of this will make a difference to my club's next purchase decision, but I'm sure not going to be getting an SX50 for myself. If and when I decide to spring for a digital projector of my own, I now know which of my shows I'm going to test on it before I buy it That's useful for me.

[Conflow]

Xaver and Ed,

Xaver with respects, that Canon SX-50 has one of the best reputations with Commercial Business and Camera Clubs Worldwide.

'Technical Reviews' put it way ahead of other comparable machines perhaps your Camera Clubs' machine had an 'intermittent'

Hardware defect.

That particular problem of Ed's is rather unique, particularily the 'intermittency' problem and strangely enough a Google-Search

has revealed nothing, no Help Questions at all. If this is a 'desigh-fault' I doubt if Canon would be of much help.

Ed, you mentioned that this happened within a 'run' of 18 Shows - could you cast your memory backwards and by any chance is

this happening when the Projector is red hot ??....from memory, older Pentium Processors used to go 'Awol' when the Cooling Fan

became clogged...could this possibly be the problem with the SX-50?...its just a last ditch idea ?

Regards,

Brian.Conflow.

[xahu34]

Xaver with respects, that Canon SX-50 has one of the best reputations with Commercial Business and Camera Clubs Worldwide.

'Technical Reviews' put it way ahead of other comparable machines perhaps your Camera Clubs' machine had an 'intermittent'

Hardware defect.

Brian,

Regarding the presentation of photgraphs, the SX50 really is first class. On the other hand: I tested two examples, the club projector, and another one belonging to a club member, both purchased in January 2006. Both behaved in the same way with respect to PTE and m-objects presentations.

From a test at "www.projectorcentral.com": Canon Realis SX50, Evan Powell, April 15, 2005: ... Furthermore, with standard definition video material being fed through the DVI port we noticed what appeared to be an occasional loss of frame synchronization in the buffer, which manifested itself as a transient separation of the image along the horizontal plane (so for example, the upper half of a telephone pole would be momentarily shifted to the left of the bottom half). This artifact was limited to standard definition material on the DVI port; we did not see it occur on the VGA component input, nor did we see it with HD material on either port...

Regards,

Xaver

[cjdnzl]

I'm not sure if this is relevant to your problem, but it's worth mentioning. I did a peruse of Google for sync problems with the SX-50, and found this from a technical review of the SX-60, successor to the 50. Quote:

"Most importantly, the SX50 I tested more than a year ago (see svconline.com/mag/avinstall_canon_realis_sx) clearly had a preference for digital inputs. Thankfully, that is no longer the case with the SX60. There is no more of the noise and sync problems of the older model's analog inputs".

As you are almost certainly using analog video to the projector, that may well be your problem.

If the desktop machine has perchance a digital output, you might try that; or perhaps the purchase of a graphics card with digital output would do the trick.

Colin

[xahu34]

Hi Colin,

It is interesting to read your comment. My experience is that for animated slide shows (PTE and m.objects) the sync problems appeared on both inputs (analog and digital). So, there seems to be no hope for Ed.

Regards,

Xaver

[Ed Overstreet]

Xaver and Ed,

Ed, you mentioned that this happened within a 'run' of 18 Shows - could you cast your memory backwards and by any chance is

this happening when the Projector is red hot ??....from memory, older Pentium Processors used to go 'Awol' when the Cooling Fan

became clogged...could this possibly be the problem with the SX-50?...its just a last ditch idea ?

Regards,

Brian.Conflow.

Hi Brian. Thanks for the thought, but the heat isn't likely the issue. The first time I saw the problem on Monday was during the sound-level test, when the projector had been on only for a few minutes. The second time was after it had been on an hour or so, and the problem looked identical to me in both instances.

JRR (who is also in my club) has kindly agreed to meet with me over the holidays in that room with his Panasonic projector and his laptop, and I'll bring my laptop (which has no trouble at all running that pan on my monitor), and we'll test various configurations of computer/video card (including the desktop system that was used on Monday), projector and monitor with my show. That should help us isolate which hardware component(s) or combinations are the main factor. Will let you all know what we find out, it may not be for several weeks though. That room is hard to get between now and Christmas, it's in a big rec centre and it gets booked a lot for holiday office parties, so we may have to wait our turn.

I've bookmarked this thread in my browser so I'll post our findings in this thread, not start a new one.

[Ed Overstreet]

As promised in my last post two months ago, JRR and I have conducted some extensive tests on the problem described above, comparing the Canon SX-50 projector with Jim’s Panasonic PT-LB104 projector. The problem of the projector not displaying smoothly animations which the monitor does display smoothly, even using the same computer, is NOT unique to the Canon projector.

I have started a new thread on this forum providing some more information about what we found, and also providing a link to a 30-mb zip file that anyone can download to learn more of the details of our testing, as well as copies (both PTE and EXE) of the relevant test shows, for others to test on other projectors if they wish and are able to do so. You will find the new thread at http://www.picturestoexe.com/forums/index.php?showtopic=9354

I am cross-posting here mainly to direct those who might have bookmarked this thread, to the new thread. Please make all replies or queries to that new thread, not here, after reading my post on that new 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.

[JRR]

Colin:

Ed has asked that discussion on this thread be moved to here as we have done more investigating and it was felt to be a new topic

[cjdnzl]

Colin:

Ed has asked that discussion on this thread be moved to here as we have done more investigating and it was felt to be a new topic

Roger that, Jim. Post removed and put into current thread, thanks.

Colin

[fh1805]

After discussion with Ed via e-mails and an exchange of PMs, he has asked me to Lock this thread from any further posting. Posts #18, #19 and #20 all include active links to the new thread in which the discussion is continuing.

regards,

Peter