Hi,

	I'm not sure this is a bug at all, much less a PovWin bug. (Maybe it's
just a setting <G>:

With a fairly large image 3000x3000 I'm getting a blank white render
window with the red message... StretchDIBits () Failed! (Windows Error).

The image is rendering to file, so no data is being lost; it's just a
display problem.

Yes, I admitt I usually have several applications running, and I am
currently starting this rendering from within Moray.

I have a Dell 400mhz PII with 256 ram. The video card is the Permedia 2
chip Diamond fire gl pro with 8 meg set to 1152 by 864 24bit.

I shouldn't be up against memory or even video memory limits, and I
certainly have other applications as resource intensive as POV-ray.

So far PovWin is the only application doing this.

I can't be sure, but I think I rendered images this big with earlier
versions of POV-ray and I never encountered this error.

Thanks.

Kevin Odhner
jko### [at] homecom

Post a reply to this message

>I can't be sure, but I think I rendered images this big with earlier
>versions of POV-ray and I never encountered this error.

This happens when you have a scene that maxes out your memory, AND you try to
display a large image. The windows API function call that we use to draw to the
render window (called StretchDIBits) fails in an unusual manner (it doesn't
return an error code, but indicates that it copied a total of 0 lines).

As far as we can tell there's no workaround, so we display the message you see.

Post a reply to this message

Well, I restarted my computer, and that fixed it for now. I think if
it's memory then it's part of another potential bug in PovWin 3.1 that
Lutz already reported.

I have 256 meg of RAM and the scene just isn't big at all. My system has
been frequently getting semi-paralyzed when using Moray and the new
PovWin.  Everything's fine for a while, and then Povray won't start.  I
go into task manager and there is "Povray (Not Responding)" needing to
be shut down.  After that Moray always becomes super slow with every
action delayed by 10 seconds or so. (I have a peppy 400mhz PII) and I am
lucky to be able to save and close. Sometimes restarting Moray seems to
clean everything up. Other times, I have to restart.

If this problem is tying up my memory, then maybe it is also causing the
StretchIDBits problem as well.

Lutz said he was waiting for the PovWin Source to take a look.

povray.org admin team wrote:
> 
> >I can't be sure, but I think I rendered images this big with earlier
> >versions of POV-ray and I never encountered this error.
> 
> This happens when you have a scene that maxes out your memory, AND you try to
> display a large image. The windows API function call that we use to draw to the
> render window (called StretchDIBits) fails in an unusual manner (it doesn't
> return an error code, but indicates that it copied a total of 0 lines).
> 
> As far as we can tell there's no workaround, so we display the message you see.

Post a reply to this message

jko wrote in message <361AB131.5EA35A66@home.com>...
>Well, I restarted my computer, and that fixed it for now. I think if
>it's memory then it's part of another potential bug in PovWin 3.1
that
>Lutz already reported.



I've gotten that fairly frequently, if I have more than 2-3 windows
running. Sometimes it recovers, sometimes I have to exit POV and come
back in. I thought it was a video card conflict, so I didn't mention
it previously (have not had it prior to 3.1, but all prev versions
were run on a different card too)

My card is an ATI All in Wonder Pro with 8 meg ram.

Post a reply to this message

povray.org admin team wrote in message
<3619f400.291602452@news.povray.org>...
>This happens when you have a scene that maxes out your memory, AND you try
to
>display a large image. The windows API function call that we use to draw to
the
>render window (called StretchDIBits) fails in an unusual manner (it doesn't
>return an error code, but indicates that it copied a total of 0 lines).
>
>As far as we can tell there's no workaround, so we display the message you
see.

I got this message when I tried to maximize a 30x5 render window (I was
rendering a df3 layer). It was not a memory problem, I am sure, since it
happens any time I try to stretch such a small render window. Rather, I
think that the StretchDIBits() function has some limitations, and as you
mention, it is kind of buggy, too. The best "workaround" is already
implemented -- POV indicates that this is a Windows error. Maybe you should
make the "Windows error!" part flash :)

Seriously, maybe if POV itself performs the zooming,, then it can do a
bilinear or bicubic interpolation as well, of course only if specified. Just
an idea.

Peter

Post a reply to this message