Printable View
I just was wondering
I read a tech tip from I believe it was Tim Grivas
About starting image content development by setting your resolution to 1024 x 768 because that is what the monitors are set at.
But what do you set the resolution at?
I know that the web is 72 dpi and printing is 300 dpi. What should I set the resolution at for Video?
what do you mean video?Quote:
Originally Posted by Urb
you are outputting at final completed movie at standard definition?
Well, maybe I asked this wrong
I was just wondering when I create an image file in photoshop what should I set the resolution at.
Width: ?
Height: ?
Resolution: ?
Mode: ?
For Example if I wanted to make some white text on a black background
Thanks
there are several technical discussions in this apparently simple question, and there is some historical and performance detail:
digital video on a computer has always always had to trade image resolution and quality against performance.
where audio playback reached close to perfection 10 years ago, the demands of doing video on computers, have not yet reached the same degree of ease.
for example doing uncompressed 16bit stereo audio takes 600MB/hour, doing the same in uncompressed video takes 60GB/hour at standard definition.
Doing video is 100 times more intensive than doing audio.
and at higher resolution than 720x480 the demands are still largely too great for most systems.
What has happened over the last 10 years, in digital video, is that the various subsystems required to do video have slowly increased in capacity - but not all at the same time, or at the same speed.
Disc speed, computer speed, internal architecture speed, graphics speed, all play a role.
What we have to do in digital video is a compromise- because doing things close to perfection is still too hard for all the sub-systems - or just too expensive.
So compromises are made on image size or compression of the image, to make the whole thing work.
These compromises mean that doing digital video can sometimes be quite difficult, as the person creating the content needs to have some idea of the capacity of the system that will be used to deliver the content.
What you have to do is understand and work within the capacity of the system you are using, and at this time ( early 2005 ), the best way to get the most - at most times - for an affordable cost - though not always - is to render at 720x480 or 720x576 and use a codec that compresses the ammount of information in the file like the dv-pal or dv-ntsc codec.
This gives usable performance on most systems.
Your screen might have a resolution of 1024x768 or even 1280x1024 but
going to higher resolutions increases the demands on the system to such an extent that other compromises have to be made - or lots more money needs to be spent.
If you feel comfortable with doing an experiment - create text at 1024x768 and 720x480- and see if you can see any difference:
you can create a still at any resolution you want preferably equal to or greater than the final intended output.
1024x768 is fine for text or you can do it at the output resolution.
it doesnt matter very much in the end.
as almost all graphics programs do their own anti-aliasing of the edges of text which looks great at any resolution.
it can be easier for editing programs if you do it at 720x480 - if your final output is a standard defintion movie.
video does not have a dpi. unlike print.
the output resolution is a fixed number of pixels depending on the output device.
if you want some technical background.
the idea you are looking for is called downsampling.
do a google search for 'image downsampling'
http://www.lassekolb.info/gim35_downsampling.htm
this is quite a comic thread on this issue around the mars images:
http://www.markcarey.com/mars/discus...his-image.html
DPI is only relevant to printing.
1024 x 768 is what you are displaying. It does not matter what DPI it is as it will always be that size on screen. When you go to print that setting comes into play.
Cheers
Toby
what would you do in this case?
Toby?
tim's advise is reasonable - but there are caveats.
I'd just add -
1024 x 768 is more than acceptible for most purposes. You can get away with 720 x 480 - depends on what compression you use.
In theory you can go to a maximum of 2048 x 2048, but these size images take a long time to load.
If you are doing a range of stills it makes sense to render these as a Photo-Jpeg movie. Photo-Jpeg is not resolution specific (like DV NTSC/PAL) and has much better and cleaner colours. Photo-Jpeg movies load a lot quicker than still images. Use the Inframe function to select the still you want.
For high quality I sometimes use Photoshop (.psd). This is useful as Photoshop is a lossless codec and it saves alpha channel information, which you can make use of in Catalyst.
As mentioned ealier - dpi is of no consequence. It's for printing.
Hugh
As I do a lot of work that involves stills I invariably use different resolutions depending on the use of the graphic.
In some cases I have used the full 2048 maximum in order to be able to do A Hi Res zoom in and out of an image. Or if I am projecting across multiple screens you need to push the limits to get the resolution for stills. For example I recently did a show where my screen width was 4 projectors with a 20% overlap giving a total width of 3484 pixels so I needed every bit of resolution I could get. The way I program it gets me around the time it takes to load images this size.
On other occasions I have saved the file at say 640 x 480 when I know I will never blow it up to full screen.
I have also used the Animation Codec to speed up load times but only ever to pull up individual frames never for playback.
Lately I have been using the PNG format for transparency as the file sizes are considerably smaller than PSD files.
Conversely with video playback I have blown up PAL 720 x 576 files to cover 3 screens and the result has been more than acceptable. All depends on what your content is. If it is the heavily detailed Company Logo with an animation it will look really crappy if you blow up a DV file. Photo Jpeg is much friendlier for this sort of thing.
On the other hand a moving background without too many hard edges will look fine.
I think with all these things it is horses for courses. The corporate world really cares about stunning logos and product shots whereas the entertainment world can be far more forgiving.
The beauty of the Catalyst is that you have all these options and the only compromise is really working within the constraints of the hardware. Only some experimentation will give you a definitive answer.
Cheers
Toby
thats a good answer toby.
thanks
Thanks for all the great advice