Test of image resizing

Let's do a practical experiment. I'm posting three 16:9 images in different resolutions:

1920x1080:



1280x720:



960x540:



I'm looking at them on a 1920x1080 monitor using Win7, IE11 and also a recent version of Chrome.

I don't see any aspect ratio issues in any of them under any conditions. If you have issues viewing them, what are you seeing, and what exactly is it that you want to see instead? What hardware/OS/browsers are you using?

no problems on IE11 (view scaled to 250%) with 4K display.

Cheers... Brian

P.S. Same result... first 2 look the same and third is slightly smaller (no horizontal scroll bar)

QuietOC wrote: The first two look to be identical and the third is slightly smaller. No horizontal scroll bar. Zooming to 200% shows much more detail in the top image while the bottom image just becomes heavily aliased. The middle image is less aliased at 200% than the bottom, but clearly has less detail than the top image.

Yes, there should be only two differences among them as they are seen here in the forum.

1. The first two are resized by your browser to the dyxum limits while the third gets no resizing at all.

2. Because the first two have been resized on the fly, the possibility of visible aliasing exists. This shows up mostly in the upper region of the image. On my monitor, the first image has the worst aliasing; the second is not too bad.

If your browser allows you to also view them outside of dyxum, then you should see them at their full resolutions and without any aliasing.

I'm using Firefox on a two monitor HD portrait/landscape PC system.

The first two appear the same size on the landscape monitor (which will be 1024 wide). On the portrait monitor, the smallest resolution just goes off the edge of the screen. I'm guessing that maybe about 900 width would fit on a portrait HD monitor without the scroll-bars.

sybersitizen wrote: QuietOC wrote: I am viewing the images in the Dyxum forum. I am using the zoom option in Chrome ... At 200% zoom the image is now 2048 pixels wide and if it has 2048 pixels it will be at 1:1 on the display. Yes, as expected.

Actually, I wonder if there might be a difference between using the zoom feature in Chrome (or IE11) while within dyxum vs. viewing the image in a browser outside of dyxum at its full resolution.

In 200% zoom mode, the browser might only be upsampling the image seen within dyxum, which spreads the data in one image pixel across multiple screen pixels; but a separate browser window for the image by itself can map each individual image pixel to an individual screen pixel.

The difference might be subtle, but it should be visible if there is one. Here I'm posting another version of that image at a resolution of 2048x1152. Within the dyxum post it should display at 1024x576. Look at that using your browser of choice at 200% and carefully compare it to a separate browser window with just the image itself at 100%.

2048x1152:



Later: Both methods seem to look the same in Chrome. Testing IE11 next.

Later still: Okay, much the same story with IE.

But either way, with an image of that resolution, I wouldn't want to use the 200% zoom option because I can only see a portion of it at a time with my 1920x1080 monitor, and a big scroll bar appears. The separate window for the image file by itself is the way I'd go.

Edited by sybersitizen - 11 January 2018 at 00:25

QuietOC wrote: I can see the whole width of that image at 200% on my 2560x1080 monitor.

Ah - you didn't mention your monitor resolution earlier.

It looks like it is 1:1 on the Dyxum page. It looks good. Just the same as viewing the image at 100% by itself in its own tab/window in Chrome.

Yes, I agree with that.

sybersitizen wrote: If your browser allows you to also view them outside of dyxum, then you should see them at their full resolutions and without any aliasing.

Doesn't work that way for me on Chrome / win7. Opening the image in a new tab copies the Dyxum size (ie, choosing 100% shows it the same size). Except that I can zoom in a long way before I see pixellation, and inspecting the source shows the image to be the size claimed, so I am guessing that opening the image again keeps the '100% display size' unchanged while maintaining the full image.

Edited by Miranda F - 11 January 2018 at 13:27