Resize oopsy?

[#1]

I made a composite in 8x5 for a postcard.  I kept all my objects in smart objects thinking this would allow me to create larger prints from this.  But I seem to be wrong on that.  When I attempt to scale up, the resolution drops.  Is there a way around this?  Like I said, I thought using smart objects would prevent the loss of resolution but I guess I was mistaken.  Thanks in advance for any suggestions you have.

[#2]

I made a composite in 8x5 for a postcard.  I kept all my objects in smart objects thinking this would allow me to create larger prints from this.  But I seem to be wrong on that.  When I attempt to scale up, the resolution drops.  Is there a way around this?  Like I said, I thought using smart objects would prevent the loss of resolution but I guess I was mistaken.  Thanks in advance for any suggestions you have.

[#3]

You can't enlarge pixel-based raster images without dropping resolution. That's what "pixels per inch" means. Just read it literally. It's a standard equation, one goes up, the other goes down.

A smart object allows multiple transforms committed once. There is ultimately nothing non-destructive about it, but it means the destruction happens only once. As long as it remains a smart object it's "suspended", but sooner or later you have to rasterize it, and at that point normal physical laws apply again.

It is always possible to resample to higher pixel dimensions, as long as it's understood that this will never improve image quality. Quite the opposite: it will usually introduce artifacts and disrupt the pixel structure. The result usually looks worse than the original. The only time it's justified is if it's important to avoid visible pixels. There aren't many situations where that is a real consideration.

[#4]

Thanks D Fosse, that makes sense.  But, if I haven't yet rasterized the image and everything is still in a smart object, and I go to image size and increase it, shouldn't that work without decreasing resolution since everything is still in smart object form?

[#5]

The smart object is still limited by the original resolution of the image.

The main benefit of a SO, in this case, is that you can downsize it and upsize it without losing quality. 

This assumes two things:

1. You make it a SO before downsizing it (or placed it, which makes it a SO automatically)
2. You don't upsize it beyond its native size. For example, if you have a 2000x1000 pixel image and you upsize it beyond 2000x1000 pixels, it will pixelate regardless of whether it is a SO or not.

[#6]

Smart Objects retain the original state of the file, whether it be RAW or overall file size; however, it sounds as though you downsized the entire file. A future workaround would be to highlight all your layers, convert to Smart Object, and then resize. This would allow you to return to that previous state. I hope that helps.

warmly/j

[#7]

Perfect.  Thanks.

[#8]

When you say “the resolution drops,” how are you seeing this? Do you mean visually, or according to the Image Size dialog box, the Info panel, or somewhere else? The reason I ask is that Photoshop doesn’t provide any way to know the resolution of images nested in Smart Objects (except to open them up).

An image in a Smart Object does retains its original resolution, but is still limited by it, so if an image in a Smart Object was imported at 300 ppi but the document was scaled up 200%, then the image in the Smart Object is now 150 ppi.

If you wanted to retain some headroom for scaling up, images included as Smart Objects should have extra ppi. If you want a document to be able to be scaled up 200% in the future, and everything in it must be 300 ppi at output time, then images used as Smart Objects should be added at 600 ppi so that they are still 300 ppi if the document is scaled up 200%.