Before I spend a fair amount on a 24" x 36" metal print, I'm looking for some guidance on processing. Have never done a metal print before and there isn't a whole lot of info out there.

I know some of this depends on the contents of the photo, so here's the shot I'm planning to use:


According to Bay Photo's file prep guidelines, they seem to recommend files be at 250 ppi. For 24" x 36", that would translate 6000 x 9000 pixels. My photo is 3264 x 4896 pixels, so that works out to 136 ppi. Well above the "minimum for good quality" value, but a fair bit short of the recommended.

If I'm exporting the file from LR3:

1) Do I have LR3 upscale the image to 24" x 36" @ 250 ppi or do I leave it at the native size and let Bay Photo use it as-is?

2) In Bay Photo's process, will they upscale it or would they print it at 136 ppi since that is above their "minimum for good quality"?

3) For output sharpening within LR, obviously there is no Metal option. Do I use the Matte or Glossy settings? Stick with Standard amount or use High?

Appreciate any guidance you guys can give here based on your experience with Bay and/or Metal Prints.

I also have CS5 if that helps, but my Photoshop skills are pretty basic. I use PS primarily to remove poles from peoples heads, etc. :)

Thanks!
Jay

Bay's guidelines answer your first two questions:

Originally Posted by rolette 1) Do I have LR3 upscale the image to 24" x 36" @ 250 ppi or do I leave it at the native size and let Bay Photo use it as-is?

"We prefer that you leave your files at the original file size ...." I would do so.

2) In Bay Photo's process, will they upscale it or would they print it at 136 ppi since that is above their "minimum for good quality"?

"Keep in mind that our printer will resize each print at the optimum resolution for that size." Let them do any upsizing that's needed.

3) For output sharpening within LR, obviously there is no Metal option. Do I use the Matte or Glossy settings? Stick with Standard amount or use High?

I wouldn't trust the presets for sharpening. In Photoshop, I would use an edge mask and sharpen (with unsharp mask) to the point that the image looks just slightly oversharpened on the screen. The printer will soften that a bit, so the print shouldn't look oversharpened.

EDIT: You might also consider straightening the verticals on the left. If you use free transform on just the bottom left corner (rather than the perspective tool, which would adjust the right side as well), you'll lose less of the image.

Originally Posted by Peano "We prefer that you leave your files at the original file size ...." I would do so.

I saw that, but from the context I thought it was more about not reducing the file size for smaller prints. Probably the engineer in me over analyzing it, but wasn't sure it applied to files with less than the ideal number of pixels.

"Keep in mind that our printer will resize each print at the optimum resolution for that size." Let them do any upsizing that's needed.

Another one where I thought the context was for something else. I thought this one was relative to providing different crops and print sizes from the same image.

I wouldn't trust the presets for sharpening. In Photoshop, I would use an edge mask and sharpen (with unsharp mask) to the point that the image looks just slightly oversharpened on the screen. The printer will soften that a bit, so the print shouldn't look oversharpened.

Ok, will give that a try. My Photoshop-fu is considerably weaker than my Lightroom-fu, so it's not the path I tend to go. One of these days I need to make myself learn PS better. Probably need to break down and try making a sports poster... That was the reason I bought PS in the first place (well, that and content-aware fill).

EDIT: You might also consider straightening the verticals on the left. If you use free transform on just the bottom left corner (rather than the perspective tool, which would adjust the right side as well), you'll lose less of the image.

Will definitely give that a go. Thanks much for the feedback!

Jay

To fix the tilt, hit Ctrl-T to get a free transform box on the image.
Hold Ctrl-Shift and pull the bottom-left handle in until the pilings are
plumb. Then hit Enter. (And then crop, of course.)

Originally Posted by Peano To fix the tilt, hit Ctrl-T to get a free transform box on the image. Hold Ctrl-Shift and pull the bottom-left handle in until the pilings are plumb. Then hit Enter. (And then crop, of course.)

Thanks, Peano. Very helpful.

I had hoped that the lens profile corrections in LR would help, but the distortion adjustments seem to be primarily aimed at barrel and pin-cushion distortion. Doesn't really help perspective distortion at all.

I mostly shoot sports, so this sort of thing is not exactly my forte. Probably landscape 101 stuff

I used the free transform as you described and it helps quite a bit. It's interesting though... I can't get all of the verticals of the bungalo square. By the time the edge of the main building is square, the covered deck portion is "leaning in" to the left.

Jay

In terms of sizing and output sharpening, they have to be done in combo (see http://www.creativepro.com/story/feature/20357.html this is how the LR capture and output sharpening are based upon).

You’d therefore have to upsize on export and then apply the output sharpening which in LR is based on output to an ink jet printer. The three settings are just differing strengths and subtle in terms of differences. And you’d have to set size and therefore output sharpening to get it, in the Print module and save out a JPEG. But the bottom line is, you can’t output sharpen, even in Photoshop until you size the image. If you decide not to upsize (and I’d do this in LR, it will do a slightly better job than after in Photoshop due to the data and processing engine), you can’t output sharpen. That is a downside to this lab’s approach to just send the data ‘as is’ and let them upsize. Unless they are applying appropriate output sharpening.

You might want to look out how upsizing and accounting for sharpening as well as adding ‘noise’ can be done on your end by reading this piece:

http://www.digitalphotopro.com/techn...s.html?start=1

Originally Posted by rolette I used the free transform as you described and it helps quite a bit. It's interesting though... I can't get all of the verticals of the bungalo square. By the time the edge of the main building is square, the covered deck portion is "leaning in" to the left. Jay

That can still be fixed, but it gets tricky. You have to select part of that corner, copy that part to a new layer, adjust it (possibly with the warp function inside free transform), and blend with a layer mask.

If you're not up to that, the next best thing is to get the verticals as close to plumb as possible using the one corner adjustment. At least it'll look closer to plumb than the original.

If you are going to get the print made though SmugMug you should send a note to help@smugmug.com about what you want to do. They will let you know how well the resolution you have will work for a Bay Photo metal print.

As a side note one metal print I did through SmugMug didn't come out right because I messed up the colors. I sent a note to SmugMug and told them what I did wrong and the just reprinted it for me.

Originally Posted by rolette Before I spend a fair amount on a 24" x 36" metal print, I'm looking for some guidance on processing. Have never done a metal print before and there isn't a whole lot of info out there.

Originally Posted by arodney In terms of sizing and output sharpening, they have to be done in combo (see http://www.creativepro.com/story/feature/20357.html this is how the LR capture and output sharpening are based upon). You’d therefore have to upsize on export and then apply the output sharpening which in LR is based on output to an ink jet printer. The three settings are just differing strengths and subtle in terms of differences. And you’d have to set size and therefore output sharpening to get it, in the Print module and save out a JPEG. But the bottom line is, you can’t output sharpen, even in Photoshop until you size the image. If you decide not to upsize (and I’d do this in LR, it will do a slightly better job than after in Photoshop due to the data and processing engine), you can’t output sharpen. That is a downside to this lab’s approach to just send the data ‘as is’ and let them upsize. Unless they are applying appropriate output sharpening. You might want to look out how upsizing and accounting for sharpening as well as adding ‘noise’ can be done on your end by reading this piece: http://www.digitalphotopro.com/techn...s.html?start=1

Thanks for the links, Andrew. I've got Fraser and Schewe's "Real World Image Sharpening" book, so I'm at least familiar with the multiple sharpening pass concept. It's been a while since I read it though and with most of my photos being high-iso indoor sports (4K on the low end, 8-10K fairly typical), most of my processing is in LR due to volume and material so I've not used their techniques in PS.

I know LR applies the same techniques, but the details are somewhat hidden.

Good reminder about the output size and output sharpening relationship!

The big question after reading the link to Fraser's article is the unknowns around the metal print process. It's clearly not an inkjet process. Probably I should just do as Peano suggested on sharpening and ignore the output device... my lack of experience with PS puts definite limits on what I can do right now.

Jay

Originally Posted by Dan7312 If you are going to get the print made though SmugMug you should send a note to help@smugmug.com about what you want to do. They will let you know how well the resolution you have will work for a Bay Photo metal print. As a side note one metal print I did through SmugMug didn't come out right because I messed up the colors. I sent a note to SmugMug and told them what I did wrong and the just reprinted it for me.

Good suggestion, Dan. Also great to hear that level of customer service, particularly for more expensive items like this. Thanks!

Jay

Originally Posted by rolette The big question after reading the link to Fraser's article is the unknowns around the metal print process. It's clearly not an inkjet process.

It could be, I’d ask them. It sounds like they are using some kind of dye sublimation process in which case the Contone output sharpening in PKS II would be the target.

Rolette, there is a better and simpler method to fix simpler geometric discrepancies like yours, but for all straight planes in the entire photo. In CS5, use the crop tool and check the 'perspective' option up top:

https://tv.adobe.com/watch/learn-pho...the-crop-tool/


As for sharpening of the print, I have ordered a handful of high resolution 30 and 36 inch wide metal prints (300 to 350ppi) and in technical terms, it doesn't hold quite as much micro detail as a high resolution paper print, since the type of material used for the ink tends to blend together (but it does it very accurately and beautifully still). Think of it as a slightly larger particle size of sand vs. a finer sand. However, it only makes a significant difference if your face is nearly pressed onto the print since you have to be closer to see the difference. Depending on how fine the detail is in your shot, I'd recommend a 1 to 3 pixel unsharp mask. It will work best with the type of sharpness the metal printing results in. In CS5: Filter -> sharpen -> unsharp mask and you'll see what I'm talking about.

Originally Posted by Overfocused Rolette, there is a better and simpler method to fix simpler geometric discrepancies like yours, but for all straight planes in the entire photo. In CS5, use the crop tool and check the 'perspective' option up top:

That's really not better or simpler than using free transform. When you click "Perspective" on the crop tool, it just allows free transform of the crop window. You'll still get the odd tilt that the OP mentioned earlier. Try it.

Originally Posted by Peano That's really not better or simpler than using free transform. When you click "Perspective" on the crop tool, it just allows free transform of the crop window. You'll still get the odd tilt that the OP mentioned earlier. Try it.

Correcting tilt is why the perspective option even exists. I've used it hundreds of times to correct this exact problem with photos... and typically on distortions in more than just 1 part of the image. Although it didn't save any more information than the transform method, it can work on multiple fronts very quickly.

What odd tilt do you mean? I'm pretty sure that's how the house is, lol. I don't think this house is exactly square being in the water, and the camera angle is not squarely facing the house either. Plus that deck is not on the same level as the house. Combining that might enhance leading our eyes into seeing it that way.

Originally Posted by Overfocused What odd tilt do you mean?

Right here. It's the same result whether you use free transform or crop with
"perspective" ticked. They both work the same way.

I think this is what the OP was referring to. If you get the verticals plumb
at 2, they're not quite plumb at 1. On the full-size image I'm sure it's more
noticeable.

I'm not saying there's anything at all wrong with making the adjustment with
the crop tool. My only point is that it won't do any better than free transform.

This happens because the lens causes more distortion near the edge of the
image than it does further in. For me, personally, the difference here isn't
important. I would correct it this much and let it go.



That's your edit, above. Here is my fix using free transform:

Originally Posted by Peano Right here.

I see! As far as it goes, we're just picking hairs now, lol.

Originally Posted by Overfocused I see! As far as it goes, we're just picking hairs now, lol.

Right. One advantage of using the crop tool to correct for keystoning is that it compensates for the "squashing" effect by automatically stretching the image vertically and enlarging the canvas. If you use the transform tool, you have to do that manually.

In the OP's image that effect isn't very noticeable, but with tall buildings and a lot of "leaning," it can be quite noticeable. In that case the crop tool wins, hands-down.

Originally Posted by arodney It could be, I’d ask them. It sounds like they are using some kind of dye sublimation process in which case the Contone output sharpening in PKS II would be the target.

You were right about the process. It's a dye-sub process.

Jay

Wanted to tell everyone thanks for their input. Much appreciated and I learned quite a bit.

Regards,
Jay