Blue grain in black areas after export

[kimw999]

Hello! I am exporting a photo and noticed that after export, the black portion of the photo has blue grain. This blue grain does not appear in LR itself before export, and am wondering how to fix this issue?

I used various brush tools in those areas so maybe that has something to do with it? Not sure as to what could be the cause.

Thank you!

[Ussnorway]

looks like the highlights is set too high to me... was the image taken in doors | poor lighting?

[kimw999]

Yes to both! I noticed in LR the image doesn't show the grain unless I zoom into the photo. When it is 'fit to screen' the blue goes away and in-turn, why I didn't notice until after exporting when it became glaring.

Is there a way to adjust in LR? I cannot retake the photo unfortunately.

[Ussnorway]

if its a raw image you can adjust it in Photoshop | Lightroom editor... both the same thing

bring down the highlights then up the contrast or move blacks & whites closer together but small adjustments are going to give you the best results

[Per Berntsen · Accepted solution]

This is high ISO noise, mostly color noise it seems, but probably also some luminance noise.

You are using a camera with a small sensor at ISO 3200, so noise is inevitable.

You need to apply some noise reduction in Lightroom, and you have to do this at 1:1 view. Any other view (like Fit view) will be inaccurate and misleading because the image has been scaled. Only 1:1 gives you a true representation of the image.

You will find noise reduction in the Detail panel in Develop.

Do color noise first, drag the slider until the colored speckles disappear, then do Luminance noise.

Luminance noise reduction should not be overdone, it will give the image a "plastic" look, so it's better to have some noise in the image.

And remember, use 1:1 view.

[calm_Zest5F94]

Thank you that worked for me

[whitefox13]

It probably is color noise, but the reason for the blue in the export is you're missing at least the blue channel.  Go back to Lightroom, open the point curve of the Tone Curve panel, select the blue channel, place your cursor at the very bottom corner of the blue grid.  The cursor should be a double pointed arrow cursor, not a crosshair.  Make a very slight movement upward.  The triangle inside the clipping warning box will be blue, stop moving upward when the triangle grays out.  If the triangle is magenta then the red channel is also lost and after you recover the blue, the triangle will be red, recover that the same way.  If the triangle is cyan then blue and green are lost, the triangle will be green after recovering the blue, recover that the same way.  If the triangle is white, then all three channels are lost and need to be recovered.  After the triangle is grayed out, Option/Alt click on the blacks slider in the Basic panel.  The screen should turn white indicating there is no loss of detail in the image.  If you also magnify the image there should be no blue in the image.  Now export the image.

[Per Berntsen]

Viewing Kim's screenshot at 1:1 reveals that the noise isn't just blue – it's red, green and blue.

And using a curve will not remove it, that will just change the color balance in the entire image.

Setting the Color slider to 100 in Lightroom removes most of it, and setting Smoothness to 100 makes it even better.

The image still has a lot of luminance noise, and it is so coarse that it's impossible to improve with the Luminance slider. I have a feeling that the image has been sharpened with the Masking slider set to 0. Using a high value here (70-90) would at least prevent the noise from being sharpened.

[whitefox13]

Of course it is color noise, that's the first thing I said.  If you recover the lost detail in the color channels first, you'll find you don't need to move the color noise reduction slider as much, sometimes not at all.  By the way, recovering the lost detail in the color channels with the point curve section of the tone curve panel will not "change the color balance of the entire image", it will correct it.