coranda

Don't forget that you can buy it from the App Store and Apple will give you a no questions asked refund within the first couple of weeks if you really don't like it.

The tools in AP are only grey when you haven't got an image loaded. Nonetheless, professional graphics programmes need to have bland grey interfaces, it's best practice. In fact, one of the criticisms I've read of AP is that it uses coloured icons in the toolbar. Similarly, a professional graphics workstation needs to be in a bland environment and should never be placed in front of a coloured wall, in a brightly lit room or have bright light directly incident on the screen. It's because human beings have very flexible colour perception. Cameras need white balance adjustments because our eyes compensate for the ambient light in an environment. A white object under tungsten light will actually be quite yellow but to our eyes it will look white. Because we automatically alter our colour perception to suit the surrounding light it's very important to not distract ourselves with colour or bright light on the screen and walls of the room. That would influence our perception of the colours in the image and that must be avoided.

I can't comment on the new 12" but I'm running Photo on a mid-2011 13" Air (1.8 GHz i7, 4GB RAM, 384 MB VRAM) and performance is excellent. No noticeable lag in painting, transforming or moving. Real-time preview is very snappy.

I had to make a small adjustment on a HP Laserjet 4050 for printed circuit boards but never anything like 4%.

It would good to hear from Andy on this issue.

I also have that checkbox greyed out for Canon 5D III. Focus points show up in Aperture ok.

Have a look at this video. It'a about a slightly different topic but it shows stacking one image on top of another as a separate layer, aligning them and masking the upper image. https://vimeo.com/135570308

It could be but it could also be a printer driver issue. I assume from what you have written that the scaling is set to 100% in the printer's setting dialogue. Have you got some other programme you could try to see if it has a similar problem. I have some software that I use to print out circuit board layouts, which must be accurate to a fraction of a millimetre, and it has scaling parameters to correct for printers not being 100% accurate. Your print out seems to be about 4% out which is larger than I'd expect but it could be printer tolerance I suppose.

Miguel, I'm a little disappointed with this because it seems to me, intuitively, that the selection tool is applied to the image view and so should be selecting based on what is visible in that view. This is related to the bug report I made, which seems not to be seen as a bug, where section tools ignore the effects of child filters. I'd be happier if, at least as an option, selection tools looked at what is visible rather than just working on the selected pixel layer. Perhaps I need to stop seeing this as a bug and make it a feature request.

@MEB Yes, but that's a useful work around until the bug is fixed.

Yes, I can get it to happen if you have the pixel layer selected but if you select the B&W adjustment layer before refining the edge then you get the red overlay.

Can you give us some more specific instructions. When I add a B&W adjustment, make a selection and the choose refine edge I'm still getting the red overlay.

You can just stack them all as layers, put a black mask on all but the bottom layer then, with a white paint brush, reveal just the moving figure in each of the upper layers image. If the images are perfectly aligned (shot off a tripod for example) the painting doesn't need to be that accurate.

Thanks. At least I now know it isn't possible yet. On a positive note, I notice that the soft proof adjustment has gamut warnings. Actually, I really like the way AP implements soft proofing.

I can see how to do this in the Develop persona but is there a way to view clipped regions in the Photo Persona of AP?

What Serif is offering are free "updates". That means that any changes within a particular version are free but, as with most other software, "upgrades" to new major version releases need to be paid for. What will probably happen (although I don't know this for sure) is that all upgrades to version 1 will be free to anyone who bought version 1. At some time in the future they will release version 2 and that will need to be purchased. The way that Mac App Store currently works there is now way to charge current users a different price than people buying for the first time so version 2 will probably be a new purchase with free updates until version 3 is released. However, the Affinity programmes are currently still in their first version and so if you have purchased one you qualify to upgrade to the latest version. If you log into the App Store and click on the Updates icon the latest version should be there. Alternatively, click on the Purchases icon and find your programme in the list.

The AP beta hasn't been released yet be they are saying it is not far off. Keep an eye on this forum for it's release. https://forum.affinity.serif.com/index.php?/forum/19-affinity-photo-beta/

It appears not to recover clipped highlights the way the Lightroom 2012 process does. You will get a similar clipped area in AP as you get in Aperture, rather than the annoyingly reduced clipping in Lightroom.

It looks as though you are just randomly applying one image to another. That's fine but what are you trying to achieve? Apply image basically takes one image and composites it on top of another (adjusting its size to fit) with the chosen blend mode.

I don't think there's a way to do that in AP, but perhaps someone can correct me. A workaround is to make your selection, from the Select menu choose Outline and you'll be able to create a selection of variable thickness that runs along the boundary of the original selection. To stroke it, create a new pixel layer and use the flood fill tool in the selection. It's not ideal but it's a workaround until a better way becomes available.

Yes, it does also work if the background is white rather than transparent.

Here's one way to do it but it's harder than it should be. Perhaps someone can come up with a better way. It would be easier if the Resize Canvas process allowed selection of the colour of the extended canvas area. 1. Get your image to the size you want it. 2. Resize the canvas adding 3 pixels on each side. 3. Create a new fill layer, drag it to the bottom of layer stack and set its colour to white. 4. Right click on the fill layer and select Rasterise. 5. Resize canvas to the final desired size (1400 x 1050). 6. Create a new fill layer, drag it to the bottom of the layer stack and set its colour to black. Note that step 4 is important, otherwise the white fill layer will expand to the final size and hide the black layer. This works if you have created a file with a transparent background. I haven't tested if it works with a white background.

I could be completely wrong but something in the back of my mind makes me think that there are public betas and developer betas. Galilo mentions beta 5 but I think the public betas are only up to version 3. Perhaps he has a developer beta that is later in the development process than the current public beta. But, as I said, I could have this completely wrong.

I believe Serif are aware of this problem and are working on it.

Actually, I don't think the X100 uses an X-trans matrix. The X100S does but not the X100. I agree that when we are dealing with a file as seriously underexposed as the example r10k has produced there are bound to be limits and, ideally, you wouldn't produce that sort of exposure. On the other hand, it's not unreasonable to compare AP's performance against other convertors on such a file. I too am, at times, critical of pixel peepers but when you want to print A3-A2 at 360 dpi pixel peeping can sometimes be important. In that case you'd rather have a smoothed image than that maze pattern. But, to cut to the chase, I'd like to know why AP produces that maze pattern when I could not reproduce it with any of these eight demosaicing algorithms: amaze, igv, lmmse, eahd, hphd, vng4, dcb and ahd. I'd just like to know what it's doing and why.