marcdraco

Someone else might have already asked this, but I hit this bumper yesterday after spending a couple of days typesetting a book for a mate of mine. Despite her seeing all the spreads with printers marks etc. (I won't bother you all with the usual boilerplate we're all professionals here) she emailed back asking for it in ... DOCX or EPUB. My heart sank. The Word 365 she'd sent me was riddled with all sorts of typographic errors and bug-u-likes from some random binary that she'd entered probably hitting the alt key or some such. Publisher (to its credit) did a much better job of filtering these nasties leaving me with a nice clean master of her work to typeset. As it turns out, what I produced could have been done in Word or something compatible but I wanted to see how Publisher 2 was fairing. Had I known that she was targeting KDP (Kindle) a priori as it were I'd obviously not have used Publisher since Kindle is really just a proprietary set of HTML extensions so the machine can lay it out using internal fonts and flow it as it needs to across devices - a la HTML. I don't know (I haven't bothered to look too close since I come from a print background in this side of things) if it even supports things like left/right pages with nicely formatted footer, page numbers and so on. I've only ever done one book for Kindle in .mobi format and the same manuscript in EPUB which was a massive chore even in some dedicated software I bought to do the job. Anyway, that's a long story for a simple request.

I hear ya. In fact, it's neat to get an assessment from someone who has used both. We could use a side-by-side feature profile really but that's a big ask given the amount of time it entails. I only use raw for video since that's my thing (so to speak). Run and gun which is the only way I'm using my 5D for stills is as easy with JPG and the quality is quite sufficient. It would be different for portraiture or landscape of course, when you have time to set up a shot as we used to do back when I started in film... I miss that sometimes. I'm hearing this a lot from LR users - but not so much from users of the other parts of the Adobe suite - which is probably why they moved into SAAS (subscription model). I'm very big in open source. The machine I'm on right now is running a version of GNU+Linux called "Void" although it can dual boot into a pure Unix (FreeBSD) should I need to. I do all my 3D work including some stuff like Matchmoving with Blender 3D. I can't use it for editing video because it's too clumsy. For that I've got the full version of DaVinci Resolve. @SrPx correctly notes that learning Blender is an uphill and sometimes Sisyphean task but once you get there it's just fabulous. The latest versions are much more beginner friendly but are still a world away from the ease of use of the more established commercial offerings. I first used Photoshop when it came with a scanner bundle at V2.x for the Mac. It wasn't cheap then as I recall, but the bundle price made it affordable. I still have the manuals too. My last version was CS4 which irritated the crap out of me being STILL 32-bit. That pushed me over the edge and I refused to buy Adobe ever again. InDesign (which I used to lay out the print version of my last book) is bugged to hell and even After Effects would choke on some of the work I had done due to lack of RAM· In Open Source we have Gimp which still looks like the 1990s are calling demanding their GUI back. MyPaint which is... well, for painting really and Krita which is amazingly powerful but like Blender, is a pig to learn. A LOT of professionals I know have pirated Photoshop (even the CC versions) but I won't go there. Affinity does everything I need and it does it well. Even on this little i5 4000U - that's the dual-core, mobile version - not the beefier quad-core desktop. I have a 16-core Xeon workstation for 3D work but as I'm a bit of a climate change nut (the sort that does something every day rather than protest loudly one day a year to feel smug) I only fire that up when I *need* to. I'm so anal about this, I declined an invite to talk in the US because it means flying. We're working on some form of Skype (but not Skype) way of getting me there "virtually". This is the way a lot of us are going to have to work in future if we're going to beat the incoming tsunami. I have other software which only works on the Mac - and thanks to Vendor Lock In, I would have to buy NEW licenses to run it on the PC (under *Nix or Windows). This is where FOSS comes in again - you can usually find it works on something you have. Pragmatic choice is the key: use what works for you that's most efficient and affordable. It's a tough call knowing which way to jump - but I suspect a lot of people coming into the field would be better served by Serif's offerings and upgrade at such point that they are no longer capable to perform the duties asked of them.

I guess it would be silly to ask if you've tried the open source offerings? There are a couple - Darktable I think is the main one. Not being a Lightroom aficionado, I can't say how good it is or how close it matches the workflow but it IS by photographers for photographers. Darktable certainly models itself on Lightroom and, I think, is the same on that has all manner of Camera Raw decoders - (debayers) way over my head to be honest as I'm really primarily video and I use Resolve for that. https://www.darktable.org/ It's free to try so the only thing we have to lose is a few hours of time to give it a test-drive.

Multiple sub-frames? Is that one of those new fangled cameras that shifts the sensor a hair for three exposures one for each colour? (Not joking, I hear they can do that now)

This is from another thread - looking at the demosaicing feature in Raw. I noted that AP on Windows "only" has the Serif engine which bears a striking resemblance to AMaZE (which would make sense as it's so highly regarded). However, there are multiple other algorithms out there including ones to deal with high noise images like this one. I wonder if we could have some alternative engines built in so we can take a crack to see what produces the best initial results off the sensor before we start fiddling?

That would be a (short-term) fix that doesn't cost anything. I use Resolve Studio for my editing work (and even my VFX now when I get time to learn that UI). Most of my current stuff doesn't need me to fire up a 16 core Xeon and since electricity it's free I'm sticking with an i3 laptop for a daily driver which is fine for most Affinity Photo work. (More props to the developers for making this possible!)

Does the free version of Resolve have that feature? It's been so long since I used it that I've forgotten and my daily driver isn't anywhere near powerful enough to run it.

I did some tinkering and tried a few different Raw engines on this image (this is extreme magnification) and uses RawTherapee so you can see the actual pattern as recorded by the camera's sensor. No noise reduction has been applied here - this really is exactly how the camera saw this section of a Microsoft mouse. The inset goes in even closer still so you can see exactly the problem these convertors have and that's before they start applying noise reduction, colour correction, curves, sharpening … you name it. I have to say that I was very pleasantly surprised with Serif's convertor - which performed as well as Aliasing Minimization and Zipper Elimination did (to my eye) even on this fairly noisy image which isn't where AMaZE is know to shine. I wonder if Serif can be persuaded to include alternative engines in future? RawTherapee suggests some algorithms for Sony cameras and others that work well with high-ISO (noisy images). Fascinating stuff.

I was just talking a pal of mine (a pro photographer) about this. He likens it to a hammer. There's more than one hammer design (I lost count actually) it's case of finding the right one. On Windows I don't get a choice of which engine to use but I can switch to Open Source software like Darktable or RawTherapee to get other options. @firstdefence is spot on. It's a case of finding the right engine for (a) your camera and (b) the shots you take. Not every shot will work well on any given camera/image because of this: https://en.wikipedia.org/wiki/Bayer_filter Perhaps the most important part of the process is demosaicing (https://en.wikipedia.org/wiki/Demosaicing) which converts the human-eye biased sensor pattern into a proper RGB image. The Wikipedia article is quite easy to follow and it will help you decide if you want to go down that path or stick with what you have. Ultimately, a lot of this is personal choice and as an artist you're the only person to decide what looks right to you. Personally, I think Affinity Photo rocks and it's the best £50 I spent on software. Period.

I don't know if this has been suggested but I use curves a fair bit and particularly LAB for those really powerful wide gamut adjustments. The current graph only has a four by four grid which isn't anywhere near enough to lock in really fine changes. It would be really handy if it were possible to size these graphs too. It's something I'm used to in other software so I was expecting to see it here and (I assume) that changing the reticule is a five minute job for an intern. Resizing is a different matter because that carries other problems … but pretty, please with knobs and bells on (and a cherry on top)?

I'm not a Python programmer - it's just another technology that I don't have time to learn. (Remember the Woodcutter problem? No? Take me too long to explain. Repeat.) I prefer compiled Python simply because the interpreted version is a bit too slow. There's a least one project that uses it, in Blender IIRC, that makes a huge difference but it's a couple of years since I used that. Damn I'm getting old fast.

@SrPx completely with you on what you said about Blender and the UI (as I've said before) still looks like it was put together with all the charm and forethought of sneeze. The post 2.5 releases (from memory) replaced the real rip-snorter with a slightly more suppressed ach-cho. It's still awful and because of the "Not invented here" attitude of many Blenderheads (most of whom, unlike me, don't remember when Blender was an in-house project) it's rather hard to get a meaningful discussion on rectifying that. As you've said, it's a tool. I'm going to give LuxCoreRender a go soon - that combines the full compute power of the nVidia/AMD chips with the full core potential of whatever you happen to be running on. My 6gb card just isn't quite big enough to render large/complex scenes and even with 32 threads, things are still slow. Combining the two might just give me the edge I need. I'll be interested to see what happens with Eeveee and if it proves to be something more than just a glorified GL renderer. (I know it's a trick, but it's damn good trick.) In film it's largely a case of how long something is static on the screen - it's amazing what you can hide in a little bit of motion blur - god I love 24FPS! Although I use Affinity for my everyday "memes" (they're infographics actually) I have a toolbox with Krita, Inkscape (which is better than Designer right now, just not as pretty) MyPaint and even Artrage for that really fine art look. It comes down to this "When all you have is a hammer" thing. I hope Serif don't "pull" tools from Photo, but for the sort of design work I do which is mainly text flow and stroke-based shapes, it's plenty powerful enough. The only thing I need Designer could do (that Photo couldn't) was text on a path. I know Designer does a lot more but I just don't need those features. So between all the art software I've had to buy for support and features, I've only spent about 100UKP, which is a deal in anyone's currency. Max/Maya are just out of my budget not to mention the 1-2 years I'd have to spend cramming even more into my noggin - which is already well past its use-by date. My only misgiving with Blender is the weak particle system but now Davinci has a professional VFX compositor built in with a decent (if sluggish) particle pipline, I'm golden. But I'll say this again, Photo is bloody amazing, particularly at such a low price and I can't see why anyone would want to be chained to Adobe these days. Ironically, some Photoshop people I've spoken to think it's too cheap! I guess that's what happens when you get used to paying through the nose: you get the idea that money=quality and that's clearly not the case. Some people will never realise that, but that's their loss and our gain.

Oh, not seen that before @PedroOfOz. I've used a couple of others like Darktable which also does RAW of course. It's another one for the toolbox, so thanks! Oh wait, it's not actually FOSS, there's a small licence cost for commercial use. Still not bad I guess.

James Ritson has a rather jolly tutorial on this - although he didn't so much align the stars as take a sequence of high-ISO shots in fairly quick succession so they earth's rotation didn't go an spoil everything.

Of course and as Woven has alluded to me separately, this is just version 1 (beta) so Serif's team is playing catchup. This may be a case of them not being able to see the wood for the trees and leaving out this "feature" because they weren't happy with the results. I gather that's why Designer doesn't have a Trace Bitmap facility (c.f. Potrace in Inkscape). Quality control like this is rare these days as publisher rush to get software out there. I don't think it's unfair to mention Adobe and Microsoft as two of the worst offenders in this regard. Props to Serif if this is the case and either way, I appreciate their willingness to share beta software so publicly so we can all have at it. It's a good business practise that others could learn from.

I've grabbed it Walt, I was just hanging on until V1.0 hit. However, Jens does have a very valid point which took me by surprise. There's no asset/resource/clip manager where you can drop (say) 100 photos into draw and just pull and place them as you need. I just assumed it would have one - I thought the iOS 12 assets would be a sample of clip art but I was clearly mistaken. Now it does have a placed asset manager, which is fine, but I can't find "bin" (as they're called in some of my video software). When I'm editing video, I have 100s (or more ugh) clips in a bin or sets of bins, or nested bins... and I can grab and use them as needs be. Quite an oversight really - I'm pretty sure InDesign CS4 (which is what I was using two years ago, and was old then) has this facility.

Drat my beta ran out... I don't use Publisher for this sort of thing but I'm intrigued now. I'm gonna buy it anyway because our last book was typeset in InDesign and that's stuck on a Mac that's draw its last breath! (Just out of warranty too, thanks Apple.) So it's going to be "Here! Take my money" (or rather Mr Bank's money) as soon as the final hits the ground.

You're right of course - I'll have to spend more time in future - I'm still learning as my primary "art" is 3D and that sorts those problems for me (thanks Blender). I did try the lighting filter but it's never as good as getting a nice flat image and hand-painting the highlights. At least Affinity Photo put this within reach of people like me who have all the artistic bent of that ant. I love it and I've already got most of my pals (who are more artistic than I am) moving from Photoshop. EDIT: Turns out that a quick bit of burning did the trick for the cone - it's "good enough" for what it's communicating, but honestly, thank you for your comment - I do learn from everyone here.

This is a simple composite shot of an ant, bee, traffic cone and a few other elements (all public domain) with an extended areas to the top (to accommodate the cone) and right to accommodate the caption about the ant complaining the bee is drunk and won't quit singing. It's not going to win any prizes but it's for a blog entry.

Beautiful!

Yeah 2.8 is a huge leap - I'd say a quantum jump but that would only be ironic and most people wouldn't get the irony. I do (did) a lot of high-end rendering for low-budget film work with Blender but even with some fairly serious hardware, and sending some jobs off to server farms, the results from Cycles took too long. Things have improved somewhat with the latest versions and like Blender, it just keeps getting better, but, as you're alluding WigglePixel, Eevee is a game-changer. It might not be 100% correct, but it's the difference between rendering 1-2 seconds (24-48 frames) in seconds or waiting 24-48 hours for better results. In VFX at low-budget, people just won't notice. Hell, some of the recent films out of Hollyweird have had some questionable VFX in places.

Blender is divine isn't it - although the UI (to misquote my friend) "looks like it was designed with all the charm and forethought of a sneeze". There is a "freemium" version of Resolve of course. The way Serif is going right now though, it's going to be the one to watch. I'm just waiting on Publisher being released so I can give them the money! I just can't justify Designer right now as Photo does everything I need on a day-to-day basis. I would hope if Serif does decide to put pressure on Hitfilm/Premier they stick with a node-based approach. Layers are all fine and nice until things start to get a little complicated and then it all goes to hell in a handcart.

The only one I can't agree with here is Hitfilm (although it's free). There are some better FOSS (Open Source) apps. Hitfilm is geared to home users and that starts to show as soon as you push it into serious projects. Closed source Blackmagic Resolve toasts Premier and After Effects although as a professional app it has a very steep learning curve.

Serif doesn't have one (yet) but right now Blackmagic Resolve isn't too expensive and it toasts Premier and Affect Effects, although it has a fairly steep learning curve.