"A Pixel Artist Renounces Pixel Art" - thoughts?

[Gil]

I would argue that:

1) I can't even remember the last time I opened Photoshop, if it was not specifically for combining and recombining different layers to achieve a single effect (last project I did had 65 layers for just the background). Photoshop is optimized for a workflow with high layer count.
2) As a programmer, combining and recombining pieces of code to get a singular effect is my job, so I don't get your programmer analogy either
3) I disagree that most pixel art is not larger stuff? Sprites are too small for index painting, but tiles are already big enough to reap the benefits and then there's a significant part of pixel art that just deals with single scenes, in which case, a decent index strategy is vital
4) The most time intensive part of Aura was probably the portraits and the splash screen, both of which would benefit from an index painting strategy?

[Joe]

I'm unlikely to pirate PS, install it in a VM

It's worth mentioning this is possible in Krita now. Haven't tried it yet but I'm glad it's an option for us Linux users.

[Ai]

1) I'm certainly not arguing that. I think people are not optimized for working with a high layer (or anything) count
2) Dunno what you mean there. It's a factoring problem as far as I can see. Maybe you view it as unimportant. It looks like the kind of problem I generally get annoyed enough at to script away.
3) Yes and no? I don't really see why you would use HD indexpainting in preference to plain indexpainting for non-huge tiles, since PS lacks real tile autocompletion.
4) portraits and splash screen are fair points.

I'm unlikely to pirate PS, install it in a VM

It's worth mentioning this is possible in Krita now. Haven't tried it yet but I'm glad it's an option for us Linux users.

Yeah, actually I remember somebody mentioning that before. Think I tried it out at that time, unless they've changed it they need an option to banish or otherwise control the dithering.
Actually am currently working on rebuilding Krita  -- starting with KDE Framework 5, which is now a prerequisite and requires 50 or so packages to be compiled and installed.

[Indigo]

It seems like Indigo's opinion may have changed since then, if we're not misunderstanding each other, though..

My opinion hasn't changed.  You were just positing that it wasn't better than traditional index painting - you weren't comparing it with standard manual pixeling.

HD index painting is not particularly better or faster than normal indexpainting in a program like GrafX2 or ProMotion; it's just different.

Traditional index painting has a lot of the same drawbacks as the HD method (not good for smaller sprites, etc), but without the added flexibility.  When comparing apples to apples, the HD way is generally better and faster. You can do a heck of a lot more with it.  Despite the complexity of the layer setup (and it's not bad to work with once it's set up), I only use it because it saves time.  The moment it doesn't have that potential, I choose another technique.

With that said, I'm a strong believer in the right tool for the job.  There is no *right* answer.  I'm sure there are instances where traditional index painting in ProMotion might be the better option, or even not index painting at all.  I switch between tools all the time.  I also agree that a dedicated program with the HD method built-in would be much better!  Unfortunately i wouldn't consider it financially viable to make one.  It would take re-implementing a huge foundation of work that photoshop already does amazingly well before you could make it a useful tool on it's own.  So the trade-off here is a bit of added complexity in exchange for using a well-developed fleshed out and proven tool

[Ai]

It seems like Indigo's opinion may have changed since then, if we're not misunderstanding each other, though..

My opinion hasn't changed.  You were just positing that it wasn't better than traditional index painting - you weren't comparing it with standard manual pixeling.  Traditional index painting has a lot of the same drawbacks as the HD method (not good for smaller sprites, etc), but without the added flexibility. 

Are we talking about the same stuff here? cause things like QShade work fine for sprites and are generally classified as index painting.

I also agree that a dedicated program with the HD method built-in would be much better!  Unfortunately i wouldn't consider it financially viable to make one.  It would take re-implementing a huge foundation of work that photoshop already does amazingly well before you could make it a useful tool on it's own.

Eh, I don't see why you have to go that far. For example, implementing it as a GMIC filter would allow it to be used in Krita as a layer effect.

[Indigo]

Are we talking about the same stuff here? cause things like QShade work fine for sprites and are generally classified as index painting.

You simply don't understand the added benefits, of which there are plenty (yes, that QShade doesn't provide).  Things like multi-layer compositing, FX animation, blend modes, re-indexing a painting with more or less colors dynamically (auto-distributes the colors), better brush support (full photoshop brush options), adjustments support.  The list goes on.  Most of the extreme benefits come in the form of tools which allow you to control/manage/tune the composition as a whole.  The closer you get to the pixel-level polish stage, the less benefits it retains.  If you'd like to know more about the benefits in detail, bring it up in the HD-index thread.

Eh, I don't see why you have to go that far. For example, implementing it as a GMIC filter would allow it to be used in Krita as a layer effect.

and that's exactly what people have done.  I prefer working in photoshop, though, and I wasn't a fan of their implementation.  If it works for you though, right on.

[Ai]

Are we talking about the same stuff here? cause things like QShade work fine for sprites and are generally classified as index painting.

You simply don't understand the added benefits, of which there are plenty (yes, that QShade doesn't provide).  Things like multi-layer compositing, FX animation, blend modes, re-indexing a painting with more or less colors dynamically (auto-distributes the colors), better brush support (full photoshop brush options), adjustments support. 

On the contrary, this is an area of image processing that I've extensively explored, and I've reread your post several times also; you may be mistaking obtuseness or bad communication for ignorance.
The separation of content for different color ramps is just sufficiently frustrating that I don't really care about those benefits, especially since many of those benefits are available without any special actions in other software.

QShade is what is generally relevant for sprites, though I'll grant that FX animation and certain complex light effects are exceptions.

Eh, I don't see why you have to go that far. For example, implementing it as a GMIC filter would allow it to be used in Krita as a layer effect.

and that's exactly what people have done.

Eh, no. I mean, that sort of achieves it, but it's crappy -- I don't consider it a real implementation. A proper implementation is needed, one that does everything with a single layer (colormap layer, with associated ramp info, and proper dither control).

I'm working on that.

EDIT: anyway, I'm probably getting too annoyed at what to others seems to be a minor workflow wart. Sorry for any offense caused.

[Gil]

anyway, I'm probably getting too annoyed at what to others seems to be a minor workflow wart. Sorry for any offense caused.

In general, I highly disagree with stuff like "everyone is entitled to their own opinion", but in the very specific case of workflows, there is a bottom layer of discussion you can hit where it's futile to discuss. When someone claims MSPaint just fits their workflow better, that's probably just a matter of the person never having properly tried to change to a proper tool and thus an area of discussion. In this case though, we are nearing that bottom layer where what you perceive as a workflow wart isn't even on the radar for the people using HD index painting, as it saves them like 30% of time working on something. Your feeling that, for you, it IS a problem, is valid though.

[Ai]

anyway, I'm probably getting too annoyed at what to others seems to be a minor workflow wart. Sorry for any offense caused.

In general, I highly disagree with stuff like "everyone is entitled to their own opinion", but in the very specific case of workflows, there is a bottom layer of discussion you can hit where it's futile to discuss. When someone claims MSPaint just fits their workflow better, that's probably just a matter of the person never having properly tried to change to a proper tool and thus an area of discussion. In this case though, we are nearing that bottom layer where what you perceive as a workflow wart isn't even on the radar for the people using HD index painting, as it saves them like 30% of time working on something. Your feeling that, for you, it IS a problem, is valid though.

Yeah, I'm not saying 'my concerns are invalid because nobody agrees with me'. My sentiment is more like "We have sufficiently different priorities that we're gonna keep talking past each other, so let's just not go there."

And also a little of this, ie. I may really not have my priorities straight:

[Gil]

And also a little of this, ie. I may really not have my priorities straight:

The best fix for avoiding that pitfall is writing compositional software, aka small, combinable parts. When you start writing those and throw them in PS or Krita or whatever, what do you get? Lots of layers :p. Best practices have been moving towards composition for ages now. Remember, Randall from XKCD is a terrible programmer, from what we've seen.

What we need in PS is more akin to Ableton style macro's. You group a bunch of layers and are then given the option to expose certain controls to macro knobs. That way, I can create a highly complex layer situation, group it, and hide everything except for two or three sliders. To the user, it looks like just another adjustment layer.