So, what's the current status of everything?


Are those matte/coated black or are they normal black without coating?
Post automatically merged:



Until you turn on lighting LEDs and it looks really blinding or simply a torch :D

Yeah we (those who like clear) don't care. Clear is awesome. And yeah I get it it shouldn't be the standard case thats totally fair.

IMO dishwater boring black case should be standard, clear should be limited edition only for the 1337 hackers, the other colours I have very little opinion on.
 
I hope you washed your hands after typing that.

Do you recognise your new baby from the photo @Grench?

Actually, I am not sure which one is mine. One of my early ideas had been marbled/blended. I love the one with the multi-colored marbled red lid. Most recently I'd said grey with black buttons is fine (based on the colors of the cases I'd seen to that point) because I really didn't want the colors to be what held this up.

So, most likely it will be one of the grey ones, but kudos to whoever gets that multi-colored blended red lid one - that turned out awesome! If I had a pick of the lot, that one would be it. But - I am perfectly fine with one of the grey ones as well. I'm glad to see that there is such cool variety in these - great examples!

Great to see progress on these - hope the testing goes well!
 
Actually, I am not sure which one is mine. One of my early ideas had been marbled/blended. I love the one with the multi-colored marbled red lid. Most recently I'd said grey with black buttons is fine (based on the colors of the cases I'd seen to that point) because I really didn't want the colors to be what held this up.

So, most likely it will be one of the grey ones, but kudos to whoever gets that multi-colored blended red lid one - that turned out awesome! If I had a pick of the lot, that one would be it. But - I am perfectly fine with one of the grey ones as well. I'm glad to see that there is such cool variety in these - great examples!

Great to see progress on these - hope the testing goes well!

Weeeell, I still have a multicolor-blended red lid, so I can fix that for you :)
 
Are those matte/coated black or are they normal black without coating?
Post automatically merged:



Until you turn on lighting LEDs and it looks really blinding or simply a torch :D

I think transparent case would be nice as accessory, a replacement case on shop, but no as standard case.

Even I continue thinking it would be nice as an option to paint your case the way you want (painting internal plastic transparent case).

This. In fact, some minutes ago my GMMK arrived and I have been assembling it. Each time I assembly a key row, I had to connect it to check if every keys are working correctly. I tried several times to install the keys with the leds on and was a no/no situation.
 
I looked at my Switch Lite, and it’s blue :(
68ed13b1175a45fe85be1bc38154da86.jpg
 
cool new protos! i love the marble red one. looks like leather somehow.
personally i like the grey ones best, as they remind me of my first breadbin c64 mixed with an sgi granite sparkle look. <3

(i had access to an sgi indy in the 90's and loved the granite look so much.)
iu
 
Last edited:
its hard to compare even CPUs with the same ISA as simple change such as bigger pipe or bigger/faster cache or L1/L2 can make huge difference, even frequency break point is a thing, some CPUs even throttle more easily due to lower resistance and high voltage, which may give us the illusion that two very similar looking and spec CPUs would preform similarly, but actually they won't

The most accurate choice would be a proper benchmark such as cinebench, pi test or blend test...
I actually think cinebench scales really good, especially better than PassMark which is really a biased benchmark that favors apple fanboyism (though apple does modify ARM chips really good)

As for sound waves - whats the target sound device sample rate and bit depth? Maybe I can write NEON functions quickly and use SIMD to accelerate software decoding, it will save both battery with cpu load and time, the best solution imo is hardware

imo 20% usage bursts on sound play is equal to device planning failure, if the device is idle but playing sound, it will constantly switch its CPU governor between performance and idle profiles, this should result in quicker death of CPUs unless the governor is propely set up for the delays of the decoding agent, which is a bad choice for general purpose device

maybe its not possible to add hardware decoding now, but should be good to add it as hardware extension since the pyra is modular (love evildragon for that)
 
@Akko Maybe I misread something, but to me it looks like Confuzzled said "If you look very close, DACs can't reproduce square waves." and then you said "If you don't look too close, DACs can reproduce square waves." but word it, as if you were contradicting C. *scratches his head*

Admittedly I phrased my reply poorly so it can be read like that. What I was actually trying to contradict was his statement that DACs produce their output by adding up sinewaves of different amplitudes and frequencies which is just plain wrong (at least I have never come across or heard of a DAC that works that way). I'd also like to add that - while admittedly outputting square waves perfectly is not really possible - most DACs have less problems outputting a square wave than a sine wave.

I didn't intend to start a lengthy discussion on the topic though ^^"

If the DAC has a DC blocking cap placed on its output, then that could certainly be a problem.

Yes but I'd argue that a DC blocking cap is not part of the DAC itself and as such does not factor into this.
 
cool new protos! i love the marble red one. looks like leather somehow.
personally i like the grey ones best, as they remind me of my first breadbin c64 mixed with an sgi granite sparkle look. <3

(i had access to an sgi indy in the 90's and loved the granite look so much.)
iu
This looks like the little blue disinfectant blocks you see in urinals!
 
Ditnt care much the look of the Prototyph, as long as the Prototyph Guys are posting pictures, and Experience and Revievs

Which collor do have the Ordered MP Cases??
I think this Time, i will keep my Release Cases, i only changed my CC Pandora to Platin Case, because these where meant to be a bit better..
 
The sound playback hardware expects 92kHz sample rate, 24 bits of depth.

Great, I will start working on NEON optimised and naive functions next week

Kind of weird that this sound device is using such unique sample rate
en.wikipedia.org/wiki/Sampling_(signal_processing)#Audio_sampling

actually this job is going to be easy compared to image processing since the sample rate does not variant by much more than 44.1, 48 and 96(48 to 96 is upsampling exactly twice, 96 to 48 is downsampling at factor 2, 44.1 to 44/96 is a bit trickier but can be done, the rest are rarely used so they can be unoptimised for now
 
Last edited:
Great, I will start working on NEON optimised and naive functions next week

Kind of weird that this sound device is using such unique sample rate
en.wikipedia.org/wiki/Sampling_(signal_processing)#Audio_sampling

actually this job is going to be easy compared to image processing since the sample rate does not variant by much more than 44.1, 48 and 96(48 to 96 is upsampling exactly twice, 96 to 48 is downsampling at factor 2, 44.1 to 44/96 is a bit trickier but can be done, the rest are rarely used so they can be unoptimised for now

96kHz I think.

...
The Pyra has a high quality 96kHz chip in there, which means the conversion is more complex.
That's why it has the hardware conversion included - but we can't use it right now because the driver for it needs to be updated.
 
That's nitpickery, *nothing* can output *perfect* square waves. A maximum step from say 0V to 5V at 44kHz is ~0.22V/µs (if I didn't mess that up, I'm still not quite awake ^^) which is definitely doable with a good DAC. The first one I found right now can do 35V/µs for example.

What *I* was mainly trying to say is that (common) DACs do not output several added sinus waves but rather a voltage defined by a series of input values directly converted to a series of output values.

A real DAC, configured to reproduce a sampled, bandlimited signal (as in audio) will bandlimit its output, exactly as if it *did* add all of the sinewaves below the Nyquist frequency of its sampling rate. If you sample a 20khz sinewave at 44.1 khz, you'll get a pointy zigzag - most pairs of adjacent samples will have opposite signs. But if you feed that into a soundcard running at 44.1khz, the output will not be a horrible sawtooth - you will get a perfect reconstruction of the input sine wave.
Post automatically merged:

You really, really need to watch at least the second video I posted links to above earlier. Filtering is actually far easier in the digital domain*, and the output of a DAC is not square waves - essentially the digital numbers give the coefficients of different frequencies and magnitudes of sine waves that need to be added together to give the intended output. This has the side effect that a DAC cannot reproduce a square wave, as it would need to sum an infinite series of sine waves to get a step function.

e.g. High quality Notch filters are expensive in the analogue domain, but trivial to do digitally.

I think you might have misinterpreted the (excellent!) videos a little. DACs don't "literally" operate in the frequency domain - usually the digital numbers represent a time series of amplitude values, just as Akko says. But for audio you want to apply a "reconstruction filter" which bandlimits the output, since the input is known to be similarly bandlimited - any frequency you output that is above the Nyquist frequency (say, the higher harmonics of the sharp edge of a square wave) is *noise* that definitely wasn't in the input signal - so you don't want to do that! It's not a "side effect" that audio DACs won't output a "perfect" square wave - it's a feature!
 
Last edited:
Back
Top