Newbie with crazy idea

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

jamesson

Active member
Joined
Aug 12, 2013
Messages
34
So, I went back to school a bit back to get some formal electronics coursework done and between my present halfway-there skillset and my backup (no need to go into lurid detail about my friendly neighborhood EE) I feel confident in taking on the following challenge; I wish to propose a 51x 1u a/d converter/recorder, sd card slot, and hdd bay (all in separate units of course). Now, the first problem is that afaik there is no direct access to the output of any of the units from any of the other units unless you patch them together. Is this correct? If so, do I have the legal/moral authority to make alterations to the existing backplane design?

In general, how much interest is there in this? How about, for instance, a kickstarter project?

Many thanks in advance

Joe
 
We're all rooting for you.  8)

Good luck. The hardware won't be too much of  a challenge, there are some really smart folks here that know hardware inside and out. The software of getting the darn audio to and from SD... that will be your main difficulty. There may be a few software guys on this forum, but I think you'll struggle to find folks with that skillset.

My thoughts are to find a path to get the data to get USB and use a Rasberry Pi as your method to get the USB audio to storage. It'll also allow for random things like streaming over wifi etc. The software support network in the linux environment is also much more varied.

A few google keywords.... :)
XMOS
PCM4204
PCM4104
RKN80
 
Many thanks for the suggestions, however my original questions remain unanswered. Am I correct in my assessment of the backplane? Am I allowed to mess with the core spec?

Re raspberry pi, an interesting proposition, but to my knowledge nobody has succeeded in getting latency below 4ms over usb2. Also, call it a matter of personal preference but somehow I'd rather not keep my OS and my audio data on one partition, especially if I intend for it to be hotswappable.

Edit:
RE software, thats where my backup comes in, that and my own code-fu is pretty strong :)
 
Who uses hard-drives today?

I see SD-cards a lot but I'd rather have dual thumb-drives. (Dual for redundancy.)

> Am I allowed to mess with the core spec?

It's your soldering iron, do what you want. But _I_ would have little interest in a non-standard backplane.

> to my knowledge nobody has succeeded in getting latency below 4ms over usb2

What is this, a "Tape Recorder"? Then the true latency is minutes, days, decades. Who cares if it takes 4mS or 4 Seconds to get from the mike to the tape?

> I'd rather not keep my OS and my audio data on one partition, especially if I intend for it to be hotswappable.

Really not clear what your product is. Can you put an O/S interface on a 51X faceplate? Yes, some cellphones have tiny interfaces, but many of us are over 30 and do not have petite thumbs (or eyes). I'd kinda assumed a REC button and files named 0001.wav 0002.wav etc. And that's not much of an O/S, would fit in a small ROM.

> all in separate units of course

No "of course" about it to me. ADC is a half-inch in the back. SD/USB is an inch in the front. A desktop drive in a tray is a big chunk, but laptop drives are less, and IMHO the music goes on some form of flash. I once had a ADC-USB with REC button, the box was the size of a cigarette pack. It wudda been useful except the REC button was unreliable (half-cent switch and poor molding), audio connectors not much better; also the "meter" was less than I like. I don't think it was at all "tightly packed", and this was 8+ years ago and surely smaller today.
 
As a full time pro I'd ask what this proposed product is for?  Location work?  In a studio we have our DAWs.  Also struggling to see the suitability of the 500/51x platform for a project like this.

 
Who uses hard-drives today?

I must admit to a lack of familiarity with usb3. If it's fast enough to do the job then probably a better choice. However;

A) Still no thumbdrives as big as the largest harddrives
B) sata/esata is much faster

_I_ would have little interest in a non-standard backplane

Even if it's fully backward-compatible?

BTW, nobody has told me if my assessment of the current design was correct or not.

Who cares if it takes 4mS or 4 Seconds to get from the mike to the tape?

Well, long term goal is a full modular interface. This is maybe phase 1.

Really not clear what your product is.

I was responding to the suggestion that I use raspberry pi to write to sd card. That system stores various os-related data on the sd in addition to whatever. A 51x pi is an interesting idea but I dont think it really meshes with my project.

No "of course" about it to me.

Imagine I had the chops to stick all this in one box. I could probably design it but I could never build it, and a prototype of that complexity is out of my budget.

As a full time pro I'd ask what this proposed product is for?

Location work was what first came to mind, yes. But, the more I think about it the more I realize that my true goals for this system could only be met with an actual interface, a/d/a conversion connected either to a computer or some kind of console controller. As it stands (again, nobody has told me if my assessment is correct yet) the individual units have no access to each others signal without crosspatching. If you want to route them to each other you have to give up outputs. IMO a line mixer would even be a big improvement, so a unit that would freely route signal between the slots would be even better. Do some kind of dedicated zero-latency routing architecture, now you're cooking with gas :). So consider a recorder system as phase 1.
 
By the way, We're not trying to be aholes and shoot down your idea. This is just idea validation. (it happens at every company that actually makes something  ;) )
Things to consider...

- What's _worth_ making
- What and how you'd really use it (re: ruairi)
- How can it be broken down into usable stand-alone modules
- How long will it take to develop - do you have the patience?

Way too many folks burn time on projects that they'll never complete because it's not worth enough to them, they don't have the skillset or they don't have the patience. I'm one of them. Expat Audio has an entire box of projects that are busy feeling sorry for themselves... including an ADC for the API500 standard.
 
Rochey's latest post

I'm completely up to speed on these concerns, and have been around EEs long enough to be extremely sensitive to them. When I first started doing this stuff, many moons ago, I worked with an E-mu 1616 and I still consider this basically an ideal model - zero-latency effects and monitoring + precisely-adjustable asio latency. Now, if i can get analog effects in there in a small footprint with no patching, how awesome is that?

I was, originally, going to build my own enclosure/cassette system, but the available parts were so shoddy that I kept looking til I found you guys. Now, regardless of which approach I take (a/d recorder or a/d/a interface) IMHO the fact that none of the units can get signal from any other units right from the pcb (without patching) is a disadvantage. Completely free analog routing would need a matrix  mixer, but zero-latency monitoring is a feature of many interfaces. Surely we can leverage that tech to handle our routing from a computer?

I'm okay, frankly, with the idea that some of the features will stay in the planning stage. Right now I am far more concerned with learning the system and the market, and finding out which of my assumptions are bad. That said, I clearly hope people will look at my designs and tell me about my errors so it makes sense for me to ask what people want me to do.
 
I just don't get what you are trying to achieve?  Can you describe the problem that your proposed system solves (or does better than existing solutions).

What are these "analog effects with no latency"?  500 series modules?  If so then these are available already to everyone with a DAW..

I've read and reread your posts and I'm totally unclear about what you are trying to do.

 
I think the idea is that if you have invested in 500/51x,  then if you want to take your studio out on the road, you just slot this thing in your lunchbox, and off you go.

Or am I missing the point?
 
You are mostly correct, in that they are available to everyone with a daw. But this has 3 problems

1) From the entry-level perspective, now you are spending $600-1200 for a good computer, then another 2-400 for a good interface, then if you are very lucky  you get into 51x series for under $600. You havent even started buying mics yet. After you do all this probably you cannot look for field recording work because causing all this to travel is a pain in the ass. Any part of this process we can eliminate will save time and expense

2)When it comes to processors you dont care about dry signal, but with effects alot of the time you want to be able to track the dry signal and the wet signal. The only other way to do this is to put 2 outputs on the pre/processor but I think my suggestion is preferrable.

3)Zero latency is a relative term. Minimum asio latency is 2ms but EMU was able to achieve monitoring even lower than that. Regardless, this is well into phase interaction territory. Clearly pure zero latency is impossible but if a/d/a conversion is involved we want to exercise fairly careful control in this situation.

Again, to be real specific about my proposal; I want to work toward designing a backplane with full a/d/a interface functionality but start by making a/d converters and recorders. If you think that's the wrong path to take towards that goal I'm fully open to suggestions.

Aphex already did this, http://www.rspeaudio.com/Aphex-500-Series-USB-Rack-p/apx-500-rck.htm which has many problems but is at least a start. Another problem I have is that I realize we need big transformers to make linear psus but I still cant quite buy the idea that we need these giant psu boxes. Surely we can at keast reduce the size of the PSU to one module?

Thanks again for everyones' help, and I look forward to more of your input

Joe
 
All criticisms I offer are in the spirit of the forum, nothing personal intended....

jamesson said:
You are mostly correct, in that they are available to everyone with a daw. But this has 3 problems

1) From the entry-level perspective, now you are spending $600-1200 for a good computer, then another 2-400 for a good interface, then if you are very lucky  you get into 51x series for under $600. You havent even started buying mics yet. After you do all this probably you cannot look for field recording work because causing all this to travel is a pain in the ass. Any part of this process we can eliminate will save time and expense

I doubt very much that you could bring this unit with the required functionality for less than the price of a laptop and interface.  The laptop has many more uses as does the interface as opposed to your very specific tool.  I can't mix on your tool so I'll need a laptop/desktop/DAW anyhow so no savings there.

Your solution still needs mics and it it really no more portable than a laptop and interface.

2)When it comes to processors you dont care about dry signal, but with effects alot of the time you want to be able to track the dry signal and the wet signal. The only other way to do this is to put 2 outputs on the pre/processor but I think my suggestion is preferrable.

Can I ask how much live/location recording you've done?  I can't think of any situation where I would want to commit effects as I track, with the substandard monitoring we often suffer through on location this seems like a bad idea. 

3)Zero latency is a relative term. Minimum asio latency is 2ms but EMU was able to achieve monitoring even lower than that. Regardless, this is well into phase interaction territory. Clearly pure zero latency is impossible but if a/d/a conversion is involved we want to exercise fairly careful control in this situation.

Low latency is only required if you plan on giving the artists headphone feeds directly from your DAW/recorder.  This is not typical in location work.  If you are doing standard recording ghetto style in someone's house then you are going to need the features of a DAW to get any real work done.


Again, to be real specific about my proposal; I want to work toward designing a backplane with full a/d/a interface functionality but start by making a/d converters and recorders. If you think that's the wrong path to take towards that goal I'm fully open to suggestions.

I'm missing the whole point of the exercise, clearly I'm not the target audience but I have done my share of location recording.
 
Ruairi, please have no fear of offending me. While I definitely appreciate your politeness and civility, I am a big boy and am far from being so egotistical as to want to make something nobody will use.

Look at it this way; I am being offered the opportunity to make a product for the pro audio market with the help of some extremely experienced people. It would be extremely rude and counterproductive of _me_ to be excessively picky.

Your arguments are by and large persuasive; would they hold, do you think, for a recorder or matrix mixer?

Thanks again for all your help, and looking forward to learning more from you

Joe

EDIT: the above aside, let me clarify just a bit further; I propose to remove the computer from the equation only as a "factory option", so to speak. Instead the core system would include an interface within the backplane. This would reduce the required patching and schlepping at the very least, and clearly in order to be competitive I would have to make something that is cheaper than existing preassembled racks. But, somehow I am not afraid of this. For instance the radial stuff is $1k and up.

 
jamesson said:
Ruairi, please have no fear of offending me. While I definitely appreciate your politeness and civility, I am a big boy and am far from being so egotistical as to want to make something nobody will use.

Duly noted!  Right now your product sounds like an answer to a problem nobody has.  A laptop and interface is not difficult to carry and does everything your proposed box does as well as a whole lot more.  very few people who are serious about audio (and hence even curious about your proposed product) do not already have a decent laptop.

Look at it this way; I am being offered the opportunity to make a product for the pro audio market with the help of some extremely experienced people. It would be extremely rude and counterproductive of _me_ to be excessively picky.

I would think that those experienced people would expect pickiness, it's part of good design IME.  If this is a commercial product it needs to solve a problem we already have a create a possibility we did not know existed.

Your arguments are by and large persuasive; would they hold, do you think, for a recorder or matrix mixer?

Again my recorder and matrix mixer of choice is Pro Tools HD.  The JoCo box covers the dedicated recorder space well, if I want more features I'll use Pro Tools HD or Radar.


EDIT: the above aside, let me clarify just a bit further; I propose to remove the computer from the equation only as a "factory option", so to speak. Instead the core system would include an interface within the backplane. This would reduce the required patching and schlepping at the very least, and clearly in order to be competitive I would have to make something that is cheaper than existing preassembled racks. But, somehow I am not afraid of this. For instance the radial stuff is $1k and up.

Again, I don't understand.  You are proposing a 500 series style rack with an audio interface built in?  And pres?  To save patching?

Maybe you should draw a flow chart or schemo of what you are proposing.  I could well be knocking something I just don't understand.
 
No, your analysis is correct, I am proposing exactly what you describe, to save patching and some $. I have some pretty large PT experience and am just not a fan (I am a Reaper guy). That's by the by, however. More importantly, I must again agree re pickiness, with the caveat that I used the word "excessive" :).

2 questions;

A) why do you think Aphex chose to go this route? I mean, they won award(s) at NAMM with what I would consider to be a half-assed product (no firewire support, for example). Would you not say that is at least some argument against your analysis? Are there perhaps some sales statistics we could look at?

B) That aside, let us say for the moment that we conclude that my idea is not sufficiently in demand. What avenues of research would you suggest I pursue? Again, I am really interested in a/d/a conversion (perhaps unwisely since EMU is basically dead at this point and I want to do some similar stuff) but if you think there are analog devices that I should look into, by all means suggest them. I'm sure, worst comes to worst, that there are still things out there that you would like to see emulated that haven't been done yet. Clearly I would prefer to do original work but in-demand beats original any day of the week in my book :).
 
jamesson said:
A) why do you think Aphex chose to go this route? I mean, they won award(s) at NAMM with what I would consider to be a half-assed product (no firewire support, for example). Would you not say that is at least some argument against your analysis? Are there perhaps some sales statistics we could look at?

Can you remember the last time Aphex made anything innovative, or even relevant?  If I had an opportunity to develop something Aphex would not be my inspiration.  Unless you know people in pro audio sales (I do) figures are impossible to come by.

B) That aside, let us say for the moment that we conclude that my idea is not sufficiently in demand. What avenues of research would you suggest I pursue? Again, I am really interested in a/d/a conversion (perhaps unwisely since EMU is basically dead at this point and I want to do some similar stuff) but if you think there are analog devices that I should look into, by all means suggest them. I'm sure, worst comes to worst, that there are still things out there that you would like to see emulated that haven't been done yet. Clearly I would prefer to do original work but in-demand beats original any day of the week in my book :).

I make part of my living helping several high end pro audio companies with these questions so I'll have to keep my ideas to myself.  That said there are two approaches to the question

- Study the marketplace carefully and identify a need or developing need
or
- Build what you are passionate about and let the market be damned.

If you are working with other's time or capitol the second route is not viable and the first is not a way to make money easily (pro audio is coming into a difficult patch IMO).

 
To gauge interest ,
If it was a product that was cheap and or easy to make , would it sell ?
would the chinese knock it off [ because it was so popular ? ]
or is it enough to do it for yourself, regardless ?
 
I have seen huge differences between what professions who do this for a living predict a new product will sell, and what they actually do, some rather disastrous sure-fire winners that went over like a lead ballon (not a Led Zeppelin). :-[

To be successful a product is like a three legged stool. First leg is it must do what the customer needs (or thinks he needs). #2 it must do this for a price the customer is willing to pay, and #3 it must be available when he wants to buy it. Implicit in this is a fourth leg, that the prospective customer must know that the product exists, easier said than done.

We love our own children and think all of them are pretty or handsome...  When just starting out, design something that you want/need and you think others will be willing to pay for. Having people say they would buy one is good, but talk is cheap....In business, money talks and BS walks. 

If you have an idea that you really like, just do it... if others like it you can sell some, if not you just got your first lesson from the school of hard knocks.

JR

PS: for extra credit count how many cliches I threw in.
 
Thanks so much to everybody for all the input. I'll keep on truckin', i guess. At the very least I've finally gotten the point that there is really no stepping-stone to a full a/d/a system and I'm far better off designing exactly what i want than going through incremental precursors in the hope that somebody somewhere will have some use for them.

John, allow me to congratulate you on a truly badass product, although I'm sure it's only the first of many I'll see here.

Thanks again to everybody, and looking forward to learning much more

Joe

 
That sounds great , being able to make exactly what you [ think you ] want .
my designing skills [ lack of ] leave me in the vampire category. 
 
Back
Top