Routing possibilities

This could be a solution for introducing true 24 bit / 192khz recording. The inputs are distributed over ethernet to a PCIe card that functions as a DAS. Up to 64 in/ 64 out channels at max record rate. Data throughput roughly 74 Mb/sec maximum, without overheads.

Therefore, this is how the studio would function. Each Red Net 4 box (shown above) costs £2000 currently, while the PCIe adapter is £800... Lets hope in 10 years there are second hand units about or they have reduced in price!

Audio inputs

So, the problem of inputs rears its ugly head. With only one ADAT input, the previously preferred audio interface - the 1616m from EMU falls short in the digital girth department.

The scenario which might use the most channels is recording a drum kit in the dry room. Lets say at maximum, 8 channels of 24bit 192khz, for the sake of the best downsampling later in the train, probably to redbook audio. I can get a measly 2 channels of that spec into the 1616m.

This presents a problem. Do I use an alternative soundcard? The Lynx AES16e-SRC looks to be the best digital contender, with an even lower noise floor of 141 dB (this must be due to the pc's actual power supply). However, monitoring the results would be difficult without some form of DAC after a digital output.

A simple way around this may be to use the 1616m Audiodock without using the ASIO drivers. Two audio interfaces can be used simultaneously if they do not use the same access method. ASIO could be dedicated to the drum inputs, whereas the remaining inputs could be handled through WDM.

I'm largely unhappy with WDM taking even a portion of the load. It has a significantly higher latency and WDM audio drivers are known to provide inefficiency to recording situations. Whilst I would gain the ability to access the digital recording of the dry room using the Lynx, I would not be able to use the 1616m to do anything else other than be a monitor out. MIDI in/out, balanced XLR for guitar performances not applicable.

Somehow I feel there must be another solution.

Dry room v1.0b

 

Grid and rubberised float floor feet laid out.

Air in duct laid out in sillcone-rubber wrapped foam insulation.

 First layer of boards laid down, including ventilation grate space.

 75mm raised section supports.

 Insulation laid between supports.

 Finish floor level complete.

 Suspended ceiling supports installed.

 Suspended ceiling supports connected via 50 x 50mm, 2mm thick box steel. Holes cut at regular intervals to reduce weight.

Insulation wrapped around suspended ceiling supports and support ties installed. This gives structural rigidity fit for purpose.

First 12 steps of Dry Room construction (alpha)

Dry room v1.0a

here is a basic dry room design, 4x3 metres outer wall to outer wall. Outer MDF skin 25mm thick, no linings to create insular sound trap. Secondary air space leading to 100mm insulation (type to be specified). Minimum internal dimension: 1.9 metres.

Prototype Pump Housing

Introduction and getting this far

This blog is intended to document the process of beginning a small-scale home studio and the associated computer equipment, instruments and discoveries found along the way.

First off, let's start with why I am writing this now, before I have even begun to purchase the necessaries, or have a place to begin construction: The timeline I have for this project is around 10 years. By that point, I want to have started a mortgage somewhere and have a space which I can actively modify to be 'my own'.

Until then, as far as possible, I am to lay the computing/ability/systems foundations to enable that studio to be built. Ultimately, I will not be able to come up with a sufficient design for the physical layout until I actually have secured a property.

Secondly, some information on my background. I am a 26yo guy that learned piano (play by ear) fairly intensely from 15-19. In that time, I developed an appreciation for performing. That has since developed into a desire to begin to write my own tracks.

I loathe using the term 'songs', because I aim to explore the space somewhere in between the long works of Mike Oldfield, and some of the more popular tracks of Zero 7.

I currently own no instruments of my own other than a slightly bent Yamaha 6 string guitar, which is not up to much acoustically. Learning guitar is a chore for me, but since multi-instrumentalism is something I wish to explore, its a necessary evil at this point. I suspect I'll enjoy it a lot more when I actually have some capability under my belt.

Some form of drumming will also be explored further down the line, but I foresee this being the last instrument to be explored.

Ok, studio asside for a second, lets concentrate on computing elements.

2 major systems will be required for the studio. 1, the Digital Audio Workstation, a.k.a. A computer with a good soundcard in it, and 2, a file server to handle the mixdowns and recordings. Currently, I will be looking to have the main box of the DAW in the control room with me. After umming and aahing about the merits of locating elsewhere, perhaps using some kind of KVM over ethernet solution, I have decided that the lag is not something I could personally cope with. Therefore, acoustic contribution to the ambient in the control room must be minimised.

This will be achieved through the use of an entirely passive radiation water cooled rig. The two pumps involved will be used in series to maximise flow throughput. This will be routed through CPU, GPU, possibly PSU (additonal loop?) and out to an external large surface area radiator. As nearly every case is specifically designed for airflow, modifications may be necessary to entirely enclose the case interior. This could marry with internal acoustic treatment, something I have done to my last 3 main builds.

Acoustically isolating the pumps internally should present no problem. The bigger question is to the degree of acoustic reduction really necessary. Should the entire case sit within a wooden hollow cube, internally lined with acoustic treatment to further reduce system noise? In theory, with the majority of the heat being transferred externally this should present no problem thermally, however I am still concerned about interior air temperatures from low heat emissive components such as processor resistors and the southbridge. An entirely enclosed system may not therefore provide the best longevity.

A custom pc case may provide the necessary benefits. With an absolutely fanless design, no vibration from hard drives using only SSDs, interior acoustic treatment may be unnecessary. In such a situation, the pumps could be enclosed in an acoustic treatment - the only devices that would necessarily cause noise.

Alternatively, if acoustic treatment with the pumps remaining inside the box, a radiator in reverse within the case may provide an additional internal heat levelling, similar to the way a fridge works. This would have to be the last device in the chain to prevent condensation internally.

Getting to the server, I have recently spent a great deal of thought to the best way to watercool and accoustically treat the server box. Theorhetically, the same principles could be used to isolate the server and draw its heat externally, so the server would not have to be located somewhere else in the house. There are several problems with this however.

The specification I am looking to build involves 16 conventional hard drives in 4 x 4 0+1 raid arrays. So if one block of 4 drives used 2TB drives, it would make available 2TB x 4 drives / 2 for mirroring. It goes without saying that this system would be run behind a UPS.

Conventional drives generate a lot of heat, and the space required for 16 drives using conventional mounts makes a prebuilt case unrealistic for use in a watercooled solution, especially when sonic isolation is a paramount, as it would be if placed in my control room.

The impracticality of actually developing a system to adequately remove the heat whilst simultaneously sonically isolating the system means that simply locating the server to a different room, using conventional low RPM active cooling solutions seem the better option. By the time this issue rears its head in 10 years, SSDs may be of significant density and cost effectiveness to replace the conventional drive specification here. If that happens, the heat output of the server significantly decreases, and a full tower system combined with the pre-mentioned watercooling system may be a viable solution.

In terms of practicality, there is no real benefit of having a server in the same room as the one you work in. Ultimately, this means the more practical low RPM passive solution is more likely to be satisfactory.

In terms of the control room itself, I anticipate the space to have part performance properties to minimise dry rooms. I will specify 2 screens and 2 monitors only. I am accustomed to working this way and the key to successful mixing and balancing is knowing your own setup. Multi-monitor setups therefore would simply cause problems for me as a user.

MIDI inputs and electric and bass guitar inputs can be done in the control room. The only dry room necessary will house the drum kit and acoustic performance ability. This means the whole setup can be accomodated in 2 and a half rooms.